Scott-Brown's Otorhinolaryngology - Head and Neck Surgery PDF [PDF]

Citation preview

Scott-Brown’s Otorhinolaryngology, Head and Neck Surgery



Scott-Brown’s Otorhinolaryngology, Head and Neck Surgery 7th edition Lead editor: Michael Gleeson



Volume Part 1 Part 2 Part 3 Part 4 Part 5 Part 6 Part 7 Part 8 Part 9 Part 10 Part 11 Part 12



1 Cell biology, edited by Nicholas S Jones Wound healing, edited by Nicholas S Jones Immunology, edited by Nicholas S Jones Microbiology, edited by Nicholas S Jones Haematology, edited by Nicholas S Jones Endocrinology, edited by Nicholas S Jones Pharmacotherapeutics, edited by Martin J Burton Perioperative management, edited by Martin J Burton Safe and effective practice, edited by Martin J Burton Interpretation and management of data, edited by Martin J Burton Recent advances in technology, edited by Martin J Burton Paediatric otorhinolaryngology, edited by Ray Clarke



Volume 2 Part 13 The nose and paranasal sinuses, edited by Valerie J Lund Part 14 The neck, edited by John Hibbert Part 15 The upper digestive tract, edited by John Hibbert Part 16 The upper airway, edited by John Hibbert Part 17 Head and neck tumours, edited by John Hibbert Volume 3 Part 18 Plastic surgery of the head and neck, edited by John C Watkinson Part 19 The ear, hearing and balance, edited by George G Browning and Linda M Luxon Part 20 Skull base, edited by Michael Gleeson Index CD-ROM



George G Browning MD FRCS Professor of Otorhinolaryngology, MRC Institute of Hearing Research, Glasgow Royal Infirmary, Glasgow, UK Martin J Burton MA DM FRCS Senior Clinical Lecturer, University of Oxford; and Consultant Otolaryngologist, Oxford Radcliffe NHS Trust Oxford, UK Ray Clarke BSc DCH FRCS FRCS (ORL) Consultant Paediatric Otolaryngologist, Royal Liverpool University Children’s Hospital, Alder Hey, Liverpool, UK Michael Gleeson MD FRCS Professor of Otolaryngology and Skull Base Surgery, Institute of Neurology, University College London; and Consultant, Guy’s, Kings and St Thomas’ and the National Hospital for Neurology and Neurosurgery, London UK; and Honorary Consultant Skull Base Surgeon, Great Ormond Street Hospital for Sick Children, London, UK John Hibbert ChM FRCS Formerly Consultant Otolaryngologist, Department of Otolaryngology, Guy’s Hospital, London, UK Nicholas S Jones MD FRCS FRCS (ORL) Professor of Otorhinolaryngology, Queen’s Medical Centre, University of Nottingham, Nottingham UK Valerie J Lund MS FRCS FRCS (Ed) Professor of Rhinology, The Ear Institute, University College London, London, UK Linda M Luxon BSc MBBS FRCP Professor of Audiovestibular Medicine, University of London at University College London, Academic Unit of Audiovestibular Medicine; and Consultant Physician, National Hospital for Neurology and Neurosurgery; and Honorary Consultant Physician, Great Ormond Street Hospital for Children, London, UK John C Watkinson MSc MS FRCS (Ed, Glas, Land) DL0 Consultant Head and Neck and Thyroid Surgeon, Department of Otorhinolaryngology/Head and Neck Surgery, Queen Elizabeth Hospital, University of Birmingham NHS Trust, Birmingham, UK



Scott-Brown’s Otorhinolaryngology, Head and Neck Surgery 7th edition



Volume 1



Edited by



Michael Gleeson George G Browning, Martin J Burton, Ray Clarke, John Hibbert, Nicholas S Jones, Valerie J Lund, Linda M Luxon, John C Watkinson



Hodder Arnold www.hoddereducation.com



First published in Great Britain in 1952 by Butterworth & Co. Second edition 1965 Third edition 1971 Fourth edition 1979 Fifth edition 1987 Sixth edition 1997 This seventh edition published in Great Britain in 2008 by Hodder Arnold An imprint of Hodder Education, a part of Hachette Livre UK, 338 Euston Road, London NW1 3BH http://www.hoddereducation.com & 2008 Edward Arnold (Publishers) Ltd All rights reserved. Apart from any use permitted under UK copyright law, this publication may only be reproduced, stored or transmitted, in any form, or by any means with prior permission in writing of the publishers or in the case of reprographic production in accordance with the terms of licences issued by the Copyright Licensing Agency. In the United Kingdom such licences are issued by the Copyright Licensing Agency: Saffron House, 6-10 Kirby Street, London EC1N 8TS Whilst the advice and information in this book are believed to be true and accurate at the date of going to press, neither the author[s] nor the publisher can accept any legal responsibility or liability for any errors or omissions that may be made. In particular (but without limiting the generality of the preceding disclaimer) every effort has been made to check drug dosages; however it is still possible that errors have been missed. Furthermore, dosage schedules are constantly being revised and new side-effects recognized. For these reasons the reader is strongly urged to consult the drug companies’ printed instructions before administering any of the drugs recommended in this book. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN



978 0 340 808 931



1 2 3 4 5 6 7 8 9 10 Commissioning Editor: Joanna Koster Project Editor: Zelah Pengilley Production Controller: Lindsay Smith / Andre Sim Text and Cover Designer: Amina Dudhia Cover photograph & MEHAU KULYK/SCIENCE PHOTO LIBRARY Typeset in 10 pt Minion by Macmillan India Printed and bound in India. What do you think about this book? Or any other Hodder Arnold title? Please send your comments to www.hoddereducation.com



This eBook does not include the ancillary media that was packaged with the printed version of the book.



Contents



Contributors Preface How to use this book Abbreviations



PART 1 CELL BIOLOGY – EDITED BY NICHOLAS S JONES



xi xxix xxxi xxxiii



1



1



Molecular biology Michael Kuo and Richard Irving



2



Genetics Karen P Steel



15



3



Gene therapy Scott M Graham and John H Lee



23



4



Mechanisms of anticancer drugs Sarah Payne and David Miles



34



5



Radiotherapy and radiosensitizers Stewart G Martin and David AL Morgan



47



6



Apoptosis and cell death Michael Saunders



56



7



Stem cells A John Harris and Archana Vats



66



PART 2 WOUND HEALING – EDITED BY NICHOLAS S JONES



3



85



8



Soft and hard tissue repair Stephen R Young and Melissa Calvin



9



Skin flap physiology A Graeme B Perks



110



Biomaterials, tissue engineering and their applications Colin A Scotchford, Matthew Evans and Archana Vats



118



10



PART 3 IMMUNOLOGY – EDITED BY NICHOLAS S JONES



87



131



11



Defence mechanisms Ian Todd and Richard J Powell



133



12



Allergy: basic mechanisms and tests Stephen R Durham and Graham Banfield



144



vi ] Contents 13



Evaluation of the immune system Elizabeth Drewe and Richard J Powell



167



14



Primary immunodeficiencies Elizabeth Drewe and Richard J Powell



174



15



Rheumatological diseases Adrian Drake-Lee



183



PART 4 MICROBIOLOGY – EDITED BY NICHOLAS S JONES



193



16



Microorganisms Vivienne Weston



195



17



Viruses and antiviral agents Paul Simons and Karl G Nicholson



204



18



Fungi Juliette Morgan and David W Warnock



213



19



Antimicrobial therapy Vivienne Weston



228



20



HIV and otolaryngology Thomas A Tami and Jahmal A Hairston



238



PART 5 HAEMATOLOGY – EDITED BY NICHOLAS S JONES



251



21



Blood groups, blood components and alternatives to transfusion Fiona Regan and Ian Gabriel



253



22



Haemato-oncology Clare Wykes and Fiona Regan



265



23



Haemostasis: normal physiology, disorders of haemostasis and thrombosis and their management Fiona Regan



278



PART 6 ENDOCRINOLOGY – EDITED BY NICHOLAS S JONES



293



24



The pituitary gland: anatomy and physiology John Hill



295



25



The pituitary: imaging and tests of function Alan P Johnson, Swarupsinh Chavda and Paul Stewart



303



26



The thyroid gland: anatomy and physiology Julian A McGlashan



314



27



The thyroid gland: function tests and imaging Susan Clarke



327



28



The thyroid: nonmalignant disease Lorraine M Albon and Jayne A Franklyn



338



29



The parathyroid glands: anatomy and physiology Mateen H Arastu and William J Oweny



367



30



Parathyroid function tests and imaging David Hosking



379



31



Parathyroid dysfunction: medical and surgical therapy E Dinakara Babu, Bill Fleming and JA Lynn



387



32



Head and neck manifestations of endocrine disease Jonathan M Morgan and Thomas McCaffrey



398



Contents PART 7 PHARMACOTHERAPEUTICS – EDITED BY MARTIN J BURTON



] vii 405



33



Drug administration and monitoring Geraldine Gallagher



407



34



Corticosteroids in otolaryngology Niels Mygind and Jens Thomsen



418



35



Drug therapy in otology Wendy Smith and Martin Burton



429



36



Drug therapy in rhinology Wendy Smith and Grant Bates



436



37



Drug therapy in laryngology and head and neck surgery Wendy Smith and Rogan Corbridge



446



PART 8 PERIOPERATIVE MANAGEMENT – EDITED BY MARTIN J BURTON



455



38



Preparation of the patient for surgery Adrian Pearce



457



39



Recognition and management of the difficult airway Adrian Pearce



467



40



Adult anaesthesia Andrew D Farmery and Jaideep J Pandit



488



41



Paediatric anaesthesia Alistair Cranston



507



42



Adult critical care Gavin G Lavery



526



43



Paediatric intensive care Helen Allen and Rob Ross Russell



542



PART 9 SAFE AND EFFECTIVE PRACTICE – EDITED BY MARTIN J BURTON



549



44



Training, accreditation and the maintenance of skills Paul O’Flynn



551



45



Communication and the medical consultation Damian Gardner-Thorpe and Richard Canter



559



46



Clinical governance: Improving the quality of patient care Debbie Wall, Patrick J Bradley and Aidan Halligan



568



47



Medical ethics Katherine Wasson



581



48



Medical jurisprudence and otorhinolaryngology Maurice Hawthorne and Desmond Watson



594



PART 10 INTERPRETATION AND MANAGEMENT OF DATA – EDITED BY MARTIN J BURTON



613



49



Epidemiology Jan HP van der Meulen and David A Lowe



615



50



Outcomes research Iain RC Swan



633



51



Evidence-based medicine Martin J Burton



645



52



Critical appraisal skills Martin Dawes



649



viii ] Contents PART 11 RECENT ADVANCES IN TECHNOLOGY – EDITED BY MARTIN J BURTON



673



53



Functional magnetic resonance imaging: Principles and illustrative applications for otolaryngology Paul M Matthews



675



54



Positron emission tomography and integrated PET/computed tomography Wai Lup Wong



684



55



Image-guided surgery, 3D planning and reconstruction Ghassan Alusi and Michael Gleeson



701



56



Ultrasound in ear, nose and throat practice Keshthra Satchithananda and Paul S Sidhu



711



57



Interventional techniques James V Byrne



731



58



Laser principles in otolaryngology, head and neck surgery Brian JG Bingham



742



59



Electrophysiology and monitoring Patrick R Axon and David M Baguley



748



60



Optical coherence tomography Mariah Hahn and Brett E Bouma



755



61



Contact endoscopy Mario Andrea and Oscar Dias



762



PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY – EDITED BY RAY CLARKE



769



62



Introduction Ray Clarke



771



63



The paediatric consultation Ray Clarke and Ken Pearman



776



64



ENT input for children with special needs Francis Lannigan



783



65



Head and neck embryology T Clive Lee



792



66



Molecular otology, development of the auditory system and recent advances in genetic manipulation Henry Pau



811



67



Hearing loss in preschool children: screening and surveillance Kai Uus and John Bamford



821



68



Hearing tests in children Glynnis Parker



834



69



Investigation and management of the deaf child Sujata De, Sue Archbold and Ray Clarke



844



70



Paediatric cochlear implantation Joseph G Toner



860



71



Congenital middle ear abnormalities in children Jonathan P Harcourt



869



72



Otitis media with effusion George Browning



877



73



Acute otitis media in children Peter Rea and John Graham



912



Contents



] ix



74



Chronic otitis media in childhood John Hamilton



928



75



Management of congenital deformities of the external and middle ear David Gault and Mike Rothera



965



76



Disorders of speech and language in paediatric otolaryngology Ray Clarke and Siobhan MCMahon



990



77



Cleft lip and palate Chris Penfold



996



78



Craniofacial anomalies: genetics and management Dean Kissun, David Richardson, Elizabeth Sweeney and Paul May



1019



79



Vertigo in children Gavin AJ Morrison



1040



80



Facial paralysis in childhood SS Musheer Hussain



1052



81



Epistaxis in children Ray Clarke



1063



82



Nasal obstruction in children Michelle Wyatt



1070



83



Paediatric rhinosinusitis Glenis Scadding and Helen Caulfield



1079



84



The adenoid and adenoidectomy Peter J Robb



1094



85



Obstructive sleep apnoea in childhood Helen M Caulfield



1102



86



Stridor David Albert



1114



87



Acute laryngeal infections Susanna Leightony



1127



88



Congenital disorders of the larynx, trachea and bronchi Martin Bailey



1135



89



Laryngeal stenosis Michael J Rutter and Robin T Cotton



1150



90



Paediatric voice disorders Ben Hartley



1167



91



Juvenile-onset recurrent respiratory papillomatosis Michael Kuo and William J Primrose



1174



92



Foreign bodies in the ear and the aerodigestive tract in children A Simon Carney, Nimesh Patel and Ray Clarke



1184



93



Tracheostomy and home care Michael Saunders



1194



94



Cervicofacial infections in children Ben Hartley



1210



95



Diseases of the tonsil William S McKerrow



1219



96



Tonsillectomy William S McKerrow and Ray Clarke



1229



97



Salivary gland disorders in childhood Peter D Bull



1242



98



Tumours of the head and neck in childhood Fiona B MacGregor



1251



x ] Contents 99



Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma Peter D Bull



1264



100



Gastro-oesophageal reflux and aspiration Haytham Kubba



1272



101



Diseases of the oesophagus, swallowing disorders and caustic ingestion Lewis Spitz



1282



102



Imaging in paediatric ENT Neville Wright



1295



103



Medical negligence in paediatric otolaryngology Maurice Hawthorne



1305



Please note: The table of contents for all three volumes can be found on the Scott-Brown website at: www.scottbrownENT.com. The index for all three volumes is included in Volume 3.



Contributors



Victor J Abdullah MBBS FRCS (Eng) FRCS (Edin) Consultant, Department of ENT United Christian Hospital; and Chief of Service in ENT Kowloon East Cluster, Hospital Authority; and Honorary Clinical Associate Professor Department of Otorhinolaryngology, Head and Neck Surgery The Chinese University of Hong Kong Shatin, Hong Kong Jose M Acuin MD MSc Professor in Otolaryngology–Head and Neck Surgery De La Salle Health Sciences Campus Dasmarinas, Cavite, Philippines Richard Adamson MB BS FRCS Consultant Otolaryngologist/Head and Neck Surgeon Fife Hospitals NHS Trust Scotland, UK David Albert FRCS Lead clinician Department of Otolaryngology Hospital for Sick Children Great Ormond Street London, UK Lorraine M Albon MSC MRCP Consultant Endocrinologist and Acute Physician Queen Alexandra Hospital Portsmouth, UK Helen Allen MBChB MRCP MRCPCH Specialist Registrar in Paediatrics Addenbrookes Hospital Cambridge, UK



Mario Andrea MD PhD Professor and Chairman Department of Otolaryngology, Voice and Communication Disorders Faculty of Medicine of Lisbon Lisbon, Portugal Jawaher Ansari MBBS MRCP FRCR Specialist Registrar in Clinical Oncology Queen Elizabeth Hospital Birmingham, UK Mateen H Arastu MBBS BSc MRCS (Eng) Specialist Registrar in Trauma and Orthopaedic Surgery South West Thamas Region, UK Sue Archbold MPhil Education Co-ordinator The Ear Foundation Nottingham, UK Stig Arlinger PhD Professor of Technical Andiology Department of Clinical and Experimental Medicine Division of Technical Audiology Linko¨ping University Linko¨ping, Sweden Marcus Atlas Professor of Otolaryngology University of Western Australia; and Director, Ear Science Institute Australia Sir Charles Gairdner Hospital Nedlands, Western Australia Patrick R Axon MD FRCS (ORL-HNS) Consultant Otologist and Skull Base Surgeon Department of Otolaryngology Cambridge University Hospitals Cambridge, UK E Dinakara Babu MS FRCS (Eng) FRCS (Ire) FRCS (Glas) FRCS (Inter collegiate) Diploma



Ghassan Alusi PhD FRCS (ORL-HNS) Consultant Otolaryngologist ENT Department Barts and the London NHS Trust London, UK



in Laparoscopy (France)



Consultant and Clinical Head Breast and Endocrine Surgery Hillingdon Hospital Uxbridge, UK



xii ] Contributors Claus Bachert MD PhD Department of Oto-Rhino-Laryngology Ghent University Hospital Ghent, Belgium



Graham Banfield MD DLO FRCS Ed (ORL-HNS) Consultant ENT Surgeon The Great Western Hospital Swindon, UK



Lydia Badia FRCS (ORL-HNS) Consultant ENT Surgeon Rhinology and Facial Plastics Royal National Throat, Nose and Ear Hospital London, UK



Jane A Baran PhD Professor and Chair Department of Communication Disorders University of Massachusetts Amherst Amherst, MA, USA



Jose V Sebastian Bagan MD DDS PhD Hospital General Universitario de Valencia Valencia, Spain



Michael E Basery Formerly of Department of Environmental Health Sciences Johns Hopkins School of Hygiene and Public Health Baltimore, MD, USA



Dan Bagger-Sjo¨ba¨ck MD PhD Professor in Ear, Nose and Throat Diseases Karolinska Institute Stockholm, Sweden David M Baguley BSc MSC MBA PhD Consultant Audiological Scientist Cambridge University Hospitals NHS Foundation Trust Cambridge, UK S Bahadur MS FAMS PhD Professor of Otolaryngology and Head/Neck Surgery All India Institute of Medical Sciences New Delhi, India Martin Bailey BSc FRCS Consultant Paediatric Otolaryngologist Great Ormond Street Hospital for Children London, UK John Bamford BA PhD Ellis Llywd Jones Professor of Audiology and Deaf Education School of Psychological Sciences Faculty of Medical and Human Sciences University of Manchester Manchester, UK Doris-Eva Bamiou MD MSc PhD Consultant in Audiological Medicine Department of Neuro-otology National Hospital for Neurology and Neurosurgery; and Honorary Senior Lecturer Academic Unit of Audiological Medicine University College London Institute of Child Health Great Ormond Street Hospital London, UK Dev Banerjee BSc MBChB MD MRCP Consultant Respiratory and Sleep Physician Sleep and Ventilation Unit Department of Respiratory Medicine Birmingham Heartlands Hospital Birmingham, UK



Grant Bates BSc (Hons) BM Bch FRCS Consultant ENT Surgeon John Radcliffe Hospital Oxford, UK Nigel Beasley FRCS (ORL-HNS) Consultant in Otolaryngology Queen’s Medical Centre Nottingham, UK Michael S Benninger MD Chairman Department of Otolaryngology-Head and Neck Surgery Henry Ford Hospital Detroit, USA Barry KB Berkovitz MSc PhD BDS FDSRCS (Eng) Department of Anatomy and Human Sciences School of Biomedical and Health Sciences King’s College London, UK Thanos Bibas MSc (Lond) DrMed FRCSI (Otol) Consultant and Lecturer in Otolaryngology Hippokrateion Hospital University of Athens Athens, Greece Carsten Bindslev-Jensen MD PhD DSc Head, Allergy Center Department of Dermatology Odense University Hospital Odense, Denmark Brian JG Bingham MBChB FRCS Ed Glas Consultant ENT Surgeon Department of Otolaryngology Victoria Infirmary; and Honorary Senior Lecturer in Otorhinolaryngology University of Glasgow Glasgow, UK y Deceased



Contributors Martin A Birchall MD (Cantab) FRCS FRCS (Oto) FRCS (ORL) Professor of Laryngology University of Bristol Bristol, UK; and Consultant in Otolaryngology Head and Neck Surgery Royal United Hospital Bath NHS Trust Bath, UK Ian D Bottrill BM FRCS (ORL) Consultant Otolaryngologist; and Honorary Senior Lecturer ENT Department University of Oxford John Radcliffe Hospital Oxford, UK An Boudewyns MD PhD Professor of Otorhinolaryngology Department of Otorhinolaryngology, Head and Neck Surgery University of Antwerp Hospital – University of Antwerp Antwerp, Belgium Brett E Bouma PhD Associate Professor, Department of Dermatology Member of the Faculty of the Harvard-MIT Division of Health Science and Technology Harvard Medical School Boston, MA, USA Jean Bousquet MD Service des Maladies Respiratoires Hoˆpital Arnaud de Villeneuve Montpellier, France Patrick J Bradley MBA FRCS FRACS (Hon) FRCSLT (Hon) Professor of Head and Neck Oncologic Surgery Department of ORL-HNS Nottingham University Hospitals Queen’s Medical Centre Campus Nottingham, UK Stefan Brew MB ChB MHB (Hons) MSc FRANZCR Consultant Neuroradiologist The National Hospital for Neurology and Neurosurgery London, UK Steven M Bromley MD Clinical Assistant Professor of Neurology (Medicine) and Attending Neurologist University of Medicine and Dentistry of New Jersey Robert Wood Johnson Medical School Cooper University Hospital, Camden, NJ; and Bromley Neurology PC, Audubon, NJ; and Smell and Taste Center University of Pennsylvania Medical Center Philadelphia, PA USA



] xiii



Adolfo M Bronstein MD PhD FRCP Professor of Clinical Neuro-otology; and Head, Neuro-otology Unit Division of Neuroscience and Mental Health Charing Cross Hospital, Imperial College London; and Honorary Consultant Neurologist Charing Cross Hospital; and National Hospital for Neurology and Neurosurgery Queen Square, London, UK Gerald Brookes FRCS Consultant Otologist and Neuro-Otologist The National Hospital for Neurology and Neurosurgery; and The Royal National Throat, Nose and Ear Hospital London, UK George G Browning MD FRCS Professor of Otorhinolaryngology MRC Institute of Hearing Research Glasgow Royal Infirmary Glasgow, UK Peter D Bull MB FRCS Consultant Otolaryngologist Royal Hallamshire Hospital and Sheffield Children’s Hospital; and Honorary Senior Clinical Lecturer University of Sheffield Sheffield, UK Martin J Burton MA DM FRCS Senior Clinical Lecturer University of Oxford; and Consultant Otolaryngologist Oxford Radcliffe NHS Trust Oxford, UK James V Byrne MD FRCS FRCR Professor of Neuroradiology University of Oxford; and Consultant Neuroradiologist John Radcliffe Hospital Oxford, UK Melissa Calvin MRCOG PhD BSC (Hons) Department of Obstetrics and Gynaecology Lister Hospital Stevenage, UK Richard Canter PhD FRCS FRCS (Otol) Hon FRCS (Edin) Consultant Otolaryngologist Royal United Hospital; and Honorary Senior Lecturer University of Bath Bath, UK Paul Carding FRCSLT Professor of Voice Pathology The Medical School, Newcastle University; and Head of Speech, Voice and Swallowing Department Otolarygology Directorate, Freeman Hospital Newcastle on Tyne, UK



xiv ] Contributors A Simon Carney BSc (Hons) MBChB FRCS FRACS MD Associate Professor and Head of ENT Unit Flinders University and Flinders Medical Centre Adelaide, South Australia Anna Cassoni BSc FRCP FRCR Consultant in Clinical Oncology University College Hospitals NHS Foundation Trust London, UK Helen Caulfield MBBS FRCS (ORL-HNS) Consultant ENT Surgeon Department of Otolaryngology The Royal Free Hospital London, UK Roderick Cawsony MD FRCPath Formerly Emeritus Professor of Oral Medicine Guy’s Hospital London, UK Borka Ceranic MD ENTspec PhD Consultant in Audiological Medicine Department of Audiology St George’s Hospital London, UK Swarupsinh V Chavda MBChB DMRD FRCR Consultant Neuroradiologist and Honorary Senior Lecturer Queen Elizabeth Hospital University Hospital Birmingham Foundation Trust Birmingham, UK Elfy B Chevretton BSc MBBS FRCS MS Consultant Otolaryngologist Department of ENT Surgery Guy’s and St Thomas’ NHS Trust London, UK Peter Clarke BSc FRCS Consultant ENT Surgeon Charing Cross Hospital London, UK Ray Clarke BSc DCH FRCS FRCS (ORL) Consultant Paediatric Otolaryngologist Royal Liverpool University Children’s Hospital Alder Hey, Liverpool, UK Susan Clarke MSc FRCP FRCR Senior Lecturer and Consultant Physician Department of Nuclear Medicine Guy’s and St Thomas’ Hospital London, UK y Deceased



Rogan Corbridge MBBS BSc FRCS FRCS (ORL) Consultant ENT Surgeon Oxford Centre for Head and Neck Oncology John Radcliffe Hospital Oxford, UK Robin T Cotton MD Director, Pediatric Otolaryngology–Head and Neck Surgery Children’s Hospital Medical Center; and Professor of Pediatric Otolaryngology Department of Otolaryngology, Head and Neck Surgery University of Cincinnati College of Medicine Cincinnati, OH, USA Graham J Cox BDS FRCS Consultant ENT Surgeon; and Macmillan Head and Neck Surgical Oncologist John Radcliffe Hospital Oxford, UK Alistair Cranston MBBS FRCA Consultant Anaesthetist Birmingham Children’s Hospital Birmingham, UK Cor WRJ Cremers Department of Otorhinolaryngology University Medical Center St Radboud Nijmegen, The Netherlands Ian S Curthoys PhD Emeritus Professor School of Psychology University of Sydney Sydney, Australia Rosalyn A Davies FRCP PhD Consultant in Audiovestibular Medicine Department of Neuro-Otology The National Hospital for Neurology and Neurosurgery; and Honorary Senior Lecturer Institute of Neurology Queen Square, London, UK Martin Dawes MB BS MD (Lond) FRCGP Chair, Family Medicine McGill University Quebec, Canada Ranit De MPhil FRCS (ORL-HNS) Consultant ENT Surgeon University Hospital North Staffordshire NHS Trust; and Stoke-on-Trent and Mid-Staffordshire NHS Trust Stafford, UK



Contributors Sujata De MBBS FRCS (ORL-HNS) Consultant Paediatric Otorhinolaryngologist Alder Hey Children’s Hospital Liverpool, UK Charles Diamond FRCS (Glas) Dip Pall Med Honorary Consultant Otolaryngologist Freeman Hospital and St. Oswald’s Hospice Newcastle upon Tyne, UK Oscar Dias MD PhD Associate Professor Department of Otolaryngology, Voice and Communication Disorders Faculty of Medicine of Lisbon Lisbon, Portugal Harvey Dillon B Eng (Hons I) PhD Director of Research National Acoustic Laboratories Chatswood, Australia Richard L Doty PhD Smell and Taste Center University of Pennsylvania Medical Center Philadelphia, PA, USA M Stephen Dover FDSRCS FRCS Consultant Oral and Maxillofacial Surgeon, and Honorary Senior Lecturer Department of Maxillofacial Surgery University Hospital Birmingham NHS Foundation Trust Birmingham, UK Wolfgang Draf Prof Dr Med Dr HC FRCS Director, Department of Ear, Nose and Throat Diseases, Head and Neck Surgery International Neuroscience Institute Hanover, Germany Adrian Drake-Lee MMEd PhD FRCS Consultant ENT Surgeon Queen Elizabeth Hospital University Hospital, NHS Trust Birmingham, UK



Sunil Narayan Dutt MS DNB PhD FRCS Ed (ORL-HNS) DLO (Eng) DORL Senior Consultant and Clinical Coordinator Department of Otolaryngology and Head and Neck Surgery Apollo Group of Hospitals Bangalore, India Charles East FRCS Consultant Otolaryngologist, Head and Neck Surgeon The Royal Free Hampstead NHS Trust London, UK Ronald Eccles BSc PhD DSc Director, Common Cold Centre and Healthcare Clinical Trials Cardiff School of Biosciences Cardiff University Cardiff, Wales, UK D Gareth R Evans MD FRCP Professor, Department of Medical Genetics St Mary’s Hospital Manchester, UK Matthew Evans PhD Editorial Manager Caudex Medical Oxford, UK Johannes J Fagan MBChB, FCS (SA) MMed (Otol) Professor and Chairman Division of Otolaryngology University of Cape Town and Groote Schuur Hospital Cape Town, South Africa Andrew D Farmery BSc MA MD FRCA Senior Lecturer in Anaesthetics Nuffield Department of Anaesthetics University of Oxford Oxford, UK Neil Fergie FRCS MD Consultant in Otorhinolaryngology Head and Neck Surgery Kings Mill Hospital, Mansfield; and Queens Medical Centre Nottingham, UK



Elizabeth Drewe MBBS PhD MRCP MRCPath Consultant Clinical Immunologist Nottingham University Hospitals NHS Trust Nottingham, UK



John Fleetham MB BS FRCP(C) Professor of Medicine Respiratory Division University of British Columbia and Vancouver Hospital Vancouver, BC, Canada



Stephen R Durham MA MD FRCP Professor of Allergy and Respiratory Medicine Imperial College School of Medicine National Heart and Lung Institute London, UK



Bill Fleming FRACS FRCS Consultant Endocrine Surgeon Hammersmith Hospital Imperial Healthcare NHS Trust London, UK



] xv



xvi ] Contributors Liam M Flood MB BS FRCS ENT Consultant James Cook University Hospital Middlesbrough, UK Adrian Fourcin PhD FIoA Emeritus Professor, Department of Phonetics and Linguistics University College London London, UK Jayne A Franklyn MD PhD FRCP FMedSci Professor of Medicine Division of Medical Sciences University of Birmingham Queen Elizabeth Hospital Birmingham, UK Nicole JM Freling MD PhD Department of Radiology Academic Medical Centre Amsterdam, The Netherlands David N Furness BSc PhD School of Life Sciences Keele University Staffordshire, UK Ian Gabriel MBBS BSC (Hons) MRCP (UK) DipRCPath Department of Haematology St Mary’s Hospital Campus Imperial College School of Medicine London, UK Geraldine Gallagher FRCSI Antrim Area Hospital Belfast, Northern Ireland Damian Gardner-Thorpe General Practitioner The Pulteney Practice Bath, UK



MRCP (UK) MRCS (Eng) MRCGP (UK)



David Gault MB ChB FRCS Consultant Plastic Surgeon London Centre for Ear Reconstruction The Portland Hospital London, UK Garrick A Georgeu MB ChB FRCS (Ed) FRCS PLAS MSc Plastic Surgery Department Selly Oak Hospital University Hospital Birmingham Birmingham, UK Kevin P Gibbin Consultant Otolaryngologist University Hospital Nottingham, UK



Ralph W Gilbert MD FRCS (C) Professor of Otolaryngology/Head and Neck Surgery University of Toronto University Health Network Princess Margaret Hospital Toronto, Ontario, Canada John Glaholm BSc FRCP FRCR (Clin Oncol) Consultant Clinical Oncologist Cancer Centre, Queen Elizabeth Hospital Birmingham, UK Michael Gleeson MD FRCS Professor of Otolaryngology and Skull Base Surgery Institute of Neurology University College London; and Consultant Guy’s, Kings and St Thomas’ and the National Hospital for Neurology and Neurosurgery London UK; and Honorary Consultant Skull Base Surgeon Great Ormond Street Hospital for Sick Children London, UK Kees Graamans MD PhD Professor and Chairman Department of Otorhinolaryngology University Medical Centre Nijmegen Nijmegen, The Netherlands John Graham MA BM BCh FRCS Consultant Otolaryngologist The Royal National Throat, Nose, and Ear Hospital Gray’s Inn Road London, UK Scott M Graham MD Professor Department of Otolaryngology – Head and Neck Surgery The University of Iowa; and Director of Rhinology, University of Iowa Hospital and Clinics Iowa City, IA, USA Luisa F Grymer MD Grymer Private Hospital Aarhus, Denmark Carole M Hackney BSc PhD Department of Physiology, Development and Neuroscience University of Cambridge Cambridge, UK Mariah Hahn PhD Assistant Professor Department of Chemical Engineering Texas A&M University Texas, USA



Contributors Jahmal A Hairston MD Department of Otolaryngology University of Cincinnati College of Medicine Cincinnati, OH, USA



Peter Haughton BSc PhD Formerly Clinical Scientist and Head of Audiology Department of Medical Physics Royal Hull Hospitals Kingston upon Hull, UK



Aidan Halligan FRCP FRCOG MA MD MRCPI FFPHM Chief Executive, Elision Health Ltd; and Deputy Chief Medical Officer, England (2003–2005); and Director of Clinical Governance for the NHS (1999–2006) Leicester, UK



Maurice Hawthorne FRCS Consultant Otolaryngologist, Head and Neck Surgeon James Cook University Hospital Middlesbrough, UK



G Michael Halmagyi MD FRACP Clinical Professor Department of Neurology Royal Prince Alfred Hospital Sydney, Australia



John Hibbert ChM FRCS Formerly Consultant Otolaryngologist Department of Otolaryngology Guy’s Hospital London, UK



John Hamilton FRCS Department of Otolaryngology Gloucestershire Royal Hospital Gloucester, UK



John M Hilinski MD Facial Plastic and Reconstructive Surgery San Diego Face and Neck Specialties University of California, San Diego Medical Center San Diego, CA, USA



Ravinder PS Harar FRCS (ORL-HNS) Specialist Registrar Otolaryngology, Head and Neck Surgery The National Hospital for Neurology and Neurosurgery; and The Royal National Throat, Nose and Ear Hospital London, UK



John Hill FRCS FRCSEd Department of ORL-HNS The Freeman Hospital Newcastle upon Tyne, UK



Jonathan P Harcourt MA FRCS Consultant ENT Surgeon Charing Cross Hospital London, UK Meredydd Harries FRCS MSC (Voice) Consultant ENT Surgeon The Royal Sussex County Hospital Brighton, UK A John Harris PhD Developmental Biology Laboratory Department of Physiology University of Otago Dunedin, New Zealand Douglas Harrison FRCS Consultant Plastic Surgeon The Wellington Hospital London, UK Ben Hartley MBBS BSc FRCS Consultant Paediatric Otolaryngologist Great Ormond Street Hospital for Children London, UK Richard J Harvey BScMed MB BS FRACS Nuffield Fellow, University of Oxford, UK; and Rhinologist and Endoscopic Skull Base Surgeon St Vincent’s Hospital Sydney, Australia



Malcolm P Hilton MA BM BCh FRCS (ENG) FRCS (ORL-HNS) Consultant Otolaryngologist Royal Devon and Exeter Hospital; and Honorary Clinical Lecturer Peninsula Medical School University of Exeter Exeter, UK Lisa J Hirst BSc PhD Cert MCRSLT Head of Service Speech and Language Therapist Salisbury District Hospital Wiltshire, UK Simon Holmes BDS MBBS (Hons) FDS RCS Eng FRCS (OMFS) Consultant Oral and Maxillofacial Surgeon Barts and The London NHS Trust London, UK David Hosking MD FRCP Consultant Physician Division of Mineral Metabolism City Hospital Nottingham, UK David J Howard BSc FRCS FRCS (Ed) Emeritus Senior Lecturer University College London; and Consultant Head and Neck Surgeon Royal Throat, Nose and Ear Hospital; and Charing Cross Hospital London, UK



] xvii



xviii ] Contributors SS Musheer Hussain MB MSc (Manc) FRCS (ORL) Consultant Otolaryngologist and Head ENT and Audiology Services Ninewells Hospital and Medical School; and Honorary Senior Lecturer and Director Temporal Bone Laboratory; and Licenced Teacher of Anatomy Department of Otolaryngology University of Dundee Dundee, UK



Petros D Karkos MPhil AFRCSI Specialist Registrar in Otolaryngology Mersey Deanery Chester, UK



Richard M Irving MD FRCS (ORL-HNS) Consultant in Neurotology University Hospital Birmingham NHS Trust and Diana Princess of Wales (Birmingham Childrens) Hospital; and Honorary Senior Lecturer University of Birmingham Birmingham, UK



Andras Armand Kemeny MD FRCS The National Centre for Stereotactic Radiosurgery Royal Hallamshire Hospital Sheffield, UK



Mark E Izzard MB BS FRACS Senior Lecturer in Otolaryngology University of Auckland; and Consultant Head and Neck Surgeon Auckland District Health Board Auckland, New Zealand Jean-Pierre Jeannon MB ChB FRCS (OTO) FRCS (ORL) Consultant Ear Nose and Throat/Head and Neck Surgeon Guy’s and St Thomas’ Hospital London, UK Chris R Jennings Department of Otolaryngology The Queen Elizabeth Hospital Birmingham, UK Dan Jiang PhD FRCSI (Otol) FRCS (ORL-HNS) Consultant Otolaryngologist Department of Otolaryngology, Head and Neck Surgery Guy’s, St Thomas’ and Evelina Children’s Hospitals London, UK Alan P Johnson FRCS Department of Otolaryngology Queen Elizabeth Hospital Birmingham, UK



Gerard Kelly MD FRCS (ORL-HNS) Consultant Ear, Nose and Throat and Skull Base Surgeon; and Clinical Director of Otolaryngology The Leeds Teaching Hospitals NHS Trust Leeds, UK



David W Kennedy MD FACS FRCSI Department of Otorhinolaryngology-Head and Neck Surgery University of Pennsylvania Philadelphia, PA, USA Richard SC Kerr BSc MS MBBS FRCS Consultant Neurosurgeon Oxford Skull Base Unit Oxford Radcliffe NHS Trust; and Honorary Senior Lecturer, University of Oxford Oxford, UK Dean Kissun FRCS (OMFS) Consultant Maxillofacial Surgeon NHS Lothian Edinburgh, UK Jean Michel Klossek MD ENT Professor, University of Poitiers; and ENT and Head and Neck Surgery Department University Hospital Jean Bernard Poitiers, France Gary Kroukamp MBChB FCORL (SA) Faculty of Health Sciences University of Stellenbosch, Tygerberg Hospital Tygerberg, South Africa Haytham Kubba MPhil MD FRCS (ORL-HNS) Consultant Paediatric Otolaryngologist, Head and Neck Surgeon The Royal Hospital for Sick Children Glasgow, UK



Andrew S Jones MB BCh MD FRCSE FRCS Professor, School of Cancer Studies Division of Surgery and Oncology Royal Liverpool University Hospital Liverpool, UK



Michael Kuo PhD FRCS (Eng) FRCS (ORL-HNS) DCH Consultant Otolaryngologist – Head and Neck Surgeon Birmingham Children’s Hospital Birmingham, UK



Nicholas S Jones MD FRCS FRCS (ORL) Professor of Otorhinolaryngology Queens Medical Centre University of Nottingham Nottingham, UK



Francis Lannigan MB ChB MD FRCS (Eng) Ed (ORL) FRACS Department of Otolaryngology – Head and Neck Surgery Princess Margaret Hospital for Children; and Clinical Professor, The University of Western Australia Perth, Western Australia



Contributors Gavin G Lavery MB BCh BAO FCARCSI MD Director of Critical Care Services Royal Hospitals, Belfast, UK; and Visiting Professor, Faculty of Life and Health Sciences University of Ulster Northern Ireland Brian Leatherbarrow BSc MBChB DO FRCS FRCOphth Consultant Ophthalmic, Oculoplastic and Orbital Surgeon Manchester Royal Eye Hospital Manchester, UK John H Lee MD Assistant Professor Department of Otolaryngology – Head and Neck Surgery University of Iowa Iowa City, IA, USA T Clive Lee MA MSc MD PhD FRCSI FRCSEd CEng FIEI Professor of Anatomy Royal College of Surgeons in Ireland Dublin, Ireland Susanna Leightony BSc FRCS (ORL-HNS) Formerly Consultant Paediatric Otolaryngologist Great Ormond Street Hospital for Children London, UK Paula Leslie PhD Cert MRCSLT Associate Professor Communication Science and Disorders University of Pittsburgh Pittsburgh, PA, USA Tristram HJ Lesser AKC FRCSEd MS Otolaryngology/Head and Neck Surgery University Hospital Liverpool, UK James W Loock MB ChB (UCT) FCS (SA) ORL Professor and Head Department of Otorhinolaryngology University of Stellenbosch Tygerberg Hospital Cape Town, South Africa David A Lowe BSc FRCSEd FRCS Research Fellow Clinical Effectiveness Unit The Royal College of Surgeons of England London, UK Valerie J Lund MS FRCS FRCS (Ed) Professor of Rhinology The Ear Institute University College London London, UK y Deceased



] xix



Linda M Luxon BSc MBBS FRCP Professor of Audiovestibular Medicine University of London at University College London Academic Unit of Audiovestibular Medicine; and Consultant Physician, National Hospital for Neurology and Neurosurgery; and Honorary Consultant Physician Great Ormond Street Hospital for Children London, UK JA Lynn MS FRCS Consultant Surgeon Cromwell Hospital London, UK Fiona B MacGregor MBChB FRCS (ORL HNS) Consultant Otolaryngologist Royal Hospital for Sick Children Glasgow, UK Ian S Mackay FRCS Consultant ENT Surgeon Royal Brompton Hospital and Charing Cross Hospital London, UK Kenneth MacKenzie MB ChB FRCS (Ed) Consultant Otorhinolaryngologist and Honorary Senior Lecturer Glasgow Royal Infirmary University of Glasgow Glasgow, UK Marcelle Macnamara MA MBBS FRCS MPhil FRCS (ORL-HNS) Retired Consultant Otolaryngologist, Head and Neck Surgeon Heart of England Foundation Trust Birmingham, UK Arnold GD Maran MD DSc FRCS (Ed, Eng, Glasg) FRCP FDS Emeritus Professor of Otolaryngology University of Edinburgh Edinburgh, UK Andrew H Marshall Bsc MBBS FRCS Consultant Otolaryngologist Department of Otorhinolaryngology and Head and Neck Surgery University Hospital Nottingham, UK Stewart G Martin BSC (Hons) MSC PhD Associate Professor of Oncology MSc Course Director and Head of Translational Radiation Biology Research Group University of Nottingham Nottingham, UK Robert C Mason BSc ChM MD FRCS Consultant Upper GI Surgeon Guy’s and St Thomas’ Hospitals London, UK



xx ] Contributors Lesley Mathieson FRCSLT Visiting Lecturer in Voice Pathology The Ear Institute University College London; and Honorary Research Adviser Speech and Language Therapy Department Royal National Throat Nose and Ear Hospital London, UK Paul M Matthews MA (Oxon) MD DPhil FRCP Vice-President for Imaging Genetics and for Neurology; and Head, GSK Clinical Imaging Center Clinical Pharmacology and Discovery Medicine GlaxoSmithKline; and Professor of Clinical Neurosciences Department of Clinical Neurosciences Imperial College, London; and (Hon.) MRC Clinical Research Professor Department of Clinical Neurology University of Oxford Oxford, UK Paul May MBBS FRCS FRCPCH Consultant Paediatric Neurosurgeon Craniofacial Unit, Alder Hey Children’s Hospital Liverpool, UK Thomas McCaffrey MD PhD Professor and Chair Department of Otolaryngology Head and Neck Surgery University of South Florida Tampa, FL, USA Leo McClymont MBChB MD FRCSEd FRCSGlas Raigmore Hospital Highland Acute Hospitals NHS Trust Inverness Inverness, UK Andrew MCCombe MD FRCS (ORL) Consultant ENT Surgeon Frimley Park Hospital Camberley, UK Gerald W McGarry MD FRCS (RCPSGlasg) FRCS(Ed) FRCS (ORL-HNS) Consultant Otorhinolaryngologist Glasgow Royal Infirmary; and Honorary Senior Lecturer University of Glasgow Glasgow, UK Julian A McGlashan MBBS FRCS (ORL) Special Lecturer and Consultant Department of Otorhinolaryngology Queen’s Medical Centre Campus Nottingham University Hospitals Nottingham, UK



Mark MCGurk MD BDS FRCS FDSRCS DLO Consultant in Oral and Maxillofacial Surgery Guy’s Hospital London, UK Stephen McHanwell BSC PhD MIBiol CBiol Professor of Anatomical Sciences; and National Teaching Fellow 2007; and Director of Stage 1 & 2 BDS School of Dental Sciences Dental School Newcastle upon Tyne, UK Michael J MCKenna MD Professor, Department of Otology and Laryngology Harvard Medical School; and Surgeon, Department of Otolaryngology Massachusetts Eye and Ear Infirmary Boston, MA, USA William S MCKerrow MB ChB MRCGP (exam) FRCSEd & Glasg Consultant Otolaryngologist Department of ENT/Head and Neck Surgery Raigmore Hospital Inverness, UK Siobhan MCMahon BSc MRCSLT Speech and Language Therapy Department Alder Hey Hospital Liverpool, UK Brent A MCMonagle MBBS FRACS Department of Otolaryngology Guy’s Hospital London, UK Hisham Mehanna BmedSc (Hons) MBChB (Hons) FRCS (ORL-HNS) Consultant ENT – Head and Neck and Thyroid Surgeon; and Honorary Senior Lecturer University Hospitals Coventry and Warwickshire Walsgrave Hospital Coventry, UK Saumil N Merchant MD Gudrun Larsen Eliasen and Nels Kristian Eliasen Professor of Otology and Laryngology Harvard Medical School; and Surgeon in Otolaryngology and Director of Otopathology Laboratory Department of Otolaryngology Massachusetts Eye and Ear Infirmary Boston; and Affiliate Faculty Member Harvard University-Massachusetts Institute of Technology Division of Health Sciences and Technology Cambridge, MA, USA David Miles FRCP MD Consultant in Medical Oncology Mount Vernon Cancer Centre Middlesex, UK



Contributors Christopher A Milford FRCS Consultant Otolaryngologist Oxford Skull Base Unit John Radcliffe Hospital Oxford, UK Robert Mills MS MPhil FRCS (Eng) FRCS (Ed) Otolaryngology Unit University of Edinburgh Royal Infirmary of Edinburgh Edinburgh, UK Steven Ross Mobley MD Director of Facial Plastic and Reconstructive Surgery Division of Otolaryngology-HNS University of Utah School of Medicine Salt Lake City, UT, USA David Moffat BSc MA FRCS Consultant Neuro-Otologist Department of Otoneurotological and Skull Base Surgery Addenbrookes Cambridge University Teaching Hospital NHS Foundation Trust; and Associate Lecturer, Cambridge University Cambridge, UK



Frank E Musiek PhD Professor and Director of Auditory Research Department of Communication Sciences; and Professor of Otolaryngology School of Medicine University of Connecticut Storrs, CT, USA Niels Mygind MD Formerly Consultant in Lung Medicine Department of Respiratory Medicine University Hospital of Aarhus Aarhus, Denmark Karl G Nicholson MBBS MRCS FRCP Professor of Infectious Diseases Department of Infectious Diseases and Tropical Medicine Leicester Royal Infirmary Leicester, UK Andrew J Nicol MBChB, FCS (SA) Associate Professor General Surgery; and Head of Trauma Unit Groote Schuur Hospital Cape Town, South Africa



Brian CJ Moore MA PhD FMedSci FRS Professor of Auditory Perception Department of Experimental Psychology University of Cambridge Cambridge, UK



Gilbert J Nolst Trenite´ MD PhD Professor of Otorhinolaryngology Academic Medical Center University of Amsterdam The Netherlands



David AL Morgan FRCR Consultant Clinical Oncologist Department of Clinical Oncology Nottingham University Hospitals Nottingham, UK



Desmond A Nunez FRCS (ORL) MD Director, Department of Otolaryngology North Bristol NHS Trust; and Honorary Senior Lecturer University of Bristol Bristol, UK



Jonathan M Morgan MD Instructor Department of Otolaryngology Head and Neck Surgery University of South Florida Tampa, FL, USA Juliette Morgan MD Division of Foodborne Bacterial and Mycotic Diseases National Center for Zoonotic, Vector-Borne and Enteric Diseases Centers for Disease Control and Prevention Atlanta, GA, USA Gavin AJ Morrison MA MBBS FRCS Consultant ENT Surgeon Guy’s, St Thomas’ and Evelina Hospitals London, UK Randall P Morton MB MSc FRACS Professor of Otolaryngology University of Auckland; and Consultant Otolaryngologist–Head and Neck Surgeon Counties Manukau and Auckland District Health Boards Auckland, New Zealand



Michael O’Connell BSc, MPhil, FRCS Consultant Otorhinolaryngologist, Facial Plastic Surgeon and Honorary Senior Lecturer Brighton and Sussex University Hospitals NHS Trust Brighton, UK Alec Fitzgerald O’Connor FRCS Consultant Otolaryngologist St Thomas Hospital London, UK Paul O’Flynn FRCS Consultant ENT/Head and Neck Surgeon University College Hospitals; and Honorary Consultant The Royal National Throat, Nose and Ear Hospital London, UK Stephen O’Leary MB BS BMedSc PhD FRACS The Department of Otolaryngology Royal Victorian Eye and Ear Hospital East Melbourne, Australia



] xxi



xxii ] Contributors Morten Osterballe MD Allergy Center, Odense University Hospital Odense, Denmark Peter O’Sullivan Bsc MPhil FRCSI (ORL-HNS) Clinical Fellow, Neurotology Department of Otolaryngology Sir Charles Gairdner Hospital Nedlands, Western Australia William J Oweny MS FRCS Formerly Oesophageal Investigation Unit Department of Surgery, St Thomas’ Hospital London, UK Jaideep J Pandit MA BM DPhil FRCA Consultant Anaesthetist Nuffield Department of Anaesthetics University of Oxford Oxford, UK Andrew J Parker MBChB (hons) DLO ChM FRCS Consultant ENT Surgeon Department of Otolaryngology Royal Hallamshire Hospital Sheffield, UK Glynnis Parker MB ChB FRCP DCH MSc Audiovestibular Physician Sheffield Children’s Hospital Sheffield, UK Nimesh Patel MBChB FRCS FRCS (ORL-HNS) Consultant Otolaryngologist Southampton General Hospital Liverpool, UK John P Patten BSc MB FRCP Consultant Neurologist (retired) South West Thames Regional Health Authority London, UK Henry Pau MD MBChB FRCSEd FRCS Ed (ORL-HNS) FRCS Consultant Otorhinolaryngologist; and Honorary Senior Lecturer University Hospitals of Leicester Leicester, UK Santdeep H Paun FRCS (ORL-HNS) Consultant Nasal and Facial Plastic Surgeon St Bartholomew’s Hospital London, UK Sarah Payne BSC (Hons) MRCP SpR in Medical Oncology Centre for Tumour Biology Institute of Cancer and the CR-UK Clinical Centre Barts and the London Queen Mary’s School of Medicine and Dentistry London, UK y Deceased



Adrian Pearce FRCA Consultant Anaesthetist Department of Anaesthesia Guy’s and St Thomas’ Hospital London, UK Ken Pearman FRCS Consultant Paediatric Otolaryngologist Children’s Hospital Birmingham, UK Chris Penfold FDSRCS FRCS Consultant Oral and Maxillofacial Surgeon Alder Hey Childen’s Hospital Liverpool, UK A Graeme B Perks FRCS FRCS (Plast) FRACS Consultant Plastic Surgeon The City Hospital Nottingham, UK Alison Perry PhD FRCSLT Chair, School of Human Communication Sciences Faculty of Health Sciences La Trobe University Melbourne, Australia James O Pickles MA MSC PhD DSC Head of Hearing Unit Vision, Touch and Hearing Research Centre Department of Physiology and Pharmacology University of Queensland Brisbane, Australia Lisa Pitkin BSC MSC FRCS ORL-HNS Specialist Registrar in Otolaryngology South (West) Thames Otolaryngology Training Region Royal Marsden NHS Foundation Trust London, UK Laysan Pope BSc MB BS MRCS Specialist Registrar in Otolaryngology, Head and Neck Surgery John Radcliffe Hospital Oxford, UK Stephen R Porter BSc MD PhD FDS RCS FDS RCSE Professor of Oral Medicine UCL Eastman Dental Institute London, UK Richard J Powell MBBS DM FRCP FRCPath Consultant and Professor in Clinical Immunology University of Nottingham Nottingham, UK Paul Pracy BSc MBBS FRCS (Glas) FRCS (ORL-HNS) Consultant Head and Neck Surgeon Department of Otorhinolaryngology/Head and Neck Surgery Queen Elizabeth Hospital University Hospital Birmingham NHS Trust Birmingham, UK



Contributors Hillel Pratt PhD Evoked Potentials Laboratory Technion – Israel Institute of Technology Haifa, Israel Tim Price Bsc MBChB MRCS DLO FRCS (ORL-HNS) Consultant Otolaryngologist, Head and Neck Surgeon Dorset County Hospital Dorchester, UK William J Primrose MB FRCS Consultant Otolarnyngologist/Head and Neck Surgeon Royal Victoria Hospital, Belfast Northern Ireland, UK Matthias Radatz MD FRCS The National Centre for Stereotactic Radiosurgery Royal Hallamshire Hospital Sheffield, UK Ullas Raghavan FRCS (ORL-HNS) Consultant Ear Nose and Throat and Facial Plastic Surgeon Doncaster Royal Infirmary Doncaster, UK Gunesh P Rajan MD FMH (Ch) FRACS Senior Lecturer of Otolaryngology, Head and Neck Surgery Department of Otolaryngology, Head and Neck Surgery University of Western Australia Fremantle, Australia James Ramsden PhD FRCS Specialist Registrar in Otolaryngology/Head and Neck Surgery John Radcliffe Hospital Oxford, UK Richard Ramsden FRCS Manchester Royal Infirmary Manchester, UK Sheila C Rankin FRCR Consultant Radiologist Guy’s and St Thomas’ Hospital NHS Trust London, UK Helge Rask-Andersen MD PhD Professor in Experimental Otology Department of Otolaryngology Uppsala University Hospital Uppsala, Sweden Peter Rea MA FRCS (Eng) FRCS (ORL-HNS) Consultant Otolaryngologist Leicester Royal Infirmary Leicester, UK Fiona Regan MBBS FRCP FRCPath Consultant Haematologist Department of Haematology Imperial College School of Medicine; and Honorary Senior Lecturer and Consultant Haematologist National Blood Service London, UK



] xxiii



Claud Regnard FRCP (Lon) Consultant in Palliative Care Medicine St. Oswald’s Hospice, Newcastle-upon-Tyne; and Freeman Hospital (Newcastle Hospitals NHS Trust) Newcastle-upon-Tyne and Northumberland Tyne and Wear NHS Trust Northumberland, UK Evan Reid BSc MB ChB PhD FRCP Wellcome Trust Senior Research Fellow in Clinical Science; and Honorary Consultant in Medical Genetics Department of Medical Genetics and Cambridge Institute for Medical Research Addenbrooke’s Campus, University of Cambridge Cambridge, UK Gerhard Rettinger Prof Dr Med Head ENT-University-Department Ulm, Germany David Richardson FRCS FDSRCS Consultant Maxillofacial Surgeon Supra Regional Paediatric Craniofacial Unit Royal Liverpool Childrens Hospital; and Maxillofacial Unit University Hospital Aintree Liverpool, UK Peter J Robb BSc (Hons) MB BS FRCS FRCSEd Epsom and St Helier University Hospitals NHS Trust Surrey, UK David Roberts FRCS St Thomas and Guy’s Hospital NHS Trust London, UK Philip J Robinson MB ChB FRCS FRCS (Otolaryngology) Consultant Otolaryngologist ENT Department, Southmead Hospital Bristol, UK Nicholas J Roland MBChB MD FRCS Consultant ENT/Head and Neck Surgeon University Hospital Aintree Liverpool, UK Geoffrey E Rose DSc MS MRCP FRCS FRCOphth Consultant Orbital Surgeon Moorfields Eye Hospital London, UK Rob Ross Russell MD FRCPCH Consultant in Paediatric Intensive Care and Respiratory Medicine Addenbrooke’s Hospital Cambridge, UK Mike Rothera MBBS FRCS Consultant Paediatric ENT Surgeon Royal Manchester Childrens’ Hospital Manchester, UK



xxiv ] Contributors Jeremy Rowe MA DM FRCS (SN) The National Centre for Stereotactic Radiosurgery Royal Hallamshire Hospital Sheffield, UK Julian Rowe-Jones MB BS FRCS (ORL) Consultant Rhinologist and Nasal Plastic Surgeon Department of Otorhinolaryngology – Head and Neck/ Facial Plastic Surgery Royal Surrey County Hospital Guildford, UK Claudia Rudack PD Dr Med ENT-University-Department Mu¨nster, Germany Michael J Rutter FRACS Division of Pediatric Otolaryngology/Head and Neck Surgery, Cincinnati Children’s Hospital Medical Center; and Associate Professor of Pediatric Otolaryngology Department of Otolaryngology, Head and Neck Surgery University of Cincinnati College of Medicine Cincinnati, OH, USA Shakeel R Saeed MBBS (Lon) FRCS (Ed) FRCS (Eng) FRCS (Orl) MD (Man) Consultant ENT and Skull Base Surgeon University Department of Otolaryngology– Head and Neck Surgery Manchester Royal Infirmary and Hope Hospital Manchester, UK Hesham Saleh MBBCh FRCS FRCS (ORL-HNS) Consultant Rhinologist/Facial Plastic Surgeon Charing Cross Hospital and the Royal Brompton Hospital; and Honorary Senior Lecturer Imperial College of Medicine London, UK Robert J Sandersony MB ChB FRCS (Ed) FRCS (Eng) FRCS (ORL-HNS) Formerly Consultant Otolaryngologist/Head and Neck Surgeon Western General Hospital Edinburg, UK Keshthra Satchithananda Consultant Radiodogist Department of Radiology Charing Cross Hospital London, UK



BDS FDSRCS MB BS FRCS FRCR



Michael Saunders MD FRCS Consultant Otolaryngologist Department of Otorhinolaryngology, Head and Neck Surgery St Michael’s Hospital Bristol, UK Glenis Scadding MA MD FRCP Consultant Immunologist, Rhinologist and Allergy Specialist Royal National Throat Nose and Ear Hospital London, UK y Deceased



Jochen Schacht PhD Professor and Director Kresge Hearing Research Institute Department of Otolaryngology University of Michigan Ann Arbor, MI, USA Rodney J Schlosser MD Department of Otolaryngology Medical University of South Carolina Charleston, SC, USA Stephan Schmid MD Professor of Otolaryngology Department of Otorhinolaryngology, Head and Neck Surgery Universitatsspital Zurich Zurich, Switzerland Colin A Scotchford PhD Associate Professor School of Mechanical, Materials and Manufacturing Engineering University of Nottingham Nottingham, UK Andrew Scott FRCS (ORL-HNS) MPhil The Royal Shrewsbury Hospital Shrewsbury, UK Crispian Scully



CBE MD PhD MDS MRCS FDSRCS FDSRCPS FFDRCSI FDSRCSE FRCPath



FMedSci DSc



Professor of Oral Medicine, Pathology and Microbiology University of London; and Professor of Special Care Dentistry UCL-Eastman Dental Institute London, UK Su-Hua Sha MD Research Investigator Kresge Hearing Research Institute Department of Otolaryngology University of Michigan Ann Arbor, MI, USA Naomi Sibtain MBBS MRCP FRCR Consultant Neuroradiologist King’s College Hospital London, UK Paul S Sidhu BSC MB BS MRCP FRCR DTM&H Senior Lecturer and Consultant Radiologist Department of Radiology King’s College Hospital London, UK Richard Sim MD FRCS (OTO) Department of Ear, Nose and Throat Royal United Hospital Bath, UK



Contributors Paul Simons MBBS BSc MRCP MRCGP DCH DRCOG DFFP Marcham Road Health Centre Abingdon, UK Robert Slack BSc MB ChB FRCS (Ed) FRCS (Eng) Department of Ear, Nose and Throat Royal United Hospital Bath, UK Wendy Smith BPharm MBBS DLO FRCS (ORL-HNS) Locum Consultant Otorhinolaryngology The Leeds Teaching Hospitals NHS Trust Leeds, UK Lewis Spitz PhD FRCS Institute of Child Health (University College London) and Great Ormond Street Hospital for Children London, UK Jacob Bertram Springborg MD PhD University Clinic of Neurosurgery The Neuroscience Centre Copenhagen University Hospital Copenhagen, Denmark Nicholas D Stafford MB FRCS Director, Postgraduate Medical Institute University of Hull Hull, UK H Stammberger MD Hon FRCS (Ed) Hon FRCS (Engl) Professor and Head Department of General ORL, H & NS Medical University Graz, Austria Michael Stearns BDS MB BS FRCS The Royal Free Hospital London, UK Karen P Steel Phd FMedSci The Wellcome Trust Sanger Institute Hinxton, UK Paul Stewart FRCP Department of Medicine Queen Elizabeth Hospital Birmingham, UK Iain RC Swan MB ChB MD FRCS (Ed) Department of Otolaryngology North Glasgow University NHS Trust Glasgow, UK Elizabeth Sweeney FRCP DRGOC MD Consultant Clinical Geneticist Craniofacial Unit Alder Hey Children’s Hospital Liverpool, UK



] xxv



Andrew C Swift ChM FRCS FRCS (Ed) Consultant in Otorhinolaryngology University Hospital Aintree Liverpool, UK Andra E Talaska BS Kresge Hearing Research Institute Department of Otolaryngology University of Michigan Ann Arbor, MI, USA Thomas A Tami MD Professor of Otolaryngology Department of Otolaryngology University of Cincinnati College of Medicine Cincinnati, OH, USA Rinze A Tange MD PhD UHD Associate Professor of Otology Department of ORL, Head and Neck Surgery Academic Medical Centre University of Amsterdam Amsterdam, The Netherlands A Thakar MS FRCS Associate Professor of Otolaryngology and Head/Neck Surgery All India Institute of Medical Sciences New Delhi, India J Regan Thomas MD Francis L. Lederer Professor and Head University of Illinois at Chicago Department of Otolaryngology – Head and Neck Surgery Chicago, IL, USA Jens Thomsen MD DMSc FRCS Professor of Otolaryngology Department of Otorhinolaryngology, Head and Neck Surgery Gentofte Hospital, University of Copenhagen Hellerup, Denmark Matthew J Thurtell MSc (Med) MBBS FRACP Neuro-Opthalmology Fellow Department of Neurology University Hospitals of Cleveland Cleveland, OH, USA Bo Tideholm MD PhD ENT Specialist Department of Otorhinolaryngology University Hospital Malmo+ , Sweden Paul Tierney BA BM BCh (Oxon) FRCS (Eng.) FRCS (ORL-HNS) Consultant Otolaryngologist – Head and Neck Surgeon North Bristol NHS Trust; and Honorary Senior Lecturer Bristol University Bristol, UK



xxvi ] Contributors Ian Todd PhD Associate Professor and Reader in Cellular Immunopathology University of Nottingham Nottingham, UK Joseph G Toner MB MA FRCS Consultant/Honorary Senior Lecturer, Otolaryngology Belfast HSC Trust Queens University Belfast, UK Michael Chi Fai Tong MBChB (CUHK) MD (CUHK) FRCS (Ed) FHKAM (ORL) Professor and Head of Academic Divisions Department of Otorhinolaryngology, Head and Neck Surgery The Chinese University of Hong Kong Hong Kong Dean M Toriumi MD Division of Facial Plastic and Reconstructive Surgery Department of Otolaryngology – Head and Neck Surgery University of Illinois at Chicago Chicago, IL, USA Mirko Tos Prof MD DMSc Dr Hc Emeritus Professor, Ear, Nose and Throat Department Gentofte Hospital University of Copenhagen Hellerup, Denmark; and Professor of Otolaryngology University of Maribor Maribor, Slovenia Stephen C Toynton MB FRCS (ORL) Consultant Otorhinolaryngologist Derriford Hospital, Plymouth Hospitals NHS Trust; and Otology Advisor to Diving Diseases Research Centre and Hyperbaric Medical Unit Plymouth, UK Kai Uus MD PhD Lecturer in Audiology School of Psychological Sciences Faculty of Medical and Human Sciences University of Manchester Manchester, UK Peter Valentine BSc FRCS (ORL-HNS) Consultant Otologist and ENT Surgeon Royal Surrey County Hospital NHS Trust Guildford; and Ashford and St Peter’s Hospitals NHS Trust Chertsey, UK Jan HP van der Meulen PhD FFPH Reader in Clinical Epidemiology Health Services Research Unit London School of Hygiene and Tropical Medicine London, UK



C Andrew van Hasselt MBChB FRCS FRCS (Edin) FCS (SA) Chairman, Department of Surgery; and Professor of Surgery (Otorhinolaryngology) Department of Otorhinolaryngology, Head and Neck Surgery The Chinese University of Hong Kong Shatin, Hong Kong Adriaan F van Olphen MD PhD ENT Surgeon University Medical Centre Utrecht Utrecht, The Netherlands Archana Vats MA (Cantab) FRCS (Eng) FRCS (Oto) PhD Imperial College and St. Mary’s NHS Trust London, UK Antonio M Vignolay Formerly of Istituto di Fisiopatologia Respiratoria Universita` Palermo Palermo, Italy Alexander C Vlantis MBBCh FCS (SA) FCSHK Associate Professor Department of Otorhinolaryngology, Head and Neck Surgery The Chinese University of Hong Kong Shatin, Hong Kong Sherryl Wagstaff FRACS Consultant Otologist Royal Victorian Eye and Ear Hospital Melbourne University Teaching Hospital East Melbourne, Australia Debbie Wall BEd (Hons) MA Senior Researcher NHS Clinical Governance Support Team Leicester, UK David Ward MBBS FRCS FRCS (Ed) Consultant Plastic Surgeon Leicester Royal Infirmary Leicester, UK David W Warnock PhD FRCPath Division of Foodborne Bacterial and Mycotic Diseases National Center for Zoonotic, Vector-Borne and Enteric Diseases Centers for Disease Control and Prevention Atlanta, GA, USA Katherine Wasson BA PhD MPH Chief, Clinical Ethics Service; and Assistant Professor, Critical Care The University of Texas M.D. Andersson Cancer Center Houston, Texas, USA John C Watkinson MSc MS FRCS (Ed, Glas, Lond) DLO Consultant Head and Neck and Thyroid Surgeon Department of Otorhinolaryngology/Head and Neck Surgery Queen Elizabeth Hospital University of Birmingham NHS Trust Birmingham, UK y Deceased



Contributors Desmond Watson BM BCh MA FRCS Former Consultant Ear Nose and Throat Surgeon and Advisor Medical Protection Society Leeds, UK



Peter-John Wormald MD FRACS FRCS (Ed) FCS (SA) MBChB Department of Otolarnyngology, Head and Neck Surgery Adelaide and Flinders Universities Adelaide, Australia



Keith Webster MMedSci FRCS FRCS (OMFS) FDSRCS Consultant Oral and Maxillofacial Surgeon University Hospital Birmingham NHS Foundation Trust; and Honorary Senior Lecturer Faculty of Dentistry and Medicine University of Birmingham Birmingham, UK



Steve Worrollo FIMPT Consultant Maxillofacial Prosthetist Department of Maxillofacial Surgery University Hospital Birmingham NHS Trust Birmingham, UK



Vivienne Weston MBBS FRCP MSc FRCPath Consultant Medical Microbiologist Nottingham University Hospitals NHS Trust Nottingham, UK Richard Wight



MB BS FRCS Eng (Otol) FRCS Ed (Otol)



Consultant Head and Neck Surgeon James Cook University Hospital Middlesbrough, UK Janet A Wilson BSc MD FRCSEd FRCS Eng Professor of Otolaryngology, Head and Neck Surgery Newcastle University Freeman Hospital Newcastle Upon Tyne, UK Wai Lup Wong BA (Hons) MRCP FRCR Paul Strickland Scanner Centre Mount Vernon Hospital Northwood, UK John Kong Sang Woo MBBS FCSHK FRCSEd FHKAM (Otorhinolaryngology) Consultant, Department of ENT, Prince of Wales Hospital; and Chief of Service in ENT New Territories East Cluster, Hospital Authority; and Honorary Clinical Associate Professor Department of Otorhinolaryngology, Head and Neck Surgery The Chinese University of Hong Kong Hong Kong Tim J Woolford MD FRCS (ORL) Consultant in Otorhinolaryngology Manchester Royal Infirmary University of Manchester Manchester, UK



Neville Wright DMRD FRCR Consultant Paediatric Radiologist Central Manchester and Manchester Children’s Hospitals NHS Trust Department of Radiology Royal Manchester Children’s Hospital Manchester, UK Tony Wright LLM DM FRCS Tech RMS Professor of Otolaryngology UCL Ear Institute London, UK Floris L Wuyts PhD Professor of Medical Physics University of Antwerp; and Head of AUREA (Antwerp University Research Centre for Equilibrium and Aerospace) Department of ENT University Hospital of Antwerp Antwerp, Belgium Michelle Wyatt MA FRCS (ORL-HNS) Consultant Paediatric Otolaryngologist Great Ormond Street Hospital London, UK Clare Wykes BSC MRCP DipRCPath Haematology SpR Hammersmith Hospitals NHS Trust London, UK Stephen R Young BSC (Hons) PhD Faculty of Science The American International University in London Surrey, UK



] xxvii



This page intentionally left blank



Preface



Fifty-five years have passed since the first edition of ScottBrown’s Otorhinolaryngology: Head and Neck Surgery was published. Many otorhinolaryngologists have read at least one edition, committed it to memory and passed their specialist examinations because of it. All will have kept referring to it throughout their careers and remember it with affection. Looking back it is apparent that a radical change in structure and format has taken place every 15 to 20 years. It is 20 years since Alan Kerr made the last radical change with the publication of the 5th edition, 20 years that have seen an information technology explosion. The internet, on-line libraries, e-delivery of journals and increasingly books, computerised search engines, CDROMs, DVDs, digital photography; the list goes on. These technological advances have transformed medical education, influenced significantly the way the current generation learns and the methods by which their competencies and knowledge are assessed. Certainly sufficient time has elapsed for Scott-Brown to evolve dramatically once more. This edition has been completely re-written. It bears little resemblance to its predecessors other than by title, and in its philosophy to provide a complete resume of the knowledge base that underpins modern ORL practice and which will guide clinicians in their everyday patient care for years to come. The number of chapters has almost doubled, with large topics dissected into more digestible parts. This reflects the expansion of our specialty such that it is now a group of subspecialties linked by the common thread, each concerned with, and committed to, the care of patients with disorders of the head and neck. Our authors are the leading experts in their respective fields of interest and have been selected from all over the world. Almost all the text is illustrated in colour and it comes with its own CD-ROM, containing all the text and illustrations in an accessible and searchable form, with references linked to PubMed. So what else could the trainee or practising otorhinolaryngologist want from the definitive reference to the field at the beginning of the new millennium? Quite simply, the level of evidence for the advice we offer and the practice we undertake. Nowadays specialties need to define best clinical practice, if only to guide and remind health care providers of their duty to their patients to practice in accordance with accepted evidence and to



strive for excellence in clinical standards at all times. Surgeons also need to know how their actions might be viewed by the courts and the areas of practice that are currently exercising the legal profession. This edition has tried to provide that information. It has not been an easy task for our contributors, some of whom were not writing in their mother tongue. That they were able to write to a structured format was much to their credit. I was fortunate to recruit, and am extremely grateful to, my team of section editors all of whom worked tirelessly with a common purpose. George G Browning, Martin J Burton, Ray Clarke, John Hibbert, Nicholas S Jones, Valerie J Lund, Linda M Luxon and John C Watkinson represent some of the very best and most respected clinicians in the United Kingdom, each one an international authority, each one with a heavy professional commitment. Alan Kerr’s advice and encouragement throughout was always welcome and extremely useful. Marcelle McNamara came to my aid and assistance numerous times during the project. She gave tirelessly of her time and energy during a very serious illness, writing chapters and putting others into format and a more readable form. She was an example and inspiration throughout. The creation of this edition has also been an interesting experience for the publishing staff. During a lengthy period of gestation, this text has changed ownership several times as the publishing houses traded and realigned their lists. Without the drive and perseverance of Zelah Pengilley and Jo Koster from Hodder Education it would surely have fallen by the wayside. Words cannot express my gratitude to them adequately. Understanding when clinical work overwhelmed me, they hid their frustrations over slow progress or irritatingly incomplete manuscripts. They buoyed us all up when the end seemed so far away. Sadly, some of our contributors will never see their chapters in print as they have died during the preparation of this text. Some had long, unpleasant illnesses but wrote despite them. Others were cut down unexpectedly in their prime but have now left a legacy, and a few were my close friends and colleagues. I am proud to have my name linked permanently through this publication with Michael Baser, Roderick Cawson, Susanna Leighton,



xxx ] Preface William Owen, Robert Sanderson and Antonio Vignola. We hope that their families will draw some comfort also by seeing their loved ones live on in this book. Finally, there are four very special people whose constant love and affection drives me on through life. They are of course my wife, Ann, and our children,



Andrew, Clare and Mark. They too will breathe a deep sigh of relief with the publication of this text and I thank them with all my heart. Michael Gleeson September 2007



How to use this book



This new edition of Scott-Brown’s Otorhinolaryngology, Head and Neck Surgery incorporates some special features to aid the readers’ understanding and navigation of the text. These are described below.



SEARCH STRATEGY The majority of the chapters feature a search strategy indicating the key words used by the author when conducting their literature review in order to prepare the chapter, so that the reader can repeat and develop the search.



EVIDENCE SCORING For the major sections in each chapter, the authors have used a hierarchical system to indicate the level of evidence supporting their statements. This is shown in the text in the form [***], with the number of stars indicating the level of evidence. The key to this system is shown in the table below. Level



Category of evidence



****



Systematic reviews, meta-analyses of randomized controlled trials and randomized controlled trials



***



Non-randomised studies



**



Observational or non-experimental studies



*



Expert opinion



Where no level is shown, the quality of supporting evidence, if any exists, is of low grade only (for example, case reports, clinical experience etc.). For more information on evidence scoring, please refer to Chapter 304, Evidence-based medicine; and 305 Critical appraisal skills.



CLINICAL RECOMMENDATIONS The authors have indicated the basis on which they have made clinical recommendations by grading them according to the level of the supporting evidence. This is shown in the text in the form [Grade A], with the grade indicating the level of evidence supporting the recommendation. The key to this system is shown in the table below.



xxxii ] How to use this book Grade



Nature of supporting evidence



A



Recommendation based on evidence from meta-analyses of randomized controlled trails



B



Recommendation based on evidence from high quality case-controlled or cohort studies



C



Recommendation based on evidence from low quality case-controlled or cohort studies



D



Recommendation based on evidence from clinical series or expert opinion



Recommendations are graded where the author is satisfied that the literature supports such a grading; otherwise a grading may not be given.



REFERENCE ANNOTATION The reference lists are annotated with an asterisk, where appropriate, to guide readers to key primary papers and major review articles. We hope that this feature will render the lists of references more useful to the reader and will encourage self-directed learning among both trainees and practicing physicians.



Abbreviations



2,3DPG 2D 3,4-DAP 3D 5-FdUMP 5-FU 5-FUMP 5-HT 6MP 18-FDG A AABR AAHL AAOHNS AAV ABC ABEP ABG ABI ABLB ABPA ABR ABRS AC ACC ACE ACF ACh AchR ACT ACTH A/D AD ADA ADAM-33 ADCC ADH ADHD ADR



2,3-diphosphoglycerate two-dimensional 3,4-diaminopyridine three-dimensional 5-fluoro-2 deoxyuridine monophophate 5-fluorouracil 5-fluorouridine monophosphate 5-hydroxytryptamine 6-Mercaptopurine 2-18-fluoro-2-deoxy-D-glucose adenine; or anterior automated auditory brainstem response age-associated hearing loss American Academy of Otolaryngologists/ Head and Neck Surgeons adeno-associated virus aspiration biopsy cytology auditory nerve and brainstem evoked potential air–bone gap auditory brainstem implant alternate binaural loudness balance allergic bronchopulmonary aspergillosis auditory brainstem response; or acoustic brainstem evoked response acute bacterial rhinosinusitis air conduction; or alternating coupled adenoid cystic carcinoma; or American College of Cardiology angiotensin-converting enzyme anterior cranial fossa Acetylcholine acetyl choline receptor Aid for Children with Tracheostomies adrenocorticotropic hormone analogue-to-digital Alzheimer’s disease adenosine deaminase A disintegrin and metalloprotease 33k antibody-dependent cellular cytotoxicity antidiuretic hormone attention deficit hyperactivity disorder adverse drug reaction



Ad-VEGF AECRS AED AEDS AEF AF AFB AFRS AgNOR AHA AHCPR AHI AI AICA AIDS AIRE AJCC ALD ALEP ALL a2b2 a2d2 a2g2 ALPS ALS ALT ALTB ALTE AML AN ANA AN/AD ANCA AND ANUG AOAE AoCD AOM



adenovirus-encoding vascular endothelial growth factor acute exacerbation of chronic rhinosinusitis aerodynamic equivalent diameter atopic eczema dermatitis syndrome auditory-evoked cortical magnetic field atrial fibrillation; or anterior fontanelle acid-fast bacilli allergic fungal rhinosinusitis silver staining nucleolar organizer region American Heart Association Agency for Health Care Policy and Research (USA) apnoea/hypopnoea index apoptotic index anterior inferior cerebellar artery acquired immunodeficiency syndrome autoimmune regulator gene American Joint Committee on Cancer assistive listening device auditory long-latency (or late) evoked potential acute lymphoblastic leukaemia two a and two b globin chains HbA2 foetal haemoglobin autoimmune lymphoproliferative syndrome amyotrophic lateral sclerosis alternative lengthening of telomere; or alternating chemoradiotherapy acute laryngotracheobronchitis apparent life-threatening event acute myeloid leukaemia acoustic neuroma; or auditory neuropathy; or audiovestibular nerve anti-nuclear antibody auditory neuropathy/auditory dyssynchrony antineutrophil cytoplasmic antibody allow a natural death acute necrotizing ulcerative gingivitis automated otoacoustic emission anaemia of chronic disease acute otitis media



xxxiv ] Abbreviations AON AP APB APC



APD APECED APHAB APL APMET APQ APTT APUD ARF ARIA ARR ARS ART ARTA ASA



a-SCC ASIC ASL ASPO ASSR AST AT ATD ATIII ATN ATP ATRA AUC AV AVCN AVM aVOR AZT



BAC BACDA BADS BAES BAHA



anterior olfactory nucleus anterior–posterior; or action potential ALT-associated promyelocytic leukaemia body antigen presenting cell; or activated protein C; or argon plasma coagulation; or adenomatous polyposis coli auditory processing disorder autoimmune polyendocrinopathy– candidiasis–ectodermal dystrophy Abbreviated Profile of Hearing Aid Benefit anti-phospholipid aggressive papillary middle ear tumour amplitude perturbation quotient activated partial thromboplastin time amine precursor uptake and decarboxylation acute renal failure allergic rhinitis and its impact on asthma absolute risk reduction acute rhinosinusitis advanced rotating tomograph; or antiretroviral therapy age-related typical audiogram aspirin-induced asthma; or aspirin-sensitive asthma; or American Society of Anesthesiologists anterior semicircular canal application specific integrated circuit American sign language; or arterial spin labelling American Society of Pediatric Otolaryngologists auditory steady state response arterial spin tagging ataxia telangiectasia; or auditory therapy or training ascending tract of Deiters antithrombin III auriculotemporal nerve adenosine triphosphate all-trans retinoic acid area under the curve apical vesicles; or arteriovenous anteroventral cochlear nuclei arteriovenous malformation angular VOR 30 azido30 deoxythymidone zidovudine; or azothiaprine bacterial artificial chromosome British Association of Community Doctors in Audiology British Association of Day Surgery British Association of Endocrine Surgeons bone-anchored hearing aid



BAHNO BAO-HNS



BAPO BCC BCG BCHA BCSH BDP BE BF BFU-E BiPAP BIPP BL BMA BMI BMP BMS BMT/SCT BOA BOLD BOR BP BPD BPPV BPV BS BSE BSL BTE BVF C CAD CADCAM CAGE cAMP CANS CAP CAPD CaR CAS CATCH-22



CB CBF



British Association of Head and Neck Oncologists British Association of Otorhinolaryngologists – Head and Neck Surgeons British Association for Paediatric Otolaryngology basal cell carcinoma Bacillus Calmette–Gue´rin bone conductor hearing aid British Committee for Standards in Haematology beclomethasone dipropionate bulla ethmoidalis biofeedback burst-forming unit erythroid bilevel positive airway pressure bismuth and iodoform paraffin paste Burkitt’s lymphoma British Medical Association body mass index bone morphogenetic protein; or bone morphogenic protein burning mouth syndrome bone marrow stem cell transplantation behavioural observation audiometry blood oxygenation level-dependent brachio-oto-renal blood pressure bronchopulmonary dysplasia benign positional paroxysmal vertigo benign paroxysmal vertigo; or benign positional vertigo Behc- et’s syndrome bedside swallowing examination; or bovine spongiform encephalopathy British sign language behind the ear bilateral vestibular failure cytosine caspase-activated DNase computer-aided design, computer-aided manufacture cerebral air gas embolism 30 ,50 -monophosphate central auditory nervous system compound action potential; or category of auditory performance central auditory processing disorder calcium sensing receptor computer-assisted surgery cardiac defects, abnormal facies, thymic hypoplasia, cleft palate and hypocalcaemia-22 concha bullosa; or critical bandwidth ciliary beat frequency



Abbreviations CBT CCA CCDU CCR CCW CD CDA CDC CDK CDP CE-CT CEA CEPOD CER CERA CEVMP CF CFD CFR CFTR CFU CFU-GM CFU-Mk CG CGD CGH CGRP CGST CHARGE



CHART CHI CI CID CJD CL CL/P CLL CM CMAP CME CMI CML CMT CMV CN



cognitive-behavioural therapy common carotid artery colour-coded duplex ultrasonography chemokine receptor counter-clockwise cluster of differentiation; or colloid droplets; or compact disk cold dry air Centers for Disease Control and Prevention cyclin-dependent kinase computerized dynamic posturography contrast-enhanced computed tomography carcinoembryonic antigen Confidential Enquiry into Perioperative Deaths control event rate cortical evoked response audiometry click-evoked vestibular myogenic potential cystic fibrosis; or characteristic frequency colour-flow duplex Doppler craniofacial resection cystic fibrosis transmembrane conductance regulator colony-forming unit colony-forming unit, granulocytemacrophage colony-forming unit, megakaryocyte clinical governance chronic granulomatous disease comparative genomic hybridization calcitonin gene-related peptide Clinical Governance Support Team coloboma, heart defects, atresia choanae, retardation of growth, genital anomalies and ear abnormalities continuous, hyperfractionated, accelerated radiotherapy Commission for Healthcare Improvement (UK) cochlear implant; or cardiac index; or confidence interval; or concha inferior Central Institute for the Deaf Creutzfeldt–Jakob disease cleft lip cleft lip with or without cleft palate chronic lymphatic leukaemia; or chronic lymphocytic leukaemia concha media; or cochlear microphonic; or cricothyroid muscle compound muscle action potential continuing medical education cell-mediated immunity chronic myeloid leukaemia Charcot-Marie-Tooth Cytomegalovirus cranial nerve; or cochlear nuclei; or cochlear nerve



CNO CNS CO2 COAD COM COPD COR COSI COX-2 CP CPA CPAP CPD CPG CPO CPPIH CPR CQI CREST CRF CRH CROS CRP CRRT CRS CRSS CS CSCI CSF CSM CSOM CT CTA CTL CTLA CTLL CTM cTNM CTR CTZ Cu-ATSM CUP CUSA CVA CVD CVI CVP CW CXR CYP



] xxxv



chronic nasal obstruction central nervous system carbon dioxide chronic obstructive airway disease chronic otitis media chronic obstructive pulmonary disease conditioned orientation reflex Client Oriented Scale of Improvement cyclo-oxygenase 2 cleft palate cerebellopontine angle continuous positive airway pressure citrate phosphate dextrose; or continuing professional development central pattern generator cleft palate only Commission for Patient and Public Involvement in Health (UK) cardiopulmonary resuscitation continuous quality improvement calcinosis, Raynaud’s, oesophageal involvement, sclerodactyly, telangiectasis corticotrophin-releasing factor corticotropin-releasing hormone contralateral routing of signal or sound C-reactive protein; or canalith repositioning procedure continous renal replacement therapy chronic rhinosinusitis; or congenital rubella syndrome chronic rhinosinusitis corticosteroid Commission for Social Care Inspection (UK) cerebrospinal fluid Committee on Safety of Medicines chronic suppurative otitis media computed tomography composite tissue allograft cytotoxic T-lymphocyte cytotoxic T-lymphocyte-associated antigen cytotoxic T-lymphocyte leukaemic cricothyroid membrane clinical tumour, nodes, metastases cricotracheal resection chemoreceptor trigger zone Cu(II)-diacetyl-bis-N4methylthiosemicarbozone carcinoma of unknown primary origin cavitational ultrasonic surgical aspirator cerebrovascular accident central vestibular disorder common variable immunodeficiency central venous pressure clockwise chest x-ray cytochrome P450



xxxvi ] Abbreviations DACH DAHANCA DAHNO dB dB SPL DBPCFC DCIA DCN DCR DD DDHS DFF DFN3 DFO-H DHA-S DHE DHI DHTR DIC DIEP DILS DIT DLE DM DMD DMSA DMSO DNA DNAR DNL DNR dNTP DP DPA DPOAE DR DRS DSA DSI DSL DTD DTIC dTMP DTPA dUMP DVB DVLA DVN DVT DWI EA EAACI



diaminocyclohexane Danish Head and Neck Cancer Study Data for Head and Neck Oncology (UK) decibel decibel sound pressure level double-blind placebo-controlled food challenge deep circumflex iliac artery dorsal cochlear nucleus dacryocystorhinostomy death domain Direct Drive Hearing System DNA fragmentation factor deafness type 3 deferoxamine-hespan dehydroepiandrosterone sulphate dihaematoporphyrinether dizziness handicap inventory delayed haemolytic transfusion reaction disseminated intravascular coagulation deep inferior epigastric perforator diffuse infiltrated lymphocytosis syndrome diiodotyrosine discoid lupus erythematosus diabetes mellitus Duchenne muscular dystrophy dimercapto succinic acid dimethylsulfoxide deoxyribonucleic acid do not attempt resuscitation nasolacrimal duct do not resuscitate deoxynucleoside triphosphate directional preponderance Data Protection Act (UK) distortion product otoacoustic emission death receptor; or drug resistance Dysphagia Research Society digital subtraction angiography Dysphonia Symptom Index desired sensation level DT-diaphorase dimethyl triazeno imidazole carboxamide deoxythymidine monophosphate diethylene triamine pentacetic acid deoxyuridine monophophase degree of voice break Driver and Vehicle Licensing Authority (UK) descending vestibular nuclei deep vein thrombosis diffusion weighted image episodic ataxia; or early antigen European Academy of Allergology and Clinical Immunology



EAC EAL EAM EB EBM EBNA EBP EBV EC ECA ECAL ECAP ECC ECG ECM ECMO EcochG ECog ECOG ECP ECR EDGT EDN EDS EDTA EDV EE EEG EER EFS EG EGF EGFR EIA ELDCR ELG ELISA ELST EM EMEA EMG EMI EN ENA ENG ENoG ENT EOG EORTC EP



external auditory canal; or external acoustic canal ethmoidal artery ligation external auditory meatus epidermolysis bullosa evidence-based medicine Epstein–Barr virus-associated nuclear antigen evidence-based practice Epstein–Barr virus embryonic carcinoma external carotid artery external carotid artery ligation electrically evoked compound action potential extracorporeal circuit electrocardiogram extracellular matrix extracorporeal membrane oxygenation electrocochleography electrocochleogram Eastern Cooperative Oncology Group (USA) eosinophil cationic protein extracapsular rupture early goal-directed therapy eosinophil-derived neurotoxin excessive daytime sleepiness ethylenediaminetetraacetic acid end diastolic velocity external frontoethmoidectomy electroencephalography; or electroencephalogram experimental event rate event-free survival embryonic germ epidermal growth factor epidermal growth factor receptor enzyme immunoassay endonasal laser dacryocystorhinostomy electrolaryngography enzyme-linked immunosorbent assay endolymphatic sac tumour erythema multiforme European Agency for the Evaluation of Medicinal Products electromyography elective mucosal irradiation enteral nutrition extra nuclear antigen electronystagmography electroneurography ear, nose and throat electroolfactogram; or electrooculography European Organisation for Research and Treatment of Cancer endolymphatic potential



Abbreviations EPO EQ-5D ER ERB ERM ERP ERT Er:YAG ES ESPAL ESR ESS ET ET-1 ETT EU EUA EVAS EXIT F0 FAAF Fab FACS FACT FAMM Fas-L FBC Fc FD FDA FDG



FDG-PET



FEES FEESST FESS FETNIM FFP FFT FGF FHH FISH FIV FLAIR FMISO fMRI



erythropoietin EuroQol enhancement ratio; or endoplasmic reticulum equivalent rectangular bandwidth ezrin, radixin, moesin event-related potential external radiotherapy erbium:yttrium-aluminium-garnet embryonic stem; or endolymphatic sac endonasal ligation of the sphenopalatine artery erythrocyte sedimentation rate endoscopic sinus surgery; or Epworth Sleepiness Scale essential thrombocytosis; or endotracheal tube endothelin-1 endotracheal tube European Union examination under anaesthesia enlarged vestibular aqueduct syndrome extrauterine intrapartum treatment fundamental frequency four alternative auditory feature fragment antigen binding fluorescence-activated cell sorter functional assessment of cancer therapy facial artery myomucosal flap Fas ligand full blood count fragment crystallizable fibrous dysplasia Food and Drug Administration (USA) fluorodeoxyglucose; or 2-[18F] fluoro-2deoxy-D-glucose; or F18-fluoro-2-deoxy-Dglucose 2-[18F] fluoro-2-deoxy-D-glucose–positron emission tomography; or fluorine-18labelled deoxyglucose positron emission tomography fibreoptic endoscopic evaluation of swallowing fibreoptic endoscopic evaluation of swallowing with sensory testing functional endoscopic sinus surgery fluorine-18 fluoroerythronitroimidazone fresh frozen plasma fast Fourier transform fibroblast growth factor familial hypocalciuric hypercalcaemia fluorescence in situ hybridization feline immunodeficiency virus fluid attenuated inversion recovery fluorine-18 fluoromisonidazole functional magnetic resonance imaging



FN FNA FNAB FNAC FOAR FOI FPANS FS FSH FT FTA FTA-ABS FTC FTP G G6PD Ga-67 GABA GABHS GAG GALT GAS G&S GBI GBLC GCS G-CSF GD GERD GH GHABP GHRH GI GIA GIC GIST GLM GMC GM-CSF GN GNE GnRH GOR GORD GOSH gp GP GPN GPP G protein GRB2



] xxxvii



facial nerve fine-needle aspiration fine-needle aspiration biopsy fine-needle aspiration cytology fronto-orbital advancement and remodelling fibreoptic orotracheal intubation fluticasone propionate aqueous nasal spray folliculostellate follicle-stimulating hormone fibrous tissue fluorescent treponemal antibody fluorescent treponemal antibody test frequency threshold curve Fitness to Practise guanine glucose-6-phosphate deficiency gallium gamma-aminobutyric acid group A beta-haemolytic streptococcus glycosaminoglycan gut-associated lymphoid tissue Goal Attainment Scaling group and screen Glasgow Benefit Inventory geometric broken line closure Glasgow Coma Score granulocyte-colony stimulating factor Graves’ disease gastrooesophageal reflux disease growth hormone Glasgow Hearing Aid Benefit Profile growth hormone-releasing hormone gastrointestinal gravitoinertial acceleration glass ionomer cement gastrointestinal stromal tumour ground lamella of middle turbinate, middle (frontal) portion ganglion mother cell; or General Medical Council (UK) granulocyte-macrophage colonystimulating factor glossopharyngeal nerve glottal-to-noise excitation gonadotropin-releasing hormone gastro-oesophageal reflux gastro-oesophageal reflux disease Great Ormond Street Hospital (UK) glycoprotein general practitioner glossopharyngeal neuralgia gingivo-periosteoplasty guanine nucleotide-binding regulatory protein growth factor receptor binding protein 2



xxxviii ] Abbreviations GSH GSPN GST GSW GTN GTR GVHD



glutathione greater superficial petrosal nerve glutathione S-transferase gun shot wound nitroglycerin guided tissue regeneration graft-versus-host disease



H&E H&N H2 HA HAART HAE HAEM HAPI HB Hb HbA HBO HBOT HBsAg HC HCA HCG HCSU Hct HCV



haematoxylin and eosin head and neck histamine receptor type 2 hydroxyapatite highly active antiretroviral therapy hereditary angioedema HSV-associated erythema multiforme Hearing Aid Performance Inventory House–Brackmann haemoglobin adult haemoglobin hyperbaric oxygen hyperbaric oxygen therapy hepatitis B surface antigen Healthcare Commission (UK) hydroxycarbonate apatite human chorionic gonadotrophin Health Care Standards Unit (UK) haematocrit hepatitis C virus; or human T-lymphocytic virus 1 haemodialysis high-density lipoprotein house dust mite high-density polyethylene high dependency unit helium-neon human epithelial type 2 hereditary familial telangiectasia hepatocyte growth factor Hearing Handicap Inventory Hearing Handicap Inventory for the Elderly hereditary haemorrhagic telangiectasia human herpesvirus 6 human herpesvirus 8 hearing impaired Haemophilus influenzae B heparin-induced thrombocytopenia heparin-induced thrombocytopenia with thrombosis human immunodeficiency virus HIV-associated salivary gland disease high jugular bulb hearing loss; or hearing level; or hairy leukoplakia human leukocyte antigen history of migraine; or hemifacial microsomia high molecular weight



HD HDL HDM HDPE HDU He-Ne HEp-2 HFT HGF HHI HHIE HHT HHV-6 HHV-8 HI HiB HIT HITT HIV HIV-SGD HJB HL HLA HM HMW



HMWC HNC HNR HNRQ HNSCC HPA HPC HPD HPL HPT HPV HRA HRCT HRM HRQOL HRT HS h-SCC HSCT HSMN HSPG HSV HSV-1 HSV-2 HSV-TK HT hTERT hTR HU HUI HUS Hz HZV IAC IAM IBP IC ICA ICAM ICAM-1 ICD ICM ICP ICRA ICU ID IDA IDT IDU IF IFN IFN-a IFN-b IFN-g



high molecular weight compound head and neck cancer harmonics-to-noise ratio Head and Neck Radiotherapy Questionnaire head and neck squamous cell carcinoma hypothalamic–pituitary–adrenal haemangiopericytoma haematoporphyrin derivative horizontal partial laryngectomy hyperparathyroidism human papillomavirus; or human herpes virus 8 Human Rights Act high-resolution computed tomography high-resolution manometry health-related quality of life hormone replacement therapy hiatus semilunaris horizontal semicircular canal haemopoietic stem cell transplant hereditary sensory-motor neuropathy heparin sulphate proteoglycan herpes simplex virus herpes simplex virus type 1 herpes simplex virus type 2 herpes simplex thymidine kinase hydroxytryptamine human telomerase reverse transcriptase human telomerase RNA Hounsfield unit Health Utilities Index haemolytic uraemic syndrome hertz herpes zoster virus internal auditory canal internal auditory meatus invasive monitoring of blood pressure inferior colliculus; or immunochemistry internal carotid artery intercellular adhesion molecule intercellular adhesion molecule 1 International Classification of Disease intensive care medicine intracranial pressure International Collegium of Rehabilitative Audiology intensive care unit inferior dental iron deficiency anaemia infant distraction test intravenous drug user intrinsic factor interferon interferon-alpha inteferon-beta interferon gamma



Abbreviations IFNP Ig IgE IGF IGFI IGFII IgG IGS IHAFF IHC IHS IL IL-1 IL-2 IL-3 IL-6 ILMA IMA IMAL IMF IMRT IMSPAC INC INE INO iNOS INR IOC IOPI IP3 IPSS IRMA IRS ISAAC ISEL ISJ ISO ISS ISSNHL IT ITA ITE ITP ITU IUCC i.v. IVIg JFC JNA JORPP



] xxxix



idiopathic facial nerve paralysis immunoglobulin immunoglobulin E insulin-like growth factor insulin-like growth factor I insulin-like growth factor II immunoglobulin G image-guided surgery International Hearing Aid Fitting Forum immunohistochemistry; or inner hair cell International Headache Society interleukin interleukin-1 interleukin-2 interleukin-3 interleukin-6 intubating laryngeal mask airway internal maxillary artery internal maxillary artery ligation intermaxillary fixation intensity-modulated radiation therapy imitative test of speech pattern contrast perception immunonuclear chemistry intranasal ethmoidectomy internuclear ophthalmoplegia inducible nitric oxide synthase international normalized ratio; or interventional neuroradiology Interim Orders Committee (UK) Iowa Oral Performance Instrument 1,4,5-inositol triphosphate inferior petrosal sinus sampling immunoradiometric assay Intergroup Rhabdomyosarcoma Study International Study of Asthma and Allergies in Childhood in situ end labeling incudostapedial joint International Standards Organization immunostimulatory DNA sequence idiopathic sudden sensorineural hearing loss inferior turbinate inferior thyroid artery in the ear idiopathic thrombocytopenic purpura intensive therapy unit International Union against Cancer intravenous intravenous immunoglobulin



K KAR keV KIR KS KSS KTP



Kirschner killer activating receptor kilo electron volt killer inhibitory receptor Kaposi’s sarcoma Kearns–Sayre syndrome potassium titanyl phosphate



LA LAD LAP LARP LAUP LB LCH LCM LD LDH LDL



just-follow-conversation juvenile nasopharyngeal angiofibroma juvenile-onset recurrent respiratory papillomatosis



LTC4-S LTR LTRA LVA



lymphangioma leukocyte adhesion defect left anteroposterior left anterior–right posterior laser-assisted uvulopalatoplasty lateral bundle Langerhans’ cell histiocytosis laser capture microdissection lymphocytic depleted lactate dehydrogenase low-density lipoprotein; or loudness discomfort level low-dose unfractionated heparin light-emitting diode lymphocyte-function associated antigen loudness growth in octave bands luteinizing hormone leuteinizing hormone-releasing hormone leukaemia-inhibitory factor Local Involvement Networks (UK) lateral lemniscus laryngeal mask laryngeal mask airway lower motor neuron low molecular weight low molecular weight compound low molecular weight heparin logarithm to the base 10 of the odds that the markers are linked at a recombination distance of N centimorgans loss of heterozygosity length of stay; or lower oesophageal sphincter lamina papyracea; or lichen planus; or lymphocyte predominant linear predictive coding laryngopharyngeal reflux likelihood ratio local research ethics committee lateral semicircular canal leukotriene long-term average spectrum long-term average speech spectrum laryngotracheobronchitis; or laryngotracheobronchoscopy leukotriene C4 synthase laryngotracheal reconstruction leukotriene receptor antagonists large vestibular aqueduct



LDUH LED LFA LGOB LH LHRH LIF LINks LL LM LMA LMN LMW LMWC LMWH LOD



LOH LOS LP LPC LPR LR LREC LSCC LT LTAS LTASS LTB



xl ] Abbreviations LVAS LVN LVOR



large vestibular aqueduct syndrome lateral vestibular nuclei linear vestibulo–ocular reflex



M MAb MABP MAC



metastases monoclonal antibodies mean arterial blood pressure membrane attack complex; or Mycobacterium avium complex magnetic-activated cell sorter; or minimal access cranial suspension minimum audible field mucosa-associated lymphoid tissue monoamine oxidase inhibitor minimum audible pressure mitogen-activated protein kinase mandibular advancement splint medial bundle mannose-binding lectin major basic protein modified barium swallow monocyte chemotactic protein monocyte chemotactic protein-1 mental component summary macrophage-colony stimulating factor mean corpuscular volume macrophage derived chemokine multiple drug resistance multidrug resistant tuberculosis myelodysplastic syndrome multidisciplinary team Multidimensional Voice Program middle ear medial edge epithelium magnetoencephalography MAPK/extracellular signal related kinase mitochondrial encephalopathy, lactic acidosis and stroke-like episode microelectromechanical system multiple endocrine neoplasia myoclonic epilepsy and ragged red fibre medical subject heading microscopic endonasal sinus surgery middle ear transducer middle fossa multifluor FISH mean airflow rate medial geniculate body melanoma growth stimulating activity monoclonal gammopathy of uncertain significance major histocompatibility complex myocardial infarction metaiodobenzylguanidine; or iodine-123metaiodobenzylguanidine sestamibi; or technetium-99m minimum inhibitory concentration



MACS MAF MALT MAOI MAP MAPK MAS MB MBL MBP MBS MCP MCP-1 MCS M-CSF MCV MDC MDR MDRTB MDS MDT MDVP ME MEE MEG MEK MELAS MEMS MEN MERRF MeSH MESS MET MF M-FISH MFR MGB MGSA MGUS MHC MI MIBG MIBI MIC



MIDD MIP



MIP-1a MISS MIT MIVAP ML MLD MLF MLR MLTB MM MMC MMN MMP MMR MMS MND MNG MOC MODS MOT MPA MPL MPO MPT MPTP MR MRA MRC MREC MRI MRL mRNA MRND MRS MRSA MRV MS MSA MSBOS MSG MST MT MTC MTD mtDNA MTHFR mTHPC MUS



maternally inherited diabetes and deafness minimally invasive open parathyroidectomy; or maximum intensity projection; or macrophage inflammatory protein macrophage inflammatory protein-1a minimally invasive sinus surgery monoiodotyrosine minimally invasive video-assisted parathoidectomy mixed cellularity masking level difference medial longitudinal fascicle or fasciculus middle latency response microlaryngotracheobronchoscopy malignant melanoma mitomycin C mismatch negativity mucous membrane pemphigoid; or matrixmetalloprotease measles, mumps and rubella Moh’s micrographic surgery motor neurone disease multinodular goitre medial olivocochlear multiple organ dysfunction syndrome malignant odontogenic tumour microscopic polyangiitis monophosphoryl lipid A myeloperoxidase maximum phonation time 1-methyl-4-phenyl-1,2,3,6tetrahydropyridine magnetic resonance magnetic resonance angiography Medical Research Council (UK) multicentre regional ethics committee magnetic resonance imaging minimal response level messenger ribonucleic acid modified radical neck dissection Melkersson–Rosenthal syndrome; or magnetic resonance sialography methicillin-resistant Staphylococcus aureus migraine-related vestibulopathy multiple sclerosis multiple systems atrophy maximum surgical blood ordering schedule monosodium glutamate maximal stimulation test maxilloturbinal; or middle turbinate medullary thyroid carcinoma muscle tension dysphonia mitochondrial DNA methylenetetrahydrofolate reductase meso-tetra (hydroxyphenyl) chlorin medically unexplained symptom



Abbreviations MV MVN



mechanical ventilation medial vestibular nuclei



NMCC NMDA



N NA NADP



nodal noradrenaline nicotinamide adenine dinucleotide phosphate reduced form of nicotinamide adenine dinucleotide phosphate National Acoustic Laboratories (Australia) National Ambulatory Medical Care Survey (USA) nonallergic noninfectious perennial rhinitis nonallergic rhinitis with eosinophilia syndrome National Anonymous Tonsil Archive n-butyl-2-cyanoacrylate; or N-butylcyanoacrylate nitro blue tetrazolium National Clinical Assessment Authority (UK) National Clinical Assessment Service (UK) National Clinical Audit Support Programme (UK) non-consultant career-grade National Comprehensive Cancer Network National Cancer Data Base (USA) National Confidential Enquiry into Patient Outcome Death (UK) National Cancer Institute of Canada nose, ear, eye and temple North of England and Scotland Study on Tonsillectomy and Adenoidectomy in Children nerve excitability test nuclear factor kappa B neurofibromatosis type 1 neurofibromatosis type 2 nonfunctioning pituitary adenomas; or nasofrontal approach nasogastric normal hearing non-Hodgkin’s lymphoma National Health Service (UK) Newborn Hearing Screening Programme automatic noninvasive blood pressure National Institute for Health and Clinical Excellence (UK) nonimmunological contact urticaria; or neonatal intensive care unit noninsulin dependent diabetes mellitus National Institutes of Health (USA) noise-induced hearing loss nasal inspiratory peak flow Na+/I– symporter natural killer Newtons/square metre



NNE NNT NO NO2 NOE non-REM NOS NP NPC



NADPH NAL NAMCS NANIPER NARES NATA NBCA NBT NCAA NCAS NCASP NCCG NCCN NCDB NCEPOD NCIC NEET NESSTAC



NET NFkB NF1 NF2 NFA NG NH NHL NHS NHSP NIBP NICE NICU NIDDM NIH NIHL NIPF NIS NK N/m2



] xli



NT NTD NTM NTS NYHA



nasal mucociliary clearance N-methyl-d-aspartate; or National Minimum Data Set (UK) normalized noise energy number needed to treat nitric oxide nitric dioxide naso-orbito-ethmoid nonrapid eye movement sleep not otherwise specified nasopharynx; or nasopharyngeal nasopharyngeal cancer; or nasopharyngeal carcinoma National Patient Safety Agency (UK) National Prospective Tonsillectomy Audit (UK) negative predictive value neuropeptide Y nucleus retroambigualis National Reporting and Learning System (UK) neural response telemetry nodular sclerosing nonsteroidal antiinflammatory drug National Specialist Commissioning Advisory Group (UK) national service framework neonatal severe hyperparathyroidism nonsyndromic recessive auditory neuropathy nasoturbinal neural tube defect non-tuberculous mycobacteria nucleus tractus solitarius New York Heart Association



O3 OAE OAN OAS OB OCB OCFC OCT ODI ODT OEC OFG OGTT OHC OHL OHS OKN OM OMC OME OMENS



ozone otoacoustic emission olfactory neuroblastoma oral allergy syndrome olfactory bulb olivocochlear bundle open controlled food challenge optical coherence tomography oxygen desaturation index olfactory detection threshold olfactory ensheathing cell orofacial granulomatosis oral glucose tolerance test outer hair cell oral hairy leukoplakia obesity hypoventilation syndrome optokinetic nystagmus occipitomental ostiomeatal complex otitis media with effusion orbit, mandible, ears, nerves and soft-tissue



NPSA NPTA NPV NPY NRA NRLS NRT NS NSAID NSCAG NSF NSHPT NSRAN



xlii ] Abbreviations OMIM OPCS OPG OR OREP ORL OS OSA OSAH OSAHS OSAS OSC OSPH OSPL OTOF OVAR P PA PAC PAD PAF PAG PAI-1 PALS PA-RT PAS PBP PCA PCC PCD PCHI PCNA PCR Pcrit PCS PCT PCTR PD PD-ECGF PDGF PDGFR PDL PDR PDS PDT PE PEEP PEG PEMA/THFMA



Online Mendelian Inheritance in Man Office for Population Censuses and Surveys (UK) orthopantomogram occupational rhinitis olfactory event-related potential otorhinolaryngology osteosarcoma obstructive sleep apnoea obstructive sleep apnoea/hypopnoea obstructive sleep apnoea/hypopnoea syndrome obstructive sleep apnoea syndrome overview and scrutiny committee ostium of sphenoid sinus output sound pressure level otoferlin off-vertical axis rotation phosphate; or posterior pernicious anaemia P1 artificial chromosome; or pulmonary artery catheter preoperative autologous deposit platelet-activating factor periaqueductal grey matter plasminogen activator inhibitor type 1 Patient Advice and Liaison Service (UK) partly accelerated radiotherapy periodic acid–Schiff progressive bulbar palsy patient-controlled analgesia prothrombin complex concentrate; or Professional Conduct Committee (UK) primary ciliary dyskinesia permanent childhood hearing impairment proliferating cell nuclear antigen polymerase chain reaction critical pressure physical component summary primary care trust partial cricotracheal resection Parkinson’s disease platelet-derived endothelial cell growth factor platelet-derived growth factor platelet-derived growth factor receptor pulsed dye laser Physicians’ Desk Reference polydimethylsiloxane photodynamic therapy polyethylene; or pulmonary embolism; or pharyngo–oesophageal positive-end expiratory pressure percutaneous endoscopic gastrostomy poly (ethylmethacrylate)/tetrahydrofurfuryl methacrylate



PET PET-CT PF PF4 PFAPA PFC PFG PGA PGE1 PGI2 PGL pHPT PI PI3-K PICA PICU PIF PIFR PIHA PIII PIV PIVC PLA PLD PLF PLG PLMD PLS PM PMS PNP PNS POGO PONV PORP PP PPC PPD PPI PPRF PPS PPV PR3 PRCT PRL PRP PRPP PRS PRV



polyethylene terephthalate; or positron emission tomography positron emission tomography/computed tomography posterior fontanelle; or cisplatinum/5fluorouracil platelet factor 4 periodic fever, aphthous stomatitis, pharyngitis and cervical adenitis perfluorocarbon percutaneous fluoroscopic gastrostomy polyglycolic acid prostaglandin-E1 prostacycline; or prostaglandin I2 persistent generalized lymphadenopathy primary hyperparathyroidism pulsatility index phosphotidyinositol 3 posterior inferior cerebellar artery paediatric intensive care unit prolactin release inhibiting factor peak inspiratory flow partially implantable hearing aid parathyroid III parainfluenza virus; or parathyroid IV parietoinsular vestibular cortex polylactic acid potentially lethal damage congenital perilymphatic fistula polylactide-coglycolide periodic limb movement disorder primary lateral sclerosis particulate matter pharyngeal mucosal space purine nucleoside phosphorylase; or paraneoplastic pemphigus peripheral nervous system; or postnasal space prescription of gain and output postoperative nausea/vomiting partial ossicular replacement prosthesis pyrophosphate Preliminary Proceedings Committee (UK) purified protein derivative proton pump inhibitor; or patient and public involvement parapontine reticular formation; or paramedian pontine reticular formation parapharyngeal space positive predictive value proteinase 3 prospective randomized controlled trial prolactin platelet-rich plasma 5-phospho-alpha-D-ribose 1-diphosphate persistent rhinosinusitis polycythaemia rubra vera



Abbreviations PSA



PVS PZT



prostate-specific antigen; or pleomorphic salivary adenoma; or persistent stapedial artery posterior semicircular canal polysomnography polarization-sensitive OCT progressive supranuclear palsy peak systolic velocity prothrombin time pure tone average; or peritonsillar abscess psychophysical tuning curve polytetrafluoroethylene parathyroid hormone parathyroid hormone-related protein; or parathyroid hormone-related peptide pathological tumour, nodes, metastases post-transfusion purpura permanent threshold shift propylthiouracil uncinate process pemphigus vulgaris polyvinyl alcohol polyvinyl chloride posteroventral cochlear nuclei pause vestibular position; or position vestibular pause persistent vegetative state lead zirconate titanate



QALY QOL QTL



quality adjusted life year quality of life quantitative trait loci



RA RAE RAI RALP RAM RANTES



retinoic acid Ring, Adair, Elwyn radioactive iodine right anterior–left posterior Rahmonic amplitude regulated on activation, normal T-cell expressed and secreted right anteroposterior retinoic acid receptor a gene recurrent acute rhinosinusitis recurrent aphthous stomatitis radioallergosorbent test rapid antigen testing retinoblastoma red blood cell regional cerebral blood flow Royal College of Paediatrics and Child Health randomized controlled trial respiratory disturbance index real-ear aided gain Revised European American Lymphoma real ear to coupler difference real-ear insertion gain



p-SCC PSG PS-OCT PSP PSV PT PTA PTC PTFE PTH PTHrP pTNM PTP PTS PTU PU PV PVA PVC PVCN PVP



RAP RARa RARS RAS RAST RAT RB RBC rCBF RCPCH RCT RDI REAG REAL RECD REIG



REM rEPO RET RFS RFTVR RFVR RHD RI RIA riMLF RLN RLS RMS RNA RND RNID RNP ROC ROI ROM ROOF ROS RP RPA RPT RR RRP RRR RS RSDI RSOM RSTL RSV RT rT3 RTK RTL rTMS RT-PCR RUDS



SACE SAD SAGM SALT SANS SAP SAPALDIA SBS



] xliii



rapid eye movement recombinant erythropoietin rearranged during transfection rhinofrontal sinuseptotomy radiofrequency tissue volume reduction radiofrequency volumetric reduction Reported Hearing Disability resistance index radioimmuno assay rostral interstitial nucleus of the medial longitudinal faciculus recurrent laryngeal nerve restless leg syndrome root mean square; or rhabdomyosarcoma ribonucleic acid radical neck dissection Royal National Institute for Deaf and Hard of Hearing People (UK) ribonucleoprotein receiver operating characteristic region of interest; or reactive oxygen intermediate range of motion retro-orbicularis orbital fat reactive oxygen species rapid prototyping retropharyngeal abscess rapid pull through relative risk recurrent respiratory papillomatosis relative risk reduction retrosigmoid Rhinosinusitis Disability Index rhinosinusitis outcome measure relaxed skin tension line respiratory syncytial virus radiotherapy reverse triiodothyronine receptor tyrosine kinase right thyroid artery repetitive low-frequency transcranial magnetic stimulation reverse transcriptase-polymerase chain reaction reactive upper airways dysfunction syndrome serum angiotensin converting enzyme supraglottic airway device saline-adenine-glucose-mannitol speech and language therapist subacute necrotizing sialadenitis signalling lymphocyte activation molecule associated protein Swiss Study on Air Pollution and Lung Diseases in Adults sick building syndrome



xliv ] Abbreviations s.c. SCBU SCC SCCA SCCHN SCD SCF SCID SCN SCUBA ScvO2 SEAC SEM sEMG SF-36 SfBH SFF SFOAE SGC Shh SHO SHOT SIADH SIDS sIg SIGN SIMEHD SIP SIR SIRS SL SLD SLE SLIT SLN SLNB SLP SLT SMAS SMOFIT SMR SMS S/N SNC SNHL SNOMED



subcutaneous special care baby unit squamous cell carcinoma or cancer; or semicircular canal squamous cell carcinoma antigen squamous cell carcinoma of the head and neck sickle cell disease stem cell factor severe combined immunodeficiency severe congenital neutropenia self-contained underwater breathing apparatus central venous oxygen saturation Spongiform Encephalopathy Advisory Committee scanning electron microscopy surface electromyography Medical Outcome Study Short-Form 36-Item Health Survey Standards for Better Health (UK) speaking fundamental frequency stimulus frequency otoacoustic emission spiral ganglion cell sonic hedgehog senior house officer serious hazards of transfusion syndrome of inappropriate antidiuretic hormone sudden infant death syndrome surface immunoglobulin Scottish Intercollegiate Guidelines Network semi-implantable middle ear electromagnetic hearing device sickness impact profile speech intelligibility rating; or standardized incidence ratio systemic inflammatory response syndrome sensation level sublethal damage systemic lupus erythematosus sublingual immunotherapy superior laryngeal nerve sentinel lymph node biopsy superficial lamina propria speech and language therapist superficial or subcutaneous muscloaponeurotic system submucous resection of the turbinate submucosal resection short message service; or indium-111 pentetreotide speech-to-noise sinonasal cancer sensorineural hearing loss Systematized nomenclature of medicine



SNOMED CT SNOT SNR SNUC SO2 SOAE SOC SOM SOOF SOS SP SPECT SPET SPF SPI SPIO SPL SPT SRS SRS-A SRT SSC SSEP SSG SSLP SSNHL SSPE SSPL SSR SSRI ST STAT STD STIR STRP SUV SVCO SVL SVN SVV SVZ SWS T T1WI T2WI T3 T4 T/A TAGVHD



Systematized Nomenclature of Medicine – Clinical Terms sino-nasal outcome test signal-to-noise ratio sinonasal undifferentiated carcinoma sulphur dioxide spontaneous otoacoustic emission superior olivary complex secretory otitis media suborbicularis oculi fat guanine nucleotide exchange factor (son of sevenless) substance P; or summating potential single photon emission computed tomography single photon emission tomography sphenopalatine foramen soft phonation index superparamagnetic iron oxide sound pressure level skin prick test; or station pull through subacute rhinosinusitis slow reacting substance of anaphylaxis speech recognition threshold; or speech reception threshold superior semicircular canal steady-state potential split skin graft simple sequence length polymorphism sudden sensorineural hearing loss subacute sclerosing panencephalitis saturation sound pressure level steady-state response selective serotonin reuptake inhibitor superior turbinate signal transducer and activator of transcription standard deviation short time inversion recovery short tandem repeat polymorphism standardized uptake value superior vena caval obstruction strobovideolaryngoscopy superior vestibular nuclei; or superior vestibular nerve subjective visual vertical subventricular zone slow wave sleep thymine; or tumour T1-weighted images T2-weighted images triiodothyronine thyroxine tonsillectomy and/or adenoidectomy transfusion-associated graft-versus-host disease



Abbreviations TARC TARGET



thymus and activation-regulated chemokine Trial of Alternative Regimens in Glue Ear Treatment TB tuberculosis; or Mycobacterium tuberculosis TBG thyroxine-binding globulin Tc T cytotoxic Tc-99m technetium Tc-99m (v) DMSA pentavalent dimercaptosuccinic acid TC thyroid cartilage TCF tracheocutaneous fistula TCI target-controlled infusion TCP tricalcium phosphate TCR T cell receptor TdT terminal deoxynucletidyl transferase TEC Tissue Engineering and Regenerative Medicine Centre TENS transcutaneous electrical nerve stimulation TEOAE transient evoked otoacoustic emission TEP tracheo-oesophageal puncture TFG temporalis fascia graft TFT thyroid function test TG thyroglobulin TGF transforming growth factor TGF-a transforming growth factor alpha TGF-b transforming growth factor beta TGF-b1 transforming growth factor beta 1 Th T helper TIA transient ischaemic attack TIBC total iron binding capacity TICA totally implantable cochlear amplifier TKI tyrosine kinase inhibitor TM tympanic membrane TMC1 transmembrane channel-like gene 1 TMD temporomandibular disorder TMJ temporomandibular joint TMTF temporal modulation transfer function TN trigeminal neuralgia; or trigeminal nerve TNF tumour necrosis factor TNF-a tumour necrosis factor alpha TNM tumour, node, metastasis TOAE transient evoked otoacoustic emission TOE transoesophageal echocardiography; or Trichophyton, Oidiomycetes and Epidermophyton TOF tracheo-oesophageal fistula TOF-o-gram tracheo-oesophageal fistulogram TORP total ossicular replacement prosthesis TPA tissue polypeptide antigen TPF docetaxel/cisplatinum/5-fluorouracil; or temporoparietal fascia TPHA T. pallidum haemagglutination test; or treponemal haemagglutination TPI T. pallidum immobilization TPN total parenteral nutrition TPO thyroid peroxidase; or thyroperoxidase Tpot potential doubling times TQM total quality management



TRALI TRAM TRH tRNA TRP TRT TSG TSH TSHoma TSS TT TTN TTP TTR TTS TUNEL TXA2 U UADT UARS UCL UICC UK-CCSG UKCISG UMN UMP UNICEF UOS UP UPP UPPP UPSIT URT URTI US USH USH1B USPIO UV uVD UVPP UWQOL



VA VAAP VAC VAM VAS VATER VCA



] xlv



transfusion-related acute lung injury transverse rectus abdominis myocutaneous thyrotropin-releasing hormone transfer ribonucleic acid transient receptor potential tinnitus retraining therapy tumour suppressor gene thyroid-stimulating hormone; or thyrotropin TSH-secreting adenoma transitional space surgery thrombin time thalamic taste nucleus thrombotic thrombocytopeniac purpura transthyretin temporary threshold shift TdT-mediated nick end labelling thromboxane A2 uracil upper aerodigestive tract upper airway resistance syndrome uncomfortable loudness level International Union Against Cancer United Kingdom Children with Cancer Study Group UK Cochlear Implant Study Group upper motor neuron uridine monophosphate United Nations Children’s Fund upper oesophageal sphincter uncinate process uvulopalatopharyngoplasty uvulopalatopharyngoplasty University of Pennsylvania Smell Identification Test upper respiratory tract upper respiratory tract infection ultrasound; or ultrasonography Usher syndrome Usher syndrome type 1B ultra-small super paramagnetic iron oxide ultraviolet unilateral vestibular deafferentiation uvulopalatopharyngoplasty University of Washington Quality of Life Questionnaire Veterans’ Affairs; or vestibular aqueduct voice activity and participation vacuum-assisted closure variation of amplitude visual analogue scale; or visual analogue score vertebral, anal, tracheooesophageal and radial viral capsid antigen



xlvi ] Abbreviations VCAM-1 vCJD VCR VDRL VEES VEGF VEMP VEP VFSS VHI VHI-10 VHL VHQ VHT VILI VIP VLA VLA4 VLDL VMA VN VOC VOG VoiSS VOR VORP VORS VPI



vascular cell adhesion molecule-1 variant Creutzfeldt-Jakob disease vestibulocollic reflex Venereal Disease Research Laboratory video endoscopic evaluation of swallowing vascular endothelial growth factor vestibular-evoked myogenic potential vestibular evoked potential videofluoroscopic swallowing study Voice Handicap Index Voice Handicap Index-10 Von Hippel–Lindau Vertigo Handicap Questionnaire vestibular habituation training ventilator induced lung injury vasoactive intestinal polypeptide very late activation antigen very late activation antigen 4 very low-density lipoprotein vanillylmandelic acid vestibular nuclei; or vagus nerve volatile organic compound video-oculography voice symptom scale vestibulo-ocular reflex vibrating ossicular prosthesis vestibulo-ocular reflex suppression velopharyngeal insufficiency



VPQ VRA VRE V-RQOL VS VSM VSR VTE VVI vWD vWF VZV



patient questionnaire of vocal performance visual reinforcement audiometry vancomycin-resistant enterococci voice-related quality of life vestibular schwannoma velocity storage mechanism vestibulospinal reflex venous thromboembolism vocal velocity index von Willebrand disease von Willebrand factor varicella zoster virus



WAS WBC WHO WMD WOB WP WPC



Wiskott Aldrich syndrome white blood cell World Health Organization weighted mean difference work of breathing Woodruff ’s plexus WARN, PAUSE, CHECK



XHIM XLA XLP



X-linked hyper immunoglobin M X-linked agammaglobulinaemia X-linked lymphoproliferative syndrome



YAC YAG



yeast artificial chromosome yttrium aluminium garnate



PART



1



CELL BIOLOGY EDITED BY NICHOLAS S JONES



1 Molecular biology Michael Kuo and Richard Irving



3



2 Genetics Karen P Steel



15



3 Gene therapy Scott M Graham and John H Lee



23



4 Mechanisms of anticancer drugs Sarah Payne and David Miles



34



5 Radiotherapy and radiosensitizers Stewart G Martin and David AL Morgan



47



6 Apoptosis and cell death Michael Saunders



56



7 Stem cells A John Harris and Archana Vats



66



This page intentionally left blank



1 Molecular biology MICHAEL KUO AND RICHARD IRVING



Introduction Molecular genetics: DNA structure and function Key points Methods in molecular biology Key points Molecular aberrations of cellular biology Key points



3 3 5 5 8 8 10



Mapping and identification of genes associated with disease Key point Deficiencies in current knowledge and areas for future research References Further reading



11 11 11 13 14



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words molecular biology, genetics, and cell biology.



INTRODUCTION Molecular biology describes the study of the biochemical processes that govern the behaviour of cells. These processes form the fundamental mechanisms by which cell function, cell–cell interactions and cell turnover are regulated. Disruption of this regulation may lead to disease, whilst an understanding of these mechanisms allows the physician to attempt to predict disease behaviour and to explore methods of restoring this regulation at a molecular level. This chapter reviews the principles of molecular genetics and outlines aspects of the molecular biology of the cell in the context of otolaryngological disease processes and describes some of the techniques that form the backbone of current molecular biology. It should give the reader sufficient background knowledge of molecular biology to understand subsequent chapters discussing the molecular biology of specific otolaryngological conditions.



MOLECULAR GENETICS: DNA STRUCTURE AND FUNCTION Hereditary information in eukaryotes is stored in the form of double-stranded deoxyribonucleic acid (DNA)



and is referred to as the genome. DNA forms a doublehelix structure as a result of hydrogen bonds between complementary pairs of nucleotides, adenine (A) with thymine (T) and cytosine (C) with guanine (G). The nucleotides on each strand are organized linearly in triplets, known as codons. Each specific sequence determines a single specific amino acid, for example ACU specifies threonine. However, as there are more triplet combinations (64) than commonly encountered amino acids (20), some proteins may be represented by different codons (e.g. lysine by AAA as well as AAG) and some codons (UAA, UGA and UAG) are ‘stop’ codons, constituting a signal for arrest of translation. The overwhelming majority of this DNA (99.9 percent) exists in the cell nucleus as the nuclear genome, which, in the human, is estimated to be 3000 megabase pairs in physical size and encodes 30,000–35,000 genes. The remaining DNA (16.6 kilobase pairs) forms the mitochondrial genome, encoding 37 genes. The mitochondrial genome and its potential role in cancer diagnostics will be discussed later. Each DNA molecule is packaged into a chromosome by complex folding of the DNA around proteins. Diploid human cells contain 22 pairs of autosomes (1 to 22) and a



4 ] PART 1 CELL BIOLOGY pair of sex chromosomes (XX or XY) which determines the sex of the organism. One of each pair of chromosomes is maternally inherited and the other is paternally inherited. Each chromosome has a distinctive shape, size and banding pattern, but have the common appearance of two arms apparently separated by a constriction. The centromere is microscopically recognizable as the central constriction separating the chromosome into a long arm (q for queue) and a short arm (p for petit), but its biological role lies in anchoring the chromosome to the mitotic spindle for segregation during cell division. The ends of the chromosomes are capped by telomeres, which are specialized structures containing unique simple repetitive sequences. They maintain the structural integrity of the chromosome and provide a solution for complete replication of the extreme ends of the chromosome. The conventional nomenclature for chromosomal locus assignment is given by the chromosome number, followed by the arm and finally the position on the arm, for example, 3p21 indicates position 21(two-one) on the short arm of chromosome three. During normal cell division, DNA replication is achieved by the separation of the two strands by DNA helicase. Each separated single strand then acts as a template for polymerization, catalyzed by DNA polymerase, of nucleotides forming a new complementary strand and thus double-stranded DNA identical to the original dsDNA. As each daughter DNA consists of one original and one newly synthesized DNA strand, the process is known as semi-conservative replication. The specificity of the complementary relationship between the nucleotides on each strand forms the basis for many techniques of modern molecular biology and molecular cytogenetics.1 The accuracy with which DNA replication takes place is remarkable with an estimated error rate of less than one in 109 nucleotide additions. Such accuracy is of vital importance to the individual as a permanent change in DNA, or mutation may cause inactivation of a gene essential to cell survival or cell cycle control. The high fidelity of DNA sequence replication is achieved by unidirectional 50 -to-30 direction of DNA replication, a rigorous DNA proofreading mechanism which detects mismatched DNA and efficient DNA repair pathways which excise and repair DNA damage. Failure of these mechanisms, such as is encountered in xeroderma pigmentosum, Fanconi’s anaemia and ataxia telangiectasia, leads to accumulation of DNA replication errors and a high incidence of malignancies. Although the human nuclear genome is 3  109 base pairs in size, about 90 percent of it is noncoding, with all the genes being coded by the remaining 10 percent of the DNA. Within the noncoding DNA are dispersed short arrays of repeat units of pairs or triplets of nucleotides (di-/trinucleotides). The exact function of these microsatellite repeats is not entirely clear, but their existence and frequency of dispersion throughout the genome have greatly facilitated study of the genetics of



tumours and many inherited disorders, which will be discussed later. A gene is a region of the chromosomal DNA that produces a functional ribonucleic acid molecule (RNA). It comprises regulatory DNA sequences which determine when and in which cell types that gene is expressed, exons which are coding sequences and interspersed introns which are noncoding DNA sequences. These regulatory sequences often consist of CpG islands, short stretches of DNA rich in dinucleotides of cytosine and guanine. The methylation status of these CpG islands determines whether that gene is expressed in a particular cell or tissue, being unmethylated in tissues where the genes are expressed. As will be discussed later, aberration of this control is one of the mechanisms of tumour suppressor gene inactivation. Transcription is the intranuclear process driven by RNA polymerase whereby one of the two DNA strands acts as a template for the synthesis of a single RNA strand which is complementary to the DNA, except that uracil replaces thymine in RNA. This primary RNA transcript then undergoes posttranscriptional processing, or splicing.2 Traditional dogma held that one gene produces one protein and therefore splicing was considered to occur simply in order to remove the noncoding intronic sequences, producing messenger RNA (mRNA). It is now known that by ‘alternative splicing’, one gene can result in the production of several different but often related proteins in different tissues.3 The mature mRNA then migrates into the cytoplasm where it acts as a template for the synthesis of a polypeptide during translation, a process regulated and catalyzed by cytoplasmic ribosomes. Successive amino acids are added to the polypeptide chain according to the triplet code on the mRNA, which is recognized by the transfer RNA (tRNA), to which each corresponding amino acid is covalently bound. Translation is commenced upon recognition of an initiation codon (usually but not exclusively AUG/methionine) and terminated upon recognition of a stop codon. The polypeptide subsequently undergoes a variable degree of post-translational modification and/or cleavage to produce the mature protein product, which may have an intracellular role or may be exported to the endoplasmic reticulum and hence to the extracellular space to execute its function. The mitochondrial genome is considerably smaller than the nuclear genome, but it deserves mention here because of the increasing recognition of the role of mitochondrial DNA (mtDNA) mutations in human disease. The mitochondrial genome is only 16.6 kb in size, comprising 37 genes, which encode polypeptides which are principally involved in the respiratory chain. mtDNA is double-stranded but does not form a doublehelix nor does it form chromosomes, but instead it takes the form of a circular double-stranded DNA structure with a heavy and a light strand. Unlike the nuclear



Chapter 1 Molecular biology



genome, which is inherited from mother and father, the mitochondrial genome of an individual is entirely maternally inherited.



KEY POINTS  The double-stranded alpha helical structure of DNA, mainly located in the nucleus, consists of nucleotide triplets called codons which code for specific amino acids and stop signals, and forms the substrate for hereditary information in eukaryotes.  The 22 pairs of autosomes and one pair of sex chromosomes, each with their distinctive shape, size and banding pattern, represent a complex folding of DNA around proteins to give the characteristic shape of a central constriction (centromere) separating the chromosome into a long arm (q) and a short arm (p) with a telomere cap at each end to maintain structural integrity.  Chromosome locus nomenclature: chromosome number – 3p21 – position on chromosome arm.  Semiconservative replication of DNA during normal cell division results in the separation of two strands of DNA by DNA helicase, each strand then acting as a template for polymerization by DNA polymerase. High fidelity is vital to prevent permanent change or mutations.  A gene is a region of chromosomal DNA which produces functional RNA consisting of: – regulatory DNA sequences; – exons, which are coding sequences; – introns, which are noncoding sequences.  Transcription is the intranuclear process driven by RNA polymerase whereby one of the two DNA strands acts as a template for single-stranded RNA synthesis complementary to the DNA, except that in RNA U is replaced by T. Splicing refers to post-transcriptional processing of RNA.  Translation is the cytoplasmic process in which mRNA acts as a template for the synthesis of polypeptide by adding successive amino acids to the polypeptide chain, according to the triplet codon of the mRNA which is recognized by the tRNA to which the corresponding amino acid is covalently bonded. This process is regulated and catalyzed by cytoplasmic ribosomes. Posttranslational modification produces mature proteins.



]5



METHODS IN MOLECULAR BIOLOGY Basic techniques of DNA fragmentation and identification Unlike RNA, DNA is extremely stable, which is understandable from the function that each has in the cell. For purposes of studying the DNA and in order to clone specific DNA, the DNA molecule needs to be divided into manageable fragments. Although the ability to cut (and also to join up) DNA molecules now appears to be a very straightforward process, it was only 1970 when the first restriction endonuclease was identified in a strain of Haemophilus influenzae, hence its name HindII (pronounced Hin-dee-two). It is believed that this restriction endonucleases act in vivo in bacteria as an immune or host-defence system, recognizing non-self DNA in bacteriophages and cleaving them. By surveying many different bacteria, a wide range of restriction endonucleases is now available, each of which recognize specific target sites based on sequences of four to eight nucleotides. As a specific, seven nucleotide sequence (heptanucleotide) will occur less frequently than a four nucleotide sequence (tetranucleotide), statistically, endonucleases recognizing heptanucleotide targets will cut less frequently thereby yielding larger fragments than those recognizing tetranucleotides. As the DNA is doublestranded, the resultant fragments may have blunt ends or cohesive (‘sticky’) ends (Figure 1.1). The nature of the ends of DNA fragments thus generated impact upon the way in which they can be ligated (joined) into recombinant molecules. Ligation of DNA fragments with cohesive ends is more efficient than joining of bluntended fragments.



ELECTROPHORESIS



Negatively charged phosphate groups on the DNA backbone confer a net negative charge on linear DNA. This allows fragments of different sizes to be resolved within a suitable gel matrix by the application of an electric current across the matrix. The DNA will migrate toward the positive electrode with the smaller fragments travelling faster than the larger fragments.4 The size of the fragment can be estimated by the use of a graduated DNA GAATTC C T T AAG



EcoRI



G CTTAA



AATTC G



Cohesive ends



C CC GGG G GG C C C



Smal



CCC GGG



GGG CCC Blunt ends



Figure 1.1 DNA cleavage by restriction endonucleases. Derived from Ref. 11, with permission.



6 ] PART 1 CELL BIOLOGY ladder containing fragments of known molecular weight. The choice of the particular matrix depends on the fragment sizes that one is trying to resolve. Polyacrylamide gels can resolve differences of just one base pair between fragments of several hundred base pairs in size by virtue of a small pore size in the gel matrix. These gels can be used for DNA sequencing and resolution of alleles varying in only one dinucleotide repeat. Agarose gels can resolve fragment sizes from around 100 bp to 20 kb. Beyond that size, electrophoretic mobility is no longer proportional to fragment size. Resolution of fragments sizes in excess of 50 kb, such as larger bacterial artificial chromosomes (BAC) or yeast artificial chromosomes (YAC) require the use of pulsed field electrophoresis. HYBRIDIZATION



Hybridization is the specific annealing of single DNA (or RNA) strands, the probe, to a DNA sample, the target. It serves to detect the presence of a specific sequence of DNA either in the cell or on a hybridization membrane and recognition that hybridization has occurred is achieved either by radioactively labelling the probe and localizing the radioactivity by autoradiography or by labelling the probe with fluorochromes which fluoresce when excited by light of specific wavelengths (Figure 1.2). Hybridization on a membrane requires the initial transfer of DNA on to a nitrocellulose membrane from an agarose gel. This elegantly simple process is eponymously known as Southern blotting after the scientist who described the process in 1975. Two other commonly used transfer techniques have their names derived from Southern blotting as jargon terms. Northern blotting is essentially the same process used for transfer of RNA to a membrane. Western blotting is one of the mainstays of protein analysis and involves the transfer of electrophoresed protein bands from a polyacrylamide gel on to a nitrocellulose or nylon membrane to which they bind strongly. Detection of the protein is usually achieved by the use of antibodies to specific antigens presented by the protein with the antibody being labelled radioactively, enzymatically or fluorescently. Labelled probe Target tissue nucleic acid



Hybridization



= Biotin/digoxigenin in FISH or = Radioactive label = Fluorochrome in FISH



Detection (by autoradiography or fluorescence microscopy)



Figure 1.2 In situ hybridization.



CYTOGENETICS AND MOLECULAR CYTOGENETICS



Although microscopy had already reached high levels of resolution in the early 1930s, the correct number of human chromosomes was not determined until 1958. The era of classical cytogenetics was thus begun. Cytogenetics is the study of chromosomal abnormalities and rearrangements. It currently has a major role to play in prenatal diagnosis of Downs syndrome and other congenital syndromes characterized by numerical chromosomal abnormalities. In the early part of this century, Theodore Boveri proposed that cancer arose from chromosomal alterations. This hypothesis was not proven until the consistent chromosomal translocation, t(9;22), was demonstrated in chronic myeloid leukaemia. Since that time, cytogenetic analysis has been the mainstay of genetic analysis in reticuloendothelial malignancies, being responsible for the identification of consistent translocations in different leukaemias. Its use in solid tumours has been hampered by the difficulties of establishing short-term primary cultures from head and neck cancers for chromosomal analysis and the erratically acquired chromosomal changes in long-term cell lines, which may have occurred in vitro, influenced by culture conditions. Nevertheless, some studies have identified chromosomal areas consistently showing frequent breakpoints suggesting the location of putative tumour suppressor genes (including 3p21, 5p14, 8p11, 17p21, 18q21) and gain or amplification implying the presence of putative protooncogenes at other sites (including 3q, 5p, 8q, 11q13). Although the refinement of karyotyping has been radically enhanced by the introduction of 24-colour combinatorial multifluor FISH (M-FISH), the resolution and therefore utility of solid tumour karyotyping remains limited.5 Hybridization to target DNA in cells, using fluorescence detection, is known as fluorescence in situ hybridization (FISH). Fluorescence in situ hybridization allows the analysis of copy number of a known specific DNA sequence within intact nuclei. In reticuloendothelial malignancies and solid tumour-derived cell lines, the use of both single-copy probes and centromere alpha-satellite repeat probes on metaphase preparations has enhanced and refined classical karyotyping. Interphase FISH has been applied to solid tumour sections to assess the copy number of a known sequence in breast, prostate, bladder, brain, lung and head and neck tumours. Fluorescence-labelled hybridization has also been combined with cytogenetics to produce the powerful technique of comparative genomic hybridization (CGH).6 Comparative genomic hybridization permits the rapid medium resolution screening of the entire genome by comparatively hybridizing matched tumour and normal DNA from a patient, which are labelled with different fluorochromes, on to normal metaphase chromosome preparations. Under red-green dual filter fluorescence microscopy and computer-aided image analysis, areas of



Chapter 1 Molecular biology



genetic ‘neutrality’ appear yellow, under-representation appears green, and over-representation appears red. Areas of genetic under-representation suggest the possibility of a tumour suppressor gene lying within that region while areas of over-representation may indicate the location of a putative oncogene. This technique has been applied to the rapid genetic analysis of many tumour types including squamous cell carcinomas of the head and neck. The advent of molecular cytogenetics has obviated the need for primary short-term cultures and refined the location of chromosomal aberrations in solid tumours. POLYMERASE CHAIN REACTION



Perhaps the single molecular technique which has had the most dramatic impact on molecular biology has been the polymerase chain reaction (PCR). The original problem lay in obtaining sufficient quantities of a particular DNA sequence such that DNA profiling (e.g. sequencing) and DNA manipulation (e.g. cloning) could be achieved. The only ‘requirement’ is that the sequences flanking the stretch of DNA of interest is known. With that proviso, PCR achieves faithful and exponential amplification of a specific sequence of DNA by repeated cycles each consisting of dsDNA denaturation, hybridization of specific oligonucleotides (primers) and extension of the polynucleotide by rapidly altering the reaction temperature between segments of each cycle. dsDNA denaturation is achieved by raising the temperature of the reaction to 941C for 30 seconds, thus disrupting the hydrogen bonds between the strands and exposing the hydrogen bond donor and acceptor groups to allow base pairing. The oligonucleotide primers are then allowed to hybridize to the denatured DNA (annealing) at around 55–651C for 90 seconds before the reaction temperature is raised to 721C to permit extension of the DNA strand by DNA polymerase in the presence of deoxynucleoside triphosphates (dNTPs). With each cycle resulting in the



doubling of the copies of the DNA sequence, a 30-cycle PCR taking approximately two hours would amplify a single copy of a DNA sequence 268 million-fold (Figure 1.3). Although the PCR was originally described by Mullis and Faloona in 1987, one practical problem prevented its instant exploitation.7 The DNA polymerase used in the original reaction was denatured during the DNA denaturation segment and therefore had to be added after each and every cycle. The solution came in 1989 when Lawyer isolated and characterized the DNA polymerase, Taq polymerase, from the thermophilic bacterium Thermus aquaticus which normally resided in temperatures above 951C.8 This polymerase was therefore ‘heat resistant’ and did not need to be replenished between cycles. The PCR holds a central position in many molecular biological techniques as well as clinical diagnostic methods. The fundamental principle of DNA amplification has been adapted to amplify messenger RNA and to amplify areas where the initial flanking oligonucleotide sequences are not known. It is often described as a sensitive and powerful technique, but with great power comes the potential for corruption! In theory, a single copy of DNA can be amplified. Therefore, careless experimental technique may lead to contamination of the DNA sample with other DNA (e.g. from the skin of the investigator) and consequently to an artefactual result. The Taq polymerase originally described in the technique does not have proofreading properties, but newer cloned enzymes such as Pfu polymerase incorporates a proofreading function to increase amplification fidelity for sequencing reactions. The sensitivity of PCR also presented a problem for the analysis of genetic alterations in certain solid tumours. Squamous cell carcinomas of the head and neck are histologically often characterized by a large stromal element within the tumour. The genetic alterations in the tumour may not be present in the stromal



Segment 1: Denaturation of double-stranded DNA by heating to 94°C 5′



3′ 50 percent);26  indicates the probability of multiglandular disease (and the need for a bilateral neck exploration);  suggests the probability of postoperative hypoparathyroidism and the potential for autografting of removed parathyroid tissue (final PTH below the lower limit of detection). This approach seems particularly appropriate for specialist centres treating a high proportion of failed or recurrent hyperparathyroidism, but there is no evidence that it is needed for previously untreated primary hyperparathyroidism. [**]



ULTRASOUND, CT AND MRI



Local expertise with ultrasound, CT and MRI may determine its use in a particular centre. Ultrasound is widely available but highly dependent on operator skill, with the result that the reported sensitivity of this technique in parathyroid disease ranges from 30 to 90 percent.21 Particular limitations of ultrasound are the inability to identify medially situated upper parathyroid glands and those which are intrathyroidal or in the mediastinum. Ultrasound is good for identifying parathyroids behind the thyroid or below the lower poles. Identification of their vascular pedicles by Doppler ultrasound can also help distinguish thyroid and parathyroid adenomas. CT and MRI find their main role in the localization of mediastinal parathyroid glands. The major limitation with CT is the similar densities of thyroid and parathyroid tissue and, in the neck, MRI may be preferable. Even so, MRI is generally less sensitive and specific than sestamibi scanning.14



INTRAOPERATIVE GAMMA PROBE LOCALIZATION AND RAPID PTH ASSAY



In difficult cases, a combination of preoperative scintigraphy, introperative gamma probe gland localization with measurement of tracer content of removed tissue22 and confirmation by rapid PTH assay may be helpful.23 In this approach, 99mTc-sestamibi is administered two to three hours before operation and a hand-held gamma probe is used to localize abnormal parathyroid tissue (420 percent above background).24 A requirement is that the preoperative sestamibi scan has shown the presence of an adenoma (approximately 80 percent of cases), although the gamma probe may be useful where there has been a negative scan in reoperative procedures or where there is an ectopic gland.25 While helpful, the gamma probe may not provide unique information which



Indications for parathyroid imaging The need for parathyroid imaging is driven entirely by the interest in limited surgery for hyperparathyroidism. This seems reasonable for a condition which is due to a solitary adenoma in almost 90 percent of patients.18 Minimally invasive parathyroidectomy offers the prospect of shorter operation times under local anaesthetic as day cases27 with potential cost savings and may follow the trend in other areas by being driven by patient preference. The need here is to identify cases with a single adenoma, since those with hyperplasia, multigland disease or abnormal neck anatomy (goitre, previous neck surgery or irradiation) will need a bilateral operation, which remains the ‘gold standard’ against which newer approaches are judged. A recent meta-analysis suggested that 87 percent of patients with sporadic primary hyperparathyroidism would be candidates for a unilateral parathyroidectomy.18 However, while a recent systematic review of minimally invasive parathyroidectomy suggested that it may have an advantage over the bilateral exploration, the confidence intervals were wide and there may have been bias due to rigorous case selection.28 There is general agreement that patients with failed or recurrent hyperparathyroidism require detailed preoperative imaging, and this approach is increasingly being extended to the previously untreated case since localization of the adenoma allows unilateral operation which is of shorter duration. However, the experienced parathyroid surgeon has a better than 90 percent probability of success compared to the 80 percent identification rate of sestamibi and may well find the small gland missed by scanning.29 The conventional bilateral surgical approach of identifying all the parathyroid glands may succeed where scanning may fail if there are asymmetrically enlarged glands and only the largest adenoma is identified. [***]



Chapter 30 Parathyroid function tests and imaging



hyperparathyroidism is dependent on careful case selection to eliminate possible cases of hyperplasia or multiple adenomata. Development of an imaging technique which was able to distinguish multiple parathyroid glands from the normal or adenomatous thyroid would greatly increase the scope of preoperative gland localization.



KEY POINTS  No single test of parathyroid function is diagnostic.  Measurement of serum calcium, phosphate, magnesium and PTH are the main tests of function.  The two-site IRMA for intact PTH is reliable and discriminatory.  Misdiagnosis of calcium disorders is usually due to lack of attention to history and examination rather than incorrect interpretation of laboratory tests.  Intraoperative measurement of PTH may facilitate management.  99mTc-sestamibi parathyroid imaging has high sensitivity and specificity. It is increasingly used to identify a solitary adenoma and facilitate unilateral parathyroidectomy.  Ultrasound, CT and MRI may help in the localization of ectopic glands.



REFERENCES







Best clinical practice [ Dual phase single tracer scintigraphy or dual isotope







subtraction scans are the method of choice for the identification of a parathyroid adenoma. [ 99mTc-sestamibi scanning has 90 percent sensitivity and almost 99 percent specificity for a solitary parathyroid adenoma. [ SPET, ultrasound, CT and MRI may have a place where there is local expertise but they are less sensitive and specific. [ Rapid PTH assays with a turnround time of 15 minutes provide important additional information during parathyroidectomy.



Deficiencies in current knowledge and areas for future research



$



$



Refinements in the assays for PTH are needed to eliminate cross reactivity with peptide fragments. This would be important in better defining changes in parathyroid function in disease states since this would influence both diagnostic precision and facilitate management. Such assays would need validation over a wide range of disease states including renal failure. Although sestamibi scanning is now well established, its use in the management of primary



] 385







1. Portale AA. Blood calcium, phosphorus, and magnesium. In: Favus MJ (ed.). Primer on the metabolic bone diseases and disorders of mineral metabolism. Philadelphia: Lippincott Williams and Wilkins, 1999: 115–8. 2. Walton RJ, Bijvoet OLM. Nomogram for derivation of renal threshold phosphate concentration. Lancet. 1975; 2: 309–10. 3. Rude RK. Magnesium metabolism and deficiency. Endocrinology and Metabolism Clinics of North America. 1993; 22: 377–95. 4. Lepage R, Roy L, Brossard JH, Rousseau L, Dorais C, Lazure C et al. A non-(1-84) PTH circulating fragment interferes significantly with intact PTH measurement by commercial assaysin uraemic samples. Clinical Chemistry. 1998; 44: 805–9. 5. Nussbaum SR, Zahradnik RJ, Lavigne JR, Brennan GL, Nozawa-Ung K, Kim LY et al. Highly sensitive two-site immunoradiometric assay of parathyrin and its clinical utility in evaluating patients with hypercalcemia. Clinical Chemistry. 1987; 33: 1364–7. 6. Brossard JH, Lepage R, Cardinal H, Roy L, Rousseau L, Dorais C et al. Influence of glomerular filtration rate on non-(1-84) parathyroid hormone (PTH) detected by intact PTH assays. Clinical Chemistry. 2000; 46: 697–703. 7. Gao P, Scheibel S, D0 Amour P, John MR, Rao SD, SchmidtGayk H et al. Development of a novel immunoradiometric assay exclusively for biologically active whole parathyroid hormone 1-84: implications for improvement of accurate assessment of parathyroid function. Journal of Bone and Mineral Research. 2001; 16: 605–14. 8. Slatopolsky E, Finch J, Clay P, Martin D, Sicard G, Singer G et al. A novel mechanism for skeletal resistance in uremia. Kidney International. 2000; 58: 753–61. 9. Blumsohn A, Hadari AA. Parathyroid hormone: what are we measuring and does it matter? Annals of Clinical Biochemistry. 2002; 39: 169–72. 10. Chase LR, Melson GL, Aurbach GD. Pseudohypoparathyroidism: Defective excretion of 30 50 -AMP in response to parathyroid hormone. Journal of Clinical Investigation. 1969; 48: 1832–44. 11. Levine MA. Pseudohypoparathyroidism: from bedside to bench and back. Journal of Bone and Mineral Research. 1999; 14: 1255–60. 12. Hinde E, Melliere D, Jeanguillaume C, Perlemuter L, Chehade F, Galle P. Parathyroid imaging using



386 ] PART 6 ENDOCRINOLOGY



13.



 14. 15.



16.



17.



 18.



19.



20.



simultaneous double-window recording of technetium99m-sestamibi and iodine-123. Journal of Nuclear Medicine. 1998; 39: 1100–5. Carty SE, Worsey J, Virji MA, Brown ML, Watson CG. Concise parathyroidectomy: the impact of preoperative SPECT 99mTc sestamibi scanning and intra-operative quick parathormone assay. Surgery. 1997; 122: 1107–16. Giordano A, Rubello D, Casara D. New trends in parathyroid scintigraphy. European Journal of Nuclear Medicine. 2001; 28: 1409–20. Rubello D, Saladini G, Casara D, Borsato N, Toniato A, Piotto A et al. Parathyroid imaging with pertechnetate plus perchlorate/MIBI subtraction scintigraphy. A fast and effective technique. Clinical Nuclear Medicine. 2000; 25: 527–31. Tsukamoto E, Russell CF, Ferguson WR, Laird JD. The role of preoperative thallium-technetium subtraction scintigraphy in the surgical management of patients with solitary parathyroid adenoma. Clinical Radiology. 1995; 50: 677–80. Borley NR, Collins RE, O’Doherty M, Coakley A. Technetium-99m sestamibi parathyroid localisation is accurate enough for scan-directed unilateral neck exploration. British Journal of Surgery. 1996; 83: 989–91. Denham DW, Norman J. Cost-effectiveness of preoperative sestamibi scan for primary hyperparathyroidism is dependent solely upon the surgeon’s choice of operative procedure. Journal of the American College of Surgeons. 1998; 186: 293–305. Gallowitsch HJ, Mikosch P, Kresnik E, Gomez I, Lind P. Technetium-99m- tetrofosmin parathyroid imaging: results with double-phase study and SPECT in primary and secondary hyperparathyroidism. Investigative Radiology. 1997; 32: 459–65. Chen CC, Holder LE, Scovill WA, Tehan AM, Gann DS. Comparison of parathyroid imaging with technetium99m-pertechnetate/sestamibi subtraction, double-phase technetium-99m-sestamibi and technetium-99m-



21.



22.



23.



24.



25.



26.



27.



 28.



29.



sestamibi SPECT. Journal of Nuclear Medicine. 1997; 38: 834–9. Mitchell BK, Merrell RC, Kinder BK. Localization studies in patients with hyperparathyroidism. Surgical Clinics of North America. 1995; 75: 483–98. Norman JG, Jaffray CE, Chheda H. The false-positive parathyroid sestamibi: A real or percieved problem and a case for radioguided parathyroidectomy. Annals of Surgery. 2000; 231: 31–7. Irvin 3rd. GL, Molinari AS, Figueroa C, Carneiro DM. Improved success rate in reoperative parathyroidectomy with intraoperative PTH assay. Annals of Surgery. 1999; 229: 874–8. Goldstein RE, Blevins L, Delbeke D, Martin WH. Effect of minimally invasive radioguided parathyroidectomy on efficacy, length of stay, and costs in the management of primary hyperparathyroidism. Annals of Surgery. 2000; 231: 732–42. Dackiw APB, Sussman JJ, Fritsche HA, Delpassand ES, Stanford P, Hoff A et al. Relative contributions of technetium Tc 99m sestamibi scintigraphy, intraoperative gamma probe detection, and rapid parathyroid hormone assay to the surgical management of hyperparathyroidism. Archives of Surgery. 2000; 135: 550–7. Garner SC, Leight Jr. GS. Initial experience with intraoperative PTH determinations in the surgical management of 130 cases of primary hyperparathyroidism. Surgery. 1999; 126: 1132–7. Vogel LM, Lucas R, Czako P. Unilateral parathyroid exploration. American Surgeon. 1998; 64: 693–6. Reeve TS, Babidge WJ, Parkyn RF, Edis AJ, Delbridge LW, Devitt et al. Minimally invasive surgery for primary hyperparathyroidism: systematic review. Archives of Surgery. 2000; 135: 481–7. McIntyre RC, Ridgway EC. Sestamibi: opening a new era of parathyroid surgical procedures. Endocrine Practice. 1998; 4: 241–4.



31 Parathyroid dysfunction: medical and surgical therapy E DINAKARA BABU, BILL FLEMING AND JA LYNN



Normal parathyroid glands Genetics of hyperparathyroidism Primary hyperparathyroidism Secondary and tertiary HPT Key points



387 388 389 395 396



Best clinical practice Deficiencies in current knowledge and areas for future research References



396 396 396



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words parathyroid glands, hyperparathyroidism, hypercalcaemia and parathyroid carcinoma. Further relevant articles were obtained by manual screening of the reference lists of selected papers.



NORMAL PARATHYROID GLANDS Anatomy The parathyroid glands are best identified by their relationship to the plane of the recurrent laryngeal nerve: superior glands posterior and inferior glands anterior to this plane. There are four glands in 87 percent of patients, five in 6 percent and six in 0.5 percent, but up to ten have been reported in one individual, which may in fact represent chief-cell hyperplasia. In less than 1 percent of cases only three glands are identified.1, 2 There is also positional symmetry to the upper parathyroid glands in 80 percent of cases and to the inferior glands in 70 percent of cases.



Embryology Parathyroid glands are formed during the fifth week of intrauterine life. Superior glands develop from the fourth



pharyngeal pouch and are found at the junction of the upper and middle third of the posterior surface of the thyroid gland, near the cricothyroid junction in 80 percent of cases. The importance of this position is its proximity to the point where the recurrent laryngeal nerve enters the larynx. In 15 percent they are found on the posterior surface of the thyroid, while 3 percent are to be found in retro-oesophageal, retrolaryngeal or retropharyngeal positions and, finally, 1 percent lie above the thyroid upper pole. Inferior glands develop from the third pharyngeal pouch and are closely related to the developing thymus (thyrothymic ligament). The inferior glands take a longer and more variable migration through the neck, with a more widespread final resting position and ectopic glands potentially anywhere from the carotid bifurcation down to the pericardium. The inferior glands are at the lower pole of the thyroid in 50 percent of cases, within the thymus or the thyrothymic ligament in 25 percent, while approximately 10 percent are more lateral and 10 percent lie on or next to the trachea.



388 ] PART 6 ENDOCRINOLOGY



Physiology Parathyroid glands are composed of chief and oxyphil cells, and secrete parathyroid hormone (PTH), which plays a major role in calcium metabolism either by directly acting on the bone or indirectly through vitamin D. Calcium ion homeostasis is maintained by a complex of hormones and different organ systems, controlled by PTH, vitamin D and calcitonin.



PTH



PTH is synthesized and secreted by the chief cells in response to a fall in extracellular calcium concentrations. PTH is secreted in an episodic manner as pre-pro and pro-PTH, before being cleaved into fragments within the chief cells and in the Kuppfer cells of the liver. The 1–34 N-terminal fragment is the biologically active form of PTH, which has a half-life of PTH of around five minutes. Parathyroid and C–cells utilize a calcium sensing receptor (CaR), which senses the level of extracellular calcium to determine the secretion of PTH and calcitonin (see below). PTH acts on target tissues by binding to PTH receptors on the cell surface. Three PTH receptors have been described to date with PTH-1 being the most important mediator of PTH and parathyroid hormonerelated peptide (PTHrP) actions on calcium homeostasis.3 In bone, PTH induces bone resorption by indirectly acting on osteoclast cells to increase their number and activity. It acts via the PTH-1 receptor primarily on osteoblasts, which release a cytokine that acts on the osteoclasts. In the kidney PTH acts via the PTH-1 receptor to decrease calcium excretion by increasing the reabsorption of calcium in the distal convoluted tubule. It also increases phosphate excretion and stimulates production of calcitriol, the active form of 1,25 dihydroxycholecalciferol (vitamin D) in the proximal convoluted tubule. This PTH-induced increase in calcitriol has an effect on the intestine which results in enhanced gut absorption of calcium. PTHrP



PTHrP closely resembles PTH and is an important mediator regulating the hypercalcaemia of malignancy. It acts via all three PTH receptors, and the PTH-2 receptor will only bind PTHrP. The normal physiological role for PTHrP appears to be in the foetus, where it facilitates placental mineral transport. VITAMIN D



1,25-dihydroxy-vitamin D is a fat-soluble compound, synthesized in the skin from its precursor 7-dehydrocholestrol by the action of sunlight, and is also derived



from plant ergosterol in the diet. It is transported to the liver where it is converted to the major circulating metabolite 25-hydroxycholecalciferol. A further hydroxylation to the active form 1,25-dihydroxycholecalciferol takes place in the kidney, under the control of PTH and calcitonin. Vitamin D exerts its action through specific receptors present in the cell nucleus in bone and gut.



CALCITONIN



Calcitonin is a hormone secreted by the parafollicular cells of the thyroid, with a half-life of ten minutes. Calcitonin reduces the serum calcium level by its effect on specific cell surface receptors on osteoclasts and on the proximal convoluted tubules. The exact physiological role of calcitonin remains uncertain however, as patients with high circulating calcitonin levels in medullary carcinoma of the thyroid do not have any disorder of calcium homeostasis.



CALCITONIN GENE–RELATED PEPTIDE



Calcitonin gene-related peptide (CGRP) is the product of the same gene that directs the synthesis of calcitonin, and is found in thyroid, pituitary, pancreas and adrenal medulla. Its exact role is unknown.



OTHER HORMONES



Other hormones that affect calcium metabolism are only important in times of greater calcium need, such as during growth, pregnancy and lactation. These are growth hormone, oestrogens, corticosteroids, thyroxine and insulin. [***/**]



GENETICS OF HYPERPARATHYROIDISM Most of the parathyroid disorders are due to genetic abnormalities and have a familial tendency. These genetic disorders are due to a defect in either PTH or the PTH receptors producing either hyper or hypoparathyroidism.4 Parathyroid tumours are either sporadic or inherited as a part of a multiple endocrine neoplasia (MEN) syndrome. They are produced by overexpression of PRAD-1 and RET gene or due to inactivation of the MEN 1 gene, retinoblastoma (RB) gene, and the gene on chromosome 1b.



PRAD-1 Parathyroid adenoma-1, or PRAD-1, is a gene on chromosome 11 that controls the cell cycle. Overexpression of PRAD-1, resulting in overexpression of the regulatory protein cyclin D1, is responsible for at least 50 percent of sporadic parathyroid adenomas.



Chapter 31 Parathyroid dysfunction: medical and surgical therapy



MENIN The MEN-1, or MENIN, gene is a tumour suppressor gene located on chromosome 11q13. Somatic mutations of this gene cause MEN-1, but are found in approximately 25 percent of sporadic parathyroid adenomas. MEN-1 is characterized by parathyroid, pancreatic islet cell and pituitary tumours, with 95 percent having hyperparathyroidism (HPT) due to parathyroid tumours. There are over 250 different mutations of the MENIN gene found in the kindred of MEN-1 patients, and 90–95 percent of MEN-1 patients have demonstrable mutations.



RET protooncogene The cRET, or MEN-2, gene is a protooncogene located on chromosome 10. Mutations of this gene result in MEN-2, which has two subtypes. MEN-2A is the more common type, characterized by medullary carcinoma of the thyroid, phaeochromocytoma and parathyroid tumours, but the latter are only present in 20 percent, reflecting the lower virulence of the multiglandular disease. MEN-2B is characterized by medullary carcinoma of the thyroid and phaeochromocytoma, marfanoid features and mucosal neuromas, but without parathyroid tumours. The cRET genetic mutation is found in 90 percent of patients, so that mutational analysis of the cRET gene is useful in the screening and management of first-degree relatives.



RB gene The RB gene is a tumour suppressor gene present on chromosome 13q14. Allelic deletion of the RB gene is found in parathyroid tumours, especially in parathyroid carcinoma, so that with the presence of abnormal RB protein staining may play an important role in differentiating benign and malignant parathyroid disease.



Gene on chromosome 1b This is a protooncogene located on the chromosome 1q21–q31, and mutation of this gene is found in 40 percent of sporadic parathyroid adenomas. In the hereditary form there is an association with a family history of mandibular and maxillary tumours, Wilm’s tumour and adult nephroblastoma, as well as an increased risk of parathyroid cancer.



Gene on chromosome Xp11 This is a tumour suppressor gene, which may play an important role in the refractory secondary HPT seen in



] 389



chronic renal failure. Hyperplasia of parathyroid glands in these patients is due to monoclonal mechanisms and produces nodular hyperplasia.



Gene for CaR The calcium sensing receptor has a key role in the calcium-mediated regulation of PTH secretion, parathyroid cellular proliferation and probably the mRNA levels of pre-proPTH.3 Inactivating mutations of this gene produce two types of hypercalcaemic disorders that reset the ‘set-point’ of calcium-mediated PTH secretion. Heterozygous mutations result in familial hypocalciuric hypercalcaemia (FHH), with a mild, generally asymptomatic hypercalcaemia. Individuals with homozygous mutations have parathyroid glands with much more marked calcium resistance, presenting with neonatal severe hyperparathyroidism (NSHPT) and profound hypercalcaemia and HPT.



PTH-receptor gene This gene is located on chromosome 3 and activated mutation of this gene results in Jansen’s disease. It is an autosomal dominant disorder characterized by dwarfism, hypercalcaemia and hypophosphataemia, with normal or undetectable serum PTH. William’s syndrome is an autosomal dominant disorder, with a genetic defect located on chromosome 7, characterized by infantile hypercalcaemia, supravalvular aortic stenosis, psychomotor retardation and elfin facies.



PRIMARY HYPERPARATHYROIDISM Hyperparathyroidism can be caused by primary parathyroid gland dysfunction or by a variety of secondary causes that lower calcium levels or raise phosphate levels. Hypoparathyroidism is usually a complication of thyroid surgery, from removal or damage to parathyroid glands. Its presentation and management are discussed in Chapter 28, The thyroid: nonmalignant disease. Normally, PTH secretion is exceedingly sensitive to changes in serum calcium, with PTH changes occurring within seconds, sequentially and graded.3 The immediate effect of lowered calcium is the release of preformed PTH from secretory vesicles, followed by a decrease in the degradation of PTH in the parathyroid cell, so that more intact hormone is released. If the stimulus persists then there is an upregulation of the expression of pre-proPTH mRNA. If the stimulus lasts for an extensive period of time, then parathyroid chief cells begin to proliferate within days (as in chronic renal failure). In primary hyperparathyroidism (pHPT) this sensitive feedback mechanism is lost.



390 ] PART 6 ENDOCRINOLOGY



Diagnosis Diagnosis of pHPT is made when there is a high serum calcium, with an abnormally high or nonsuppressed PTH level. Occasionally, it may be difficult to identify milder forms of pHPT, as PTH is not released at an even rate, so repeated measurements of serum PTH and calcium may be required to make a biochemical diagnosis. pHPT occurs in two forms, as nonfamilial or sporadic in 95 percent, and as familial or hereditary in 5 percent. In the hereditary form, pHPT may be a part of one of the MEN syndromes.



Incidence The exact incidence of pHPT is unknown, but is increasing with the use of multichannel autoanalysers picking up hypercalcaemia more often. Overall, the incidence in men is around 0.3 percent, and in women 1–3 percent, but in autopsy examination this sex discrepancy disappears, perhaps reflecting the more likely diagnosis in women being screened for osteoporosis. In women, the incidence rises with age and could be as high as 1 in 500, with a rate of 2.6 percent in women aged between 55 and 75 years.5



Pathology The aetiology of primary HPT is unknown, but radiotherapy to the head and neck during childhood can predispose to either adenoma or hyperplasia.6 The pathology of pHPT is adenoma in 80 percent, hyperplasia in 15–20 percent and carcinoma in 1 percent of patients.



ADENOMA



Parathyroid adenomas are usually solitary, with the other glands showing evidence of suppressed activity. Double or multiple adenomas may occur occasionally, but more likely are misdiagnosed hyperplasia. Parathyroid tumour more often affects the lower glands than the upper glands, and intrathymic and intrathyroidal adenomas are not uncommon. The adenomas are generally ovoid, soft, reddish-brown tumours, usually a little darker than the normal glands. Microscopically, they are characterized by varying compositions of chief, clear and oxyphil cells, with chief-cell adenoma by far the most common. Parathyroid adenomas generally exhibit a peripheral rim of condensed normal parathyroid tissue, separated by a slender capsule.



HYPERPLASIA



Parathyroid hyperplasia affects more than one gland, as a result of an increase in parathyroid parenchyma. The



enlarged glands are rounded or grossly lobulated, and grey-brown in colour, but the process may not involve all glands or involve them uniformly. Microscopically, primary chief-cell or nodular hyperplasia is the predominant finding, but a multiplicity of cell types can be found. Nodular chief-cell hyperplasia is the characteristic pathology of the parathyroids in MEN-1 syndrome.



CARCINOMA



Parathyroid carcinomas are nearly always functioning tumours and account for approximately 1 percent of cases of pHPT, usually with a very high serum calcium and PTH. These tumours are often large and clinically palpable, usually greater than 2 cm in size, and appear firm, greyish-white and adherent to adjacent structures. Microscopically, carcinoma shows features of malignancy, such as invasion of the capsule, vascular invasion, marked fibrosis with nodule formation, focal areas of necrosis, cellular atypia and a high mitotic rate. Chief and oxyphil cells are predominantly seen. Flow cytometry studies show that an aberrant ploidy pattern is associated with a bad prognosis. It is extremely difficult for the pathologist to make the diagnosis at frozen and paraffin section. Parathyroid carcinoma spreads by local invasion and lymphatic permeation, and distant metastasis occurs to lung, liver and bones. Treatment involves en bloc resection of parathyroid and adjacent structures, including ipsilateral thyroid lobectomy and node dissection. [**]



Clinical features In asymptomatic, ambulatory patients, over 80 percent of patients with hypercalcaemia will have pHPT, but this falls to less than a third in symptomatic and/or hospitalized patients. More often in this hospital group, severe symptomatic hypercalcaemia is caused by malignant disease. The majority of patients with pHPT have no symptoms or only mild symptoms, and are picked up on biochemical screening for another reason. Symptomatic presentation due to hypercalcaemia with the classical rhyme of bones, stones, abdominal groans and psychic moans is extremely rare.



MUSCULOSKELETAL



The most common symptoms are muscular weakness, bone and joint pain; fatigue and lethargy are very common. Patients have reduced bone mineral density, but the severe form of bone disease known as osteitis fibrosa cystica is extremely rare in developed countries. Parathyroidectomy leads to incomplete improvement of bone density, but does improve muscle function and relieves the feeling of tiredness.



Chapter 31 Parathyroid dysfunction: medical and surgical therapy



RENAL



The usual symptoms are renal and ureteric stones; renal failure and nephrocalcinosis are extremely rare. Parathyroidectomy lowers the frequency of recurrent stones.



] 391



Complementary tests that may be helpful to support the diagnosis are plasma phosphate, creatinine and magnesium.



LOCALIZATION STUDIES GASTROINTESTINAL



Abdominal pain can arise from chronic constipation, but the linkage between peptic ulceration, pancreatitis and pHPT is controversial. Improvement of symptoms after parathyroidectomy is unpredictable. PSYCHOLOGICAL



Prospective analyses have suggested that even apparently asymptomatic patients with pHPT may have psychiatric symptoms, which can be reversed by active treatment.7 Anxiety and depression, lassitude, dementia and loss of concentration or memory are the common symptoms. Parathyroidectomy reverses the dementia-like symptoms in the elderly population, and significant improvements can be made in other symptoms.8 METABOLIC ‘SYNDROME’



pHPT is associated with a metabolic syndrome which carries an increased risk of premature death, even in mild hypercalcaemia, and increases with the extent of hypercalcaemia. Diabetes mellitus and an impaired glucose tolerance can be seen due to peripheral insulin resistance. Blood lipid levels show a decrease in high-density lipoprotein (HDL) and low-density lipoprotein (LDL) levels and a rise in very low-density lipoprotein (VLDL) levels. Cardiovascular disorders, such as hypertension, left ventricular hypertrophy and valvular calcification, are known to be associated with pHPT. Many of these effects can be reversed or improved after parathyroidectomy. [****/**]



Preoperative localization of parathyroid tumours can locate abnormal glands in 80–85 percent of cases, which allows minimally invasive techniques to be used, with the potential for operation under local anaesthesia. Typically, localization is achieved by a combination of highresolution ultrasound and 99mTc-sestamibi scan. Other investigations, such as selective angiography and venous sampling, CT and MRI will be needed if a second or subsequent operation is being planned. Ultrasound Ultrasound is cheap, noninvasive and can be used intraoperatively. Parathyroid adenoma is typically sonolucent (Figure 31.1) and colour Doppler will demonstrate the presence of an arterial signal at the vascular pole.9 Ultrasound can only be used to locate adenomas in the neck, with a sensitivity of around 85 percent in the unexplored neck, dropping to 40 percent in patients who have had a previous exploration. Sestamibi scintiscan Parathyroid adenoma concentrates 99mTc-labelled sestamibi because of the higher number of metabolically active mitochondria (Figure 31.2). It can be used to locate glands in either the neck or mediastinum with or without the addition of single photon emission computerized tomography (SPECT). Scanning can detect 87 percent of solitary adenomas, 55 percent of abnormal glands in patients with multiglandular disease and 75 percent of persistent or recurrent lesions in the previously explored neck.10 Diffuse hyperplasia will often lead to a negative scan as the parathyroid glands are less mitochondria-rich.



Investigations for HPT pHPT is responsible for over 80 percent of cases of asymptomatic hypercalcaemia in the ambulatory population, but accounts for only 20 percent in the symptomatic or hospitalized population with hypercalcaemia.



BIOCHEMICAL



The common biochemical tests performed are plasma calcium, albumin, vitamin D and intact PTH. Measurements may have to be repeated due to the episodic secretion of PTH, which can result in normocalcaemic periods. A 24 hour urine collection for calcium will help to exclude FHH and lithium as causes of hypercalcaemia.



Figure 31.1



Ultrasound of a typical parathyroid adenoma.



392 ] PART 6 ENDOCRINOLOGY



Figure 31.3 Selective arteriogram showing a mediastinal parathyroid adenoma.



usually identify the missed gland, with a sensitivity of 90–95 percent, especially when combined with venous sampling for parathyroid hormone assay (Figure 31.3). Almost all enlarged glands are hypervascular, with a distinctive angiographic ‘blush’, resulting in a low falsepositive rate. Computed tomography and magnetic resonance imaging Thin-slice computed tomography (CT) will visualize ectopic parathyroids in the anterior mediastinum, where ultrasound cannot see. CT and magnetic resonance imaging (MRI) have a sensitivity of 75–85 percent in an unexplored neck, which decreases to 45 percent for CT after previous surgery, but remains more accurate (50–75 percent) with MRI. Positron emission tomography scanning



Figure 31.2 adenoma.



Sestamibi scan showing right inferior parathyroid



Selective venous sampling and angiography In the patient who has had a previous failed exploration, perhaps due to a mediastinal adenoma, more invasive investigative measures may be needed in addition to sestamibi scan and ultrasound. Selective angiography will



11C-methionine positron emission tomography (PET) is clinically useful in highly preselected patients with recurrent pHPT, as well as in secondary and tertiary HPT, with a sensitivity of 83 percent, a specificity of 100 percent and an accuracy of 88 percent in successfully locating parathyroid adenomas.11, 12



Indications for surgery SYMPTOMATIC



All patients with symptomatic pHPT should have parathyroidectomy. The advantages of the operation



Chapter 31 Parathyroid dysfunction: medical and surgical therapy



are:  relief of muscle pain, lethargy and weakness;  improvement in bone pain and the bone disease of hypercalcaemia;  reduction in incidence of renal stones;  improvement in psychiatric symptoms;  reversal of many of the symptoms of ‘metabolic syndrome’. ASYMPTOMATIC



Opinions differ as to the correct management of patients without symptoms; however, the NIH Consensus Development Conference held in 1990 provided some guidelines.13 Patients with levels of calcium above 3.00 mmol/L should have parathyroidectomy with or without symptoms. Patients with mild HPT (calcium 2.85– 3.00 mmol/L) should have surgery if any of the following criteria are fulfilled:  creatinine clearance reduced by at least 30 percent;  urine calcium excretion above 400 mg/24 hours;  bone mass two standard deviations below that of age and sex-matched persons;  age below 50 years;  recent history of kidney stones;  apparent neuromuscular or psychiatric symptoms;  medical surveillance undesirable or unsuitable. However, despite these guidelines, the authors feel that parathyroidectomy should be recommended for all patients with hypercalcaemia and pHPT, for the following reasons:  the difficulty of determining symptoms preoperatively (especially psychiatric symptoms), that may be alleviated by surgery;  approximately 50 percent of conservatively-treated asymptomatic patients become symptomatic;  6 percent develop renal stones, 5 percent develop bone disease and 5 percent develop psychological problems;14  the present evidence shows untreated asymptomatic patients have increased mortality due to cardiovascular disease;15  minimally invasive surgery under local anaesthesia is now possible;  low rate of complications of parathyroidectomy in experienced hands;  high rate of postoperative cure in experienced hands.



Management of pHPT MEDICAL THERAPY OF HYPERCALCAEMIA



The treatment of pHPT is predominantly surgical. Prior to surgery, the medical treatment of hypercalcaemia will depend on its severity. The main aim is to reduce



] 393



hypercalcaemia as rapidly as possible and then to diagnose the underlying cause. Mild (o3.0 mmol/L) to moderate (3–3.5 mmol/L) hypercalcaemia can generally be managed on an outpatient basis, whereas severe (>3.5 mmol/L) hypercalcaemia requires more vigorous inpatient treatment. Acute treatment of hypercalcaemia involves rehydration with i.v. normal saline, diuretics and the use of bisphosphanates intravenously, such as pamidronate. Longer-term treatment requires maintenance of adequate hydration, a low calcium diet, oral bisphosphanates and parathyroidectomy if pHPT is present.16 An alternative for the future may be the use of a calcimimetic drug to interact with the calcium receptor on the parathyroid cell, decreasing serum PTH and ionized calcium concentrations.17 [***/**]



SURGICAL STRATEGIES



Conventional parathyroidectomy Bilateral neck exploration in the conventional operation allows access to all four glands and the ability to diagnose hyperplasia, double adenoma and microadenoma that might be missed by minimally invasive techniques.18 Frozen section biopsy is useful in confirming parathyroid tissue, but differentiating adenoma from hyperplasia is only 60 percent reliable at frozen section and is better achieved by the assessment of an experienced surgeon.19 While exploring the parathyroid glands, it is essential to understand the following principles.  The normal parathyroid gland tends to be approximately 5 mm in size, and can be seen to move independently within its fascial layer.  Parathyroids tend to be symmetrically placed on each side.  Most parathyroid glands are situated in the normal anatomical position, medial to the carotid arteries;  superior glands are located behind the plane of the recurrent laryngeal nerve usually above the inferior thyroid pedicle (Figure 31.4).  Inferior glands are found in front of the plane of the recurrent nerve, usually below the inferior thyroid pedicle, but their position is more variable. Missing superior gland Eighty percent of upper glands lie in a 2 cm radius 1 cm above the intersection of the inferior thyroid pedicle and the recurrent nerve (Figure 31.5). If not found there, it may help to divide the superior part of the sternothyroid muscle and/or take down the superior pole of the thyroid to improve exposure. It may be that a large adenoma has descended into the tracheo-oesophageal groove, so a search behind the oesophagus, higher in the retropharyngeal space and down into the posterior mediastinum should be made.



394 ] PART 6 ENDOCRINOLOGY



Figure 31.6



Parathyroid adenoma in thymus.



Indication for mediastinal exploration



Figure 31.4 Normal right upper parathyroid gland (at tip of metal forceps) in front of the plane of the recurrent laryngeal nerve (blue arrow).



If four normal glands have been found, the neck should be closed and a reassessment made of the original diagnosis. Further dissection at this point places the blood supply of the remaining parathyroids at risk, as well as increasing the potential for recurrent nerve damage. Post-operatively, the patient should be reinvestigated, including selective venous sampling. Less than 1 percent of patients need a mediastinotomy as the majority of mediastinal lesions are actually in the thymus, which can be safely removed from the neck. The advantages of staging mediastinal surgery as a second procedure are:  dissection in the neck may disrupt an adenoma’s blood supply, turning an apparently negative exploration into a successful result;  the pathologist may identify parathyroid tissue, for example in the thymus, which was missed at operation;  there is an opportunity to reconfirm or change the diagnosis;  angiographic ablation can be performed for inaccessible mediastinal glands. Scan-directed minimally invasive open parathyroidectomy



Figure 31.5 Right upper parathyroid adenoma in tracheooesophageal groove (blue arrow).



Missing inferior gland Most commonly, the inferior gland is found within a 1 cm radius of the lower pole of the thyroid, but its position is more variable than the upper gland. If the inferior gland is missing, a thymectomy should be performed which may reveal the adenoma lying within the gland (Figure 31.6). The surgeon should also consider an undescended inferior parathyroid, particularly if the ipsilateral thymus is absent from its normal position. The carotid sheath should be opened and explored from the hyoid to the thoracic inlet, and consideration also given to a possible intrathyroidal adenoma.



The technique of scan-directed minimally invasive open parathyroidectomy (MIP) has rapidly gained favour with the advent of accurate preoperative localization techniques, which allow a targeted approach to the adenoma, minimizing dissection and potential complications.20 The parathyroid adenoma is localized by sestamibi scan and ultrasound, and only the area identified is explored, either under general or local anaesthesia. Access is either through a small lateral transverse incision, or one made along the medial border of the sternocleidomastoid muscle. Ideally, the success of MIP should be confirmed by measurement of intraoperative PTH.21 If scans are concordant, then MIP should cure between 92 and 95 percent of patients. [***/**] Radio-guided parathyroidectomy Radio-guided parathyroidectomy, popularized by Norman and Chheda,22 relies on uptake into the adenoma of 99m Tc-labelled sestamibi given two hours before



Chapter 31 Parathyroid dysfunction: medical and surgical therapy



operation, then a hand-held gamma probe is used intraoperatively. Focused dissection is performed in the area of highest activity and the radioactive gland is removed. The main disadvantages are cost and the organizational difficulties of arranging and timing of the scan before operation. [**]



] 395



Minimally invasive video-assisted (MIVAP) and endoscopic parathyroidectomy are more recent techniques that result in a smaller scar in the neck and a quicker recovery, but at the expense of longer operating times and higher costs. Endoscopic parathyroidectomy uses gas insufflation and several 3 mm ports in the neck. Miccoli et al.23, 24 modified this operation, substituting a gasless, videoassisted approach, in combination with intraoperative PTH measurement. Ikeda et al.25 have gone even further to remove the scars from the neck by approaching the parathyroid tumour via an axillary or a submammary approach. [****/***]



Patients with persistent or recurrent HPT need to be reinvestigated more vigorously and the diagnosis reconfirmed prior to any second operation. Mild asymptomatic or normocalcaemic HPT may not need reoperation at all, but symptomatic patients with high serum calcium and renal or bony disease will need a second exploration. Reoperations may be relatively uncomplicated, but generally are technically difficult, with a high complication rate and low success rate in inexperienced hands. Injury to the recurrent laryngeal nerve and permanent hypoparathyroidism occur in 8 percent of patients. The technique involves a ‘back door’ approach to a focused dissection of the abnormal area identified at reinvestigation. The sites of missed adenoma are in the normal positions in 40 percent, posterior superior mediastinum in 30 percent, intrathymic in 15 percent, posterior midline in 5 percent, mediastinal in 5 percent, intrathyroidal in 2 percent and other areas 3 percent. Mediastinal glands are usually found around the aortic arch, superior vena cava and aorto–pulmonary window.28



Results of parathyroidectomy



SECONDARY AND TERTIARY HPT



CURE RATE



Secondary HPT is caused by hypocalcaemia or phosphate retention, typically seen in chronic renal failure. Less common causes are vitamin D deficiency and lithium ingestion. Chronic renal failure causes hyperphosphataemia and hypocalcaemia stimulating diffuse hyperplasia of all four glands and a rise in serum PTH. Prolonged stimulation of parathyroid glands results in nodular hyperplasia from monoclonal cell proliferation and a decrease in vitamin D receptor and CaR density. This results in resistance to medical treatment.29 In tertiary HPT, one of the hyperplastic nodules becomes autonomous due to alteration in PTH-gene transcription, resulting in hypercalcaemia and raised PTH similar to pHPT.



Video-assisted and endoscopic parathyroidectomy



Many patients will require post-operative calcium and vitamin D supplementation until the suppressed glands recover their function. At the first operation, more than 95 percent of pHPT patients will be cured, with 83 percent of those requiring reoperation. The outcome is significantly dependent on how many parathyroidectomies are carried out by a given surgeon per year, with higher failure rates for those performing less than 15 cases per year.26



COMPLICATIONS



The post-operative complication rate is under 1 percent after primary operations and less than 5 percent after reoperative surgery, with a mortality rate well below 1 percent. Approximately 8 percent of primary operations for pHPT result in recurrent or persistent disease, this is higher in multigland disease and particularly the MEN syndromes, where it may be as high as 40 percent. PERSISTENT OR RECURRENT HPT



Persistent HPT is defined as the presence of hypercalcaemia in the immediate postoperative period or within six months. Recurrent HPT is defined as the return of hypercalcaemia after six months of post-operative normocalcaemia or hypocalcaemia, which is caused by inadequate resection of diseased glands, such as regrowth of residual, unresected, hyperplastic gland after subtotal parathyroidectomy, incomplete resection of an adenoma, rupture and spillage of parathyroid cells (parathymomatosis) or recurrence of parathyroid carcinoma.27



Medical treatment The primary treatment of secondary and tertiary HPT is predominantly medical, through management of chronic renal failure by medical manipulation of phosphate levels and peritoneal or haemodialysis. Renal transplantation usually cures secondary HPT.



Parathyroidectomy Twenty-eight percent of renal patients on dialysis have secondary HPT, and 10 percent of patients who have had successful renal transplantation develop either secondary or tertiary HPT. These patients will generally need surgical parathyroidectomy.



396 ] PART 6 ENDOCRINOLOGY There are a number of different surgical strategies available: total parathyroidectomy, with or without autotransplantation, and subtotal parathyroidectomy. Some endocrine surgeons prefer subtotal parathyroidectomy, removing three-and-a-half glands, but it is extremely difficult to judge the correct amount of parathyroid tissue to achieve normocalcaemia. In addition, graft-dependent recurrence occurs in up to 75 percent of patients on dialysis who have had autotransplantation, and 10 percent develop graft failure with hypocalcaemia.30 For this reason, the authors favour total parathyroidectomy and transcervical thymectomy. [****/***/**]



KEY POINTS  Primary hyperparathyroidism and malignancy are responsible for 90 percent of hypercalcaemia.  Eighty percent of pHPT is due to single gland disease.  Sestamibi and ultrasound localization allows unilateral exploration of the neck.  Intraoperative quick PTH assay is an essential tool.  Management of secondary HPT is difficult.  Parathyroid carcinoma is rare; when suspected will require en bloc excision with ipsilateral thyroidectomy.



Best clinical practice [ The approach to pHPT is the same as for any endocrine condition, following five basic principles: confirmation of endocrine diagnosis; localization of the tumour or tumours; ensuring the patient is biochemically safe for surgery; deciding necessity for operation; choosing a suitable procedure. [ Minimal diagnostic investigations: – biochemical tests: calcium, PTH, vitamin D and 24 hour urine calcium; – localization studies: ultrasound, sestamibi scan; – before reoperation: CT or MRI, PET and selective venous sampling and angiography. [****/ ***/**] [ Indications for parathyroidectomy: – symptomatic hypercalcaemia; – asymptomatic hypercalcaemia with: calcium >3 mmol/L; calciumo3 mmol/L with renal disease, bone disease, young age (o50 years) or ‘metabolic syndrome’. [Grade A/B] [ Plan of parathyroid exploration: – find recurrent laryngeal nerve;



– look for upper parathyroid (2 cm radius from 1 cm above intersection of RLN and ITA);



– missing upper: look tracheo-oesophageal groove, retro-oesophageal and retropharyngeal spaces; – look for lower parathyroid (inferior thyroid pole to thymus); – missing lower: perform thymectomy, look in carotid sheath; – consider thyroid lobectomy; – stop, then reassess diagnosis and consider mediastinal exploration. [****/***/**] [ Revision parathyroid surgery. Recurrent or persistent HPT requires: – reinvestigation to confirm diagnosis and localize tumour; – more invasive investigation, for example selective venous sampling; – assess need for surgery; – ‘back door’ approach between sternocleidomastoid and strap muscles; – focused exploration in identified area, with meticulous dissection. [***/**]



Deficiencies in current knowledge and areas for future research The main problem in primary hyperparathyroidism is separating multiple gland disease from single gland disease. If accurately detected, single gland disease can be easily managed by minimally invasive techniques, either open, video-assisted or endoscopically, with the potential for operation under local anaesthetic. Thus, there is a great need for an accurate scan that will separate single from multiple gland disease. Our wish list for the future would therefore be: more accurate forms of scanning; a preoperative method of separating carcinoma of the parathyroid from benign disease; effective, long-acting inhibitors of PTH secretion by either hyperplastic glands or adenomas, such as calcimimetic drugs; a more effective way of controlling preoperative hypercalcaemia without the inherent risk of postoperative hungry bone disease.



$ $ $ $



REFERENCES 1. Akerstrom G, Malmaeus J, Bergstrom R. Surgical anatomy of human parathyroid glands. Surgery. 1984; 95: 14–21.



Chapter 31 Parathyroid dysfunction: medical and surgical therapy















2. Randolph GW, Urken ML. Surgical management of primary hyperparathyroidism. In: Randolph GW (ed.). Surgery of the thyroid and parathyroid glands. Philadelphia, USA: Elsevier Science, 2003: 507–28. 3. Mithal AM, Brown EM. An overview of extracellular calcium homeostasis and the roles of the CaR in parathyroid and C-cells. In: Chattopadhyay N, Brown EM (eds). Calcium-sensing receptor. Boston, USA: Kluwer Academic Publishers, 2003: 1–27. 4. Thakker RV. Molecular genetics of parathyroid disease. Current Opinion in Endocrinology and Diabetes. 1996; 3: 521–8. 5. Lundgren E, Rastad J, Thurfjell E, Akerstrom G, Ljunghall S. Population-based screening for primary hyperparathyroidism with serum calcium and parathyroid hormone values in menopausal women. Surgery. 1997; 121: 287–94. 6. Tezelman S, Rodriguez JM, Shen W, Siperstein AE, Duh QY, Clark OH. Primary hyperparathyroidism in patients who have received radiation therapy and in patients who have not received radiation therapy. Journal of the American College of Surgeons. 1995; 180: 81–7. 7. Rastad J, Joborn C, Akerstrom G, Ljunghall S. Incidence, type and severity of psychic symptoms in patients with sporadic primary hyperparathyroidism. Journal of Endocrinological Investigation. 1992; 15: 149–56. 8. Solomon BL, Schaaf M, Smallridge RC. Psychologic symptoms before and after parathyroid surgery. American Journal of Medicine. 1994; 96: 101–6. 9. Mazzeo S, Caramella D, Marcocci C, Lonzi S, Cambi L, Miccoli P et al. Contrast-enhanced color Doppler ultrasonography in suspected parathyroid lesions. Acta Radiologica. 2000; 41: 412–6. 10. Pattou F, Huglo D, Proye C. Radionuclide scanning in parathyroid diseases. British Journal of Surgery. 1998; 85: 1605–16. 11. Otto D, Boerner AR, Hofmann M, Brunkhorst T, Meyer GJ, Petrich T et al. Pre-operative localization of hyperfunctional parathyroid tissue with 11C-methionine PET. European Journal of Nuclear Medicine and Molecular Imaging. 2004; 31: 1405–12. 12. Beggs AD, Hain SF. Localization of parathyroid adenomas using 11C-methionine positron emission tomography. Nuclear Medicine Communications. 2005; 26: 133–6. 13. Consensus Development Conference Panel: Diagnosis and management of asymptomatic primary hyperparathyroidism. Consensus development conference statement. Annals of Internal Medicine. 1991; 114: 593–7. 14. Scholz DA, Purnell DC. Asymptomatic primary hyperparathyroidism: 10 year prospective study. Mayo Clinic Proceedings. 1981; 56: 473–8. 15. Dahlberg K, Brodin LA, Juhlin-Dannfelt A, Farnebo LO. Cardiac function in primary hyperparathyroidism before and after operation: an echocardiographic study. The European Journal of Surgery. 1996; 162: 171–6.











] 397



16. Swan JW, Stevenson JC. The medical management of hypercalcaemia. In: Lynn J, Bloom SR (eds). Surgical endocrinology. Oxford: Butterworth-Heinemann, 1997: 341–50. 17. Silverberg SJ, Bone III HG, Marriott TB, Locker FG, ThysJacobs S, Dziem G et al. Short-term inhibition of parathyroid hormone secretion by a calcium-receptor agonist in patients with primary hyperparathyroidism. New England Journal of Medicine. 1997; 337: 1506–10. 18. Ogilvie JB, Clark OH. Parathyroid surgery: we still need traditional and selective approaches. Journal of Endocrinological Investigation. 2000; 28: 566–9. 19. Saxe A, Baier R, Raile R, Tesluk H, Toreson W. The role of the pathologist in the surgical treatment of hyperparathyroidism. Surgery, Gynecology and Obstetrics. 1985; 161: 101–5. 20. Carty S, Worsey M, Virji M, Brown ML, Watson CG. Concise parathyroidectomy. The impact of preoperative SPECT 99mTc sestamibi scanning and intraoperative quick parathormone assay. Surgery. 1997; 122: 1107–16. 21. Gupta VK, Yeh KA, Burke GJ, Wei JP. 99m-Technetium sestamibi localized solitary parathyroid adenoma as an indication for limited unilateral surgical exploration. American Journal of Surgery. 1998; 176: 409–12. 22. Norman J, Chheda H. Minimally invasive parathyroidectomy facilitated by intraoperative nuclear mapping. Surgery. 1997; 122: 998–1004. 23. Miccoli P, Bendinelli C, Vignali E, Mazzeo S, Cecchini GM, Pinchera A et al. Endoscopic parathyroidectomy: report of initial experience. Surgery. 1998; 124: 1077–80. 24. Miccoli P, Bendinelli C, Berti P, Vignali E, Pinchera A, Marcocci C. Video-assisted versus conventional parathyroidectomy in primary hyperparathyroidism: a prospective randomized study. Surgery. 1999; 126: 1117–22. 25. Ikeda Y, Takami H, Sasaki Y, Kan S, Niimi M. Endoscopic neck surgery by the axillary approach. Journal of the American College of Surgeons. 2000; 191: 336–40. 26. Sosa JA, Powe NR, Levine MA, Udelsman R, Zeiger MA. Profile of a clinical practice: thresholds for surgery and surgical outcomes for patients with primary hyperparathyroidism: a national survey of endocrine surgeons. Journal of Clinical Endocrinology and Metabolism. 1998; 83: 2658–65. 27. Carty SE, Norton JA. Management of patients with persistent or recurrent primary hyperparathyroidism. Journal of Clinical Endocrinology and Metabolism. 1991; 15: 716–23. 28. Gaz RD. Revision parathyroid surgery. In: Randolph GW (ed.). Surgery of the thyroid and parathyroid glands. Philadelphia, USA: Elsevier Science, 2003: 564–70. 29. Hruska KA, Teitelbaum SL. Renal osteodystrophy. New England Journal of Medicine. 1995; 333: 166–74. 30. Rothmund M, Wagner PK, Schark C. Subtotal parathyroidectomy versus total parathyroidectomy and autotransplantation in secondary hyperparathyroidism: A randomized trial. Journal of Clinical Endocrinology and Metabolism. 1991; 15: 745–50.



32 Head and neck manifestations of endocrine disease JONATHAN M MORGAN AND THOMAS McCAFFREY



Introduction Pituitary disease Thyroid disease Adrenal disease Parathyroid disease



398 399 399 401 401



Pancreatic islet cell disease Pregnancy and gonadal disease Key points References



401 402 403 403



SEARCH STRATEGY AND EVIDENCE-BASE Information for this chapter was gathered by searching Medline using the key words head and neck, endocrine disease, as well as using the specific disease names. Due to the nature of the context of this chapter, studies cited here are predominantly retrospective and/or observational, unless specifically stated in the body of the text, and therefore fall into evidence levels 3 and 4.



INTRODUCTION Although endocrine disorders usually produce symptoms and findings that are primarily systemic, specific and localized manifestations of endocrine disease may occur early or predominantly in the head and neck region. Awareness of these presentations may lead to the early diagnosis of the underlying endocrine disease. This chapter presents the symptoms and findings in the head and neck that are caused by changes in hormone levels or endocrine disease. We begin with perturbations of the hormones produced in the pituitary, then progress to the thyroid, adrenals, pancreas and gonads. We should also note that altered physiologic states, including pregnancy, and paraendocrine diseases are defined by changes in hormone levels. These changes in physiologic states are also seen in the head and neck region and are addressed in this chapter. The regulation of the endocrine system is based upon a basic principle of negative and positive feedback loops. The complex interactions of the endocrine system are beyond the scope of this text and we would refer you to an endocrinology text for a more complete description.



The feedback mechanism begins with neuroendocrine regulation. The hypothalamus produces chemical mediators (growth hormone-releasing hormone (GHRH), leuteinizing hormone-releasing hormone (LHRH), thyrotropinreleasing hormone (TRH), corticotropin-releasing hormone (CRH) and dopamine (prolactin inhibiting hormone.) These mediators travel via a portal system to the anterior pituitary and in turn regulate the release of growth hormone (GH), leuteinizing hormone (LH) and folliclestimulating hormone (FSH), thyroid-stimulating hormone (TSH), adrenocorticotropic hormone (ACTH) and prolactin. The hypothalamus also produces vasopressin and oxytocin and these are stored in the posterior pituitary. The main feedback loop continues from the hormones of the anterior pituitary to the three target endocrine glands, the gonads, adrenal cortex and the thyroid, these in turn produce an additional set of hormones that effect multiple glands and organ systems. There are many possible mechanisms that can cause a disturbance in this feedback loop. The result of these disturbances being a real or apparent increase or decrease of hormone levels, which in turn cause the signs and symptoms of endocrine disease. Hypothalamic disease



Chapter 32 Head and neck manifestations of endocrine disease



] 399



can cause hypopituitarism and all of the associated symptoms, although typically head and neck manifestations are rare in diseases of the hypothalamus.



presentation of invasive pituitary adenomas is nasal obstruction.10 These tumours are typically diagnosed by transnasal biopsy of the mass.



PITUITARY DISEASE



THYROID DISEASE



Diseases of the anterior pituitary can result in the production of many different hormones, although otolaryngologic manifestations are rare. The exception is acromegaly, or the overproduction of GH (somatotropin). The overproduction of GH prior to epiphyseal closure results in gigantism. Acromegaly is an insidious chronic debilitating disease. This is a rare disease with a prevalence of 50–70 cases per million and an incidence of 3–4 per million per year. The disease presents most frequently in middle age and has no predilection for men or women. Facial features are well documented. Table 32.1 documents all of the head and neck manifestations. The pathophysiology of increased GH is the result of pituitary adenomas (in almost all cases). Typically, the tumour is found laterally in the sella, where somatotrophs are abundant. However, rare cases have been reported of tumours along the migration line of Rathke’s pouch, including the sphenoid sinus and the parapharyngeal space.7 Regional symptoms secondary to either mass effect of direct extension of pituitary adenomas most commonly present with blurred vision and visual field defects. (Figure 32.1). However, several pathways of extension of invasive adenomas have been defined:8



The thyroid is also under the control of the hypothalamic-pituitary axis. Hypothyroidism is caused by many processes including the following: thyroid agenesis, surgical removal (thyroidectomy and laryngectomy), therapeutic irradiation (I131 or external radiation), autoimmune (Hashimoto’s) thyroiditis, replacement by cancer or other disease, post-thyroiditis (acute or subacute), inhibition of thyroid hormone synthesis, iodine deficiency, excess iodine in susceptible patients, antithyroid drugs, inherited enzyme defects and post-partum. The incidence of hypothyroidism increases with age, and affects women four- to six-times as frequently as men. It occurs at all ages but gradually rises with age and peaks at 40–60 years. Up to 5 percent of persons older than age 65 have clinically overt hypothyroidism, whereas 15 percent have subclinical hypothyroidism marked only by serum TSH elevations. General symptoms include weakness, dry and coarse skin, myxedema and pallor. The hair of these



 pharyngeal extension via the floor of the sella and sphenoid sinus;  hypothalamic extension via the floor of the third ventricle;  temporal extension;  invasion of the cavernous sinus and associated cranial nerves;  posterior subtentorial extension involving the third, fourth and fifth nerves;  frontal extension. Extension of the tumour into the middle ear and external auditory canal is a very rare presentation.9 Another rare Table 32.1



Figure 32.1 MR scan showing a pituitary adenoma impinging on the optic chiasm and extending into the sphenoid sinus.



Signs and symptoms of acromegaly manifested in the head and neck.



General Headaches Visual field changes Prognathism1 Obstructive sleep apnoea2 Coarse facial features Oily skin Soft tissue and bony enlargement



Oral



Nasal



Laryngeal



Enlarged tongue Malocclusion Mandibular protrusion



Sinus congestion Rhinorrhoea Enlarged paranasal sinuses3



Hoarseness Decreases vocal pitch Asymmetric enlargement of epiglottis, aryepiglottic folds, false and true vocal cords Cricoarytenoid arthritis4 Vocal cord paresis/paralysis5 Airway obstruction6



400 ] PART 6 ENDOCRINOLOGY patients may be dry and fine, and they may suffer from alopecia. A classic symptom of hypothyroidism is the loss of the outer third of the eyebrows, also known as Queen Anne’s eyebrows. Vocal manifestations include hoarseness and a generalized slow speech rate. These patients may exhibit a low pitch and decreased vocal clarity. It was first postulated that this was due to central oedema of the vagal nuclei. Ritter noted that there was retention of proteinaceous fluid, high in mucopolysaccharides within the mucosa of the larynx in rats rendered hypothyroid by treatment with either propothiouracil, or radioactive I131.11 Rhinitis is a common complaint of patients with hypothyroidism. There is a correlation between hypothyroidism and rhinitis, however, no clear definition of the pathophysiology has been elucidated. Proetz12, 13 assessed 130 hypothyroid patients and found nasal congestion, rhinorrhea, pale wet nasal mucosa, as well as rarely a dry red (‘chapped’) mucosa. Mucosal biopsy revealed nonspecific inflammatory changes only. In a study of 66 hypothyroid patients, Gupta14 described symptoms of nasal stuffiness in 64 percent and increased colds and rhinorrhea in 55 percent. Rhinoscopy revealed pale, boggy mucosa in 41 percent and erythematous dry mucosa in 27 percent. Biopsies of 16 patients showed increased submucosal acid mucopolysaccharide ground substance and glandular hypertrophy. In general, symptoms of nonallergic rhinitis are common for patients with hypothyroidism. However, undiagnosed hypothyroidism among patients with uncomplicated rhinitis is very low.15, 16 Therefore, tests of thyroid function of patients with nonallergic rhinitis should be reserved for patients with other systemic symptoms of hypothyroidism.17 [Grade D] Vestibular and otologic symptoms of hypothyroidsim include vertigo, hearing loss and pruritic external auditory canals; the latter has been postulated to be due to the decrease of cerumen production. As many as twothirds of patients with hypothyroidism suffer from vertigo. The attacks are typically mild and brief and are not associated with changes in ENG or hearing loss.18 These patients can also experience hearing loss. Although this is much more common in congenital disease, approximately 30–40 percent of patients with myxoedema suffer from bilateral sensorineural hearing loss. These patients can also experience conductive losses secondary to Eustachian tube mucosal oedema.19 Congenital hypothroidism (cretinism) is a rare disease with a higher incidence in the underdeveloped nations. Head and neck manifestations of affected children are puffy face, yellowish skin, flat nose and enlarged tongue that can protrude from an open mouth. Fontanels and sutures may be widened, and the child’s cry is hoarse. A goitre is characteristic of endemic cretinism, but the sporadic type caused by thyroid agenesis or dysgenesis typically has no palpable thyroid. Sensorineural hearing loss is progressive and responds to early treatment with thyroid hormone replacement. Bilateral sensorineural



hearing loss is more common in endemic cretinism, whereas sporadic cretinism tends to exhibit either a conductive, mixed or sensorineural hearing loss.20 Pendred’s syndrome is an autosomal recessive disorder that is characterized by a goitre and progressive sensorineural hearing loss. The children are typically euthyroid and have an abnormal perchlorate discharge. The hearing loss is more pronounced in higher frequencies and recruitment is present, indicating a lesion of the cochlea. This syndrome results from an inborn defect in thyroid hormone synthesis, a peroxidase defect in iodine metabolism. Graves’ disease and toxic nodular goitre account for more than 90 percent of hyperthyroidism and thyrotoxicosis, both more common in women. Graves’ disease is classically manifested by an infiltrative opthalmopathy (Figure 32.2). This is a lymphocytic inflammatory reaction that infiltrates the extraocular muscles and orbital fat; the fibroblasts proliferate and deposit glycosaminoglycans, predominately hyaluronic acid.21 As enlargement of orbital muscle and fat progresses, the bony walls of the orbit will not allow any increase in volume, therefore proptosis occurs. With an average orbital volume of 26 mL in healthy adults, an increase of only 4 mL in the volume of the orbital contents will result in 6 mm of proptosis. As the proptosis progresses, the protective mechanisms of the cornea are diminished. This can cause exposure, desiccation, irritation and, ultimately, ulceration and blindness if not treated. The inflammatory reaction in turn evokes a deposition of collagen into the extraocular muscles and this results in a restrictive opthalmoplegia. Optic nerve involvement as a result of compression at the orbital apex by the enlarged extraocular muscles22 is another later finding in Graves’ disease. Lid lag and the appearance of a ‘stare’ are ocular manifestations of thyrotoxicosis and are often seen in the earliest forms of Graves’ disease. This is the result of an increased sympathetic sensitivity to catecholamines.23 Generalized tremors are common in patients with thyrotoxicosis. In the head and neck these manifest as fine tremors of the lips and the tongue (more evident when it is protruding24). Vocal changes are often noted in these patients as well. Although no true dysphonia is noted, there is a tremulous voice that is a result of the increased respiratory rate and the resultant decreased vital



Figure 32.2 A patient with Graves’ opthalmopathy showing proptosis and lid retraction characteristic of this condition.



Chapter 32 Head and neck manifestations of endocrine disease



capacity.25 Less common manifestations are osteoporosis of the mandible and maxilla and dark pigment deposits in the buccal mucosa. Goitre, or neck masses, are very common in patients with thyrotoxicosis, as well as thyroid neoplasia. Neoplasia will be addressed in another chapter in this text, however, it should be noted that dysphagia, hoarseness, haemoptysis and haematemesis are all common symptoms and should alert the otolaryngologist to possible thyroid neoplasia.



ADRENAL DISEASE Of the diseases of the adrenal gland, or disease processes that cause perturbations in the hormones produced within the adrenal gland, Cushing’s syndrome/disease is the one with the most significant head and neck manifestations. Common causes of Cushing’s syndrome/ disease include adrenal hyperplasia, adrenal adenoma/ carcinoma, ectopic production of ACTH and iatrogenic causes. A rare cause of ectopic production of ACTH is medullary thyroid carcinoma (MTC). These tumours have demonstrable adrenocorticotropin immunoreactivity and can release adrenocorticotropin. Classic signs of this disease are central obesity and ‘moon’ faces. Androgen excess causes acne and hirsutism. Patients with Cushing’s disease are at an increased risk of infectious diseases, oral candidiasis is a common problem. Generalized muscle weakness (steroid myopathy) can cause difficulties with phonation and deglutition. In contrast to Cushing’s disease, adrenocortical insufficiency (Addison’s disease) has oral mucosal hyperpigmentation and generalized muscle weakness that results in vocal weakness.



PARATHYROID DISEASE Patients with hypoparathyroidism display many common head and neck signs and symptoms. The patients who suffer from hypoparathyroidism and resulting hypocalcaemia complain of perioral tingling and numbness. Hoarseness, aphonia and stridor are the typical vocal manifestations of laryngospasm seen with tetani. Chvostek’s sign is also commonly seen with tetani (however, as many as 10 percent of normal patients will have a positive sign). Chvostek’s sign is elicited by percussing over the facial nerve, anterior to the tragus. Facial muscle twitching is a positive response. Long-term hypocalcaemia can cause enamel defects and hypoplasia of the teeth. Hyperparathyroidism and resulting hypercalcaemia is most often caused by parathyroid adenoma (in the ambulatory population), or malignancy (in the hospitalized population), although the differential for hypercalcaemia is long and getting longer. Classically, this disease process manifests as ‘stones’ (renalithiasis), ‘bones’ (bone



] 401



pain, with demineralization often seen in the skull), ‘groans’ (GI distress) and ‘psychiatric overtones’ (emotional changes including abnormal mentation). Osteitis fibrosis cystica or ‘brown tumours’ consisting of osteoblasts, osteoclasts and fibrous tissue are occasionally seen in the mandible and maxilla. In a small study, Simpson noted that auditory dysfunction, in three cases, and aphonia, in two cases, were reversed after removal of parathyroid adenomas.26 He postulated that the high calcium ion concentration inhibited neural synaptic transmission, causing both of these symptoms. Hyperparathyroidism also has a genetic component: multiple endocrine neoplasia (MEN) syndromes. MEN 1 (Wermer’s syndrome) includes pituitary adenoma/hyperplasia, parathyroid adenoma/hyperplasia and pancreatic islet cell hyperplasia/adenoma/carcinoma. MEN 2A (Sipple syndrome) includes MTC, pheochromocytoma and parathyroid hyperplasia/adenoma. The MEN syndromes are inherited as autosomal dominant traits, therefore a family history is important to ascertain and decide if further work-up is necessary. Hyperparathyroidism is often the first manifestation of MEN 1 and usually appears after the patient reaches 18 years of age.



PANCREATIC ISLET CELL DISEASE Diabetes mellitus (DM) is a disease that is caused by inadequate action of insulin, due to either resistance of end organs due to obesity reducing the insulin receptors on insulin responsive cells (type 2), or reduced levels of insulin due to autoimmune response to beta cells of the pancreas triggered by infection (type 1). Type 1 typically presents in patients under age 25 in 0.2–0.5 percent of the population, with no predilection for men or women. There is only a 50 percent concordance in identical twins. Prevalence of type 2 is 2–4 percent, and usually occurs in patients older than 40. There is a slight predilection for women. There is also a 100 percent concordance rate in identical twins. Head and neck manifestations of type 1 and type 2 diabetes are identical, and we will discuss these as a unit. Chronic manifestations of DM are secondary to micro and macrovascular disease, neuropathies (including peripheral, autonomic radiculopathy and mononeuropathy) as well as immunodeficiency secondary to impaired leukocyte/phagocyte function.27 [***] General considerations for otolaryngologists as surgeons revolve around the immunodeficiency and the decreased ability of diabetics to heal their surgical wounds and their increased susceptibility to infection. This should be considered prior to operating and glycaemic control should be optimized during the perioperative period. [Grade B] The ocular findings of DM include diabetic retinopathy where venous dilatation leads to full-blown exudates and haemorrhage in the retina and vitreous



402 ] PART 6 ENDOCRINOLOGY body. Haemorrhages in the conjunctiva and iris result in glycogen deposits, as well as depigmentation and neovascularization. Abnormal lens function leads to cataract formation. Mononeuropathy in the head and neck can be a presenting sign of DM. Bell’s palsy has a high comorbidity with DM. In a retrospective study, 45 percent of patients between the ages of 10 and 19 who presented with Bell’s palsy had DM, and in a series of 130 patients with DM, 66 percent presented with Bell’s palsy.28 Other neuropathies of the cranial nerves can result in unilateral vocal cord paralysis, dysphagia and dysphonia. Laryngeal nerve sections of moderate to severe diabetics showed segmental demyelination, axonal degeneration and concentric proliferation of Schwann cells.29 Neuropathies affecting all other cranial nerves (with the exception of the olfactory and hypoglossal) have been documented in DM. Retrospective studies also show that as many as 50 percent of patients with diabetes have some manifestations of auditory dysfunction, and analysis of diabetic temporal bones has been found to contain PASpositive lesions of the capillaries in the stria vascularis.30 This is the same lesion seen in Kimmelstiel–Wilson disease (diabetic nephropathy). Animal studies of auditory brainstem responses (ABRs) have also shown significant prolongation of wave I as well as an elevation in the ABR threshold in diabetic rats.31 Equilibrium disorders are also very common for diabetics. Aetiologies for disequilibrium include peripheral neuropathy, direct toxicity of the vestibular end organ and fluctuations in glucose concentrations in the cerebral circulation. Diabetics routinely have abnormal electronystagmography and dynamic posturography. In a recent animal study, diabetic rats were shown to have prolonged latency and decreased amplitude of the first wave of the vestibular evoked potentials (VEPs). These parameters of the first wave have been shown to correlate with vestibular end-organ function.4 Diabetics are particularly prone to certain infections. Malignant otitis externa, commonly due to Pseudomonas aeruginosa, typically presents in the older diabetic with severe otalgia, otorrhoea, fever and leukocytosis. Granulation tissue at the bony-cartilagenous junction is a classic sign, and cranial nerve findings including facial nerve paralysis, as well as involvement of V, X, XI, XII are all poor prognostic indicators. Disease may spread beyond the external auditory canal and cause osteitis, chondritis and osteomyelitis of the temporal bone. Intracranial extension and death are possible if treatment is not thorough. Early cases are responsive to oral floroquinolones and meticulous and frequent debridements under the binocular microscope, more advanced cases require hospitalization and i.v. antiPseudomonas antibiotics and possible surgical debridement. [Grade B/C] Another infection particular to diabetics and other imunocompromised hosts is invasive fungal sinusitis (rhinocerebral mucormycosis) (Figure 32.3). Presentation



Figure 32.3 A diabetic patient with invasive fungal sinusitis involving the paranasal sinuses and orbit.



in the immunocompromised patient includes fever and localization of symptoms to the paranasal sinus area, such as orbital swelling, facial pain or nasal congestion. Physical examination may show necrosis of the nasal mucosa, indicative of mucormycosis and, in rare situations, actual hyphae. However, typical examination findings are indistinguishable from nonfungal causes of sinusitis. Anaesthesia of the nasal mucosa or cheeks, independent of topical anaesthetics, is suspicious for mucormycosis. The oral cavity always should be examined for invasion through the hard palate from the sinuses. Changes seen on sinus CT or plain radiographs are usually nonspecific and usually indistinguishable from bacterial sinusitis, although they may show bony erosion or soft-tissue invasion. Biopsy and culture are critical to making the diagnosis. Appropriate antifungal therapy and aggressive surgical debridement should be initiated as soon as possible.



PREGNANCY AND GONADAL DISEASE The most common head and neck complaints during pregnancy are nasal obstruction and rhinorrhoea, typically serious and frequently accompanied by postnasal drip. Physical examination reveals hyperaemic mucosa with marked oedema and hypersecretory states. This is most common during the second and third trimesters and tends to continue for the duration of the pregnancy. Typically, resolution occurs within one week of parturition. Approximately 15–30 percent of pregnant females suffer from these symptoms.32 Although some patients have pre-existing allergic rhinitis, vasomotor rhinitis or sinus disease, and pregnancy merely exacerbates these conditions, many patients who do not have these underlying problems also suffer from this. The exact pathophysiology is not understood. It is clear that there are direct influences



Chapter 32 Head and neck manifestations of endocrine disease



of oestrogens on the nasal mucosa and resulting airflow, however no correlation has been shown between estradiol levels and nasal symptomatology.33 [***] Animals given oestrogen over a period of weeks to months show histologic changes of the mucosa, including proliferation of squamous metaplasia with increases in the cornification of the epithelium and a decrease in the number of cilia and goblet cells. Diffuse glandular hyperplasia with increased secretory activity of the acini was also seen. Vascular changes included endothelial proliferation and capillary and arteriolar swelling.34 [***] Symptomatic patients versus asymptomatic pregnant patients showed increased activity of succinic dehydrogenase indicative of increased secretory activity, and an increase in alkaline phosphatase concentration indicative of increased vascularity and a transfer of metabolites across plasma membranes.35 The reasons why some women are symptomatic and others are not is not clearly understood. Another common problem during pregnancy is an inflammatory gingival hyperplasia caused by increased gingival reactivity to local irritants. Typically seen in the patient’s first trimester and commonly maintained until termination of the pregnancy, it may evolve into a ‘pregnancy tumour’ of the gingival. Biopsies of these have shown to be pyogenic granuloma. Menopause may be associated with several oral manifestations; the most common are desquamative gingivitis and generalized atrophic alterations of the oral mucosa. Kallmen’s syndrome and Turner’s syndrome are two syndromes pertaining to the reproductive organs that deserve mention here. Kallmen’s syndrome (hypogonadotropic hypogonadism) has two inheritance patterns: the first is X-linked recessive and the second is autosomal dominant with variable expressivity. The syndrome consists of gonadotropin deficiency secondary to LHRH deficiency in association with anosmia. This is often diagnosed when males fail to achieve puberty. There are variable degrees of dysnosmia and even female carriers are shown to have decreased olfactory ability. Studies have shown that some of these patients have no olfactory bulb and stalks, while others have varying degrees of malformation of the olfactory epithelium (ranging from complete intact epithelium to complete agenesis). Other head and neck manifestations sometimes seen are midline facial deformities including cleft palate and deafness. Turner’s syndrome (gonadal dysgenesis) is characterized by an XO karyotype and occurs in 1 in 2500–10,000 live female births. This disease presents in females with gonadal dysgenesis, short stature and often a webbed neck and/or shield chest. Sensorineural, conductive or mixed hearing loss, which may be progressive, is often an early sign of Turner’s syndrome in prepubertal females. These patients can also have a high arching palate, epicanthal folds and low-set ears.



] 403



KEY POINTS  General awareness of endocrine disorders can aid in their early diagnosis as head and neck manifestations are often the presenting signs or symptoms.  Patients with DM often complain of many common head and neck symptoms; this includes cranial neuropathies, sensorineural hearing loss and vestibular dysfunction.  Patients with DM are susceptible to unique and invasive infections including malignant otitis externa and invasive fungal sinusitis.  Pregnancy should be considered as a cause of rhinorrhea in the child-bearing age woman.  Thyroid perturbations are often associated with abnormal function of the larynx, causing changes in pitch and clarity of speech, the nose, causing rhinitis, and the inner ear, causing vertigo and hearing loss.



REFERENCES











1. Maceri DR. Head and neck manifestations of endocrine disease. Otolaryngologic Clinics of North America. 1986; 19: 171–80. 2. Melmed S. Acromegaly. New England Journal of Medicine. 1990; 322: 966–77. 3. Daughaday WH. The adenohypophysis. In: Williams RH (ed.). Textbook of endocrinology. Philadelphia: WB Saunders, 1981: 31–79. 4. Lucente FE. Endocrine problems in otolaryngology. Annals of Otology, Rhinology, and Laryngology. 1973; 82: 131–7. 5. Grotting JK, Pemberton J. Fixation of the vocal cords in acromegaly. Archives of Otolaryngology. 1950; 52: 608–17. 6. Jackson C. Acromegaly of the larynyx. Journal of the American Medical Association. 1918; 71: 1787–9. 7. Sano T, Asa SL, Kovacs K. Growth hormone-releasing hormone-producing tumors: clinical, biochemical and morphological manifestations. Endocrine Reviews. 1988; 9: 357–9. 8. Jefferson G. Extrasellar extension of pituitary adenomas. Proceedings of the Royal Society of Medicine. 1940; 33: 433–58. 9. Rowe PJW, Jones TK. Malignant adenoma with extensive skull destruction. Radiology. 1966; 86: 532–4. 10. Dent JA, Rickhuss PK. Invasive pituitary adenoma presenting with nasal obstruction. Journal of Laryngology and Otology. 1989; 103: 605–9. 11. Ritter FN. The effects of hypothyroidism upon the ear, nose and throat. Laryngoscope. 1967; 78: 1427–79. 12. Proetz AW. The thyroid and the nose. Annals of Otology, Rhinology, and Laryngology. 1947; 56: 328–33.



404 ] PART 6 ENDOCRINOLOGY







13. Proetz AW. Further observations of the effect of thyroid insufficiency on the nasal mucosa. Laryngoscope. 1950; 60: 627–33. 14. Gupta OP, Bhatia MS, Agarival MS, Mehrorta MD, Mishr SK. Nasal, pharyngeal, and laryngeal manifestations of hypothyroidism. Ear Nose and Throat Journal. 1977; 56: 10–22. 15. Proctor DF, Andersen IB. The nose: upper airway physiology and the atmospheric environment. Amsterdam: Elsevier Biomedical Press, 1982: 207. 16. Settipane GA, Klein DE. Nonallergic rhinitis: the domography of eosinophils, nasal smear, blood total eosinophil counts, and IgE levels. Allergy Proceedings. 1985; 6: 363–6. 17. Gustafson RO, Knops JL. In: McCaffrey TV (ed.). Effect of thyroid disease, pituitary disease, and pregnancy on the nasal airway: systemic disease and the nasal airway. New York: Thieme Medical Publishers, 1993: 120–30. 18. Bhatia PL, Gupta OP, Agrawal MK, Mishr SK. Audiological and vestibular function tests in hypothyroidism. Laryngoscope. 1977; 87: 2082–9. 19. Debruyne F, Vanderschueren-Lodeweyckx M, Bastijns P. Hearing in congenital hypothyroidism. Audiology. 1983; 22: 404–9. 20. Fraser GR. Association of congenital deafness with goiter (Pendred’s syndrome): a study of 207 families. Annals of Human Genetics. 1965; 28: 201–49. 21. Sergott RC, Glasner JS. Graves’ ophthalmopathy: a clinical, immunologic review. Survey of Ophthalmology. 1981; 26: 1–21. 22. Neigel JM, Rootman J, Belkin RI, Nugent RA, Drance SM, Beattie SW et al. Dysthyroid optic neuropathy: the crowded orbital apex syndrome. Ophthalmology. 1988; 95: 1515–21. 23. Calcaterra TC, Thompson JW. Antral-ethmoidal decompression of the orbit in Graves’ disease: 10-year experience. Laryngoscope. 1980; 90: 1941–9.















24. Burke H. Endocrine aspects of otolaryngology. Laryngoscope. 1968; 78: 857–62. 25. Maceri D. Head and neck manifestations of endocrine disease. Otolaryngologic Clinics of North America. 1986; 19: 171–80. 26. Simpson JA. Aphonia and deafness in hyperparathyroidism. British Medical Journal. 1954; 1: 494–6. 27. Drachman RH, Root RH, Wood WB. Studies on the effect of experimental nonketotic diabetes mellitus on antibacterial defense. Journal of Experimental Medicine. 1966; 124: 227–40. 28. Korezyn A. Bell’s palsy and diabetes mellitus. Lancet. 1971; 1: 108–10. 29. Schechter GL, Kostianovsky M. Vocal cord paralysis in diabetes mellitus. Transactions-American Academy of Ophthalmology and Otolaryngology. 1972; 76: 729–40. 30. Igarashi M. Pathology of the inner ear end organs. In: Minkler J (ed.). Pathology of the nervous system. New York: McGraw Hill, 1972. 31. Perez R, Ziv E, Freeman S, Sichel JY, Sohmer H. Vestibular end-organ impairment in an animal model of type 2 diabetes mellitus. Laryngoscope. 2001; 111: 110–3. 32. Mabry RL. Rhinitis of pregnancy. Southern Medical Journal. 1986; 79: 965–71. 33. Bende M, Hallgarde M, Sjogren U, Uvnas-Moberg K. Nasal congestion during pregnancy. Clinical Otolaryngology and Allied Sciences. 1989; 14: 385–7. 34. Helmi AM, El-Ghazzawi IF, Mandour MA, Shehata MA. The effect of estrogen on the nasal respiratory mucosa. Journal of Laryngology and Otology. 1975; 89: 1229–41. 35. Toppozado H, Michaels L, Toppozado M, El-Ghazzawi I, Talaat M, Elwany S. The human respiratory nasal mucosa in pregnancy. Journal of Laryngology and Otology. 1982; 96: 613–26.



PART



7



PHARMACOTHERAPEUTICS EDITED BY MARTIN J BURTON



33 Drug administration and monitoring Geraldine Gallagher



407



34 Corticosteroids in otolaryngology Niels Mygind and Jens Thomsen



418



35 Drug therapy in otology Wendy Smith and Martin Burton



429



36 Drug therapy in rhinology Wendy Smith and Grant Bates



436



37 Drug therapy in laryngology and head and neck surgery Wendy Smith and Rogan Corbridge



446



This page intentionally left blank



33 Drug administration and monitoring GERALDINE GALLAGHER



Introduction Drug development Drug administration Drug pharmacology Monitoring of drug therapy Adverse drug reactions



407 407 408 409 412 414



Drug prescribing in special treatment groups Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



415 416 416 416 416



SEARCH STRATEGY The information in this chapter is supported by a PubMed search using the key words drug regulation and development, drug bioavailability, clinical pharmacokinetics and pharmacodynamics, drug compliance and monitoring, and adverse drug reactions.



INTRODUCTION In clinical ear, nose and throat (ENT) practice, therapeutic decision making has become increasingly complex, particularly in the realm of drug prescribing. The increasing range of possible interventions and the changing expectations of patients have been accompanied by a dramatic rise in the number of drugs available. In 1932, the British Pharmacopoeia listed 213 medical products. By 2001, this figure had increased to 2760. For clinicians, these factors have created a personal responsibility: we are now required to assess the relevant published evidence, to make a balanced therapeutic decision and to evaluate the outcome of that decision. Ideally, each drug prescription for a patient should be an individualized therapeutic plan which maximizes the clinical benefit and the patient’s well-being while achieving the lowest toxicity and the lowest cost. To implement such a logical approach to drug therapy, a number of steps are known to be important. The clinician must:  make an accurate diagnosis;  understand the pathophysiology of the disease;



 have a good knowledge of the therapeutic options;  select treatment: either no drug or an optimum drug and dosing schedule;  monitor the effect of treatment and decide a therapeutic end point;  maintain a ‘therapeutic alliance’ with the patient. In this chapter, the aim is to discuss the regulatory mechanisms involved in drug control and to examine the factors which influence the therapeutic options for individual patients. In essence, the chapter sets out the principles involved in rational prescribing.



DRUG DEVELOPMENT The discovery and development of a drug can take 15 years or more and involves many steps. Once a chemical with a potential therapeutic effect has been identified, a series of investigations in animals and humans is carried out to check whether the chemical is effective and whether it is safe to use. The first step in the development of a drug is to establish its pharmacokinetic and toxic



408 ] PART 7 PHARMACOTHERAPEUTICS profile in animals. Three stages of trials in humans establish its potential for the treatment of disease. Phase I studies are primarily concerned with assessing a drug’s safety, and are typically carried out on a small number of healthy volunteers, investigating the side effects as the dose increases. Phase II studies test for efficacy and may involve up to several hundred patients in a randomized, controlled trial. Finally, in phase III, most drugs are ideally entered into a randomized, double-blind, controlled trial involving several hundred to several thousand people, giving a more thorough understanding of the effectiveness, benefits and range of possible adverse reactions compared to an existing drug or placebo. In the UK, once a drug has been successfully trialled, the pharmaceutical company applies to the Committee on Safety of Medicines (CSM) for a licence to launch the drug into the market. When a licence is received, the drug then joins the range of therapeutic options available to the clinician.



Drug regulatory systems In most countries, a drug cannot be made legally available until a government-sponsored regulatory body has reviewed its safety and efficacy. This system was put in place in the UK in 1968 as a consequence of the thalidomide tragedy. Prior to this, there was no body in the UK which could prevent the marketing of a new medicinal product. Thalidomide was launched in Britain in 1958 as a sedative/hypnotic and for use in morning sickness. By 1961, however, it became clear that its use in early pregnancy was causally related to a congenital abnormality where the long bones in babies failed to develop. Around 1000 deformed babies were born in Britain as a result and around 10,000 worldwide. Following this tragedy, the Committee on Safety of Drugs was established, later to become the CSM following the Medicines Act of 1968. The CSM is an independent group of 34 clinicians, pharmacologists, toxicologists and pharmacists plus two lay appointees who advise the licensing authority (the health and agriculture ministers) on medicinal products. Adverse drug reactions are reported to the CSM by individual doctors, using a voluntary yellow card system. Although the voluntary system has yielded some impressive regulatory decisions, there is evidence that adverse reactions are significantly under-reported.1, 2, 3 In 1995, the European Union (EU) established the European Agency for the Evaluation of Medicinal Products (EMEA). This body coordinates the scientific evaluation of the safety, quality and efficacy of medicinal products for use throughout the EU. EMEA employs a system of ‘mutual recognition’ of drug authorizations already granted by individual member states. In addition, EMEA provides a centralized European evaluation for



new drugs and all decisions agreed under this new system are binding on member states. Herbal medicines are largely outside the current regulatory systems. Used by 80 percent of the world’s population and an increasing number of patients and practitioners in Europe and North America, herbal medicines include homeopathic, traditional herbal, Ayurvedic, naturopathic and native medicines. Most products are still unregulated, but in the UK, adverse reactions to Chinese and herbal medicine are reported to the CSM. The most comprehensive current regulatory system and information resource is provided by the German ‘Commission E’ monographs.



DRUG ADMINISTRATION Following diagnosis of the disease and a decision to choose drug therapy, the clinician has to select an appropriate drug and decide the mode of administration. There is a wealth of information available about licensed drugs, including recommendations on delivery methods, dosage and dosage frequency, as well as information on known adverse effects. In considering the mode of administration, it is important for the clinician to assimilate and understand this information in order to select the best options for the individual patient.



Delivery methods When a drug is used therapeutically the aim is to achieve an adequate concentration at the site of action and to maintain that concentration for as long as the effect is required. ‘Systemic bioavailability’ is a term used to describe the proportion of administered drug that reaches the systemic circulation and is available for distribution to the receptor site. The method of administration is a major determinant of bioavailability. When a drug is given intravenously, the total dose enters the systemic circulation. This method achieves 100 percent bioavailability. However, use is limited by the constraints of venous access and issues of safety. When a drug is given orally, the bioavailability depends on the extent of absorption from the gut and the metabolic activity of the liver as the drug circulates following absorption, known as the ‘first pass effect’. This effect can substantially reduce the bioavailability of a drug. Most drugs are given orally and are therefore subject to ‘first pass’. Transnasal, sublingual, buccal and rectal formulations all avoid the first pass effect. The bioavailability depends only on the absorption across the mucosa. This mechanism may become an important consideration in some circumstances. For example, prochlorperazine in the buccal form may help a patient suffering an acute



Chapter 33 Drug administration and monitoring



labyrinthine attack in the presence of nausea and vomiting. Transdermal formulations in the form of patches are absorbed through the skin and enter the systemic circulation directly. Absorption depends on lipid solubility and a concentration gradient being established across the dermis. Only some drugs are successful, one example being transdermal hyoscine for the treatment of motion sickness. Intramuscular and subcutaneous preparations are limited in number because the absorption of drugs given by these routes is variable and often unpredictable, hindering their usefulness. Special consideration has to be given to the formulation of the drug and to the site of application. Drugs given intramuscularly are given into the deltoid muscle rather than gluteal, as absorption in the former is much more rapid due to the higher vascularity. A reduction in the rate of local absorption can be achieved, for example, by the use of adrenaline in combination with local anaesthetics administered subcutaneously, causing local vasoconstriction. Important subcutaneous preparations include heparin and insulin.



DRUG PHARMACOLOGY The biological activity of a drug, whether therapeutic or toxic, is proportional to the concentration of the drug at the receptor site and the persistence of the drug’s effect is directly related to the length of time the drug remains at the receptor site. Therefore, the biological activity of a drug is dependent on the rate and completeness of its absorption, distribution, metabolism and elimination. Within ENT practice, getting a drug to a receptor site can either be by direct application to that site (as with nasal antihistamines), or by administration via a site remote from the target (as with oral antihistamines in the case of allergic rhinitis). Whatever the method used, the drug must then diffuse across several membranes (absorption), distribute into a variety of tissues and fluids (distribution), be subject to a wide array of metabolizing enzymes



] 409



(metabolism) and finally be eliminated from the body (elimination). Pharmacokinetics is the study of the rate of drug movement into and through body tissues. This information will help predict the time course and magnitude of the drug effect. The pharmacokinetics of a drug can be studied by regular analysis of plasma and urine samples following its administration, giving quantitative assessment of the rate of drug movement, metabolism and elimination. Plasma concentration is not necessarily the same as tissue concentration, but accurately reflects it in most therapeutic regimens and for practical clinical purposes is taken to be the same. The study of the underlying mechanism of action of the drug at the receptor site and the resultant biological response is known as pharmacodynamics. Variations in this biological response due to genetically determined factors is known as pharmacogenetics.



Pharmacokinetics and clinical practice Pharmacokinetics is the key factor in deciding on drug formulation and dosing frequency. To understand why, it is necessary to look at how drug concentration in the plasma (and by implication, at the receptor site) changes with time. After a drug is administered to the body in any form, the concentration of the drug in the plasma rises in keeping with absorption, then falls as a result of distribution, metabolism and elimination. Figure 33.1 shows a typical profile of drug concentration in nasal tissue following instillation of nasal drops. The curve contains three features of interest. Cmax This is the maximum level of drug in the plasma after a single administration (reflecting the maximum level of drug in the tissue). Cmax dictates the initial therapeutic and toxic responses. Most drug applications aim to have Cmax falling within the therapeutic range. The position of Cmax along the time-axis reflects the rate at which the drug becomes bioavailable.



Therapeutic range Plasma concentration



Toxic level



C max



Min effective level



0



T max



AUC Time



Figure 33.1 Change in drug concentrations in the plasma/tissue following nasal administration.



410 ] PART 7 PHARMACOTHERAPEUTICS Tmax This is the time taken to reach the maximum level of drug in the tissue. Area under the curve The area under the curve (AUC) is a measure of the total amount of drug absorbed from a single dose. The bioavailability of the drug – that is, amount instilled versus amount absorbed – can be computed from the AUC.



Choosing the drug formulation Figure 33.2 shows two curves representing the theoretical plasma/tissue concentration over time, following oral administration of equal doses of the same drug using different formulations. The total bioavailability, as indicated by the AUC, is the same in each case. However, the rate at which the drug becomes available depends on the formulation. For drugs such as analgesics, where a rapid plasma concentration is required after single dosing, choice of formulation may therefore be very important. For example, a soluble Diclofenac formulation giving a curve A would be preferable in the treatment of acute pain (such as postoperative tonsillectomy) rather than an enteric-coated formulation giving rise to a curve B. However, the latter would be more useful in the longterm treatment of chronic pain or for an antiinflammatory effect, allowing a reduction in the frequency of administration.



the half-life of the drug. The resultant plasma level shows a series of peaks and troughs as in Figure 33.3. The peak and trough levels rise over time because each subsequent dose given at t1/2 is additional to the quantity of the drug left behind from the earlier administration. Eventually, the size of the fall in plasma concentration after the peak equals the preceding rise and a steady state is reached. At steady state, the amount of drug eliminated from the body in a single dose interval is the same as the amount which enters it. Strictly speaking it is not a true steady state because the plasma concentration fluctuates. However, for practical purposes, a steady state is reached within a finite number of doses, usually between four and six. The concentration of drug in the plasma after multiple doses will be greater than the single dose, Cmax, but at all times should be within the therapeutic range. If the half-life is long and it takes several days to reach a steady state, then an initial loading dose can be prescribed to overcome the delay in establishing the therapeutic effect. In Figure 33.4, a standard oral regime (Oral regime A) should ideally maintain a therapeutic plasma range throughout the drug administration. It can be seen that shortening the dosing interval and giving the drug more frequently (Oral regime B) may result in toxic plasma levels. Furthermore, a decrease in the dosing frequency could result in a subtherapeutic effect. Choice of dosing interval is therefore extremely important in achieving and maintaining therapeutic effectiveness with minimal toxicity.



Deciding on dosing frequency Metabolism The frequency with which drugs are administered is governed by the rate at which they are metabolized and by the rate of clearance of the unmetabolized drug by the body. The overall rate of loss of the drug is described as the half-life (t1/2): the time for the plasma concentration to fall to one half of the maximum value. The half-life in turn dictates the timing interval for delivering the second and subsequent doses; the recommended interval equals



Most drug metabolism occurs in the liver, but can occur elsewhere, for example in the lungs and kidneys. The end result is inactivation of the drug. The enzymatic processes involved in drug metabolism are described as phases I and II. Phase I involves chemical alteration of the basic structure of the drug by a number of versatile groups of enzymes in the liver, one of the most important being



B Min effective level



Plasma drug concentration



Plasma concentration



A



Steady state concentration



Time Time



Figure 33.2 Drug bioavailability curves for differing formulations.



Oral doses



Figure 33.3 Achieving steady state concentration following initiation of oral therapy.



Chapter 33 Drug administration and monitoring



Plasma drug concentration



B



] 411



Toxic



Therapeutic



A



Subtherapeutic (Oral regime A) (Oral regime B) Time



cytochrome P-450. Cytochrome P-450 consists of a large number of genetically related enzymes. (The name is derived from the fact that a light wavelength of 450 nm is absorbed by the enzymes.) Where a patient is on two or more drugs, competition between the drugs for attachment to cytochrome P-450 can result in a rise in the concentration of one of these drugs to a toxic level. This problem has been experienced in ENT in the past, due to a practice of coprescribing antihistamines, such as terfenadine, with macrolide antibiotics, such as erythromycin. Erythromycin competes with terfenadine for a binding site, and will then inhibit the isoenzyme within the P-450 system that degrades terfenadine. This leads to toxic levels of terfenadine resulting in a characteristic pattern of ventricular tachycardia known as torsades de pointes, which is potentially fatal. Phase II metabolism introduces hydrophilic molecules to the drug resulting in end products which are water soluble and can more easily be eliminated from the body.



Drug elimination This overwhelmingly takes place through kidney excretion. Minor routes include lungs, breast milk, sweat and tears.



Pharmacodynamics and clinical practice Pharmacodynamics are the key factor in determining which drug and dosing schedule to prescribe. When a drug is administered, it enters the pharmacokinetic process and eventually reaches its target tissue. The next stage, the pharmacodynamic stage, will lead to a therapeutic effect. This stage involves attachment of the drug to receptors. Receptors are proteins situated either in the cell membranes or within the cellular cytoplasm of target organs. For each receptor there is a specific group of



Figure 33.4 Drug concentration in plasma/tissue following change in dosing schedule.



ligands (either drugs or endogenous substances) that can bind to the receptor and produce pharmacological effects. There are two types of ligands of interest in ENT prescribing – agonists and antagonists. Agonists are ligands that bind to a receptor, stimulate it and produce a response. The drugs ephedrine and hyoscine are examples. Antagonists are ligands that bind to a receptor and thereby prevent an endogenous agonist from binding to and stimulating the receptor. The drugs ipratropium bromide and cetirizine hydrochloride are examples. The intensity of the therapeutic response that drugs produce, through agonist or antagonist activity, can be plotted against the dosage. Most drugs produce graded dose-related effects and Figure 33.5 is a dose–response curve for two similar analgesics (A and B) used in clinical ENT practice. There are a number of important points to note in the figure.  A log conversion is used for dosage to allow a greater range of dosage to be represented.  There is a straight line relationship between log dose and response over a large part of the curve which allows easier comparison of drugs with similar therapeutic actions.  There is a ‘therapeutic range’ within which most patients will achieve a therapeutic effect. The recommended dosing schedule for drugs is based on the median effective dose or ED50, which is the dose effective for 50 percent of the population studied. For drug A, the therapeutic range is between 8 and 25 mg and the ED50 is 15 mg.  There is a ceiling effect represented by the plateau, beyond which increasing the dose achieves no increase in the therapeutic effect.  The potency of a drug, a term often used by pharmaceutical companies in their marketing, describes where the curve lies along the x-axis. So, in Figure 33.5, drug A is more potent than drug B: less



412 ] PART 7 PHARMACOTHERAPEUTICS



% Patients achieving satisfactory analgesia



100



A



B



50



ED50



1



2



5



10



20



50



100



200



Dose (mg)



of A is needed for a given therapeutic effect. A difference in potency is not an important issue in itself, unless the actual volume or mass to be given is so large that it causes difficulty in administration, or the drugs differ significantly in cost.  The therapeutic index is a measure of the safety of a drug. It is the ratio of the dose producing the desired therapeutic effect, to the dose which produces a toxic effect. A low therapeutic index indicates a potentially dangerous drug. In Figure 33.5, the therapeutic index is represented by the slope of the curve. If the slope is steep, then increasing the dosage by a small amount can increase the effect to levels where toxic side effects can occur. We have now seen how a drug formulation and drug pharmacokinetics affect the bioavailability and how pharmacodynamics relate to a therapeutic response. The next step in the therapeutic process is to monitor the therapeutic response.



Pharmacogenetics From the previous discussion on pharmacokinetics and pharmacodynamics, it can be seen that the resultant therapeutic response to a drug could be modified by genetic factors at many different stages of the pharmacological process, from transit through the body to effector action. Sensitivity to succinyl choline is one example of an inherited autosomal recessive trait. Administration of this drug during anaesthesia in patients carrying this trait can lead to prolonged periods of apnoea, which may result in unplanned mechanical ventilation. In these patients, the cholinesterase enzyme which converts succinyl choline to succinyl monocholine has altered properties resulting either in an unexpected



500



1000



Figure 33.5 Dose response curves for analgesic drugs A and B.



sensitivity to the drug or on occasions an abnormal resistance to its effects. Biochemical variations can occur unpredictably with many drugs, giving rise to occasional idiosyncratic responses. Often these genetic variations are undetected and the effects misinterpreted.



MONITORING OF DRUG THERAPY Once a drug regime is decided and treatment is underway, it is important to monitor the effect of the treatment and provide feedback to the patient. As well as gauging the therapeutic effect of the drug, monitoring is essential to identify adverse drug reactions.



Methods of monitoring These fall into one of three categories. 1. Physical monitoring. Certain medical problems lend themselves to serial physical measurements to assess the effects of treatment, for example blood pressure in hypertensive patients, forced expiratory volume in asthmatic patients or the extent of nodal disease in lymphoma. 2. Patient self-monitoring. This is probably the most common type of monitoring in use in ENT practice. Patients report back on their changing symptomatology following a trial of therapy. Because such reports are subjective, problems arise in quantifying the improvement or deterioration in symptoms. It is therefore often difficult to decide whether change of dosage is required unless the expected outcome of therapy is the complete alleviation of symptoms. Short of this aim, using a written record of a problem



Chapter 33 Drug administration and monitoring



list and setting intermediate goals for therapy can be of help. 3. Plasma concentration monitoring. This method gives objective measurements of the bioavailability of the drug and reflects the tissue concentration. In clinical practice, such monitoring is important mainly for those drugs with a low therapeutic index (that is, a high risk of toxicity). Examples include aminoglycosides, antiarrythmics, cytotoxic therapy and anticoagulants. Plasma monitoring is also used in cases of therapeutic failure, for example anticonvulsant treatment, to establish an objective level of compliance with the treatment regime. It can be used in a medicolegal context to ascertain the plasma level of recreational or illegal drugs, and is of course essential during clinical pharmacological trials.



Monitoring and compliance Meaningful monitoring of the therapeutic effect of a drug depends on the compliance of the patient. In clinical terms, compliance is defined as the extent to which a patient’s actual history of taking medication corresponds to the prescribed treatment.4 Many factors are involved in noncompliance. They include failure to take a prescription to the pharmacy, failure to follow the doctor’s instructions, taking an incorrect dose, taking the medication erratically, stopping the treatment too early or even failure to attend for follow-up appointments. Whatever the factor, noncompliance or poor compliance is a constant and complex issue which causes major problems in the delivery of health care. Noncompliance adversely affects:  the clinical outcome of prescribed regimes;  the effectiveness of healthcare budgets;  the conclusions drawn from clinical research, especially drug trials. The rational prescription of medication is the most common intervention in primary care and thus forms a large part of the health care costs. Poor compliance with prescribed treatment can have a major impact on clinical outcome, especially if the drug is needed to maintain vital physiological functions. This pattern is often seen with cardiac or hypertensive medication. It is also a problem in established preventative health care programmes, for example aspirin for patients with atherosclerotic disease. Drug trials rely on an intention-to-treat analysis and unless specifically structured, do not reliably account for the effects of poor compliance. Thus results are analysed on the basis of the treatment to which the patients were assigned, regardless of whether the drugs were actually taken. This approach has created a number of problems, including biased estimates of the effectiveness of some drugs. Compliance is difficult to measure, but studies



] 413



suggest that poor compliance is to be expected in 30–50 percent of all patients, irrespective of their disease or prognosis.5, 6, 7 If this extrapolation were applied to drug trials, then the sample size would often need to be increased five-fold in order to yield the same number of ‘complying’ patients as for a sample where 100 percent compliance is assured. Numerous methods of measuring compliance have been introduced, such as patient diaries and blister packs. These depend on the patient being motivated and truthful. Newer methods such as microelectromechanical systems (MEMS) are at present used only during research trials. MEMS are devices containing microchips which are attached to a bottle and are stimulated on opening the bottle. They record the number of times the patient opens the bottle – not necessarily when the patient takes the medication. Although much research has been carried out on compliance over the last three decades, no single factor or group of factors has emerged as the dominant influence on compliance. Therefore, few if any changes introduced into therapeutic regimes have led to a significant improvement in compliance. One of the reasons for the lack of progress in compliance research may be the difficulty in achieving meaningful input from patients themselves.8



Factors affecting compliance NATURE OF THE TREATMENT



It has been shown that poor compliance is related to the duration of the treatment, with long-term therapies being most problematic. Other factors related to low compliance are the number of medications prescribed and the frequency of administration. The more frequently patients have to take a drug, the less likely they are to take it, and the more drugs they are prescribed the lower the overall compliance.9 THE TYPE OF ILLNESS



Low compliance is universally associated with psychiatric disorders and the more symptoms reported, the lower the compliance. Elderly patients with memory disorders have difficulty following complex instructions and this contributes to poor compliance. Disability can also be a factor, with patients finding difficulty in managing tablet containers. On the other hand, high compliance is found in patients who are severely disabled, and in young children, probably due to increased supervision by health care workers and parents. BEHAVIOURAL ISSUES



Patients have been shown to have poor compliance if they are unsure about their diagnosis, if they are



414 ] PART 7 PHARMACOTHERAPEUTICS asymptomatic, if there is a significant time lag between taking the drug and its effect, or if they fear the adverse effects. There may also be social or physical reasons why they do not pick up their medication from the pharmacy – for example, financial difficulties or mobility problems – in addition to the everyday inconveniences involved in carrying or taking medication.



pharmacodynamic action other than that which produces a therapeutic effect; such effect may or may not be dose related. It is estimated that up to 5 percent of hospital admissions per year are due to adverse reactions, mostly dose related.11, 12 [****] More than 10,000 hospital beds in the UK are constantly occupied by sufferers, at a cost to the National Health Service (NHS) of up to £1 billion per year. Adverse drug reactions can be classified as:



BEHAVIOUR OF THE DOCTOR



   



Compliance appears to be related to the quality of the interaction between the patient and the doctor. The enthusiasm and confidence with which treatment is prescribed has been shown to influence compliance, as does the frequency of interaction and the duration of medical appointments.



Methods of improving compliance Research has shown that compliant patients generally have better clinical outcomes than noncompliant patients. Therefore measures aimed at improving patients’ understanding of their illness, their treatment and their doctor’s instructions, should improve compliance. Supervised administration of tablets and practical aids, such as blister packs and dosage counters, do help to improve compliance. It is advisable to develop a menu of complianceenhancing strategies that can then be tailored to the individual patient. Examples include:        



involving the patient in determining the treatment goal; reducing the complexity of the treatment; tailoring the treatment to the patient’s lifestyle; using reminders to the patient; encouraging family support; informing the patient about possible side effects; monitoring compliance; giving the patient feedback.



The patient’s input is now considered such an important aspect of compliance that the Royal Pharmaceutical Society of Great Britain has changed its terminology from compliance to concordance.10 At the centre of the concordance model is the patient as decision maker, with professional support forming a cornerstone of the medical treatment.



ADVERSE DRUG REACTIONS Adverse drug reactions are ‘unwanted effects of drugs occurring at therapeutic doses’. They may be due either to toxic effects or to side effects. A toxic effect is an adverse effect that arises because of an exaggeration of the therapeutic action and is therefore dose related. A side effect is an adverse effect that arises through a



non-dose related; dose related; long-term effects; drug interactions.



Non-dose-related adverse reactions These are caused by allergic responses to drugs and are unrelated to the drug’s therapeutic action or dosage. The ability to mount an allergic response may be inherited or may arise spontaneously. Allergic reactions are subdivided into four types: 1. Type 1 reactions (anaphylactic): The drug couples to cell-bound IgE leading to release of immune mediators. Type 1 reactions are caused by antibiotics such as penicillin, nonsteroidal antiinflammatory drugs such as indomethacin, plus other chemicals such as iodinated contrast media used in radiography. 2. Type 2 reactions (cytotoxic): The drug engages with a protein on the surface of cells, in particular hepatocytes and blood cells. Together they form a surface antigen which attracts and binds circulating IgE or subclasses of IgG, resulting in complement fixation and cell lysis. Examples of drugs causing this reaction are penicillin, heparin and methyldopa. 3. Type 3 reactions (immune complex): The drug binds directly to circulating IgG or IgM. The resulting immunoglobin–drug complex becomes trapped in small capillaries, causing release of complement with destruction of the underlying capillary endothelium. Drugs causing type 3 reactions include penicillin, aspirin and cephalosporins. 4. Type 4 reactions (cell mediated): The drug binds to Langerhans or dendritic cells in the skin. Together, they sensitize T lymphocytes resulting in an inflammatory response. Examples are contact dermatitis due to latex, topical antibiotics or antifungal drugs.



PSEUDOALLERGIC REACTIONS



‘Pseudoallergy’ is a term applied to reactions that clinically resemble allergic reactions, but for which no underlying immunological basis can be currently



Chapter 33 Drug administration and monitoring



established. For example, aspirin and nonsteroidal antiinflammatory drugs can trigger asthma attacks in patients who have coexisting nasal polyposis and asthma.



Dose-related adverse reactions Dose-related adverse reactions are a consequence of the drug’s normal therapeutic effects. They may be due to inaccurate prescribing or to inadvertent administration. They may on occasions result from a change in the generic formulation of a drug which can in turn lead to an increase in its bioavailability. This can be critical for drugs with a low therapeutic index and can be a risk when using different generic formulations. An example occurred in Australia with the drug phenytoin. The drug diluent was changed from calcium sulphate to lactose, resulting in toxic effects and death. Dose-related adverse reactions can also be related to pharmacokinetic or pharmacodynamic variation among individuals. Such variation can be genetic in origin, as seen in a Gaussian distribution, where some patients display toxic effects with even a small ‘normal’ dose of a drug. Variation in patients’ responses may also be caused by disease such as cirrhosis or by environmental factors such as smoking, diet or alcohol consumption.



Long-term therapy causing adverse reactions Some drugs cause adverse reactions which only develop after a prolonged period of treatment. The adverse reaction may or may not be reversible on stopping the treatment. For example, a patient may develop drug tolerance. This is a state of decreased responsiveness to a drug and can be seen following prolonged application of ephedrine nasal drops. Ephedrine causes release of noradrenalin from sympathetic nerve endings but leads to its depletion in prolonged treatment. Another example is the development of abnormal facial movements with neuroleptic drugs, such as chlorpromazine. Both these reactions are reversible on stopping the drugs. Some long-term therapies result in irreversible adverse reactions. A well-known example is hypothyroidism occurring many years after treatment with radioactive iodine. Other irreversible adverse affects are more serious, for example, the increased risk of bladder cancer in patients taking cyclophosphamide.



Adverse reactions caused by drug interactions When two drugs are present one may alter the effects of the other and the interaction may produce an adverse drug reaction. Drug interactions are responsible for about 7 percent of all adverse drug reactions. The drugs most likely to produce harmful interactions are those with a steep dose–response curve and therefore a low therapeutic



] 415



index. The adverse reactions may arise from many different causes, such as stimulation or inhibition of metabolism, as in the case of terfenadine and erythromycin (see above under Metabolism) or through altering renal function, thereby effecting drug elimination.



DRUG PRESCRIBING IN SPECIAL TREATMENT GROUPS Drugs are prescribed because of their potential to benefit the patient, but in every case this is accompanied by the risk of adverse effects. This risk is greater during certain stages of life: pregnancy, childhood, old age and also in certain diseases such as hepatic or renal failure.



Pregnancy In pregnancy, total body water increases by 6–8 L, resulting in a fall in the concentration of plasma albumin. For drugs that bind to albumin (such as anticonvulsants) the result is an increase in the free drug fraction. At the same time, the high level of endogenous progesterone during pregnancy stimulates hepatic enzymes and renal blood flow increases, which can lead to rapid clearance of drugs excreted by the kidney. In addition, there is a risk to the foetus from drugs crossing the placenta, leading to spontaneous abortion, foetal abnormalities or behavioural abnormalities.



Neonatal, infancy and childhood In the neonatal and infancy period, major physiological changes take place in the respiratory, cardiovascular and renal systems which have important implications for drug therapy. The childhood period lasts from 1 to 12 years, and when prescribing drugs for children it is important to remember that they are not ‘miniature adults’. Research suggests the incidence of adverse drug reactions occurring in hospitalized children is as high as 9.5 percent.13, 14



Old age Equally, in later years, there is a pattern of change in the physiological functioning of the body with age which can affect drug pharmacokinetics and pharmacodynamics. Greater numbers of drugs are prescribed in the elderly and the resulting polypharmacy increases the risk of drug interactions. If there is a concomitant disease such as hepatic or renal failure, then drug pharmacology is even further affected due to decreased metabolism and elimination. These cumulative factors can have profound results: it has been estimated up to 18 percent of deaths in the elderly are related to adverse drug reactions.15



416 ] PART 7 PHARMACOTHERAPEUTICS



Renal and hepatic failure The treatment of patients with renal or hepatic failure is outside the scope of this chapter. It is important to recognize that the kidney is often a major determinant of drug kinetics, drug efficacy and drug toxicity. Renal function must be considered in the planning of any therapeutic strategy. As discussed above, the liver plays a major role in the pharmacodynamics of almost all drugs. In the event of liver failure, the metabolism of a drug may be radically altered, resulting in toxic levels, leading to adverse effects. Liver function must be taken into consideration when prescribing and appropriate rational therapeutic strategies implemented.



[ The quality of the interaction between patient and doctor has a significant effect on prescription compliance. [ A menu of compliance-enhancing strategies should be devised for individual patients.



Deficiencies in current knowledge and areas for future research



$



KEY POINTS  The drug prescription should – maximize clinical benefit; – maximize patient well-being; – achieve lowest toxicity; – achieve lowest cost.  The method of administration of a drug is the major determinant of drug bioavailability at the target organ.  The plasma half-life of a drug determines the dosing frequency.  The choice of dose interval is important in achieving and maintaining a steady state of therapeutic effectiveness and minimal toxicity.  Most drugs are metabolized in the liver.  Virtually all drugs are excreted via the kidney.  If a drug has a narrow therapeutic range, ineffective dosing or drug toxicity are more likely to occur.  The recommended dosing schedule for a drug is based on the median effective dose for 50 percent of the population, the ED50.  Poor compliance with prescribed treatments has a major negative impact on clinical outcomes.  A therapeutic concordance model places the patient at the centre, as a decision maker, with surrounding professional support.  Adverse drug reactions account for 5 percent of all hospital admissions.



$ $ $ $



In clinical practice, ENT surgeons use a range of oral medication, topical medication, sprays, drops and inhalers. Thus, for ENT, the mechanisms of drug administration and drug absorption are particularly important. However, fundamental absorption issues such as the patterns of the mucosal distribution of nasally administered drugs, both in health and disease, are still incomplete. Further study may lead to improvements in the delivery systems for drugs to the nasal mucosa, with enhanced absorption and clinical effect. Patient compliance with prescribed medication and the related issue of the identification and recording of adverse drug reactions are problems shared by all medical specialities, but progress on specific ENT solutions to these problems should be sought. Respiratory medicine, ophthalmology and ENT all share the need for better mechanisms to assess the dose delivery of sprays, drops and inhalers. More work is required before reliable, effective and affordable mechanisms can emerge. The next edition of Scott-Brown’s otorhinolaryngology, head and neck surgery may well see progress in: – mapping the distribution and absorption of nasal drugs in health and disease; – understanding the factors affecting patient compliance and identifying proven strategies for improvement; – progress towards an effective and comprehensive reporting system for adverse drug reactions (ADR); – the continued development towards optimal drug delivery systems for ENT mucosal administration.



REFERENCES Best clinical practice [ In monitoring the effect of a drug, the use of a written symptom list and the setting of intermediate treatment goals can be effective.



1. Smith CC, Bennett PM, Pearce HM, Harrison PI, Reynolds DJ, Aronson JK et al. Adverse drug reactions in a hospital general medical unit meriting notification to the Committee on safety of Medicines. British Journal of Clinical Pharmacology. 1996; 42: 423–9.



Chapter 33 Drug administration and monitoring







2. Heeley E, Riley J, Layton D, Wilton LV, Shakir SA. Prescription-event monitoring and reporting of adverse drug reactions. Lancet. 2001; 358: 1872–3. 3. Eland IA, Belton KJ, van Grootheest AC, Meiners AP, Rawlins MD, Stricker BH. Attitudinal survey of voluntary reporting of adverse drug reactions. British Journal of Clinical Pharmacology. 1999; 48: 623–7. 4. Haynes RB. Introduction. In: Haynes RB, Taylor DW, Sackett DL (eds). Compliance in Health Care. Baltimore: The Johns Hopkins University Press, 1979: 1–7. 5. Lassen LC. Patient compliance in general practice. Scandinavian Journal of Primary Health Care. 1989; 7: 179–80. 6. Griffith S. A review of the factors associated with compliance and the taking of prescribed medicines. British Journal of General Practice. 1990; 40: 114–6. 7. Dunbar-Jacob J, Mortimer-Stephens MK. Treatment adherence in chronic disease. Journal of Clinical Epidemiology. 2001; 54: S57–60 (Review). 8. Donovan JL. Patient decision making. The missing ingredient in compliance research. International Journal of Technology Assessment in Health Care. 1995; 11: 443–55 (Review). 9. Vermeire E, Hearnshaw H, Van Royen P, Denekens J. Patient adherence to treatment: three decades of research. A comprehensive review. Journal of Clinical Pharmacy and Therapeutics. 2001; 26: 331–42 (Review).



10.



11.



 12.



13.



14.



15.



] 417



Major review of literature on factors affecting patient compliance. Royal Pharmaceutical Society of Great Britain. From compliance to concordance: towards shared goals in medicine taking. London: Royal Pharmaceutical Society of Great Britain, 1997. Lagnaoui R, Moore N, Fach J, Longy-Boursier M, Begaud B. Adverse drug reactions in a department of systemic diseases-oriented internal medicine: prevalence, incidence, direct cost and avoidability. European Journal of Clinical Pharmacology. 2000; 56: 181–6. Lazaron J, Pomeranz BH, Corey PN. Incidence of adverse drug reactions in hospitalized patients. A meta-analysis of prospective studies. Journal of the American Medical Association. 1998; 279: 1200–5. Weiss J, Krebs S, Hoffmann C, Werner U, Neubert A, Brune K et al. Survey of adverse drug reactions on a paediatric ward: a strategy for early and detailed detection. Pediatrics. 2002; 110: 254–7. Impicciatore P, Choonara I, Clarkson A, Provasid, Pandolfini C, Bonati M. Incidence of adverse drug reactions in paediatric in/out patients: a systematic review and metaanalysis of prospective studies. British Journal of Clinical Pharmacology. 2001; 52: 77–83. Beijer HJ, de Blaey CJ. Hospitalisations caused by adverse drug reactions (ADR): a meta-analysis of observational studies. Pharmacy World and Science. 2002; 24: 46–54.



34 Corticosteroids in otolaryngology NIELS MYGIND AND JENS THOMSEN



Pharmacodynamics Pharmacokinetics Side effects Allergic rhinitis Idiopathic rhinitis (perennial nonallergic noninfectious rhinitis) Rhinitis medicamentosa Infectious rhinitis (common cold) Nasal polyposis Hyposmia and anosmia Sinusitis Angioedema Adenoids Pharyngitis Epiglottitis Tonsillectomy



418 418 419 419 420 421 421 421 421 421 421 422 422 422 422



Mononucleosis Croup External otitis Secretory otitis media Acute otitis media Chronic otitis media Sudden deafness Menie`re’s disease Facial palsy Sarcoidosis Wegener’s granulomatosis Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



422 422 422 423 423 423 423 423 424 424 424 424 424 425 425



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words corticosteroids and the specific diseases cited throughout the text.



PHARMACODYNAMICS Corticosteroids (CS) are small, lipophilic molecules that readily diffuse across the cell membrane into the cytoplasm, where they bind to the glucocorticoid receptor.1 The CS–glucocorticoid receptor complex interacts with proteins that serve as transcription factors. This results in a reduced synthesis of inflammatory cytokines and subsequently in a reduced number and activity of inflammatory cells in inflamed tissue. Due to this mode of action, there is a time delay between the administration and the clinical activity of CS. However, studies have shown that the clinical effect of an intranasal CS starts as early as three hours after application.2 The density of CS molecules at receptor sites in the surface epithelium, lamina propria, blood vessels,



circulating cells and the bone marrow varies markedly between topically and systemically administered drugs. This difference in target cell accessibility to CS has an influence on the effect of topical versus systemic therapy. In the nose, for example, topical treatment is most effective on itching and sneezing, while systemic treatment is most effective on blockage and anosmia.



PHARMACOKINETICS Intravenous corticosteroids One hour is gained when a CS is given intravenously instead of orally. Methylprednisolone or dexamethasone are preferred, as they have a minimal minerallocorticoid



Chapter 34 Corticosteroids in otolaryngology



] 419



effect. When adrenal insufficiency is suspected in a patient on long-term oral treatment, hydrocortisone is the drug of choice.



two to three weeks are few and minor,3 unless treatment is given to patients with diabetes or glaucoma.



Oral corticosteroids



Intranasal corticosteroids



Prednisolone is, in principle, preferred to prednisone, which is an inactive prodrug, metabolized to prednisolone in the liver. Dexamethasone is more potent and has minimal minerallocorticoid effects.



When nasal CS is first sprayed into a hyperreactive nose, it can result in sneezing. Patients should be encouraged to continue treatment, as this symptom usually disappears after a few days.6 The condition is more a symptom of the disease than a true side effect. A sensation of dryness in the nostrils, associated with slight blood-stained crusts, is frequently encountered, but it does not seem to worsen during long-term treatment. Usually patients can continue therapy if they reduce the dosage, use the medication only once daily and use a neutral lotion in the nostril. The patient should be told to return for examination if frequent bleeding or crusting develops. Rarely, treatment has to be temporarily stopped because of epistaxis. In very rare cases, a septal perforation can develop during treatment.7 Studies have shown that long-term treatment with an intranasal CS has no effect on the growth rate in children, when the drug is given once in the morning in an ordinary recommended dose, i.e. half the adult dose.8, 9 However, one study has indicated a slight effect on growth, when a high adult dose is given twice daily.10 Long-term studies of inhaled CS have shown that although there might be a temporary growth inhibition, the children will reach normal final height.11 Consequently, the risk of a reduced final height from correct nasal medication seems negligible. Theoretically, even a small systemically active dose of CS, given for a prolonged period of time, may give an increased risk of osteoporosis and cataract. Although studies have not shown such adverse effects from intranasal CS treatment,12 a small risk cannot be completely excluded.



Intramuscular depot-injection of corticosteroids When a depot-injection, e.g. methylprednisolone acetate, is given the maximum CS effect on the hypothalamic– pituitary–adrenal (HPA)-axis, measured by its effect on the plasma cortisol level, occurs after three days and the effect lasts for up to three weeks.3 The CS dose in one injection of 80 mg methylprednisolone corresponds to about 100 mg prednisolone orally.



Intranasal corticosteroids When an aqueous spray is used, 50 percent is deposited in the nostril and in the nonciliated anterior part of the nose, while 50 percent will reach the ciliated mucous membrane, where it is either absorbed or removed by mucociliary clearance within 30 minutes.4 Intranasal CS are not deactivated locally and it has to be assumed that the active drug is also absorbed once it reaches the inflammatory cells. Thus, only the dosage which has been absorbed to the nasal mucosa will have an antiinflammatory effect. The measured bioavailability, demonstrated by routine tests, depends upon the CS molecule used. Highly lipophilic molecules, such as fluticasone propionate and mometasone furoate, have a large tissue distribution volume, while less lipophilic molecules, such as beclomethasone dipropionate and budesonide, are quickly absorbed into the circulation.5 Consequently, highly lipophilic drugs have a long elimination time, while the less lipophilic drugs have a shorter elimination time. Use of these drugs for 10–30 years has shown that both types of drug have a minimal systemic effect without a risk of clinical side effects, when used once daily in the morning at the recommended dosage.



Corticosteroid ear drops There is no risk of systemic side effects from this very small CS dose. It is unknown whether ear drops can reach the middle ear mucosa and whether this implies a risk of side effects, but it seems highly unlikely. Theoretically, long-term use may cause atrophy of the skin in the external ear canal.



SIDE EFFECTS Systemic corticosteroids The side effects from long-term systemic CS therapy are well known. However, side effects following treatment for



ALLERGIC RHINITIS As rhinitis is a disease confined to 0.1 percent of the body weight, local intranasal treatment is, in principle, preferable to systemic treatment.



420 ] PART 7 PHARMACOTHERAPEUTICS



Intranasal corticosteroids Since the introduction of beclomethasone dipropionate in 1973, a series of other molecules has been developed and marketed (flunisolide, budesonide, fluticasone propionate, mometasone furoate, triamcinolone acetonide). As these drugs do not differ significantly with regard to effect and risk of side effects, it is recommended to choose the cheapest drug. The drug is usually given as an aqueous spray. In some countries, budesonide is also available as a powder (Turbuhaler) and fluticasone propionate as nose drops. The clinical effect has been demonstrated in more than 100 placebo-controlled studies.13, 14, 15, 16 [****] The effect on the nasal symptoms is pronounced. It begins after a few hours and the maximum effect is achieved after several days. When treatment is stopped, the effect disappears over a number of days. The effect is equally good on all nasal symptoms (sneezing, rhinorrhea, blockage). A series of placebocontrolled studies has clearly shown intranasal CS to be more effective on nasal symptoms than oral and intranasal antihistamines, especially on nasal blockage.17, 18 [****] Treatment is usually started with a daily dose equivalent to 400 mg beclomethasone dipropionate in an adult. When control of symptoms has been achieved, dose reduction can be attempted. A study has shown that even 64 mg budesonide is more effective that placebo.2 In some patients, continuous treatment can be replaced by periodic treatment for two to four weeks.19 Some physicians are reluctant to prescribe topical nasal CS in children because of concern about long-term systemic and local adverse effects. However, used for 30 years, topical nasal CS have resulted in virtually no adverse reports. Consequently, these agents may be used in children with persistent moderate to severe allergic rhinitis, particularly when nasal blockage is a pronounced symptom. The dosage is half the adult dose or less and the medication is given once daily in the morning. In our opinion, allergic rhinitis in children is more often undertreated than overtreated. It is the task of the clinician to balance the minimal theoretical risk of side effects against the increased quality of life for a child who can breathe again through the nose and be rid of disturbing itching, sneezing and runny nose. Pregnancy is not a contraindication as intranasal CS treatment has not been associated with any teratogenic or adverse effects. However, as there are no adequate and well-controlled studies in pregnant women for any drug, treatment should be used only if the expected benefit justifies the potential risk to the foetus.



Systemic corticosteroids Treatment can be given to adults either orally or as an intramuscular injection of a slow-release formulation.



There is no reason to believe that these two therapies have different therapeutic indices. Surprisingly, there is only a single placebocontrolled study of oral CS (three different doses of methylprednisolone given three times daily). The study showed an effect on nasal blockage at 6 mg (equipotent to 7.5 mg prednisolone) and on blockage and rhinorrhoea at 24 mg (30 mg prednisolone).3 [***/**] Three placebo-controlled studies of a depotinjection have shown a marked effect on nasal blockage, lasting for more than four weeks, but merely a mild to moderate effect on sneezing and rhinorrhea.3 [***] The treatment can be added, when an intranasal CS is insufficient. However, there are no studies to demonstrate added efficacy of these two types of CS administration. [*] The optimum dose of oral CS is not known. The CS dose of a depot-injection (e.g. 80 mg methylprednisolone acetate) corresponds to 100 mg prednisolone. Systemic CS treatment should not be given to children, pregnant women or to patients with diabetes mellitus, severe osteoporosis, glaucoma or cataract. The risk of side effects from a short-term, two- to three-week treatment is very small and should be balanced against the benefits of therapy.



IDIOPATHIC RHINITIS (PERENNIAL NONALLERGIC NONINFECTIOUS RHINITIS) It is more difficult to treat perennial rhinitis than simple hay fever. Tarlo et al.20 found that only 54 percent of patients with perennial rhinitis achieved acceptable symptomatic improvement from a CS spray. However, after administration of short-term oral prednisone, 73 percent of the patients obtained moderate or marked symptomatic improvement from topical treatment. In severe cases, it would be advisable to conduct a therapeutic trial with short-term systemic CS, when the effect of nasal CS is insufficient. In patients suffering from idiopathic rhinitis, intranasal CS have shown a significant effect in most,20, 21, 22, 23 but not in all studies.24 The general impression, supported by a few small studies,25, 26 is that the treatment is particularly effective when nasal eosinophilia is present. However, a recent large study of idiopathic rhinitis has shown equal efficacy in patients with and without nasal eosinophilia.23 While the effect is marked in patients with eosinophildominated inflammation and intolerance to acetylsalicylic acid,27 the effect in other cases of idiopathic rhinitis cannot be predicted. It would be advisable to conduct a one-month therapeutic trial with intranasal CS in all patients with persistent idiopathic rhinitis. [***]



Chapter 34 Corticosteroids in otolaryngology



RHINITIS MEDICAMENTOSA A placebo-controlled study of topical CS in rhinitis medicamentosa has shown that patients on intranasal CS have a faster onset of symptom reduction and of improvement of mucosal swelling,28 indicating that an adequate treatment of such patients consists of a combination of vasoconstrictor withdrawal and a topical CS to alleviate the withdrawal symptoms. Another study has shown an effect of intranasal CS on oxymethazolineinduced rebound congestion.29 A short course of intranasal CS is recommended, when a vasoconstrictor is withdrawn. [***]



INFECTIOUS RHINITIS (COMMON COLD) The common cold is a virus-induced inflammation, characterized by neutrophil accumulation in the nasal mucous membrane. Theoretically, a clinical effect could be expected from an antiinflammatory CS. However, at least two placebo-controlled trials have failed to show any effect.30, 31 Thus, CS are highly effective in eosinophildominated inflammation and allergy, but not in neutrophil-dominated inflammation and infection. CS are not recommended for the treatment of common cold. [***]



NASAL POLYPOSIS The aetiology of nasal polyps is unknown. Eosinophils are commonly found in polyps, but the association with allergy appears to be coincidental. Topical CS treatment reduces nasal symptoms, polyp size and the number of recurrences after polypectomy. [****] A short course of systemic CS may help combat obstruction, improve intranasal spray distribution, improve the sense of smell and facilitate intranasal surgery. [***] Management of patients with severe polyposis is still unsatisfactory but it has been improved by the combination of long-term topical CS, short-term systemic CS and surgery, as described in Chapter 121, Nasal polyposis.



] 421



patients with nasal polyposis and when it is indicated to make a distinction between a potentially reversible (rhinitis, polyposis) and an irreversible (viral respiratory infection, toxic chemical exposure, head trauma) olfactory dysfunction. [***]



SINUSITIS Clinically it is difficult to define sinusitis, especially in those patients with chronic disease where a distinction needs to be made between the symptoms of rhinitis and those of sinusitis.33 Information received from studying the patient’s symptoms and the imaging examination may not agree. Based on a CT scan, a simple common cold is, as a rule, sinusitis, and patients with nasal polyposis invariably show sinusitis on imaging, whether they have sinusitis symptoms or not. In patients known to be steroid responders, an antirhinitis effect may masquerade as an antisinusitis effect. Therefore, the inclusion of some steroid responders (nasal polyposis, perennial eosinophilic rhinitis, acetylsalicylic acid intolerance) in sinusitis studies may give a false-positive result. While a few small studies of intranasal CS treatment of acute exacerbations in patients with chronic recurrent sinusitis have been negative,34, 35, 36, 37 two recent large placebo-controlled studies have been more positive, showing a significant effect on symptoms, but not on imaging.38, 39 At present, an unqualified recommendation for the use of intranasal CS treatment for chronic recurrent sinusitis cannot be made. However, a therapeutic trial for three months could be made in selected patients. [***] As a nasal spray will not reach the sinuses nor the middle meatus, with the openings to the paranasal sinuses, it seems possible that a short course of systemic CS may be effective, but placebo-controlled studies have not been performed. [*] In conclusion, many clinicians believe that intranasal CS are important in the treatment of sinusitis. However, only scant controlled studies support this conjuncture and the use of CS remains controversial.



HYPOSMIA AND ANOSMIA ANGIOEDEMA Hyposmia is a common finding in patients with allergic rhinitis, especially when the disease is chronic. Severe hyposmia or anosmia frequently occurs in nasal polyposis. While intranasal CS treatment has a beneficial effect on this symptom in allergic rhinitis, the effect in polyposis is poor. Systemic CS are invariably effective, but the effect is short-lasting.32 A short course of systemic CS can be used occasionally, when olfactory dysfunction impairs the quality of life in



Due to their antiinflammatory and antioedema effects, systemic CS, in conjunction with epinephrine and antihistmine, are generally used in the treatment of angioedema involving the mouth, throat and the upper respiratory tract40 and as such should be considered an acceptable treatment.41 However, this practice is not supported by any placebo-controlled study, which are unlikely to be undertaken. [**/*]



422 ] PART 7 PHARMACOTHERAPEUTICS



ADENOIDS



MONONUCLEOSIS



Lymphocytes are sensitive to CS and, by mucociliary clearance, a nasal spray will be carried to the adenoidal region. There has been only one placebo-controlled study of intranasal CS treatment of children with adenoidal hypertrophy. This study, involving only 17 children, showed a highly significant and long-lasting effect on both adenoid/choanal ration and symptom scores.42 Further studies are warranted before this treatment can be recommended. [***/**]



In upper airway obstruction due to infectious mononucleosis, it is tempting to resort to some antioedematous drug treatment. The use of CS in this disease is not investigated in any larger study, not to mention any metaanalysis. Tynell et al.48 demonstrated that acyclovir combined with prednisolone inhibited oropharyngeal Epstein–Barr virus (EBV) replication without affecting the duration of clinical symptoms or development of EBV-specific cellular immunity, while Brandfonbrener et al.49 found a more beneficial effect, although the results were not striking. Furthermore, there is one report of presumably CS-induced lethal disseminated intravascular coagulation and hepatic necrosis.50 We must conclude that the use of CS in EBV infection is still a moot point. [**/*]



PHARYNGITIS A single placebo-controlled study of an intramuscular CS injection has shown an improvement of pain score in adults with severe pharyngitis.43 However, in our opinion, this does not justify the use of CS in pharyngitis in general, but it may be tried in selected cases with severe and threatening oedema, compromising breathing. [***/**]



EPIGLOTTITIS There is a remarkable absence of documentation of the use of CS in acute epiglottitis. Due to the severity of this condition, placebo-controlled studies should not be carried out. However, such investigations should be undertaken, since it is common practice in intensive care units to use CS in patients with severe oedema. [*]



TONSILLECTOMY Tonsillectomy is one of the most commonly performed surgical procedures. The effect of a single preoperative intravenous dose of CS (dexamethasone, 0.15–1.0 mg/kg) has been studied in a number of placebo-controlled trials. A recent meta-analysis of randomized studies found that the effect is clinically relevant with regard to nausea, vomiting and intake of solid food on the first postoperative day.44 Given the frequency of tonsillectomy, the relative safety and low cost of a single injection of CS, and the reduction in postoperative morbidity, the authors recommend routine use of a single intravenous dose during paediatric tonsillectomy. In contrast to the above studies, Palme et al.45 concluded that a seven-day course of CS plays a limited role in patients’ recovery from tonsillectomy. We find that the use of CS in tonsillectomy is open for debate and reference should be made to an article expounding the merits of CS use46 and another against their use.47 [****]



CROUP Croup (acute laryngotracheobronchitis) is a frequent disease in children. It is characterized by hoarseness, a dry barking cough and inspiratory stridor. These symptoms occur as a result of oedema in the larynx and trachea due to a viral infection. The disease is usually benign and selflimiting, but it may be the cause of death due to sudden obstruction of the airways. There is ample evidence from randomized, placebocontrolled trials, summarized by a meta-analysis51 that CS are effective in croup. [****] Symptoms are improved as early as six hours and for up to at least 12 hours. There is no demonstrable effect after 24 hours. Dexamethasone (orally or intramuscularly) has been evaluated in 17 trials and nebulized budesonide in nine studies. In most studies, dexamethasone has been given as a single dose of 0.3 mg/kg. This is a high dose, corresponding to 150 mg prednisolone in an adult. However, the risk of side effects from a single dose is very small. In mild to moderate cases, it is preferable to start treatment at home with a palatable soluble CS (e.g. betamethasone tablets dissolved in a drink). In hospitalized patients, treatment is usually given as an intramuscular injection of dexamethasone (0.3 mg/kg). If available, nebulized budesonide can be used instead (2 mg).



EXTERNAL OTITIS The symptoms of external otitis vary considerably based on whether the otitis is caused by slight eczema or severe inflammation caused by a bacterial infection. Consequently, the standard treatment is either CS alone, antimicrobial agents alone or a combination of the two. Considering the widespread use of CS ear drops, there is a remarkable absence of placebo-controlled studies. Only one study52 has demonstrated a significant effect



Chapter 34 Corticosteroids in otolaryngology



of budesonide compared to placebo. This CS molecule is, however, not marketed as ear drops. Several animal studies have been published demonstrating the availability of an animal model for studying external otitis.53, 54 In these studies the potency of the CS is discussed and it is most likely that the commonly used CS, hydrocortisone, is too weak to have the desired effect. Further controlled studies are needed. [**] It is strange that perforation in the drum is listed as a contraindication for using ear drops containing CS. This must be a misconception, since litres of steroidcontaining ear drops have been poured into middle ears without causing any inner ear damage, and we believe this warning is not based on solid evidence.



] 423



CHRONIC OTITIS MEDIA As with acute otitis media, evidence for supporting the notion that CS have a place in the treatment in chronic suppurative otitis media is scarce. Using monkeys inoculated with Pseudomonas aeruginosa in the middle ear, it was found that dexamethasone enhanced the efficacy of tobramycin,64 whereas a 0.2 percent solution of ciprofloxacin was as effective as a combination preparation with CS.65 Further placebo-controlled studies are needed to shed light on this issue, and for the present time CS treatment in suppurative otitis media is not recommended. [**/*]



SUDDEN DEAFNESS SECRETORY OTITIS MEDIA secretory otitis media (SOM) is one of the most frequent diseases in childhood. The demonstration of or the rejection of the efficacy of CS is therefore of utmost importance. It is likely that the addition of systemic CS to an antibiotic treatment has a beneficial effect upon the status of the middle ear.55, 56, 57, 58 On the other hand, Lambert59 and Macknin and Jones60 could not demonstrate any effect of either prednisone or dexamethasone. In a Cochrane review,61 it was concluded that there is evidence that CS combined with an antibiotic leads to a faster resolution of SOM in the short term. However, there is no evidence for a long-term benefit from treating hearing loss-associated SOM with either oral or topical nasal CS. These treatments are therefore not recommended at the present time. [***] Considering the possibility that SOM is indeed a bacterial infection, it cannot be ignored that CS inhibit the development of endotoxin-induced otitis media with effusion as demonstrated by Baggett et al.62 This adds to the confusion, but stresses the need for additional controlled studies.



ACUTE OTITIS MEDIA From a theoretical point of view, only systemic CS administration can be taken into consideration. In children with tympanostomy tubes, who experience a bout of acute otitis media, with discharge through the tubes, a borderline effect of oral prednisolone has been suggested.63 Apart from this report, there is no evidence supporting the notion that CS are of any value in the treatment of this disease entity, and such treatment is not to be recommended. [**/*]



Sudden deafness or idiopathic sudden sensorineural hearing loss (ISSNHL) is an abrupt sensorineural hearing loss of at least 30 dB in at least three contiguous audiometric frequencies, variably defined as developing in 12 hours or less,66 24 hours or less67 or in 3 days or less.68 In making the diagnosis of ISSNHL, identifiable causes of hearing loss should be excluded and the patient should not have a history of previous fluctuations of hearing. The most commonly proposed hypotheses for the aetiology of ISSNHL include disturbance of labyrinthine blood circulation, subclinical viral labyrinthitis, spontaneous labyrinthine membrane rupture and autoimmune pathogenesis.67 Among the postulated aetiologies of ISSNHL, autoimmunity and perhaps viral infection may be ameliorated by steroids. The documentation for an effect of CS is poor. A number of publications claim a beneficial effect of CS,69, 70, 71, 72 while others have come to the opposite conclusion.73, 74, 75, 76 A Medline search77 came to the conclusion that no concensus existed on the effective treatment of ISSNHL. At this point we will refrain from recommending systemic use of CS in patients with ISSNHL. [***/**] However, recent developments have rendered it possible to deliver CS directly at the round window membrane, resulting in a much higher concentration of CS in the inner ear fluids. Anecdotal reports show that such delivery results in hearing improvement in patients who had no benefit from systemic CS prior to the intratympanic, round window application.78



MENIE`RE’S DISEASE Treatment of Menie`re’s disease is hampered by the erratic course of the disease. Oral CS is one treatment that is heralded as being effective. To date, there are no existing controlled studies proving that systemic CS have any effect upon the course of the disease. In recent years, intratympanic CS application has been suggested as an



424 ] PART 7 PHARMACOTHERAPEUTICS effective treatment; however, the only existing randomized double-blind study79 concludes that intratympanic administration of dexamethasone showed no benefit over placebo. New delivery systems are now available, resulting in a more specific delivery of the drug to the inner ear. While there are only anecdotal reports of the efficacy, we cannot rule out a possible effect. However, at the present time we cannot recommend either systemic or intratympanic use of CS in Menie`re’s disease. [**/*]



FACIAL PALSY Bell’s palsy, also known as idiopathic facial nerve paralysis (IFNP) is the most common cause of facial paralysis. IFNP results in facial motor dysfunction, the degree of which ranges from minor weakness to complete paralysis. Treatment is aimed at improving recovery of facial function and the prevention of neural degeneration and its associated complications. Of the multiple treatment modalities evaluated over the past three decades, CS therapy has become the most widely accepted. However, despite the widespread clinical acceptance of CS therapy for IFNP, the efficacy of this therapy has not been clearly demonstrated in the literature. In this age of meta-analysis and Cochrane database systemic reviews, the confusion about the efficacy of CS in the treatment of INFP is striking. A meta-analysis including 47 trials80 came to the conclusion that there is a statistically significant benefit for complete IFNP patients treated with a total prednisone dose equivalent to 400 mg or more started within seven days of onset of paresis. It is, however, remarkable among the 47 trials used to reach this conclusion that only three met the criteria for a randomized controlled trial. In contrast to this meta-analysis, a Cochrane database systemic review81 came to the opposite conclusion that the available evidence from randomized controlled trials does not show significant benefit from treating Bell’s palsy with CS. More randomized controlled trials with a greater number of patients are needed to determine reliably whether there is real benefit (or harm) from the use of CS therapy in patients with Bell’s palsy. Also in this review, only three trials met the requirements of a randomized controlled trial. These trials were not the same as the trials used in the meta-analysis of Ramsey et al.80 We must conclude that using CS for IFNP is still a moot point and further placebo-controlled, blinded randomized studies are required before a recommendation for CS therapy can be made. [***]



manifestation.82 In the absence of controlled trials, the efficacy of CS needs to be deduced from pulmonary studies. [**] Only five randomized controlled trials of CS treatment (prednisolone 15–40 mg/day) for pulmonary sarcoidosis were identified in a meta-analysis.83 Oral steroids improved the chest x-ray, symptoms and lung function (diffusion capacity) over 6–24 months. There are no data beyond two years. In a consensus report,84 it was concluded that oral CS is an effective short-term therapy, but it is not known whether the treatment alters the natural history of the disease. It is recommended to start with prednisolone 20–40 mg/day. Among steroid responders, the dose is slowly tapered to 5–10 mg/day. Treatment should be continued for a minimum of 12 months. Oral CS will probably relieve the discomfort of nasosinal sarcoidosis, but the symptoms need to be severe in order to balance the expected side effects from long-term treatment.



WEGENER’S GRANULOMATOSIS Combined therapy with cyclophosphamide and systemic CS for remission induction in Wegener’s granulomatosis is a well-established and generally recommended treatment.85 [*] Placebo-controlled trials have not been performed.



KEY POINTS  Corticosteroids are widely used in otolaryngology and especially in rhinology.  Topical nasal steroids are safe and effective.  The risk of side effects from short-term systemic steroids is small.



Best clinical practice [ Allergic rhinitis, idiopathic rhinitis and nasal polyposis



[ [



SARCOIDOSIS Nasosinal involvement is rare in patients with sarcoidosis and exceptionally rare as an isolated disease



[ [



are best treated with an intranasal CS, when necessary with a supplement of a short course of systemic CS. A short course of systemic CS can be tried in anosmia. The role of intranasal CS in chronic sinusitis remains controversial. Systemic CS are often used for angioedema and epiglottitis, but the effect is unproven. CS for adenoids and for pharyngitis cannot be recommended at present.



Chapter 34 Corticosteroids in otolaryngology



[ The use of a short course of systemic CS in [ [



[ [



[



tonsillectomy is open for debate. A single dosage of systemic CS has a proven effect in croup and prompt treatment, either orally or intravenously, is recommended. CS ear drops alone or in combination with an antibiotic has a well-established, but poorly documented, place in the treatment of external otitis. A short course of systemic CS for otitis media with effusion may be helpful, but a recommendation awaits further controlled studies. At present, CS are not to be recommended for acute otitis media, chronic otitis media, sudden deafness, Menie`re’s disease and facial palsy. Systemic CS must be used for Wegener’s granulomatosis in spite of a lack of controlled evidence.



Deficiencies in current knowledge and areas for future research



$ $ $ $ $ $ $ $ $



Systemic CS are often used for a variety of diseases. However, the treatment is often not evidence based and there is a striking lack of placebo-controlled studies and in particular of dose–effect studies. In severe allergic rhinitis, a short course of systemic CS is often used, but the optimal dosage has not been defined. Intranasal CS are effective in some, but not all, cases of idiopathic rhinitis. Could responsiveness be predicted by local eosinophila or by another inflammatory marker? The role of intranasal CS in subgroups of chronic sinusitis needs to be defined by placebo-controlled studies. Are systemic CS effective? Placebo-controlled studies of intranasal CS treatment for enlarged adenoids in children are warranted. There are no placebo-controlled studies of the effect of systemic CS in angioedema, in epiglottitis and in mononucleosis. We need studies of whether more than a single dose of systemic CS is indicated in croup. CS ear drops alone or in combination with an antibiotic for external otitis media are remarkably poorly documented. A study with one formulation used in one ear and placebo/another formulation in the other ear would be easy to perform. The usefulness of intranasal and of systemic CS for otitis media with effusion needs a detailed analysis in placebo-controlled studies.



] 425



REFERENCES 1. Mygind N, Nielsen LP, Hoffman H-J, Shukla A, Blumberga G, Jacobi H. Mode of action of intranasal corticosteroids. Journal of Allergy and Clinical Immunology. 2001; 108: S16–25. 2. Day JH, Briscoe MP, Rafeiro E, Ellis AK, Pettersson E, A˚kerlund A. Onset of action of intranasal budesonide (Rhinocort Aqua) in seasonal allergic rhinitis in a controlled exposure model. Journal of Allergy and Clinical Immunology. 2000; 105: 489–94. 3. Mygind N, Laursen LC, Dahl M. Systemic corticosteroid treatment for seasonal allergic rhinitis: a common but poorly documented therapy. Allergy. 2000; 55: 11–5. 4. Newman SP, More´n F, Clarke SW. Deposition pattern of nasal sprays in man. Rhinology. 1988; 26: 111–20. 5. Szefler SJ. Pharmakokinetics of intranasal corticosteroids. Journal of Allergy and Clinical Immunology. 2001; 108: S26–31. 6. Toft A, Whil J-A˚, Taxman J, Mygind N. Double-blind comparison between beclomethasone dipropionate as aerosol and as powder in patients with nasal polyposis. Clinical Allergy. 1982; 12: 391–401. 7. Cervin A, Andersson M. Intranasal steroids and septum perforation – an overlooked complication. Rhinology. 1998; 36: 128–32. 8. Scenkel EJ, Skinner DP, Bronsky EA, Miller D, Pearlman DS, Rooklin A. et al. Absence of growth retardation in children with perennial allergic rhinitis after one year of treatment with mometasone furoate nasal spray. Pediatrics. 2000; 105: E22. 9. Mo¨ller C, Ahlstro¨m H, Henricson K-A˚, Malmqvist L-A˚, A˚kerlund A, Hildebrand H. Safety of nasal budesonide in the long-term treatment of children with perennial rhinitis. Clinical and Experimental Allergy. 2003; 33: 816–22. 10. Skoner DP, Rachelefsky GS, Meltzer EO, Chervinsky P, Morris RM, Storms WW et al. Detection of growth suppression in children during treatment with intranasal beclomethasone dipropionate. Pediatrics. 2000; 105: E23. 11. Agertoft L, Pedersen S. Effect of long-term treatment with inhaled budesonide on adult height in children with asthma. New England Journal of Medicine. 2000; 343: 1064–9. 12. Derby L, Maier WS. Risk of cataract among users of intranasal corticosteroids. Journal of Allergy and Clinical Immunology. 2000; 105: 912–6. 13. Lund VJ, Aaronson D, Bousquet J, Durham SR, Mygind N et al. International consensus report on the diagnosis and management of rhinitis. Allergy. 1994; 49: 1–34. 14. van Cauwenberge P, Bachert C, Passalacqua G, Bousquet J, Mygind N, Durham SR et al. Consensus statement on the treatment of allergic rhinitis. Allergy. 2000; 55: 116–34. 15. Bousquet J, van Cauwenberge P, Khaltaev N. Aria Workshop Group and the World Health Organization. Allergic rhinitis and its impact on asthma. Journal of Allergy and Clinical Immunology. 2001; 108: S147–333.



426 ] PART 7 PHARMACOTHERAPEUTICS 16. Nielsen LP, Mygind N, Dahl R. Intranasal corticosteroids for allergic rhinitis. Drugs. 2001; 61: 1563–79. 17. Stempel DA, Thomas M. Treatment of allergic rhinitis: an evidence-based evaluation of nasal corticosteroids versus nonsedating antihistamines. American Journal of Management Care. 1998; 4: 89–96. 18. Weiner JM, Abramson MJ, Puy RM. Intranasal corticosteroids versus oral H1 receptor antagonists in allergic rhinitis: systemic review of randomised controlled trials. British Medical Journal. 1998; 317: 624–9. 19. Rinne J, Simola M, Malmberg H, Haathela T. Early treatment of perennial rhinitis with budesonide or cetirizine and its effect on long-term outcome. Journal of Allergy and Clinical Immunology. 2002; 109: 426–32. 20. Tarlo SM, Cockroft DW, Dolovich J, Hargreave FE. Beclomethasone dipropionate aerosol in perennial rhinitis. Journal of Allergy and Clinical Immunology. 1977; 59: 232–6. 21. Hansen I, Mygind N. Local effect of intranasal beclomethasone dipropionate aerosol in perennial rhinitis. Acta Allergologica. 1974; 29: 281–7. 22. Lindkvist N, Balle V, Karma P, Karja K, Lindstrom D, Makinen J et al. Long term safety and efficacy of budesonide nasal aerosol in perennial rhinitis – 12 month multicenter study. Allergy. 1986; 41: 179–86. 23. Webb DR, Meltzer EO, Finn Jr AF, Rickard KA, Pepsin PJ, Westlund R et al. Intranasal fluticasone propionate is effective for perennial nonallergic rhinitis with and without eosinophilia. Annals of Allergy, Asthma and Immunology. 2002; 88: 385–90. 24. Lundblad L, Sipila¨ P, Farstad T, Drozdziewicz D. Mometasone furoate nasal spray in the treatment of perennial non-allergic rhinitis: a Nordic multicenter, randomised, double-blind, placebo-controlled study. Acta Oto-laryngologica. 2001; 121: 505–9. 25. Small P, Black M, Frenkiel S. Effect of treatment with beclomethasone dipropionate in subpopulations of perennial rhinitis patients. Journal of Allergy and Clinical Immunology. 1982; 70: 78–82. 26. Feiss G, Welch M, Meltzer E, Alderfer V, Smith J. The predictive value of nasal eosinophilia for therapeutic response to intranasal corticosteroid treatment in perennial allergic rhinitis. Journal of Allergy and Clinical Immunology. 1992; 89: 209 (Abstr.). 27. Mastalerz L, Milewski M, Duplago M, Nizankowska E, Szczeklik A. Intranasal fluticasone propionate for chronic eosinophilic rhinitis in patients with aspirin-induced asthma. Allergy. 1997; 52: 895–900. 28. Halle´n H, Enerdal J, Graf P. Fluticasone propionate nasal spray is more effective and has a faster onset of action than placebo in treatment of rhinitis medicamentosa. Clinical and Experimental Allergy. 1997; 27: 552–8. 29. Ferguson BJ, Paramaesvaran S, Rubinstein E. A study of the effect of nasal steroid sprays in perennial allergic rhinitis patients with rhinitis medicamentosa. Otolaryngology and Head and Neck Surgery. 2001; 125: 253–60.



30. Farr BM, Gwaltney Jr JM, Hendley JO, Hayden FG, Naclerio RM, McBride T et al. A randomised controlled trial of glucocorticoid prophylaxis against experimental rhinovirus infection. Journal of Infectious Diseases. 1990; 162: 1173–7. 31. Puhakka T, Makela MJ, Malmstro¨m K, Uhari M, Savolainen J, Terho EO et al. The common cold: effect of intranasal fluticasone propionate treatment. Journal of Allergy and Clinical Immunology. 1998; 101: 726–31. 32. Seiden AM, Duncan HJ. The diagnosis of a conductive olfactory loss. Laryngoscope. 2001; 111: 9–14. 33. Kaliner MA, Osguthorpe JD, Fireman P, Anon J, Georgitis J, Davis ML et al. Sinusitis: bench to bedside. Journal of Allergy and Clinical Immunology. 1997; 99: S829–48. 34. Qvarnberg Y, Kantola O, Salo J, Toivanen M, Valtonen H, Vuori E. Influence of topical steroid treatment on maxillary sinusitis. Rhinology. 1992; 30: 103–12. 35. Meltzer EO, Busse WW, Druce HM, Metzger WJ, Mitchell DO, Selner J et al. Assessment of flunisolide nasal spray versus placebo as an adjunct to antibiotic treatment of sinusitis. Journal of Allergy and Clinical Immunology. 1992; 89: 301. 36. Meltzer EO, Orgel HA, Backhaus JW, Busse WW, Druce HM, Metzger WJ et al. Intranasal flunisolide spray as an adjunct to oral antibiotic therapy for sinusitis. Journal of Allergy and Clinical Immunology. 1993; 92: 812–23. 37. Parikh A, Scadding GK, Darby Y, Baker RC. Topical corticosteroids in chronic rhinosinusitis. Rhinology. 2001; 39: 75–9. 38. Meltzer EO, Charous BL, Busse WW, Zinreich SJ, Lorber RR, Danzig MR. Added relief in the treatment of acute recurrent sinusitis with adjunctive mometasone furoate nasal spray. Journal of Allergy and Clinical Immunology. 2000; 106: 630–7. 39. Dolor RJ, Witsell DL, Hellkamp AS, Williams Jr JW, Califf RM, Simel DL. Comparison of cefuroxime with and without intranasal fluticasone propionate for the treatment of rhinosinusitis. Journal of the American Medical Association. 2001; 286: 3097–105. 40. Kaplan AP. Chronic urticaria and angioedema. New England Journal of Medicine. 2002; 346: 175–9. 41. Shah UK, Jacobs IN. Pediatric angioedema: ten years’ experience. Arch Otolaryngology and Head and Neck Surgery. 1999; 125: 791–5. 42. Demain J, Goetz DW. Pediatric adenoidal hypertrophy and nasal obstruction: reduction with aqueous nasal beclomethasone. Pediatrics. 1995; 95: 355–64. 43. O’Brian JF, Falk JL. Dexamethasone as adjuvant for severe acute pharyngitis. Annals of Emergency Medicine. 1993; 22: 212–5. 44. Steward DL, Welge JA, Myer CM. Do steroids reduce morbidity of tonsillectomy? Meta-analysis of randomised trials. Laryngoscope. 2001; 111: 1712–8. 45. Palme CE, Tomasevic P, Pohl DV. Evaluating the effects of oral prednisolone on recovery after tonsillectomy: a prospective double blind randomised trial. Laryngoscope. 2000; 110: 2000–4.



Chapter 34 Corticosteroids in otolaryngology 46. Heathley DG. Perioperative intravenous steroid treatment and tonsillectomy. Archives of Otolaryngology and Head and Neck Surgery. 2001; 127: 1007–8. 47. Shott SR. Tonsillectomy and postoperative vomiting. Do steroids really work? Archives of Otolaryngology and Head and Neck Surgery. 2001; 127: 1009–10. 48. Tynell E, Aurelius E, Brandell A, Julander I, Wood M, Yao QY et al. Acyclovir and prednisolone treatment of acute infectious mononucleosis: a multicenter, double-blind, placebo-controlled study. Journal of Infectious Diseases. 1996; 174: 324–31. 49. Brandfonbrener A, Epstein A, Wu S, Phair J. Corticoid treatment in Epstein–Barr virus infection. Effect on lymphocyte class, subset, and response to early antigen. Archives of Internal Medicine. 1986; 146: 337–9. 50. Shane SA, Wollman M, Claassen D. Herpes simplex dissemination following glucocorticoids for upper airway obstruction in an adolescent girl. Pediatric Emergency Care. 1994; 10: 160–2. 51. Ausejo M, Saenz A, Pham BK, Kellner JD, Johnson DW, Moher D. The effectiveness of glucocorticosteroids in treating croup: metaanalysis. British Medical Journal. 1999; 341: 595–600. 52. Jacobsson S, Karlsson G, Rigner P, Sanner E, Schrewlius C. Clinical efficacy of budesonide in the treatment of eczematous external otitis. European Archives of Otorhinolaryngology. 1991; 248: 246–9. 53. Engard P, Hellstrom S. An animal model for external otitis. European Archives of Otorhinolaryngology. 1997; 254: 115–9. 54. Engard P, Hellstrom S. A topical steroid without an antibiotic cures external otitis efficiently: a study in an animal model. European Archives of Otorhinolaryngology. 2001; 258: 287–91. 55. Tracy JM, Demain JG, Hoffman KM, Goetz DW. Intranasal beclomethasone as an adjunct to treatment of chronic middle ear effusion. Annals of Allergy, Asthma and Immunology. 1998; 80: 198–206. 56. Rosenfeld RM, Mandel EM, Bluestone CD. Systemic steroids for otitis media with effusion in children. Archives of Otolaryngology and Head and Neck Surgery. 1992; 117: 984–9. 57. Hemlin C, Carenfelt C, Papatziamos G. Single dose of betamethasone in combined medical treatment of secretory media. Annals of Otology, Rhinology, and Laryngology. 1997; 106: 359–63. 58. Zocconi E. Antibiotics and contisone in the treatment of otitis media with effusion. Pediatria Medica e Chirurgica. 1994; 16: 273–5. 59. Lambert PR. Oral steroid therapy for chronic middle ear persusion: a double-blind crossover study. Otolaryngology – Head and Neck Surgery. 1986; 95: 193–9. 60. Macknin ML, Jones PK. Oral dexamethasone for treatment of persistent middle ear effusion. Pediatrics. 1985; 75: 329–35. 61. Butler CC, van Der Voort JH. Oral or topical nasal steroids for hearing loss associated with otitis media with effusion



62.



63.



64.



65.



66.



67.



68.



69.



70.



71.



72.



73.



74.



75.



] 427



in children. Cochrane Database of Systematic Reviews 2000: CD001935. Baggett HZ, Prazma J, Rose AS, Llane AP, Pillsbury 3rd HC. The role of glucocorticoids in endotoxin-mediated otitis media with effusion. Archives of Otolaryngology and Head and Neck Surgery. 1997; 123: 41–6. Ruohola A, Heikkinen T, Jero J, Puhakka T, Juven T, NarkioMakel M et al. Oral prednisolone is an effective adjuvant therapy for acute otitis media with discharge through tympanostomy tubes. Journal of Pediatrics. 1999; 134: 459–63. Alper CM, Dohar JE, Gulhan M, Ozunlu A, Bagger-Sjobak D, Hebd PA et al. Treatment of chronic suppurative otitis media with topical tobramycin and dexamethasone. Archives of Otolaryngology and Head and Neck Surgery. 2000; 126: 165–73. Miro N. Controlled multicenter study on chronic suppurative otitis media treated with topical applications of ciprofloxacin 0.2 percent solution in single-dose containers or combination of polymyxin B, neomycin, and hydrocortisone suspension. Otolaryngology and Head and Neck Surgery. 2000; 123: 617–23. Mattox DE, Lyles A. Idiopathic sudden sensorineural hearing loss. American Journal of Otology. 1989; 10: 242–7. Stokroos RF. Summary and conclusions. In: Stokroos RJ (ed.). Idiopathic sudden sensorineural hearing loss. Groningen: Van Denderen BV, 1997: 99–104. Wilson WR, Byl FM, Laird N. The efficacy of steroids in the treatment of idiopathic sudden hearing loss. Archives of Otolaryngology. 1980; 106: 772–6. Alexiou C, Arnold W, Fauser C, Schratzenstaller B, Gloddek B, Fuhrmann S et al. Sudden sensorineural hearing loss: does application of glucocorticosteroids make sense? Archives of Otolaryngology and Head and Neck Surgery. 2001; 127: 253–8. Chandrasekhar SS. Intratympanic dexamethasone for sudden sensorineural hearing loss: clinical and laboratory evaluation. Otology and Neurotology. 2001; 22: 18–23. Parnes LS, Sun AH, Freeman DJ. Cortisteroid pharmacokinetics in the inner ear fluids: an animal study followed by clinical evaluation. Laryngoscope. 1999; 109: 1–17. Moskowitz D, Lee KJ, Smith HW. Steroid use in sensorineural hearing loss. Laryngoscope. 1984; 94: 664–6. Kitajiri S, Tabuchi K, Hiraumi H, Hirose T. Is corticosteroid therapy effective for sudden-onset sensorineural hearing loss at lower frequencies? Archives of Otolaryngology and Head and Neck Surgery. 2002; 128: 365–7. Cinamon U, Bendet E, Kronenberg J. Steroids, carbogen or placebo for sudden hearing loss: a prospective doubleblind study. European Archives of Otorhinolaryngology. 2001; 258: 477–80. Minoda R, Masuyama K, Habu K, Yumoto E. Initial steroid hormone dose in treatment of idiopathic sudden deafness. American Journal of Otology. 2000; 21: 819–25.



428 ] PART 7 PHARMACOTHERAPEUTICS 76. Wilkins SA, Mattox DE, Lyles A. Evaluation of a ‘shotgun’ regimen for sudden hearing loss. Otolaryngology and Head and Neck Surgery. 1987; 97: 474–80. 77. Haberkamp TJ, Tanyeri HM. Management of idiopathic sudden sensorineural hearing loss. American Journal of Otology. 1999; 20: 587–92. 78. Kopke RD, Hoffer ME, Wester D, O’Leary MJ, Jackson RL. Targeted topical steroid therapy in sudden sensorineural hearing loss. Otology and Neurotology. 2001; 22: 475–9. 79. Silverstein H, Isaacson JE, Olds MJ, Rowan PT, Rosenberg S. Dexamethasone inner ear perfusion for the treatment of Meniere’s disease: a prospective, randomized, doubleblind, cross-over trial. American Journal of Otology. 1998; 19: 196–201. 80. Ramsey MJ, DerSimonian R, Holtel MR, Burgess LPA. Corticosteroid treatment for idiopathic facial nerve paralysis: a meta-analysis. Laryngoscope. 2000; 110: 335–41.



81. Salinas RS, Alvarez G, Alvarez MI, Ferreira J. Corticosteroids for Bell’s palsy (idiopathic facial paralysis). Cochrane Database of Systematic Reviews 2002: CD001942. 82. Wilson R, Lund V, Sweatman M, Mackay IS, Mitchell DN. Upper respiratory tract involvement in sarcoidosis and its management. European Respiratory Journal. 1988; 1: 269–72. 83. Paramothayan NS, Jones PW. Corticosteroids for pulmonary sarcoidosis. Cochrane Database of Systematic Reviews. 2000: CD001114. 84. Costabel U, Hunninghake GW. ATS/WASOG statement on sarcoidosis. European Respiratory Journal. 1999; 14: 735–7. 85. Langford CA, Sneller MC. Update on the diagnosis and treatment of Wegener’s granulomatosis. Advances in Internal Medicine. 2001; 46: 177–206.



35 Drug therapy in otology WENDY SMITH AND MARTIN BURTON



Introduction Topical ear preparations Preparations used in the management of vertigo Drugs used in sudden sensorineural hearing loss Sodium fluoride treatment in otosclerosis



429 429 433 434 434



Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



434 435 435 435



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words otitis externa, ear drops, antibiotics, betahistine, gentamicin, Menie`res, acyclovir and sodium fluoride.



INTRODUCTION Medicines are commonly prescribed in the management of otological diseases. Whilst the efficacy of some treatments may be uncertain, it is important to understand how these drugs act, their indications, contraindications and side effects. This chapter discusses these factors in relation to a number of specific types of otological medications.



TOPICAL EAR PREPARATIONS Eardrops are solutions or suspensions of medicaments in water, glycerol, diluted alcohol, propylene glycol or other suitable solvent for instillation into the ear. A solution comprises a solute (drug) dissolved in the solvent whereas a suspension consists of an insoluble drug distributed in a liquid. Some preparations used in otology are also used as eye or nose drops. Since these drops are in multipleapplication containers, the vehicle contains a bactericidal and fungicidal agent such as benzalkonium chloride (0.01 percent). Other adjuvants in ear and eye drops include buffers such as sodium metabisulphite and disodium



edetate. The buffers are used to maintain the pH to minimize breakdown of the active constituents or to increase comfort for the patient. Both sodium metabisulphite and disodium edetate are effective at retarding oxidation reactions and the latter can enhance the bactericidal activity of benzalkonium chloride and chlorhexidine acetate. Information on excipients contained in drops is found in the British National Formulary or its equivalent. This information is useful since some patients may be allergic to topical ear medication and this allergy may not be to the primary constituent but to the excipients. Ear drops are dispensed in coloured fluted glass bottles with a plastic screw cap incorporating a glass dropper tube and rubber teat, or, more commonly, in plastic squeeze bottles fitted with a plastic cap incorporating a dropper device. Such containers are designed to prevent light degradation of the contents. Some ear preparations are available as creams or ointments where the antimicrobial and/or antiinflammatory substance is prepared in a suitable base such as liquid paraffin, wool fat and yellow soft paraffin. Both are semisolid preparations but in a cream the base is absorbed into the skin, whereas in an ointment only the medication



430 ] PART 7 PHARMACOTHERAPEUTICS is absorbed from the greasy base. In general, ointments are useful in dry scaly conditions whereas creams can also be used in weeping skin conditions. Again, warnings that the preparation contains lanolin (wool fat) are useful since some patients are allergic to lanolin.



Topical ear preparations for inflammatory and infective conditions INDICATIONS



A number of topical ear preparations in the form of drops or ointments are available for the treatment of otitis externa and discharging mastoid cavities. Many are also used in discharging ears with a perforated tympanic membrane or ears with a grommet in situ. In this latter situation this is an unlicensed indication for the use of many of these preparations. Manufacturers’ data sheets continue to state that even plain steroid drops are contraindicated in the presence of a perforation. In chronic suppurative otitis media, it has been shown that aural toilet and topical antibiotics, especially a quinolone, is effective in resolving otorrhoea and eradicating bacteria from the middle ear. Topical antibiotic preparations are more effective than oral preparations and the addition of oral therapy to topical antibiotics confers no greater benefit than the latter used alone (see Chapter 34, Corticosteroids in otolaryngology).1 SPECIFIC PREPARATIONS



Acetic acid 2 percent has antifungal and antibacterial properties and can be used to treat mild otitis externa. Aluminium acetate is an astringent that can be applied as drops or onto a gauze wick. An astringent is an agent which causes shrinkage or constriction and is usually applied topically. The hydroscopic effect reduces oedema in the inflamed ear canal, opening the meatus. Aluminium acetate has a tendency to form crystals in the ear. Regular aural toilette is required to remove both the crystals that form as well as the debris produced by the inflammatory process. This treatment can be safely used in pregnancy. Boric acid has been used in the past for its weak fungistatic and bacteriostatic activity and is used as a mild disinfectant in lotions, ointments and powders in concentrations of up to 5 percent. It is absorbed through damaged skin and may cause systemic toxicity. Acute and chronic toxicity can occur, presenting with gastrointestinal disturbance, rash, central nervous system and renal involvement that may result in death. Slow excretion of boric acid can lead to cumulative toxicity during repeated use. Some preparations contain only a steroid and are used in eczematous otitis externa. The steroid reduces inflammatory swelling and helps control irritation. The steroid stimulates the synthesis of lipocortin in leukocytes. This protein inhibits phospholipase A2 that reduces the



formation of arachidonic acid, the precursor of many inflammatory mediators. Betamethasone sodium phosphate (Betnesols) and prednisolone sodium phosphate (Predsols) are steroids available as ear drops. These steroids are combined with anti-infective agents for use in the management of infected otitis externa. Other steroids found in combination with anti-infective drugs include dexamethasone (Otomizes – with neomycin and glacial acetic acid, Sofradexs – with framycetin sulphate and gramicidin), flumetasone pivalate (Locorten-Vioforms – with flumetasone pivolate and clioquinol), hydrocortisone acetate (Gentisone HCs – with gentamicin, NeoCortefs – with neomycin sulphate) and triamcinolone acetonide (Audicorts – with neomycin undecanoate and Tri-Adcortyl Otics – with gramicidin, neomycin and nystatin). Clioquinol (found in Locorten-Vioform) is an 8-hydroxyquinolone with both antibacterial and antifungal properties and is useful when a mixed infection is suspected. The other antibiotics included in ear preparations are framycetin sulphate, gentamicin and neomycin sulphate which are all aminoglycosides. They are bactericidal, inhibiting microbial protein synthesis and are effective against aerobic Gram-negative bacteria. Bacteria may acquire resistance via plasmids against one aminoglycoside but these bacteria rarely exhibit resistance to other aminoglycosides. This accounts for the benefit in changing the ear drops if the otitis externa fails to respond to one preparation. Some of these preparations are available as ear sprays (Otomize) and there is some evidence that these cover the external meatus more effectively than traditional ear drops2 and are easier for the elderly to apply. However, in some clinical situations, drops are more likely to reach the infected and inflamed parts of the external ear, middle ear or mastoid cavity at which treatment is aimed. Prolonged use of antibiotic/steroid preparations may sensitize the skin and may lead to fungal infections. Clotrimazole is an azole derivative with a broadspectrum antifungal activity that acts by inhibiting ergosterol synthesis in the fungal cell membrane. At lower concentrations, clotrimazole merely inhibits fungal growth but at higher concentrations, fungi are killed by clotrimazole action causing direct membrane damage. Increasingly, ciprofloxacin eye drops are used as ear drops in the management of Pseudomonas spp. ear infections. This is an unlicensed use in the UK where licensed ciprofloxacin ear drops are unavailable. Ciprofloxacin drops have been used widely in the rest of Europe, North America and beyond. CONTRAINDICATIONS



The use of preparations containing only a steroid is contraindicated in untreated infections since the immunosuppressive effect of corticosteroids may exacerbate the infection. More controversial is the use of preparations with anti-infective agents in the presence of a perforated



Chapter 35 Drug therapy in otology



tympanic membrane because of the potential for ototoxicity. SIDE EFFECTS



The most common side effects with these preparations are hypersensitivity reactions and local irritation, burning and itching. The use of anti-infective ear drops in the presence of a tympanic membrane perforation The product licences do not permit the use of antiinfective ear drops in ears with perforated tympanic membranes as a result of cochlear damage that occurred when such drops were instilled into guinea pig ears.3, 4 These authors recognize that the round window niche in humans is relatively deep and often protected by a pseudomembrane whilst in the guinea pig the round window is completely exposed. In patients with active chronic suppurative otitis media there is no evidence that the use of these ear drops causes sensorineural deafness. (see Chapter 34, Corticosteroids in otolaryngology).5, 6, 7



Topical ear preparations for removal of ear wax Many cerumenolytics, including oils and aqueous preparations, are available to soften wax prior to syringing or to disintegrate or disperse the wax, to avoid the need for syringing altogether. Burton and Dore´e8 undertook a systematic review and found eight clinical trials. All had a small number of participants, and most were of poor methodological quality. The review concluded that there is insufficient evidence to favour any one particular cerumenolytic. Water and sodium chloride 0.9 percent seem to be as effective as any proprietary agent. Sodium bicarbonate 5 percent ear drops, olive or almond oil ear drops are also safe and inexpensive although their effectiveness has not been evaluated in randomized controlled trials. If the wax is impacted, these drops can be used twice a day for a few days prior to syringing. Some proprietary preparations contain organic solvents (chlorbutanol, paradichlorobenzene) that may cause irritation to the meatal skin. The rational for inclusion of these ingredients in the preparations is not clear and the vehicle alone is often effective.



SYSTEMIC ANTIBIOTICS



Indications Systemic antibiotics are prescribed in the otological conditions of acute otitis media (AOM), cellulitis associated with furunculosis, otitis externa and perichondritis. In AOM, the use of antibiotics is controversial. Galsziou et al.9 found that antibiotic usage in AOM in children varied from 31 percent in the Netherlands to



] 431



98 percent in the USA and Australia. Their Cochrane review concluded that antibiotics provide a small benefit, however, since most cases resolve spontaneously, this benefit must be weighed against possible adverse reactions. Seventeen children must be treated to prevent one child from having pain after two days. Antibiotic treatment may play an important role in reducing the risk of mastoiditis in populations where it is more common. If antibiotics are used, amoxycillin appears to be the first-line treatment, with erythromycin in those who are penicillin sensitive. Should treatment fail, second-line agents include coamoxiclav and cephalosporins. Conservative management of acute mastoiditis (without subperiosteal abscess) has been adopted since a third of patients will settle in 24–48 hours with intravenous antibiotics.10, 11, 12 Flucloxacillin is the antibiotic of choice in treating cellulitis due to Staphylococcus. Penicillin V is used in erysipelas due to Streptococcus. Ciprofloxacin is the drug of choice in malignant otitis externa and is also used in treating perichondritis due to Pseudomonas aeruginosa. Mechanisms of actions The penicillins have a bactericidal action, inhibiting cell wall synthesis by preventing the formation of peptidoglycan cross-bridges. The penicillinase-resistant penicillins, flucloxacillin, cloxacillin and methicillin, are semisynthetic penicillins, resistant to penicillinase by virtue of an isoxazolyl group on R1. Many bacterial b-lactamases are inhibited by clavulanic acid, and a mixture of this inhibitor with amoxycillin(co-amoxiclav) is available. The cephalosporins are also bactericidal. They contain a b-lactam ring and their mechanism of action is similar to the penicillins. The macrolides, for example erythromycin, act by inhibiting bacterial protein synthesis. It binds to the 50s bacterial ribosome subunit inhibiting translocation. Metronidazole was initially used in protozoal infections but was found to be very effective against anaerobic bacteria. The drug is reduced to active metabolites that interfere with nucleic acid function. Ciprofloxacin is a fluroquinolone and acts on both stationary and dividing bacteria by inhibiting DNA gyrase, an enzyme that compresses the bacterial DNA into supercoils. Cell death is thought to occur as a result of the unwinding of the supercoils. Antibacterial spectrum Table 35.1 demonstrates the antibacterial spectrum of the various antibiotics. The penicillinase-sensitive penicillins have a greater spectrum of activity than the b-lactamase resistant drugs, but the combination of clavulanic acid with amoxycillin enables protection of the b-lactam ring, allowing this essential part of the penicillin molecule to remain active.



432 ] PART 7 PHARMACOTHERAPEUTICS Table 35.1



The antibacterial spectrum of antibiotics.



Penicillins



Antibacterial spectrum



Mecillinam



Streptococcal, non b-lactamase producing staphylococcal, pneumococcal, clostridial infection, meningococcal, gonococcal, spirochaetes (syphilis), anthrax, actinomycosis As above and Strep. faecalis, most Haemophillus influenzae and many coliforms Similar to penicillin but less active. Stable to staphylococcal b-lactamase Similar to amoxyillin and in addition activity to Ps. aeruginosa, most Proteus spp. and against bacteroides Similar to ticarcillin and in addition to Klebsiella spp. and greater activity against pseudomonads Coliforms, little activity against Gram-positive bacteria



Cephalosporins



Antibacterial spectrum



First generation: cephaloridine, cephalothin, cephalexin, cephradine and cephazolin Second generation: cefuroxime, cefamandole, cefoxitin Third generation: cefotaxime, latamoxef, ceftazidime, cefsulodin Erythromycin



Broad spectrum EXCEPT against Strep. faecalis, Ps. aeruginosa, H. influenzae and Bacteroides spp. Staphylococcus is sensitive, except for MRSA Broad spectrum with stability against b-lactamases. Active against H. influenzae and Bacteroides spp. but less activity against staphylococcus As for second generation but also active against Ps. aeruginosa



Benzylpenicillin and phenoxymethylpenicillin



Ampicillin and amoxycillin Methicillin, cloxacillin, flucloxacillin Carbenicillin and ticarcillin Mezlocillin, azlocillin, piperacillin



Sulphonamides and trimethoprim: co-trimoxazole Aminoglycosides: gentamicin, tobramycin, netilmicin, amikacin, kanamycin and neomycin Vancomycin Tetracyclines: tetracycline, chlortetracycline, oxytetracycline, doxycycline, minocycline Metronidazole Ciprofloxacin Monobactams: aztreonam



See penicillins but also active against H. influenzae, Bord. Pertussis, Bacteroides spp., Campylobacter spp. Legionella pneumophilia, Mycoplasma pneumoniae and Chlamydiae Broad, active against Gram-positive and -negative bacteria except Ps. aeruginosa Coliforms, Ps. aeruginosa and staphylococci; streptomycin; mycobacteria Staphylococci including MRSA, streptococci and clostridia Broad; Gram-positive and -negative bacteria, brucellae, M. pneumoniae, rickettsia, Coxiella burneti and Chlamydiae. Some resistance to Strep. pyogenes, pneumococci and H. influenzae. Ps. aeruginosa and Proteus spp. resistant Anaerobic bacteria and protozoa Broad; Gram-negative bacteria including Ps. aeruginosa, staphylococci including MRSA, streptococci less sensitive Narrow, active against aerobic Gram-negative bacteria (less sensitive to Ps. aeruginosa)



Dosage Specific dosage recommendations can be found in the British National Formulary. Phenoxypenicillin, amoxycillin and flucloxacillin have better oral absorption than ampicillin but all should be taken at least 30 minutes before food since they are destroyed, to some extent, by gastric acid. The penicillins have good penetration to most tissues but poor entry to CSF (overcome by giving higher doses intravenously). Dosage modification is required in patients with severe renal failure. Ciprofloxacin can achieve high concentrations in bone and soft tissue, even after oral administration. A dosage of 1.5 g daily over a period of 6–12 weeks has been recommended in the treatment of malignant otitis externa.13, 14 Contraindictions In all cases the prescribing of these drugs is contraindicated when the patient is known to be allergic to the



ingredients or related compounds. Immediate hypersensitivity to penicillins occurs in 0.05 percent of patients, the severity of which ranges from urticaria or wheezing to a life-threatening anaphylactic response. Less than 5 percent of patients may develop a delayed hypersensitivity response to penicillins, usually a rash. Occasionally, haemolytic anaemia, leukopenia and interstitial nephritis may occur. There is a 10 percent hypersensitivity crossover with cephalosporins. Precautions The prolonged use of antibiotics may result in superinfection with nonsusceptible organisms. There is no evidence that the penicillins, cephalosporins or erythromycin are hazardous in pregnancy but metronidazole should be avoided in high dosages and ciprofloxacin must be avoided in all trimesters of pregnany. Antibiotics are secreted into breast milk and ciprofloxacin should be avoided in lactation. Other antibiotics, such as the penicillins and



Chapter 35 Drug therapy in otology



] 433



cephalosporins, are secreted into the breast milk but may not be harmful to the infant. Antibiotics, such as the cephalosporins and ciprofloxacin, cleared by the kidneys may require dosage adjustments in patients with renal impairment. Similarly, antibiotics metabolized by the liver, for example erythromycin, may require dosage adjustments or be avoided in patients with liver failure.



pregnancy and breastfeeding. It is contraindicated in patients with phaeochromocytoma. The side effects include gastrointestinal disturbance, headaches, rashes and pruritis, however, these are uncommon. Betahistine is prescribed initially at 16 mg three times a day, a maintenance dose of 24–48 mg has been recommended.



Interactions



Dopamine antagonists



Before prescribing a specific medication, the reader is advised to check for interactions in, for example, The British National Formulary. Antibiotics may lead to oral contraceptive failure, probably because of diminished enterohepatic circulation. The anticoagulant effect of warfarin is affected by the penicillins, macrolides, metronidazole and quinolones. Erythromycin interacts with some antihistamines and cisapride, resulting in cardiac arrythmias. It also increases the plasma level of a number of drugs. Metronidazole produces a disulfiram-like reaction with alcohol. Acetaldehyde accumulates in the body producing facial flushing, headaches, palpitations, nausea and vomiting. Indigestion, iron or zinc therapies must not be taken within two hours of ciprofloxacin since absorption of this antibiotic is significantly affected. Side effects Most antibiotics may produce diarrhoea, rashes, blood disorders, nausea and vomiting. Ampicillin and amoxycillin have a unique adverse effect, comprising a rash, in up to 90 percent of patients with mononucleosis or chronic lymphocytic leukaemia. Reversible hearing loss has been reported with erythromycin. Angioneurotic oedema and anaphylaxis are, fortunately, relatively rare.



PREPARATIONS USED IN THE MANAGEMENT OF VERTIGO Betahistine Betahistine hydrochloride is commonly prescribed in the management of vertigo, usually when associated with Menie`re’s disease or syndrome. Betahistine is thought to reduce the endolymphatic pressure through improved microvascular circulation in the stria vascularis of the cochlear15 or by inhibiting activity of the vestibular nuclei.16 There is insufficient evidence from high quality randomized trials to say whether or not betahistine has any effect on Menie`res.17 It may reduce vertigo, and possibly tinnitus, but does not seem to influence the hearing loss. Betahistine should be used with caution in patients with asthma, a history of peptic ulcer disease, in



Prochlorperazine (Stemetils) is a dopamine antagonist and acts centrally by blocking the chemoreceptor trigger zone and thereby blocking the vomiting centre. The vomiting centre also has afferent input from the vestibular apparatus. This region has a high concentration of muscarinic receptors and histamine H1-receptors. Prochloperazine belongs to group three of the phenothiazines. This means it has less sedative effects, fewer antimuscarinic effects but more pronounced extrapyramidal side effects when compared to the other phenothiazine groups. Prochloperazine is available as tablets, syrup, effervescent sachets, injection, suppositories and as a buccal preparation (Buccastems). These last two preparations are useful because patients using prochlorperazine often vomit and fail to absorb the orally ingested form. The oral dosage varies from 5 mg three times a day increasing to 30 mg daily.



Antihistamines The antihistamines are competitive antagonists of histamine at H1-receptors, and their main action is on the vomiting centre rather than on the chemoreceptor trigger zone. They have weak anticholinergic effects and may occasionally produce a dry mouth and blurred vision. Drowsiness, occasional insomnia and euphoria are side effects which have been reported. These central effects are accentuated with alcohol. Cinnarazine (Stugerons) and cyclizine (Valoids) are less sedating than promethazine teoclate (Avomines). Cinnarizine has been used in the prophylaxis and treatment of Menie`res at a dosage of 30 mg three times a day. Cyclizine may be given in the acute attack orally or parentally at a dosage of 50 mg three times a day.



Gentamicin therapy in Menie`res disease Intratympanic gentamicin therapy was described by Beck and Schmidt18 in 1978 and has been used both in Canada and the UK. Since its introduction, a variety of gentamicin dosage regimens and methods of administration have been developed and remain in clinical practice. These range from injection through the tympanic membrane, through a grommet or via intratympanic or round window catheters.



434 ] PART 7 PHARMACOTHERAPEUTICS



Other medical treatment for Menie`re’s disease The use of salt restriction and diuretic therapy (hydrochlorothiazide, acetazolamide and co-triamterzide 50/25 (Dyazides) are aimed at reducing the accumulation of endolymph. Although a study with Dyazide showed a reduction in vestibular symptoms, there was no effect on hearing loss or tinnitus. Acetazolamide may initially increase the hydrops and Brookes and Booth19 concluded that this drug has no place in treating Menie`re’s disease. Steroids and immunological therapy have also been proposed by those who believe that a disorder of the immune system underlies Menie`re’s disease or syndrome.



DRUGS USED IN SUDDEN SENSORINEURAL HEARING LOSS Corticosteroids Steroids are commonly used in patients with sudden sensorineural hearing loss (SSNHL), but evidence for their effectiveness is lacking. There are several alternative dosage regimes when a ‘short reducing course’ of steroids are required. One such consists of enteric-coated prednisolone, 60 mg on the first day, 50 mg on the second day, 40 mg daily for three days, 30 mg daily for three days and a further reduction so that therapy is discontinued after three weeks. The action, side effects, etc. are discussed in Chapter 34, Corticosteroids in otolaryngology.



Acyclovir Acyclovir (acycloguanosine) is an antiviral agent active against herpesviruses and is prescribed in patients with Ramsay Hunt syndrome (herpes zoster oticus). It acts by inhibiting nucleic acid synthesis. Herpes simplex and varicella zoster contain a thymidine kinase that converts the acyclovir to a monophosphate that is then phosphorylated by the host cell enzymes of acycloguanosine triphosphate that inhibits viral DNA polymerase and viral DNA syntheses. Selectivity for infected cells is achieved since the DNA polymerase of herpesvirus has a much greater affinity for the activated drug than the cellular DNA polymerase. Acyclovir may be administered orally at a dosage of 800 mg five times a day for five days. If treatment is commenced prior to 72 hours after the onset of the rash, acyclovir may shorten the rash duration and acute symptoms and reduce the incidence of post-herpetic neuralgia.20 Acyclovir should be used with caution in patients with renal impairment or who are pregnant or breastfeeding. Side effects include nausea, vomiting, gastrointestinal disturbances, rash, photosensitivity and, rarely, hepatitis, acute renal failure and neurological



reactions (see British National Formulary21 for further details).



SODIUM FLUORIDE TREATMENT IN OTOSCLEROSIS Sodium fluoride has been used for 35 years in an attempt to slow down or arrest sensorineural hearing loss in patients with stapedial otosclerosis or after stapedectomy. It has also been used in patients with ‘pure’ cochlear otosclerosis. Sodium fluoride is an enzyme inhibitor and reduces osteoclastic bone resorption. The clinical benefit of sodium fluoride is controversial. Causse et al.22 found that in ‘otospongiosis-otosclerosis’, sodium fluoride influenced the underlying bony changes in the labyrinth so as to arrest or prevent the onset of hearing loss. In their prospective clinical double-blind, placebo-controlled study of 95 patients, Bretlau et al.23 found that there was a statistically worse deterioration of hearing loss in the placebo group than in the active treated (40 mg sodium fluoride daily) group. Further work by Colletti24 showed benefit in 50 percent of patients five years after a two-year treatment with sodium fluoride (dosages up to 16 mg/day). Deka et al.25 suggested the use of Florical at a dosage of two capsules three times a day in active cochlear otospongiosis. Their study showed that variation in absorption occurs with the use of different preparation and also amongst individuals. The side effects of sodium fluoride need to be considered. In a prospective case-controlled study of ten patients with otosclerosis receiving sodium fluoride 30 mg/day and matched healthy volunteers, Das et al.26 found a high incidence (70 percent) of dyspeptic symptoms in those taking sodium fluoride, as well as histological and electron microscopic abnormalities.



KEY POINTS  Topical ear preparations can be used for a limited period in discharging ears with a perforated tympanic membrane or a grommet in situ.  The use of topical ear preparations containing a steroid only is contraindicated in untreated infections since the immunosuppressive action of the corticosteroid may exacerbate the infection.  Flucloxacillin is the antibiotic of choice in cellulitis due to Staphyloccus. Penicillin V is used in erysipelas due to Streptococcus. Ciprofloxacin is the antibiotic of choice in malignant otitis externa.



Chapter 35 Drug therapy in otology



Best clinical practice [ Before prescribing, ensure that the patient is not known to be allergic to the medicine or adjuvants (such as the preservatives). [ Ensure medication does not interact with patient’s established medication or medical conditions. [ Prescribe medicines at the lowest dose for the shortest time that is effective. [ Review whether medication is still required.



Deficiencies in current knowledge and areas for future research



$ $



There is insufficient evidence to favour any one particular cerumenolytic and a randomized controlled trial may be useful to determine the most clinically and cost-effective preparations for this purpose. There is insufficient evidence from high-quality randomized trials to determine whether betahistine, diuretics, steroids and immune therapy have any effect in Menie`re’s disease.



REFERENCES 1. Acuin J, Smith A, Mackenzie I. Interventions for chronic suppurative otitis media. The Cochrane Library, Issue 2, Oxford Update Software, 2002. 2. Mcgarry GW, Swan IRC. Endoscopic photographic comparison of drug delivery by ear drops and by aerosol spray. Clinical Otolaryngology. 1992; 17: 359–60. 3. Kohonen A, Tarkanen J. Cochlea damage by ototoxic antibiotics by intratympanic application. Acta Otolaryngologica. 1969; 68: 90–7. 4. Brummett RE, Harris RF, Lindgren JA. Detection of ototoxicity from drugs applied topically to the middle ear space. Laryngoscope. 1976; 86: 1177–87. 5. Browning GG, Gatehouse S, Calder IT. Medical management of active chronic otitis media: a control study. Journal of Laryngology and Otology. 1988; 102: 491–5. 6. Fairbanks DNF. Anti-microbial therapy for chronic otits media. Annals of Otology, Rhinology, and Laryngology. 1981; 90: 58–62. 7. Phillips JS, Yung MW, Burton M, Swan IRC for the Clinical Audit and Practice Advisory Group, British Association of Otolaryngologists – Head and Neck Surgeons (ENT-UK). Use of aminoglycoside-containing ear drops in the presence of a perforation: evidence review and ENT-UK consensus statement (in press).







] 435



8. Burton MJ, Dore´e CJ. Ear drops for the removal of ear wax (Cochrane Review). In: The Cochrane Library, Issue 3, Chichester, UK: John Wiley and Sons, Ltd, 2004. 9. Galsziou PP, Del Mar CB, Sanders SL, Hayem M. Antibiotics for acute otitis media in children. The Cochrane Library, Issue 2, Oxford Update Software, 2002. 10. Rubin JS, Wei WI. Acute mastoiditis: a review of 34 patients. Laryngoscope. 1985; 95: 963–5. 11. Ogle JW, Lauer BA. Acute mastoiditis. Diagnosis and complications. American Journal of Diseases of Children. 1986; 140: 1178–82. 12. Nadal S, Herrmann P, Baumann A, Fanconi A. Acute mastoiditis: clinical, microbiological, and therapeutic aspects. European Journal of Paediatrics. 1990; 149: 560–4. 13. Brody T, Pasak ML. The fluoroquinolones. Americal Journal of Otology. 1991; 17: 902–4. 14. Levenson MJ, Parisier SC, Dolitsky J, Bindra G. Ciprofloxacin: drug of choice in the treatment of malignant external otitis. Laryngoscope. 1991; 101: 821–84. 15. Martinez DM. The effect of Serc on the circulation of the inner ear in experimental animals. Acta Otolaryngologica. 1972; Suppl 305: 29–46. 16. Timmerman H. Pharmacotherapy of vertigo: any news to be expected? Acta Otolaryngologica, 1994; Suppl 573: 28–32. 17. James AL, Burton MJ. Betahistine for Menieres disease or syndrome. The Cochrane Library, Issue 2, Oxford Update Software, 2002. 18. Beck C, Schmidt CL. 10 years experience with intratympanically applied streptomycin (gentamicin) in the therapy of morbus Meniere. Archives of Otorhinolaryngology. 1978; 221: 149–52. 19. Brookes GB, Booth JB. Oral acetazolamide in Menieres disease. Journal of Laryngology and Otology. 1984; 98: 1087–95. 20. Collier J. Acyclovir in general practice. Drug and Therapeutics Bulletin. 1992; 30: 101–4. 21. British National Formulary 43. March 2002. 22. Causse JR, Causse JB, Uriel J, Berges J, Shambaugh Jr. GE, Bretlau P. Sodium fluoride therapy. American Journal of Otolaryngology. 1993; 14: 482–90. 23. Bretlau P, Causse J, Causse JB, Hansen HJ, Johnsen NJ, Salomon G. Otospongiosis and sodium fluoride. A blind experimental and clinical evaluation of the effect of sodium fluoride treatment in patients with otospongiosis. Annals of Otology, Rhinology and Laryngology. 1985; 94: 103–7. 24. Colletti V, Fiorino FG. Effect of sodium fluoride on early stages of otosclerosis. American Journal of Otolaryngology. 1991; 12: 195–8. 25. Deka RC, Kacker SK, Shambaugh Jr GE. Intestinal absorption of fluoride preparations. Laryngoscope. 1978; 88: 1918–21. 26. Das TK, Susheela AK, Gupta IP, Dasarathy S, Tandon RK. Toxic effects of chronic fluoride ingestion on the upper gastrointestinal tract. Journal of Clinical Gastroenterology. 1994; 18: 194–9.



36 Drug therapy in rhinology WENDY SMITH AND GRANT BATES



Introduction Treatment of rhinosinusitis with corticosteroids Treatment of rhinosinusitis with antibiotics Treatment of rhinosinusitis with other medicines Medication that may improve the immune response Nasal and antral irrigation with saline Antileukotrienes Agents used to block the parasympathetic nervous system



436 436 438 439 441 441 441 442



Related topics Key points Best clinical practice Deficiencies in current knowledge and areas for future research References Further reading



442 443 444 444 444 445



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words acute rhinosinusitis, intermittent rhinosinusitis, chronic rhinosinusitis, persistent rhinosinusitis, fungal rhinosinusitis and medical treatment. The chapter has also relied on a review produced by the European Academy of Allergology and Clinical Immunology (EAACI) that was published as a supplement by Rhinology in March 2005.



INTRODUCTION This chapter covers the medical treatment of all forms of rhinosinusitis. The most effective drugs for managing rhinosinusitis are corticosteroids and antibiotics. The majority of this chapter discusses these two major groups of medication. A separate section is devoted to other medicines that are used in the management of rhinosinusitis. Finally, other medical treatments used in rhinology are discussed under Related topics. Rhinosinusitis is a significant health problem which reduces the quality of life for individuals and places a large financial burden on society. Effective medical treatment is available for allergic rhinosinusitis and infective rhinosinusitis. The evidence for the effectiveness of medical treatment of rhinosinusitis has been reviewed by the European Academy of Allergology and Clinical Immunology (EAACI) and the authors of this chapter have relied on the EAACI review.1 Rhinosinusitis is an inflammatory process involving the mucosa of the nose and one or more of the sinuses.



Factors that contribute to the inflammation include mucociliary impairment, bacterial infection, allergy, swelling of the mucosa for other reasons and mechanical obstruction. It is recognized that inflammation around the osteomeatal complex is of particular importance in causing rhinosinusitis. Drug therapy in the treatment of rhinosinusitis is aimed at reducing the factors causing inflammation with the additional aim of ensuring adequate ventilation through the osteomeatal complex. Acute rhinosinusitis is now known as intermittent rhinosinusitis and the term chronic rhinosinusitis has been replaced by persistent rhinosinusitis (PRS).



TREATMENT OF RHINOSINUSITIS WITH CORTICOSTEROIDS Topically administered glucocorticoids have improved the treatment of rhinosinusitis and asthma. The efficacy of glucocorticoids may partly depend on their ability to reduce the viability and activation of eosinophils and, in



Chapter 36 Drug therapy in rhinology



addition, to reduce the secretion of chemotactic cytokines by the nasal mucosa. The biological action of glucocorticoids is mediated through activation of intracellular glucocorticoid receptors that are expressed in many tissues and cells.



Intermittent rhinosinusitis without nasal polyps There are a number of high-quality studies where intranasal steroids have been used as an additional treatment to antibiotics, but at present there are no studies where topical nasal steroids have been compared to antibiotics as a single treatment for intermittent rhinosinusitis. Studies are now underway but the results have yet to be published. In published studies, where either a placebo or a topical steroid has been added to antibiotic treatment for intermittent rhinosinusitis, the results generally show that the patients who were treated with a topical corticosteroid improved more rapidly than when the antibiotic was combined with placebo. A representative study is that conducted by Meltzer et al.2 Mometasone furoate 400 mg a day, was given to 200 patients and placebo to 207 patients, all of whom had intermittent rhinosinusitis. All patients were also treated with amoxicillin clavulanate potassium for 21 days. The symptom score, which considered congestion, facial pain, headache and rhinorrhoea, improved significantly in the mometasone group. The effect was most obvious after 16 days treatment and no side effects were seen. [****] Gehanno et al.3 looked at the effectiveness of systemic steroids in the treatment of intermittent rhinosinusitis. Eight milligrams of prednisolone was given three times a day for five days as an adjunct to ten days treatment with amoxicillin clavulanate potassium in patients with intermittent rhinosinusitis. The diagnostic criteria consisted of facial pain, purulent nasal discharge, purulent secretions from the middle meatus, together with opacities on CT scan. The conclusion was that there was no difference in the therapeutic outcome at day 14 between the steroid group and the placebo group, although four days after initiating treatment, the headache and facial pain was significantly less in the steroid group. [****]



Persistent rhinosinusitis without nasal polyps The majority of studies have compared the effects of topical steroid versus placebo as an adjunctive treatment to antibiotics. Two large trials of 407 and 967 patients, respectively, found that mometasone furoate produced a significant improvement in symptom score over the placebo. In one study,4 topical steroid did not help the symptom of post-nasal drip but it did in the second study.5 In both these studies there was no statistical



] 437



difference in the CT findings of the treatment group compared with placebo. A recent multicentre double-blind placebo-controlled randomized trial of 134 patients with persistent rhinosinusitis took a different approach in that topical budesonide was compared against placebo after all the patients had received and had not been cured by two weeks of antibiotics. The treatment period was for 20 weeks and the topical steroid produced a significant reduction in nasal congestion and discharge scores and improved the patient’s sense of smell when compared with placebo. Peak nasal and inspiratory flows also significantly improved in both allergic and nonallergic patients.6 [****]



Persistent rhinosinusitis with nasal polyps: topical steroids In patients with persistent rhinosinusitis and nasal polyps the studies have tended to consider the effectiveness of the treatment on the rhinitic symptoms and then, as a separate outcome measure, the effect of the treatment on the size of the polyps. In a landmark paper published in 1979, Mygind et al.7 showed that beclomethasone dipropionate (BDP), at a dose of 400 mg daily for three weeks, reduced nasal symptoms in 19 patients with nasal polyps when compared to a controlled group of 16 patients treated with a placebo aerosol. There was no reduction in the size of the polyps during this short treatment period. [****] Subsequent studies in which a topical steroid has been compared to placebo in patients with persistent rhinosinusitis and nasal polyposis, have generally shown that the topical steroid is more effective than placebo in reducing the patient’s symptoms and, in particular, the symptom of nasal blockage. In approximately half of the reported studies an improvement in the sense of smell occurred and, again, in approximately half the studies the topical steroid also reduced the size of the polyps. The treatment interval in the various studies varied from between 3 and 120 weeks. [****/***] A comprehensive study compared fluticasone propionate, 400 mg daily, with BDP 400 mg daily and with topical placebo over a 12-week period in a double-blind randomized parallel group that was conducted in a single centre study.8 The symptom score was significantly improved in the fluticasone group and the nasal cavity volume improved in both active treatment groups when measured with acoustic rhinometry. The peak nasal and inspiratory flow also improved in both active groups but the improvement was quicker in the FP group. After 12 weeks there was no difference statistically in the symptoms between the two active groups. [****] In conclusion, therefore, there is good evidence that topical corticosteroids improve the symptoms associated with nasal polyps, in particular nasal blockage, nasal



438 ] PART 7 PHARMACOTHERAPEUTICS secretions and sneezing, and in some patients they are effective at improving the sense of smell. In the majority of patients, but not all, topical corticosteroids also appeared to reduce the size of the nasal polyps, provided they were used for several weeks.



Persistent rhinosinusitis with nasal polyps: systemic steroids There are no studies on treatment with systemic steroids alone in patients with nasal polyps. Topical steroids have always been given as well. Placebo-controlled studies are also lacking but there is a clinical acceptance that systemic steroids have a significant effect on the symptoms of nasal polyposis. [*] In one study, oral prednisolone was given in doses of 60 mg daily to 25 patients with severe nasal polyposis for four days and, for each of the following 12 days, the dose was reduced by 5 mg daily. Antibiotics and antacids were also given. Seventy-two percent of patients experienced a symptom improvement due to the involution of polyps and in 52 percent of patients an improvement was seen on the CT. The symptoms of nasal obstruction and the sense of smell were improved. The findings of this study support the general clinical impression that systemic steroids are highly effective in treating a patient’s symptoms when the patient is actually taking the systemic steroid. Unfortunately, the beneficial effects seem to be lost once the steroid is stopped.9 [**] There have been anecdotal reports of injecting corticosteroid directly into the polyp or into the inferior turbinate. There is no evidence to support this treatment and there have been case reports of blindness following intranasal injection of steroid, and so this method is not to be recommended.



The postoperative treatment of patients with chronic rhinosinusitis and nasal polyps with steroids To date, there have been six studies into the effectiveness of topical steroids against a placebo and their ability to prevent recurrence of nasal polyps. Topical steroids do reduce the recurrence rate of nasal polyps after nasal polyp surgery. To quote from two of the representative studies, Karlsson and Rundcrantz10 treated 20 patients with BDP and 20 were followed with no treatment after polyp surgery. The follow-up period was for 30 months. After six months, there was a statistically significant difference between the groups in favour of BDP and its effect increased during the study period over the next 30 months. Hartwig et al.11 used budesonide six months after polypectomy in a double-blind parallel group on 73 patients. In the budesonide group, ‘polyp’ scores were



significantly lower than in the controls after three and six months. Interestingly, this difference was only significant for patients with recurrent nasal polyposis and not those who had had a polypectomy operation for the first time. [****]



TREATMENT OF RHINOSINUSITIS WITH ANTIBIOTICS The exact incidence of rhinosinusitis within populations is not known but it is a common condition for which antibiotics are frequently prescribed. According to the National Ambulatory Medical Care Survey in the USA, rhinosinusitis is the fifth most common diagnosis for which antibiotics are prescribed, accounting for 9 and 21 percent of all paediatric and adult antibiotic prescriptions, respectively, written in 2002.12



Intermittent (acute) acquired rhinosinusitis There are more than 2000 studies in the literature concerning the effect of antibiotic treatment on intermittent sinusitis. However, only 49 studies met the appropriate criteria for inclusion in a systematic review in the Cochrane Library. These criteria included placebo control, appropriate statistical analysis and sufficient sample size, together with an adequate description of clinical improvements or success rates.13 Amongst these 49 studies, major comparisons were between antibiotic versus control; the newer nonpenicillin antibiotics (macrolides) versus penicillins, and finally the amoxicillin/clavulanate versus other extended spectrum antibiotics. Most trials were conducted within ENT departments and only 20 of the 40 trials were doubleblinded. [****/***] In the studies looking at antibiotic versus control, in general, both penicillin and amoxicillin were more likely to effect a cure than placebo, typical figures being an 82 percent cure for amoxicillin against a 68 percent cure for placebo.13 When comparisons were made between the newer nonpencillins (cephalosporins, macrolides, minocycline) versus penicillins (amoxicillin and penicillin V), no significant differences were shown. The rates of cure or improvement appeared to be 84 percent for both antibiotic classes. Drop-outs due to adverse effects were infrequent and there was no significant difference between the two groups of antibiotics. Sixteen trials involving 4818 patients compared newer nonpenicillin antibiotics (macrolide or cephalosporin) to amoxicillin/ clavulanate. Rates for cure or improvement were 72.7 and 72.9 percent for the newer nonpenicillins and amoxicillin/ clavulanate, respectively. Drop-out rates due to adverse effects were significantly lower for the cephalosporin



Chapter 36 Drug therapy in rhinology



antibiotics. Relapse rates within one month of successful therapy did not differ between the groups. Six trials, of which three were double-blinded, involving 1067 patients, compared a tetracycline (doxycycline, tetracycline, minocycline) to a heterogeneous group of antibiotics (folate, cephalosporin, macrolide, amoxicillin) and no relevant differences were found. The Cochrane reviewers13 concluded that for intermittent rhinosinusitis, confirmed radiologically, current evidence is limited but supports the use of penicillin or amoxicillin for 14–17 days. [****/***] In the Cochrane review, local differences in susceptibility of microorganisms to antibiotics used were not acknowledged, although it is known that resistance patterns of predominant pathogens, such as Streptococcus pneumoniae, Haemophilus influenzae and Moraxella catarrhalis, vary considerably.14 The prevalence and degree of antibacterial resistance in common respiratory pathogens is also increasing worldwide, presumably because of the increase in antibiotic consumption. The choice of which antibiotic is used will not be the same in all regions. The selection may depend on local resistance patterns and the disease aetiology.



Antibiotic treatment for persistent rhinosinusitis It is significantly more difficult to evaluate the efficacy of antibiotic treatment in PRS compared to intermittent rhinosinusitis because of the difficulties of defining the clinical diagnosis of PRS in the literature. In many studies there is no radiological diagnosis and, therefore, data supporting the use of antibiotics in persistent rhinosinusitis is limited and there are no randomized placebocontrolled clinical trials. In a double-blinded prospective study by Legent et al.,15 251 adult patients with PRS were treated with ciprofloxacin or amoxicillin/clavulanic acid for nine days. Only 141 of the 251 patients had positive bacterial cultures from the middle meatus at the beginning of the study. At the end of the treatment period, the nasal discharge disappeared in 60 percent of the patients in the ciprofloxacin group and 56 percent of those in the amoxicillin/clavulanic acid group. The clinical cure and bacteriological eradication rates were 59 and 89 percent for ciprofloxacin versus 51 and 91 percent for amoxicillin/ clavulanic acid, respectively. The differences were not significant. [***] There is some evidence that long-term treatment with low-dose macrolide antibiotics may be effective in treating patients with PRS that has not been cured by surgery or corticosteroid treatment. In animal studies, macrolides have been shown to increase mucociliary transport and reduce goblet cell secretion. There is evidence in vitro, as well as clinical experience, that macrolides reduce the virulence and tissue damage caused



] 439



by chronic bacterial colonization without actually eradicating the bacteria. Clinical studies have shown that long-term treatment with macrolide antibiotics increases ciliary function.16 In a prospective randomized controlled trial (RCT), 90 patients with polypoid and nonpolypoid PRS were randomized to oral macrolide (erythromycin) for three months, or endoscopic sinus surgery, and the patients were followed for over a year. The outcome measures included a symptom score, the SNOT 22, the SF 36, nitrous oxide levels, acoustic rhinometry, saccharine clearance times and nasal endoscopy. Both the medical and surgical treatment of PRS significantly improved almost all the subjective and objective parameters with no significant difference between the two groups, or between polypoid and nonpolypoid PRS, except for the increase in total nasal volume which was greater following surgery.16 To summarize, for persistent rhinosinusitis, there are no placebo-controlled studies on the effectiveness of antibiotic treatment. Studies comparing different antibiotics did not show any significant difference between ciprofloxacin versus amoxicillin/clavulanic acid. The few available prospective studies showed a positive effect on the patient’s symptoms in 56–95 percent of patients. [***] Long-term low-dose macrolide treatment may be of use if surgery and/or steroids have failed. Placebocontrolled studies should be performed to establish the efficacy of macrolides if this treatment is to be accepted in the future. There is also an urgent need for randomized placebo-controlled studies to investigate the effectiveness of antibiotics in general for persistent rhinosinusitis.



TREATMENT OF RHINOSINUSITIS WITH OTHER MEDICINES Decongestants After corticosteroids and antibiotics, the medications that are perceived to be most useful for treating rhinosinusitis are decongestants. The rationale behind the use of decongestants is to improve sinus ventilation and drainage. Radiological studies show that topical decongestants markedly reduce the size of the inferior and middle turbinates and improve osteomeatal complex patency.17 Experimental studies suggest that topical decongestants (zylometazoline and oxymetazoline) have a beneficial antiinflammatory action that is caused by decreasing nitric oxide synthetase.18 A controlled clinical trial showed that there was improved mucociliary clearance in vivo after two weeks of oxymetazoline for persistent bacterial rhinosinusitis when compared to fluticasone and hypertonic saline. The clinical course of the disease between the three groups, however, was not significantly indifferent.19 [***]



440 ] PART 7 PHARMACOTHERAPEUTICS Topical decongestants are also recommended for individuals who have problems clearing their ears while flying or diving. The decongestants are all alpha adrenergic agonists and act on the two types of alpha receptor, one of which controls the venous capacitance vessels of the nasal tissues which are responsible for erectile function, and the other alpha receptors which mediate contraction of arterioles that supply the mucosa. [*] Ephedrine nasal drops (0.5 and 1 percent) are the weakest of the sympathomimetic preparations, whilst oxymetazoline and zylometazoline are more potent and cause intense vasoconstriction. The rebound effect that occurs after the vasoconstriction wears off is more likely to occur with the potent decongestants and, generally, patients should be advised not to use topical decongestant for more than ten days. In patients who become addicted to decongestant drops or sprays, there is a risk of developing rhinitis medicamentosa and, in this condition, the nasal mucosa becomes permanently damaged. One preparation combines the steroid dexamethasone isonicotinate with the sympathomimetic tramazoline hydrochloride (Dexa-Rhinaspray Duo). The spray is promoted for the treatment of allergic rhinitis. The suggestion is that the decongestant allows better mucosal access for the steroid. Clinical experience suggests that this is a useful spray. [*] The use of all sympathomimetic preparations is contraindicated in patients on monoamine oxidase inhibitors (MAOIs) and they should be used with caution in patients with hypertension, hyperthyroidism, cardiovascular disease, diabetes mellitus and closed angle glaucoma, and in infants of less than three months of age. The use of decongestants for persistent rhinosinusitis has not been evaluated in any RCTs. There are no controlled trials to test the efficacy of decongestants for the treatment of nasal polyps. Systemic decongestants appear to be less effective than local preparations but rebound nasal congestion on withdrawal of the drug does not arise. Pseudo-ephedrine is available over the counter in the UK, both in tablet form and as a linctus. There is little evidence to support the use of systemic decongestants in the treatment of rhinosinusitis.



Mucolytics The rationale for using mucolytics in the treatment of rhinosinusitis is that they may reduce the viscosity of the nasal secretions. There is minimal evidence to support the use of mucolytics in the treatment of rhinosinusitis. In one paediatric study, an RCT did not prove a bomhexine to be superior to saline inhalation for children with PRS.20 [****]



Antihistamines These can be used both orally and as a nasal spray. Clinical experience suggests that they reduce rhinorrhoea, sneezing and itching, but have little effect on nasal obstruction. There is no evidence that antihistamines reduce or abolish the symptoms of the common cold. The nonsedating antihistamines, acrivastine (Semprex), cetirizine hydrochloride (Zirtek), desloratadine (Neoclarityn), fexofenadine hydrochloride (Telfast) and terfenadine (Triludan) cause less sedation and psychomotor impairment because they penetrate the blood–brain barrier to a lesser degree than the older type of antihistamines. Terfenadine, however, was found to be associated with hazardous arrhythmias. The side effects of antihistamines include hypotension, hypersensitivity reactions, extrapyramidal effect, dizziness, blood disorders and liver dysfunction. In the treatment of intermittent rhinosinusitis, the beneficial effect of loratadine (the predecessor of desloratadine) for patients with allergic rhinitis was confirmed in a multicentre randomized double-blind placebocontrolled trial. Patients receiving loratadine as an adjunct to antibiotic treatment suffered significantly less sneezing and obstruction on daily visual analogue scale (VAS) scores.21 [****] In PRS, there is little evidence that antihistamines are effective. However, they are often prescribed for patients with PRS, particularly in the USA. There are no controlled trials evaluating such treatment. Sprays such as azelastine hydrochloride (Rhinolast) and levocabastine (Livostin) are topical antihistamines that are promoted for use in allergic rhinitis. Topical antihistamines are considered less effective than topical steroids and good evidence to support their use is lacking.



Sodium cromoglicate Sodium cromoglicate is available as a 4 percent aqueous nasal spray (Rynacrom) and a 2 percent nasal spray (Vividrin), and it is also combined with xylometazoline (Rynacrom) and is promoted for prophylactic use in allergic rhinitis. Depending on the preparation, it needs to be taken either four or six times a day. Its mechanism of action is debatable; it was originally thought to act by preventing mediator release from mast cells. However, agents subsequently developed with this property do not demonstrate the same anti-asthmatic affects as cromoglicate. There is some evidence that cromoglicate depresses the exaggerated neuronal reflexes generated by irritant receptor stimulation. The side effects are few and include local irritation and transient bronchospasm. It may have a role in the treatment of allergic rhinitis in children but the fact that it needs to be taken several times a day means that compliance is low. [**]



Chapter 36 Drug therapy in rhinology



Antimycotics There is increased interest in the role of fungi as a possible cause for the various types of rhinosinusitis. Antimycotics can be used topically and systemically and as an adjunct to sinus surgery in the treatment of allergic rhinosinusitis, invasive fungal rhinosinusitis and conventional intermittent rhinosinusitis. Surgery is considered the first-line treatment for allergic fungal rhinosinusitis,22 and surgery and systemic antimycotics are used in the treatment of invasive fungal rhinosinusitis.23 Although the use of antimycotics in the treatment of allergic fungal rhinosinusitis has not been tested in controlled trials, a high dose of postoperative itraconazole, combined with oral and topical steroids in a cohort of 139 patients with allergic fungal rhinosinusitis, reduced the need for revision surgery.24 [**] The optimum medical treatment for invasive fungal sinusitis is not known because the available evidence comes from small series and case reports which do not meet the criteria for a meta-analysis. [**/*] A group of researchers from the Mayo Clinic produced the hypothesis that persistent rhinosinusitis may arise because of a local immune response to fungi that are present in nasal and sinus secretions.25 Given the correct equipment, fungi can be detected in nasal secretions in virtually all patients with PRS, but also in a controlled disease-free population. As yet, there is no definitive proof that fungi are involved in the aetiology of the inflammatory response in some patients. One doubleblind randomized placebo-controlled trial has looked at this in 60 patients with PRS, giving them topical treatment with amphotericin B, comparing this with saline douching.26 Radiological and subjective scores were actually worse in the treatment group; however this is an area of considerable interest and further randomized controlled trials are awaited. [****]



MEDICATION THAT MAY IMPROVE THE IMMUNE RESPONSE It is thought that an altered immune response to bacterial infection or fungal infection may be responsible for some of the episodes of recurrent rhinosinusitis. With resistance to antibiotic treatment increasing, there is interest in medications that may alter the immune response.



Bacterial lysate preparations Bacterial lysate preparations, e.g. ribosomal fractions of Klebsiella pneumoniae, Streptococcus pneumoniae, Streptococcus pyogenes and Haemophilus influenzae, have been tested against placebo in three multicentre placebocontrolled trials and the evidence from these studies



] 441



suggested that this type of therapy may reduce the need for antibiotics in the treatment of PRS.27, 28, 29 [****] Treatment with more expensive agents that either stimulate or modulate the immune system, e.g. recombinant human granulocyte colony stimulating factor, have been tested in a randomized controlled trial in a group of patients with PRS that were refractory to other types of treatment. The studies showed there was no significant improvement with this expensive treatment. [****] A pilot study has looked at treatment with gamma interferon and it suggested that this treatment may be beneficial in treating resistant PRS but the number of patients was small.30 [**]



NASAL AND ANTRAL IRRIGATION WITH SALINE Saline irrigation A number of randomized controlled trials have tested nasal irrigation with isotonic or hypertonic saline in the treatment of intermittent and persistent rhinosinusitis. In these randomized trials, modalities of application of the saline or hypertonic saline are compared. The evidence is that nasal washout with isotonic or hypertonic saline is beneficial in alleviating symptoms and improves endoscopic findings in patients with PRS. Irrigation with saline has also shown to significantly improve nasal mucociliary clearance, as measured by saccharine tests in healthy volunteers.31 [***]



ANTILEUKOTRIENES The role of leukotrienes in the pathogenesis of bronchial asthma has been well documented and increased levels of these mediators have been detected in patients with rhinosinusitis and nasal polyps. Antileukotrienes have been evaluated in the treatment of asthmatics, especially in those with the aspirin-induced asthma (ASA) triad. When seasonal allergic rhinitis was considered, antileukotrienes were not found to be superior to placebo in reducing symptom scores in a randomized controlled trial.32 [****] For patients with PRS and nasal polyps, one study looked at antileukotriene treatment in 36 patients and found that when antileukotrienes were added to standard treatment regimes, there was a significant reduction in symptom score.33 [***] Two other studies support the use of antileukotriene treatment as an adjunct to standard treatment in patients with nasal polyps, asthma and aspirin intolerance.34, 35 [**]



442 ] PART 7 PHARMACOTHERAPEUTICS



Aspirin desensitization This treatment has been used in patients with nasal polyps. A case-controlled trial of treatment with lysine aspirin to one nostril and placebo to the other in 13 patients with bilateral nasal polyposis, resulted in delayed polyp recurrence on the lysine aspirin-treated side.36 [***] A large randomized controlled trial is underway at the Royal National Throat, Nose and Ear Hospital in London.



AGENTS USED TO BLOCK THE PARASYMPATHETIC NERVOUS SYSTEM Ipratropium bromide (Rinatec) is a muscarinic receptor antagonist and therefore blocks the parasympathetic nervous system. In theory, it should be effective in anyone with a wet dripping nose due to parasympathetic overactivity. In practice, it is effective for some elderly men with a dripping nose. It does not have any other clinical use and there are side effects which include a dry mouth, epistaxis and dryness of the nose. [*]



RELATED TOPICS Preparation of the nasal mucosa prior to surgery Cocaine solution (5 or 10 percent) or cocaine paste (25 percent) has been used to prepare the nose prior to surgery for over 100 years. The majority of rhinologists believe it to be the most effective way of ensuring good operating conditions. As with all local anaesthetic techniques, leaving sufficient time for the drug to work is paramount. The application of Moffat’s solution by a spray (Figure 36.1) is one method of preparing the nose and it is sensible to do this immediately after the induction of anaesthesia. Moffat’s solution consists of 2 mL of 5 percent cocaine, 1 mL of 1:1000 adrenalin and a small amount of bicarbonate solution. In order to avoid systemic effects, the maximum dose recommended for application to the nasal mucosa in fit adults is 1.5 mg/kg. Colouring the cocaine solution with pink dye reduces the risk of it being mistaken for other drugs in the anaesthetic room.



Antifungal agents used in rhinology Fungal infections may be superficial or systemic. Antifungal agents can be classified into polyenes, fluocytosine, imidazoles and triazoles. The polyenes, amphotericin and



Figure 36.1 The application of Moffat’s solution by a spray.



nystatin interact with ergosterol in the fungal cell membrane. Pores are formed through which the fungal cell contents are lost. Selectivity is obtained since the human cell membrane contains mostly cholesterol rather than ergosterol. Amphotericin has a wide spectrum of activity and is used parentally in severe systemic infections since oral absorption is poor. Side effects are common and include nausea and fevers. Renal impairment can occur but is reversible if detected early. Liposomal amphotericin is significantly less toxic but is more expensive. It is likely that it will become the agent of choice. Nystatin is used principally for Candida albicans infections of the skin and mucosal membranes. Flucytosine is given orally or intravenously to treat systemic candidiasis or cryptococcal infections, often in combination with amphotericin to prevent resistance. Flucytosine is converted in fungal cells to flurouracil which inhibits DNA synthesis. The imidazoles, miconazole, ketoconazole and clotrimazole, are mostly used to treat topical infections. They are broad-spectrum antifungals that prevent ergosterol synthesis. Ketoconazole is better absorbed by mouth than the other imidazoles but it has been associated with fatal hepatotoxicity. The triazoles include fluconazole and itraconazole and both can be given orally and parentally. Itraconazole is active against Aspergillus but has been associated with heptatoxicity.



Vestibulitis Noninfective vestibulitis may be treated by applying vaseline or a mild topical corticosteroid to the vestibule. In the presence of infection, however, an antibiotic ointment should be applied and the infective agent is usually Staphylococcus aureus. Two preparations are available in the UK, a combination of chlorhexidine and



Chapter 36 Drug therapy in rhinology



] 443



neomycin (Naseptin) and mupirocin (Bactroban nasal). The latter is used as a second-line agent for the eradication of methicillin-resistant Staphylococcus aureus (MRSA). The ointment is applied three times a day for five days and a swab is taken two days later to determine whether MRSA has been successfully eradicated. Naseptin, when it is used, should be applied four times a day for ten days to treat an infection. It is important to remember that naseptin includes arachis (peanut) oil and this should therefore not be used in patients who have an allergy to peanuts. Naseptin is also contraindicated in pregnancy.



Hereditary familial telangiectasia Figure 36.2 The best position for administering steroid drops.



The epistaxis associated with hereditary familial telangiectasia (HFT) is difficult to treat and a variety of medical and surgical treatments have been tried. Systemic and topical oestrogen provoke squamous metaplasia of the epithelium and in this way provide a protective coat over the blood vessels. Ethinyloestradiol may be used under supervision for the treatment of HFT in women. The side effects include nausea, fluid retention and thrombosis. It is not popular in men because of the gynaecomastia it induces.



Atrophic rhinitis Atrophic rhinitis is characterized by a dry crusting mucosa, nasal obstruction and a foul smell. Steam inhalation and humidification are useful, as is nasal douching with saline. A 25 percent glucose in glycerine solution appears to restore some moisture to the nasal mucosa. [*] To prepare the solution, 75 g of glycerine is gently warmed to which 25 g of glucose is added and the mixture stirred until the glucose has dissolved. After bottling, the drops have a three-month life span. The drops can be used several times a day and provide some relief to this distressing condition.



METHODS OF ADMINISTERING TOPICAL NASAL PREPARATIONS



Nasal sprays: If two sprays are to be administered to each nostril, one spray should be directed upwards and the other backwards whilst the patient does not breath or breaths in gently. The best position for administering nasal drops is head-down as in Figure 36.2.  If betamethasone drops are used it is difficult for the patient to put two drops only into each nostril. In order to establish an exact dose of betnesol drops, they can be decanted into an empty cophenylcaine spray bottle and the patient instructed to put one



spray in each nostril once daily. One spray is equivalent to two drops.  Topical steroid drops (fluticasone or betamethasone) give a higher systemic dose (approximately equivalent to 2 mg of prednisolone daily) than the topical steroid sprays (triamcinolone acetonide – nasacort at 55 mg per spray, mometasone furoate – nasonex at 50 mg per spray).  Fluticasone propionate nasules (400 mg) can be used to increase the topical dose of steroid. One nasule is shared between each nostril once daily; thus the dose given is precisely known.



KEY POINTS  Rhinosinusitis is one of the most common health care problems. It diminishes patients’ quality of life and consumes resources.  Corticosteroids and antibiotics are effective treatments for rhinosinusitis.  Topical decongestants are a useful adjunct.  Antihistamines reduce rhinorrhoea, sneezing and itching but not nasal obstruction.  Antimucolytics are indicated for invasive fungal rhinosinusitis and their role may extend to other types of rhinosinusitis in the future.  Saline irrigation of the nose is an effective treatment for rhinosinusitis.  Antileukotrienes, used as an adjunct, work in 50 percent of patients with the ASA triad.  Topical aspirin desensitization may improve nasal obstruction in PRS with and without polyps. Further trials are underway.  Topical cocaine remains the best way to prepare the nose for surgery.



444 ] PART 7 PHARMACOTHERAPEUTICS



Best clinical practice [ Topical steroids improve the symptoms associated with nasal polyps and reduce the recurrence rate of nasal polyps after surgery. [ The best position for administering steroid drops is head-down (Figure 36.2). [ One way of introducing cocaine into the nose is by spraying Moffat’s solution (cocaine and adrenalin) into the nose immediately after induction of anaesthesia, increasing the chances of obtaining a ‘bloodless surgical field’. [ Before prescribing naseptin, ensure the patient is not allergic to peanuts.



5.







6.



7.



8.



Deficiencies in current knowledge and areas for future research



$ $



$ $



Topical steroids reduce the recurrence rate of nasal polyps following surgery and antileukotrienes help 50 percent of patients with the ASA triad, but there is yet to be an effective long-lasting medical polypectomy. The optimum length of treatment with topical steroids is yet to be established with the different clinical scenarios. The role of fungi in the aetiology of rhinosinusitis has yet to be clarified. The current role of systemic and topical antifungal medication may be extended in the future. Placebo-controlled trials are required to study the effectiveness of antibiotics for persistent rhinosinusitis. Currently, there is considerable interest in macrolide antibiotics. It is not known why some patients with rhinosinusitis develop polyps and others do not.



9. 10.



11.



12.



 13.



14.



REFERENCES







15. 1. Fokkens W, Lund V, Bachert C, Clement P, Hellings P, Jones N et al. European position paper on rhinosinusitis and nasal polyps. Rhinology. 2005; 27–46. 2. Meltzer EO, Orgel HA, Backhaus JW, Busse WW, Druce HM, Metzger WJ et al. Intranasal flunisolide spray as an adjunct to oral antibiotic therapy for sinusitis. Journal of Allergy and Clinical Immunology. 1993; 92: 812–23. 3. Gehanno P, Beauvillain C, Bobin S, Chobaut JC, Desau¨lty A, Dubreuil C et al. Short therapy with amoxicillinclavulanate and corticosteroids in acute sinusitis: results of a multicentre study in adults. Scandinavian Journal of Infectious Diseases. 2000; 32: 679–84. 4. Meltzer EO, Charous BL, Busse WW, Zinreich SJ, Lorber RR, Danzig MR. Added relief in the treatment of acute



 16.



17.



18.



recurrent sinusitis with adjunctive Mometasone furoate nasal spray. The Nasonex Sinusitis Group. Journal of Allergy and Clinical Immunology. 2000; 106: 630–7. Nayak AS, Settipane GA, Pedinoff A, Charous BL, Meltzer EO, Busse WW et al. Effective dose range of Mometasone furoate nasal spray in the treatment of acute rhinosinusitis. Annals of Allergy, Asthma and Immunology. 2002; 89: 271–8. Lund VJ, Black JH, Szabo LZ, Schrewelius C, Akerlund A. Efficacy and tolerability of budesonide aqueous nasal spray in chronic rhinosinusitis patients. Rhinology. 2004; 42: 57–62. Mygind N, Pedersen CB, Prytz S, Sorensen H. Treatment of nasal polyps with intranasal Beclomethasone dipropionate aerosol. Clinical Allergy. 1975; 5: 159–64. Keith P, Nieminen J, Hollingworth K, Dolovich J. Efficacy and tolerability of Fluticasone propionate nasal drops 400 microgram once daily compared with placebo for the treatment of bilateral polyposis in adults. Clinical and Experimental Allergy. 2000; 39: 1460–8. van Camp C, Clement PA. Results of oral steroid treatment in nasal polyposis. Rhinology. 1994; 32: 5–9. Karlsson G, Rundcrantz H. A randomized trial of intranasal Beclomethasone dipropionate after polypectomy. Rhinology. 1982; 20: 144–8. Hartwig S, Linden M, Laurent C, Vargo AK, Lindqvist N. Budesonide nasal spray as prophylactic treatment after polypectomy (a double blind clinical trial). Journal of Laryngology and Otology. 1988; 102: 148–51. Anon JB, Jacobs MR, Poole MD, Ambrose PG, Benninger MS, Hadley JA et al. Antimicrobial treatment guidelines for acute bacterial rhinosinusitis. Otolaryngology and Head and Neck Surgery. 2004; 130: 1–45. Williams Jr. JW, Aguilar C, Cornell J, Chiquette E, Dolor RJ, Makela M et al. Antibiotics for acute maxillary sinusitis (Cochrane review). Cochrane Database of Systematic Reviews 2003(4). Hoban D, Felmingham D. The PROTEKT surveillance study: antimicrobial susceptibility of Haemophilus influenzae and Moraxella catarrhalis from community-acquired respiratory tract infections. Journal of Antimicrobial Chemotherapy. 2002; 59: 49–59. Legent F, Bordure P, Beauvillain C, Berche P. A doubleblind comparison of ciprofloxacin and amoxycillin/ clavulanic acid in the treatment of chronic sinusitis. Chemotherapy. 1994; 40: 8–15. Ragab SM, Lund VJ, Scadding G. Evaluation of the medical and surgical treatment of chronic rhinosinusitis: a prospective, randomised, controlled trial. Laryngoscope. 2004; 114: 923–30. Stringer SP, Mancuso AA, Avino AJ. Effect of a topical vasoconstrictor on computed tomography of paranasal sinus disease. Laryngoscope. 1993; 103: 6–9. Westerveld GJ, Voss HP, van der Hee RM, de HaanKoelewijn GJ, den Hartog GJ, Scheeren RA et al. Inhibition of nitric oxide synthase by nasal decongestants. European Respiratory Journal. 2000; 16: 437–44.



Chapter 36 Drug therapy in rhinology 19. Inanli S, Ozturk O, Korkmaz M, Tutkun A, Batman C. The effects of topical agents of fluticasone propionate, oxymetazoline, and 3% and 0.9% sodium chloride solutions on mucociliary clearance in the therapy of acute bacterial rhinosinusitis in vivo. Laryngoscope. 2002; 112: 320–5. 20. Van Bever HP, Bosmans J, Stevens WJ. Nebulization treatment with saline compared to bromhexine in treating chronic sinusitis in asthmatic children. Allergy. 1987; 42: 33–6. 21. Braun JJ, Alabert JP, Michel FB, Quiniou M, Rat C, Cougnard J et al. Adjunct effect of loratadine in the treatment of acute sinusitis in patients with allergic rhinitis. Allergy. 1997; 52: 650–5. 22. Schubert MS. Medical treatment of allergic fungal sinusitis. Annals of Allergy, Asthma and Immunology. 2000; 85: 90–7; quiz 97–101. 23. Kuhn FA, Javer AR. Allergic fungal rhinosinusitis: perioperative management, prevention of recurrence, and role of steroids and antifungal agents. Otolaryngologic Clinics of North America. 2000; 33: 419–33. 24. Rains 3rd BM, Mineck CW. Treatment of allergic fungal sinusitis with high-dose itraconazole. American Journal of Rhinology. 2003; 17: 1–8. 25. Ponikau JU, Sherris DA, Kern EB, Homburger HA, Frigas E, Gaffey TA et al. The diagnosis and incidence of allergic fungal sinusitis. Mayo Clinic Proceedings. 1999; 74: 877–84. 26. Weschta M, Rimek D, Formanek M, Polzehi D, Podbielski A, Riechelmann H. Topical antifungal treatment of chronic rhinosinusitis with nasal polyps: a randomized, doubleblind clinical trial. Journal of Allergy and Clinical Immunology. 2004; 113: 1122–8. 27. Habermann W, Zimmermann K, Skarabis H, Kunze R, Rusch V. [Reduction of acute recurrence in patients with chronic recurrent hypertrophic sinusitis by treatment with a bacterial immunostimulant (Enterococcus faecalis Bacteriae of human origin]. Arzneimittelforschung. 2002; 52: 622–7. 28. Serrano E, Demanez JP, Morgon A, Chastang V, Van Cauwenberge P. Effectiveness of ribosomal fractions of Klebsiella pneumoniae, Streptococcus pneumoniae, Streptococcus pyogenes, Haemophilus influenzae and the membrane fraction of Kp (Ribomunyl) in the prevention of clinical recurrences of infectious rhinitis. Results of a multicenter double-blind placebo-controlled study.



29.



30.



31.



32.



33.



34.



35.



36.



] 445



European Archives of Otorhinolaryngology. 1997; 254: 372–5. Heintz B, Schlenter WW, Kirsten R, Nelson K. Clinical efficacy of Broncho-Vaxom in adult patients with chronic purulent sinusitis – a multi-centric, placebo-controlled, double-blind study. International Journal of Clinical Pharmacology, Therapy and Toxicology. 1989; 27: 530–4. Jyonouchi H, Sun S, Kelly A, Rimell FL. Effects of exogenous interferon gamma on patients with treatmentresistant chronic rhinosinusitis and dysregulated interferon gamma production: a pilot study. Archives of Otolaryngology – Head and Neck Surgery. 2003; 129: 563–9. Talbot AR, Herr TM, Parsons DS. Mucociliary clearance and buffered hypertonic saline solution. Laryngoscope. 1997; 107: 500–3. Pullerits T, Praks L, Skoogh BE, Ani R, Lotvall J. Randomized placebo-controlled study comparing a leukotriene receptor antagonist and a nasal glucocorticoid in seasonal allergic rhinitis. American Journal of Respiratory and Critical Care Medicine. 1999; 159: 1814–8. Parnes SM, Chuma AV. Acute effects of antileukotrienes on sinonasal polyposis and sinusitis. Ear, Nose and Throat Journal. 2000; 79: 18–20, 24–5. Ulualp SO, Sterman BM, Toohill RJ. Antileukotriene therapy for the relief of sinus symptoms in aspirin triad disease. Ear, Nose and Throat Journal. 1999; 78: 604–6, 613, passim. Ragab S, Parikh A, Darby YC, Scadding GK. An open audit of montelukast, a leukotriene receptor antagonist, in nasal polyposis associated with asthma. Clinical and Experimental Allergy. 2001; 31: 1385–91. Scadding GK, Hassab M, Darby YC, Lund VJ, Freedman A. Intranasal lysine aspirin in recurrent nasal polyposis. Clinical Otolaryngology. 1995; 20: 561–3.



FURTHER READING British Society for Allergy and Clinical Immunology, ENT SubCommittee. Rhinitis Management Guidelines, 3rd edn. London: Martin Dunitz. European Academy of Allergology and Clinical Immunology. European position paper on rhinosinusitis and nasal polyps. Rhinology Supplement. 2005; 18: 1–87.



37 Drug therapy in laryngology and head and neck surgery WENDY SMITH AND ROGAN CORBRIDGE



Introduction Anticoagulants Antibiotics Preparations used to irrigate wounds Local anaesthetic sprays and lozenges Mouthwashes Throat lozenges and pastilles Solutions, suspensions and syrups Treatment of dry mouth Cough medicines Management of stridor Drug therapy in angioneurotic oedema



446 446 448 448 448 448 448 448 449 449 449 449



Use of botulinum toxins in spasmodic dystonia Collagen injection of paralysed vocal cords Drugs used in thyroid disease including the management of hypocalcaemia Treatment of reflux oesophagitis Chemotherapeutic agents Management of fungating wounds Key points Best clinical practice References Further reading



449 450 450 451 452 453 453 453 453 453



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words anticoagulant, antibiotic prophylaxis, and head and neck surgery. This chapter relied on the medicines listed in the British National Formulary.1



INTRODUCTION The laryngologist and head and neck surgeon prescribes medicines for prophylactic purposes, for example antibiotics and anticoagulants, as well as for therapeutic use in the management of infections, immune conditions and voice/throat problems. An appreciation of the pharmacology, preparations available, side effects and contraindications is required for appropriate and safe prescribing. This chapter addresses these matters.



ANTICOAGULANTS The perioperative management of patients taking anticoagulants may be complex and should be considered on



an individual patient basis, taking into account the increased risk of haemorrhagic complications in that procedure with the risk of thromboembolism if the anticoagulation therapy was stopped. Where necessary, involvement of the haematology/medical teams is advised. The anticoagulant drugs heparin and warfarin are widely used in the prevention and treatment of venous thrombosis and embolism perioperatively. Anticoagulants should not be used where there is a history of haemorrhagic disorders, peptic ulcer disease, severe hypertension and severe liver disease. Figure 37.1 illustrates the clotting pathway and the action of heparin and oral anticoagulants. Heparin is a family of mucopolysaccharides with a molecular weight of 4000–30,000. The molecules are attached to a protein backbone consisting entirely of



Chapter 37 Drug therapy in laryngology and head and neck surgery Intrinsic pathway XII



Extrinsic pathway



XIIa XI



] 447



VII XIa



IX



VIIa



Heparin stimulates antithrombin III to form antithrombin/thrombin complex



IXa X



Xa



Prothrombin



Warfarin resembles vitamin K structurally and blocks vitamin K-dependent gammacarboxylation of factors VII, IX, X and prothrombin.



XIII



X Thrombin



Fibrinogen



Fibrin



XIIIa



Stable fibrin clot



serine and glycine residues. Heparin is found in mast cells, in plasma and in the endothelial cell layer of blood vessels. This highly acidic substance is extracted from beef lung or hog intestinal mucosa for therapeutic use. Heparin’s main action is preventing fibrin formation by interacting with the protease inhibitor antithrombin III enhancing (by 1000-fold) the binding of antithrombin III with thrombin. In addition, the heparin–antithrombin III complex has inhibitory effects on factors IXa, Xa, XIa and XIIa. Heparin is not absorbed from the gastrointestinal mucosa because of its charge and large molecular weight; therefore it must be given intravenously or by subcutaneous injection. Intramuscular injection is avoided to prevent haematoma formation. The onset of action of heparin is immediate and it has a half-life of 40–90 minutes. It is inactivated by heparinase in the liver and platelet factor IV released from activated platelets may also have a role. When used for prophylaxis of deep vein thrombosis in surgery, 5000 units of heparin should be given two hours before surgery, then 8–12 hourly for seven days or until the patient is ambulant. Monitoring is not required in this situation but when used to treat a thrombosis, a regime of 5000 units intravenously followed by 15–25 units/kg/hour intravenous infusion or 15,000 units 12 hourly necessitates monitoring of the thrombin time, which should be increased by a factor of two to three. Low molecular weight heparins (certoparin, dalteparin, enoxaparin, reviparin and tinzaparin) are also effective in the prophylaxis of venous thromboembolism and their once daily subcutaneous administration is more convenient. These can also be used to treat deep vein thrombosis and monitoring is not required.



Figure 37.1 The clotting pathway and the mechanism of action of heparin and oral anticoagulants.



Platelet counts should be measured in patients receiving heparin since thrombocytopenia (an immune reaction) may occur but this does not usually happen until six to ten days after the start of treatment. Should thrombocytopenia, or a 50 percent reduction of the platelet count occur, and further anticoagulation is required, patients should be given the heparinoid, danaparoid, or the hirudin, lepirudin. Heparin inhibits aldosterone and may result in hyperkalaemia, especially in patients with diabetes mellitus, chronic renal failure, acidosis, raised plasma potassium or those taking potassium-sparing drugs. The oral anticoagulants prevent the reduction of vitamin K that is required as an active cofactor of carboxylase in factors II, VII, IX and X. The effect on fibrin formation depends upon the balance between the decreased rate of carboxylation and the unaltered rate of degradation of factors already carboxylated. Warfarin has a half-life of 40 hours and a duration of action between two and six days. It takes at least 48–72 hours for the anticoagulant effect to develop fully and so heparin is also used initially when an immediate anticoagulant effect is required. Warfarin is the most commonly prescribed oral anticoagulant; nicoumalone and phenindione are rarely used. The drugs are metabolized by the mixed function oxidases in the liver. Monitoring of the prothrombin time (usually reported as the international normalized ratio (INR) is required and dose adjustments made accordingly. Target INR values (usually between two and four) are detailed in the British National Formulary. The INR can be dramatically altered by co-prescribing other medication and the reader is again directed to the British National Formulary1 prior to co-prescribing.



448 ] PART 7 PHARMACOTHERAPEUTICS



ANTIBIOTICS Antibiotics are commonly prescribed prophylactically, even in so-called clean operations. The use of antibiotics post-tonsillectomy has failed to show a reduction in the secondary haemorrhage rate but may reduce pain at day five postoperatively.2, 3 Where there is no violation of mucosa, no preoperative inflammation and no drain, the incidence of wound infection in the head and neck is 2 percent. In surgery such as laryngectomy, prophylactic antibiotics with a broad spectrum are used to prevent wound infections. A preoperative dose (given to allow a high tissue level at the time of surgery) and three postoperative doses of a cephalosporin and metronidazole are suitable. Violaris and Bridger, using a cephalosporin at eight hours preoperatively with the premedication and three doses eight hourly postoperatively, reduced the development of pharyngocutaneous fistulae.4 Johansen et al.5 found prophylactic metronidazole resulted in a highly significant decrease in the frequency of postoperative fistulae. Antibiotic prophylaxis is also necessary in patients with artificial heart valves or cardiac valve disease. Local policies may vary but amoxycillin 3 g orally or 1 g intramuscularly can be given one hour before surgery and six hours postoperatively. The choice of antibiotic to treat an established infection is initially on a ‘best guess’ policy based on the probable pathogen until swab results and sensitivities become available.



of low potency and is only used for surface anaesthesia in noninflamed tissue in the mouth and pharynx. It is available in lozenges and sprays, sometimes combined with antiseptics. Sensitization may occur and there is little evidence for the benefit in using these combined preparations. Patients should be advised not to eat or drink until the numbness has worn off. Lidocaine belongs to the amide class, benzocaine to the ester class of local anaesthetics. Both bind to a receptor on the sodium channel in the axon preventing the opening of the channels, thereby preventing depolarization of the nerve. In overdosage, cardiac toxicity can occur.



MOUTHWASHES Mouthwashes are usually aqueous solutions in a concentrated form of substances with deodorant, antiseptic, local anaesthetic or astringent properties. Sometimes they should be diluted before use. They have a mechanical cleansing action. Hydrogen peroxide mouthwash is a 6 percent solution and 15 mL should be diluted in half a tumblerful of warm water two to three times daily. It contains an oxidizing agent and is useful in the treatment of acute ulcerative gingivitis. It froths in contact with oral debris, thereby having a mechanical cleansing effect, hence its use in the management of secondary tonsillectomy haemorrhages. Chlorhexidine gluconate, used when toothbrushing is not possible or in combating oral infection, inhibits plaque formation on teeth but has the side effect of causing reversible brown staining of the teeth.



PREPARATIONS USED TO IRRIGATE WOUNDS Before closing a wound following tumour excision it has been recommended that it should first be washed out well. Stell and Maran6 describe the use of Savlon, hydrogen peroxide and sterile water as suitable solutions since it appears that the mechanical process of washing rather than the lytic action on the cells is important. They recommend the use of at least 1 L of solution. In a xenograft model of tumour-cell wound contamination, Allegretto et al.7 found that irrigation with water, saline or gemcitabine delayed tumour development: the latter two improved rates of long-term disease control.



LOCAL ANAESTHETIC SPRAYS AND LOZENGES These preparations are used to desensitize the mouth and pharynx prior to examination, investigation or treatment. Lidocaine (lignocaine) is effectively absorbed from mucous membranes and is available as a pump spray as a 10 percent solution (maximum dose is 20 sprays in pharynx, larynx or trachea) and more commonly as a topical 4 percent solution (maximum dose is 7.5 mL). Benzocaine is a neutral, water-insoluble local anaesthetic



THROAT LOZENGES AND PASTILLES Lozenges and pastilles are both used to deliver medicaments for local effect, either to soothe or treat infections. Lozenges consist of medicaments incorporated into a flavoured base that dissolves or disintegrates slowly in the mouth. They are prepared either by moulding and cutting or by compression. Colours, flavours and sweetening agents may be incorporated. Heavy compression is used to ensure slow disintegration in the mouth. Pastilles consist of medicaments in a base containing gelatin and glycerol or a mixture of acacia and sucrose. Sodium benzoate and citric acid monohydrate may be used as a preservative and antioxidant, respectively. Flavourings such as lemon oil may be incorporated.



SOLUTIONS, SUSPENSIONS AND SYRUPS Solutions are liquid preparations containing one or more soluble ingredients, usually dissolved in water. They may be used internally, externally or for instilling into body cavities and may be sterile or unsterilized depending on



Chapter 37 Drug therapy in laryngology and head and neck surgery



the application. Solutions given orally usually result in rapid absorption. Problems with solubility, taste and stability may prevent this formulation being available for a particular drug. Suspensions are formulations in which the drug does not dissolve in the solvent but is distributed within it. It usually provides a more rapid dispersion and dissolution of the drug when compared to tablets and capsules so long as the drug has a suitable particle size and does not settle or cake on storage. The primary particle size may grow due to the action of surfactants, emulsifiers and other adjuvants. Syrups are concentrated aqueous solutions of sucrose or other sugars to which medicaments or flavourings are added. Glycerol, sorbitol and polyhydric alcohols may be added to reduce the rate of crystallization of sucrose and to increase the solubility of other ingredients. Growth of microorganisms is usually retarded by sucrose concentrations greater than 65 percent w/w.



] 449



more placebo. Demulcent cough preparations may relieve a dry irritating cough by virtue of the fact that they contain a syrup or glycerol that has a soothing effect. Simple linctus (a sugar-free preparation is also available) is harmless and inexpensive.



MANAGEMENT OF STRIDOR The treatment required depends upon the cause and severity of the stridor. Heliox is useful in an acute situation but antibiotics, adrenaline, steroids and antihistamines may be required, as may surgery. Helium is a colourless, odourless, tasteless gas that when mixed, one volume of helium to two volumes of air, diffuses more rapidly than air itself. Breathing such a mixture requires less effort and an air–helium mixture or a mixture of 21 volumes of oxygen and 79 volumes of helium (Heliox) has been used in the management of stridor.



TREATMENT OF DRY MOUTH DRUG THERAPY IN ANGIONEUROTIC OEDEMA Artificial saliva can be useful to relieve dry mouth resulting from radiotherapy or diseases affecting the salivary glands. These preparations should be of neutral pH and contain electrolytes approximating to the composition of saliva. These preparations are available as oral sprays, lozenges and pastilles. Pilocarpine (Salagen) is a muscarinic and can stimulate any residual salivary gland function. Side effects relate to the muscarinic action and its use is contraindicated in those with significant respiratory and cardiovascular disease, angle-closure glaucoma, pregnancy and breastfeeding.



COUGH MEDICINES These are divided into cough suppressants and the expectorant and demulcent cough preparations. After excluding an underlying cause of a cough, such as asthma and gastro-oesophageal reflux, cough suppressants may be used. It is thought that these drugs act by an ill-defined central action in the nervous system and may depress the ‘cough centre’ in the brain stem. The narcotic analgesics are effective as antitussives in sub-analgesic doses. Codeine phosphate is useful for dry or painful coughs but it also inhibits the secretion and mucociliary clearance of sputum, is constipating and dependence can develop. Pholcodine (related to codeine) and dextromethorphan (a non-narcotic, nonanalgesic) have lesser side effects. Over the counter preparations include sedating antihistamines, such as diphenhydramine, and may work by causing drowsiness. There is no evidence that expectorants (ammonium chloride, ipecacuanha and squill) are effective at promoting expulsion of bronchial secretions. Their action is



Swelling of the face and lips, and occasionally of the larynx, occurs in angioneurotic oedema of allergic origin. Antihistamines and corticosteroids are prescribed and, if life-threatening, 1 mL/1:1000 adrenaline can be administered subcutaneously. The nonallergic type results from a serum deficiency of the C1 esterase inhibitor protein. An acute attack is treated with an intravenous injection of 36,000 units of C1 esterase inhibitor protein. This can also be given prior to surgery for prophylaxis. Long-term prophylaxis is achieved with epsilon aminocaproic acid or its derivative tranexamic acid or with androgen methyltestosterone or its derivative danazol. These stimulate the production of C1 esterase inhibitor protein.



USE OF BOTULINUM TOXINS IN SPASMODIC DYSTONIA Botulinum is available as botulinum A toxin–haemagglutinin complex (Botox, Dysport) and botulinum B toxin (NeuroBloc). The dosage is specific to each individual preparation and therefore the product literature must be consulted prior to use. In laryngeal dystonia, localization of the involved muscles and confirmation of correct needle position can be achieved by electromyographic guidance. The effect of botulinum toxin is observed within a few days. Patients may initially have worsening of their voice, dysphagia with the potential for aspiration and occasionally airway compromise. Improvement is seen two weeks post-injections and the effects of treatment may last for three to six months. Botulinum toxin is contraindicated in patients known to be hypersensitive to ingredients, bleeding disorders,



450 ] PART 7 PHARMACOTHERAPEUTICS pregnancy and lactation, concurrent or potential aminoglycoside or spectinomycin administration (neuromuscular blockade is enhanced). It should not be used in patients with generalized muscle disorders such as myasthenia gravis. No information is available on its use in patients with renal or hepatic impairment. Adrenaline should be available in case anaphylaxis occurs.



COLLAGEN INJECTION OF PARALYSED VOCAL CORDS As a result of Teflon injection being unavailable in the UK, collagen (Contigen) is now used to medialize paralysed vocal cords. It is a purified bovine dermal gluteraldehydecross-linked collagen and has been used in genuine stress incontinence. A skin test (0.1 mL collagen and lignocaine, injected intradermally into the volar surface of the forearm) should be carried out four weeks prior to the treatment. Many patients requiring this treatment have a short life expectancy and so a compromise with the skin test performed only a week prior to vocal cord injection is practised. A positive response is defined as erythema, induration, tenderness or swelling with or without purities, persisting for more than six hours or first appearing more than 24 hours after the injection. The use of collagen injection is contraindicated in patients hypersensitive to the ingredients, with a positive skin test, in pregnancy and lactation and in patients with autoimmune disease or a history of multiple severe allergies.



DRUGS USED IN THYROID DISEASE INCLUDING THE MANAGEMENT OF HYPOCALCAEMIA Thyroid hormones The two preparations levothyroxine sodium/thyroxine sodium (Eltroxin) and liothyronine sodium (Tertroxin) are available for use in the management of hypothyroidism, diffuse nontoxic goitre, Hashimoto’s thyroiditis and thyroid carcinoma. Levothyroxine sodium (thyroxine sodium) is used for maintenance therapy, usually as a single dose before breakfast. The initial dose is 50–100 mg daily increasing at two to three week intervals by 25–50 mg increments until normal metabolism is obtained. In the elderly, patients with cardiac insufficiency or severe hypothyroidism, the initial dose is 25 mg, increased by 25 mg every four weeks until normal metabolism is achieved. The usual maintenance dose is 100–200 mg, the higher dose is used to suppress T4 in thyroid carcinoma. Liothyronine has a shorter half-life with a more rapid onset of action and shorter duration of action. Twenty micrograms of liothyronine is equivalent to 100 mg of thyroxine sodium. It can also be given intravenously. A



pretherapy ECG should be performed since hypothyroidism can produce changes resembling ischaemia. The starting dose is 10–20 mg every eight hours. The usual maintenance dose is 60 mg daily in three divided doses. Sometimes it is co-administered with carbimazole to treat thyrotoxicosis. It may be used in severe hypothyroid states when a rapid response is required and is used in patients awaiting radioactive iodine scan following thyroid surgery. The latter enables patients to remain euthyroid in this interval period and patients need to stop liothyronine only two days prior to the scan. These thyroid hormones must be used with caution in patients with hypertension, diabetes mellitus and insipidus, cardiovascular disorders, angina, the elderly, lactation, pregnancy (especially the first trimester) and in adrenal insufficiency. Interactions include sucralfate, phenylbutazone, warfarin, carbamazepine, phenytoin, rifampicin, barbiturates and propranolol. The side effects of arrythmias, insomnia, tremor, palpitations, sweating, weight loss, thyroid crisis, vomiting, diarrhoea and headache have been reported.



Drugs used in hyperthyroidism Antithyroid drugs are used for hyperthyroidism either preoperatively or for long-term management. Carbimazole (Neo-Mercazole) is the most commonly used in the UK; propyluracil is used in patients sensitive to carbimazole. Both are thionamides containing a thiocarbamide group (S = C–N) that is essential for their activity. Carbimazole is rapidly converted to methimazole in vivo. Methimazole is available in the USA. Thioamides prevent the synthesis of thyroid hormones by competitive inhibition of I– to I2 by peroxidase and also block the coupling of the iodotyrosine, especially in forming diiodothyronine. More controversial is the possibility that the thionamides have immunosuppressive properties. These drugs are administered orally and accumulate within the thyroid gland. Their delayed onset of action of three to four weeks results from the need of preformed hormones to be depleted first. The main concern with thionamides is the development of neutropenia and agranulocytosis. Carbimazole has an incidence of causing agranulocytosis in 0.1 percent of patients; propylthiouracil has an incidence of four times this (explaining the preference for carbimazole in the UK). The Committee on Safety of Medicines recommends that:  patients should be asked to report symptoms and signs suggestive of infection, especially sore throat;  a white blood count should be performed if there is any clinical evidence of infection;  the drug should be stopped promptly if there is clinical or laboratory evidence of neutropenia.



Chapter 37 Drug therapy in laryngology and head and neck surgery



Iodine is used as an adjunct to antithyroid drugs in the preoperative management of thyrotoxicosis. An aqueous iodine oral solution (Lugol’s solution) is given at a dose of 0.1–0.3 mL three times a day, taken well diluted with water or milk. Patients may develop flu-like symptoms, headache, rashes, insomnia, lacrimation, conjunctivitis, laryngitis and bronchitis. It must be used with caution in pregnancy and be avoided in lactation.



] 451



intestinal absorption of calcium, mobilize calcium from bone and inhibit renal excretion. The hypocalcaemia of hypoparathyroidism often requires doses of up to 2.5 mg (1,000,000 units) calciferol daily. Calcifediol is the main derivative of liver hydroxylation and this is further hydroxylated in the kidney to the potent calcitriol. The latter step is regulated by parathyroid hormone. A synthetic derivative of vitamin D commonly prescribed in hypoparathyroidism is alfacalcidiol.



Drugs used in hypocalcaemia Following thyroid or parathyroid surgery, patients may develop hypoparathyroidism either temporarily or permanently. Parathormone is the most important regulator of the extracellular calcium concentration. Hypocalcaemic tetany is initially managed by an initial intravenous injection of 10 mL 10 percent calcium gluconate followed by a continuous infusion of 40 mL daily or oral calcium with careful monitoring of the plasma calcium concentration. Bradycardia, arrhythmias and irritation after intravenous injection may occur as can gastrointestinal disturbances after oral administration. Failure to obtain and maintain a corrected calcium concentration within the normal range with calcium supplements alone necessitates the additional administration of a vitamin D preparation. Figure 37.2 demonstrates the control of calcium plasma concentration by vitamin D. Vitamin D is a prohormone, metabolized to hormones that increase



TREATMENT OF REFLUX OESOPHAGITIS Initial treatment depends on the severity of the symptoms and treatment response. Antacids and alginates alone may be used for mild symptoms. The alginates are said to form a raft on the surface of the stomach’s contents to reduce reflux and protect the oesophageal mucosa. Histamine H2 receptor antagonists (cimetidine, famotidine, nizatidine and ranitidine) block the action of histamine on the parietal cells and reduce acid secretion. They should be used with caution in patients with liver or renal disease, in pregnancy and breastfeeding. Side effects include diarrhoea, altered liver function tests, rashes and, more rarely, hypersensitivity reactions, A V block and blood dyscrasias. Cimetidine has been reported to cause gynaecomastia and it binds to cytochrome P-450 reducing the hepatic metabolism of drugs such as warfarin.



Low plasma calcium stimulates PTH



Vitamin D required for maximum response to PTH



Parathyroid hormone (PTH)



Vitamin D (1,23-(OH)2-D)



Stimulates calcium absorption against a concentration gradient along the intestinal tract



Stimulates bone resorption? Via osteoblast-derived factor since osteoclasts do not have receptors for vitamin D



Increases the calcium concentration in the plasma



PTH stimulates osteoclasts releasing calcium and phosphate



Augments the active absorption of phosphate across the intestinal cell membrane



No direct effect on PTH secretion in vitro



Figure 37.2 Calcium homeostasis and the role of Vitamin D.



452 ] PART 7 PHARMACOTHERAPEUTICS The proton pump inhibitors (omeprazole, esomeprazole, lansoprazole, pantoprazole and rabeprazole sodium) react with sulphydryl groups in the H1/K1 ATPase (proton pump) responsible for the transportation of H1 ions out of the parietal cells.



CHEMOTHERAPEUTIC AGENTS These drugs are used either alone or in combination with surgery and/or radiotherapy with the aim to cure or palliate a cancer. They inhibit the mechanisms of cell proliferation and rely upon malignant tumours having a greater proportion of cells undergoing division than in normal proliferating cells, especially in the bone marrow, gastrointestinal mucosa and in hair follicles.



Chemotherapeutic agents are classified with respect to their site of action, as demonstrated in Figure 37.3. Alkylating agents (mustine, cyclophosphamide, chlorambucil, cisplatin and busulphan) cross-link the two strands of the double helix of DNA. The antibiotics actinomycin D, doxorubicin, mitomycin, mithramycin and bleomycin interact with the DNA preventing RNA production. DNA synthesis is prevented by a group of antimetabolites (methotrexate, flurouracil, mercaptopurine and thioguanine) that prevent purine or pyrimidine synthesis. The vinca alkaloids (vincristine and vinblastine) bind to the microtubular proteins inhibiting mitosis. Glucocorticoids are included in regimes since they inhibit cell division by interfering with DNA synthesis. Side effects such as nausea and vomiting, intestinal ulceration, diarrhoea, alopecia and bone marrow suppression are common but may become life-threatening.



Inhibitors of DNA synthesis



Inactivate DNA Procarbazine



Cytarabine DNA Synthesis



The antimetabolites Methotrexate



Alkylating agents Mustine, cyclophosphamide, chlorambucil, busulphan and cisplatin



Fluorouracil Antibiotics Actinomycin D, doxorubicin, mitomycin, mithramycin, bleomycin



Mercaptopurine Thioguanine



RNA (transfer, ribosomal, messenger) Steroid hormones Glucocorticoids Sex hormones



Vinca alkaloids Vinblastine, vincristine Protein (enzymes, hormones)



DNA synthesis Premitotic interval Mitosis Prereplication stage Resting cells



Figure 37.3 Site of action of chemotherapy.



Chapter 37 Drug therapy in laryngology and head and neck surgery



[ In order to avoid severe hypersensitivity reactions,



MANAGEMENT OF FUNGATING WOUNDS Charcoal dressings have been used to help combact malodorous fungating tumours. When this fails, topical metronidazole gel (Anabact or Metrotop) can be applied to the cleaned wound once or twice a day and covered with a nonadherent dressing. The metronidazole has antimicrobial activity against the anaerobes likely to be responsible for the odour. Alginate dressings may stop bleeding from fungating wounds and, if moistened prior to a dressing change, the risk of bleeding at this time can be reduced.



KEY POINTS  Perioperative antibiotics have not been shown to reduce bleeding rates after tonsillectomy but may reduce pain.  The use of perioperative cephalosporins and metronidazole reduce the rates of fistula formation in head and neck operations.  Pilocarpine (used in the treatment of dry mouth) is contraindicated in patients with significant respiratory/cardiovascular disease, glaucoma, pregnancy or breastfeeding.  Rarely, carbimazole can cause agranulocytosis.  Botulinum is available as botulinum A toxin–haemogglutinin complex and botulinum B toxin. It is important to realize that the dosage is specific to each preparation.



Best clinical practice



] 453



preoperative test injections of collagen are necessary where this is used to perform vocal cord medialization.



REFERENCES







1. British National Formulary 42. British Medical Association and Royal Pharmaceutical Society of Great Britain. 2005, chapter 12. 2. Lee WC, Duignan MC, Walsh RM, McRae-Moore JR. An audit of prophylactic antibiotic treatment following tonsillectomy in children. Journal of Laryngology and Otology. 1996; 110: 357–9. 3. Telian SA, Handler SD, Fleisher GR, Baranak CC, Wetmore RF, Potsic WP. The effect of antibiotic therapy on recovery after tonsillectomy in children. A controlled study. Archives of Otolayngology and Head and Neck Surgery. 1986; 112: 610–5. 4. Violaris N, Bridger M. Prophylactic antibiotics and post laryngectomy pharyngocutaneous fistulae. Journal of Laryngology and Otology. 1990; 104: 225–8. 5. Johansen LV, Overgaard J, Elbrrond O. Pharyngo-cutaneous fistulae after laryngectomy. Influence of previous radiotherapy and prophylactic metronidazole. Cancer. 1988; 61: 673–8. 6. Watkinson JC, Gaze MN, Wilson JA (eds). Stell and Maran’s Head and Neck Surgery, 4th edn. London: Arnold, 2000: 55–6. 7. Allegretto M, Selkaly H, Mackay JR. Intraoperative saline and gemcitabine irrigation improves tumour control in human squamous cell carcinoma-contaminated surgical wounds. Journal of Otolaryngology. 2001; 30: 121–5.



[ The starting or stopping of anticoagulants should be



[ [



[



[



in accordance with guidance from the haematologist and medical colleagues, taking into account other prescribed medications which may interact. Platelet counts should be monitored in patients on heparin to detect any thrombocytopenia. Cefuroxime and metronidazole should be given at induction and for at least three postoperative doses in head and neck procedures where the mucosa is breached. Heliox is easier to breath than air and buys useful time in patients with stridor, however, it is important to take additional steps to treat the underlying cause and to secure the airway. First-line treatment of angioneurotic oedema consists of intravenous antihistamines and steroids with nebulized or subcutaneous adrenaline.



FURTHER READING Doctors.net.uk eFormulary. Hardman JG, Limbird LE, Molinoff PB, Ruddon RW, Gilman AG (eds). The pharmacological basis of therapeutics, 9th edn. New York: McGraw-Hill. Johnson JT, Myers EN, Thearle PB, Sigler BA, Schramm Jr. VL. Antimicrobial prophylaxis for contaminated head and neck surgery. Laryngoscope. 1984; 94: 46–51. Lambert HP, O’Grady FW (eds). Antibiotic and chemotherapy, 6th edn. Churchill Livingstone, 1992. Lund W (ed.). Pharmaceutical codex principles and practice of pharmaceutics, 12th edn. Pharmaceutical Press, 1994. Rang HP, Dale MM. Pharmacology. Churchill Livingstone, 1990.



This page intentionally left blank



PART



8



PERIOPERATIVE MANAGEMENT EDITED BY MARTIN J BURTON



38 Preparation of the patient for surgery Adrian Pearce



457



39 Recognition and management of the difficult airway Adrian Pearce



467



40 Adult anaesthesia Andrew D Farmery and Jaideep J Pandit



488



41 Paediatric anaesthesia Alistair Cranston



507



42 Adult critical care Gavin G Lavery



526



43 Paediatric intensive care Helen Allen and Rob Ross Russell



542



This page intentionally left blank



38 Preparation of the patient for surgery ADRIAN PEARCE



Patient pathway Preoperative assessment Preoperative tests Explanation of, and written consent for, planned surgery Explanation of, and consent for, anaesthesia Consideration of venous thromboembolism prophylaxis Consideration of special requirements



457 458 460 461 462 463 464



Planning and scheduling of theatre time and postoperative care Preparation on the day of surgery Key points Best clinical practice References



464 465 465 465 466



SEARCH STRATEGY This included recent publications in scientific journals, from professional bodies and the UK Department of Health, using key words such as consent, suitability for day surgery, preoperative assessment, preoperative preparation and venous thromboembolism prevention.



PATIENT PATHWAY The patient pathway describes the ‘route’ taken by a patient from initial referral to regaining health. A pathway includes processes and documentation, and should be amenable to audit and external review. Generally, the surgeon makes a broad judgement at the time of seeing the patient as to whether or not the patient is fit for surgery. The in-hospital segment of the pathway begins here and, within the confines of a busy clinic, the patient requiring surgery should be placed into one of five routes: 1. 2. 3. 4.



day surgery; inpatient, scheduled admission within a few weeks; inpatient, elective surgery, fit patient; planned inpatient but requiring prior medical/ anaesthetic review; 5. immediate admission.



Suitability for day or ambulatory surgery It is helpful to be able to refer to guidelines about which patients are suitable for day surgery in a particular



hospital. The National Health Service Modernisation Agency published its recommendations1 and advice on day surgery in December 2002. Traditional criteria for suitability are outlined in Table 38.1. Table 38.1



Factors indicating suitability for day surgery.



Factors indicating suitability Patient



Surgical



Anaesthetic



Access to a telephone Responsible adult available for the first 24 hours GP back-up Travelling distance home/day unit o1 hour Escort to collect Suitable home circumstances Peripheral surgery Duration of surgery o2 hours Limited blood loss No wound drain Early resumption of oral intake fluid/food No previous or predicted serious anaesthetic problems Post-operative pain control with oral agents only



458 ] PART 8 PERIOPERATIVE MANAGEMENT There are a number of conditions which either exclude day surgery or require considerable thought:  body mass index (BMI, mass kg/height m2)435–40;  diastolic blood pressure BP 4 100 mmHg, systolic 4 170 mmHg;  moderate/severe cardiorespiratory disease;  patients on haemodialysis;  advanced liver disease;  some neuromuscular disorders such as myasthenia;  opioid dependency if pain relief likely to be difficult;  limited mouth opening or difficult intubation. The head and neck surgical procedures recommended in the original Audit Commission basket of day surgery procedures 1990 were myringotomy with or without grommets, submucous resection, reduction of nasal fractures and operation for bat ears. A revised list was proposed by the British Association of Day Surgery (BADS)2 in 1999 which added tonsillectomy in children. BADS also suggested a number of procedures of which it felt that 50 percent should be possible as day cases – submandibular gland excision, partial thyroidectomy, superficial parotidectomy, rhinoplasty, tympanoplasty and dentoalveolar surgery. In 2000, the Audit Commission prepared a list or basket of 25 operations suitable for day surgery which included myringotomy, tonsillectomy, submucous resection, reduction of nasal fractures and correction of bat ears. The Healthcare Commission’s latest acute hospital portfolio review3 on day surgery was published in July 2005. Failure of the initial broad screening will direct the patient into the wrong pathway and may lead to substantial delays or frustration and patient harm. It is the pathway of medical/anaesthetic review prior to surgery that is particularly testing since it involves initial detection of problem patients, sending them for review, initiating treatment as required and waiting for optimal response before admission for surgery. It is clear that perioperative mortality is reduced by preoperative optimization, particularly of cardiac and respiratory disease.4



       



presence of significant gastro-oesophageal reflux; problems with venous thromboembolism; likelihood of postoperative nausea/vomiting (PONV); difficulties with pain control; medications, allergy, weight; smoking, alcohol and recreational drugs; current or prior drug dependency; home/family circumstances.



A structured form covers the relevant areas, improves data capture,5 aids assessment and can be used by nurse specialists. Problems detected at preadmission clinics are quite varied and need to be resolved before admission. Monoamine oxidase inhibitors present opportunities for serious drug reactions and some anaesthetists feel they should be changed, if possible, to other classes of antidepressants. Special needs patients may require intricate arrangements to attend for surgery on a particular date.



Previous problems with anaesthesia/surgery Notable problems are a history of latex allergy, anaphylaxis to an anaesthetic agent, artificial colloid or antibiotic, difficult or failed intubation, pulmonary aspiration, post-operative nausea or vomiting, awareness, prolonged action of suxamethonium due to plasma cholinesterase deficiency, malignant hyperpyrexia or unexpected admission to a high dependency or intensive care unit. Elective surgery should be postponed until the nature of the previous incident has been fully understood.



Cardiovascular disease Relevant conditions are systemic hypertension, valvular disease, arrythmias, angina (Table 38.2), myocardial infarction, ventricular failure, dyspnoea (Table 38.3), cerebro- or renovascular disease and presence of an Table 38.2 New York Heart Association (NYHA) classification of angina.



PREOPERATIVE ASSESSMENT



Classification



The general health of the patient is assessed to determine whether it can be improved, whether the risk of surgery is worthwhile and to determine the site of perioperative care and the assessment includes:



0 1



      



2



problems with previous surgery; personal or family problems with anaesthesia; cardiovascular system, usual BP; respiratory system; other diseases; bleeding/clotting tendencies; assessment of difficulty with airway management;



3 4



No angina No limitation of ordinary physical activity. Angina caused by strenuous or rapid, prolonged exertion Slight limitation of normal activity, e.g. angina with rapid walking, climbing stairs, emotional stress Marked limitation of normal activity, e.g. angina on one flight of stairs, but comfortable at rest Incapacitation with angina on mildest effort or at rest



Chapter 38 Preparation of the patient for surgery Table 38.3



Dyspnoea grading.



Grade



Table 38.4 Clinical predictors of increased perioperative cardiac risk. Severity



0 1 2 3 4



No dyspnoea whilst walking on level at normal pace Mild, restricted by speed of walking, not distance (OK if I take my timey) Moderate, specific outdoor limitation (Stop after a certain distancey) Marked dyspnoea on mild indoor exertion Dyspnoea at rest



implanted pacemaker or defibrillator. The recent recommendations6 of the American College of Cardiology/ American Heart Association (ACC/AHA) for perioperative cardiovascular evaluation give the following advice.  Hypertension with systolic blood pressure greater than or equal to 180 mmHg and diastolic greater than or equal to 110 mmHg should be controlled before surgery. Usually, an effective regimen can be achieved within several days to weeks. With urgent surgery, agents can be administered that allow effective control within hours. Beta blockers are particularly attractive agents.  Valvular disease. Symptomatic stenotic lesions are associated with high risk and consideration should be given to valvotomy or valve replacement prior to elective surgery. Mild regurgitant valve disease appears to be tolerated well but optimal medical therapy and monitoring should be employed.  The presence of arrythmias or cardiac conduction defects should provoke a careful evaluation for underlying cardiopulmonary disease, but the indications for antiarrhythmic therapy or pacing are identical to those in the nonoperative setting. Frequent ventricular extrasystoles are not associated with an increased risk of nonfatal myocardial infarction (MI) or cardiac death.  Pacemakers should be checked for correct functioning, preferably by a pacemaker clinic, and implantable defibrillators should be turned off immediately before surgery and on again postoperatively. Risk indices7 have been developed over the years and, generally, coronary artery disease, heart failure, cerebrovascular disease, elevated creatinine, insulin-dependent diabetes and high-risk surgery have all been associated with increased perioperative cardiac morbidity. Age is another risk factor8 with the overall 30-day mortality associated with surgery and anaesthesia rising from 2.2 percent (age 60–69) to 8.4 percent (age 490 years). The ACC/AHA recommendations place risk factors into three categories of predictors (Table 38.4) and provide three levels of surgery-specific risk with head and neck surgery in the intermediate group. Areas of particular concern are the evaluation of the degree of myocardial ischaemia and function.9



] 459



Clinical predictor



Major Unstable coronary syndromes



Decompensated heart failure Significant arryhthmias



Acute or recent MI (o4 weeks) Unstable or severe angina High grade AV block Symptomatic ventricular with underlying heart disease Supraventricular with fast ventricular rate



Severe valvular disease Intermediate Mild angina Previous MI by history or Q waves Compensated or prior heart failure Diabetes mellitus (particularly insulin dependent) Renal insufficiency Minor Advanced age Abnormal ECG Rhythm other than sinus Low functional capacity (unable to climb one flight of stairs) History of stroke Uncontrolled systemic hypertension



Assessment of coronary artery disease generally involves history, an exercise ECG, followed where necessary by coronary angiography. Angiography may indicate coronary stenosis amenable to stenting or severe disease requiring coronary grafting. Medical treatment of angina is usually with aspirin or clopidogrel, nitrates, beta blockers, calcium channel antagonists or angiotensinconverting enzyme (ACE) inhibitors. Whilst patients with mild, infrequent, stable angina and normal exercise tolerance do not require any additional investigation, it is generally useful to seek cardiological opinion preoperatively for:  patients with new onset angina who are not on treatment or under review;  patients with increasing frequency of angina;  patients with NYHA class 3 or 4 angina. Heart failure is a serious condition and generally should be evaluated in all patients before surgery. Relevant history is of limited exercise tolerance, dyspnoea,



460 ] PART 8 PERIOPERATIVE MANAGEMENT orthopnoea, paroxysmal nocturnal dyspnoea and ankle oedema. Medical therapy involves frusemide, spironolactone, ACE inhibitors and good control of hypertension or arrhythmias. The best investigation is echocardiography which will demonstrate size of chambers, wall thickness, wall motion, valve function and allow an estimate of ejection fraction. Ejection fraction refers to the proportion of left ventricular blood volume ejected during ventricular contraction. The normal value is 60–65 percent and values below 35–40 percent are significantly reduced and require a raised level of perioperative cardiovascular monitoring. Other investigations look at myocardial perfusion during stress; poorly perfused areas during stress may return to normal on resting (indicating ischaemia) or remain poorly perfused (infarction). Patients with valvular disease, prosthetic valves or cardiomyopathy are at risk of bacterial endocarditis and should receive antibiotic prophylaxis. The current recommendation for most head and neck surgery under general anaesthesia is for amoxycillin 1 g and gentamicin 120 mg intravenously at induction. Patients with mechanical prosthetic valves, or with some cardiomyopathies or atrial fibrillation, may be anticoagulated with warfarin. These patients require admission three to four days before surgery so that warfarin may be discontinued and anticoagulation continued with a heparin infusion at an initial rate of 1000 IU/hour adjusted according to the activated partial thromboplastin time (APTT). Aspirin therapy affects platelet function irreversibly and is a contraindication to surgery if small degrees of platelet dysfunction are unacceptable (e.g. neurosurgery). The drug should be stopped, if required, for one month prior to surgery.



Respiratory disease Common conditions are chronic bronchitis, emphysema, chronic obstructive airway disease and asthma. Generally, a patient with known, mild, stable disease under review by the GP or respiratory department will not need special preoperative assessment. Medical therapy involves inhaled or oral bronchodilators, inhaled or oral steroids, inhaled anticholinergic agents and inhaled agents influencing local immune function. The disease severity can be gauged by history of exercise capability, number of attacks, requirement for admission to hospital or ITU, need for ventilation, requirement for home support or domiciliary oxygen, chest x-ray and respiratory function testing. Asthmatic patients often know their normal peak flow measurement and this is a simple test to carry out. A respiratory physician and anaesthetic review should be sought for patients with respiratory disease who:  have limited exercise capability;  have required hospital admission, particularly ventilation;



 are nearly housebound by respiratory symptoms;  have deteriorating symptoms or signs;  require home oxygen therapy. All patients should be advised to stop smoking.



PREOPERATIVE TESTS Recommendations come from the American Society of Anesthesiologists (ASA) Task Force10 on preanesthesia evaluation and the Association of Anaesthetists of Great Britain and Ireland.11 Preoperative tests may be routine (a test ordered in the absence of specific clinical indication or purpose) or indicated (ordered for a specific clinical indication or purpose). A consensus view has been reached that preoperative tests may be ordered on a selective basis for purposes of guiding or optimizing perioperative management, but should not be ordered routinely. The Association of Anaesthetists of Great Britain and Ireland states that ‘Blanket routine preoperative investigations are inefficient, expensive and unnecessary’. Written guidelines are, therefore, hospital or department-based and should be discussed during induction of new department members. All patients should undergo ‘dipstick’ urinalysis for blood, glucose and protein, and pregnancy testing should be offered to relevant patients. Audiometry should be undertaken on all patients undergoing surgery on the middle ear, and vocal cord function visualized when surgery might damage the recurrent laryngeal nerve. Reasonable guidelines for preoperative tests in adult patients undergoing head and neck surgery are:  haemoglobin in all females, males aged over 40 years and in any patient in whom blood grouping will be undertaken;  urea and electrolytes in all patients over 40 years or when indicated by disease process or medication;  clotting studies when indicated by history;  sickle cell testing in all patients of African or AfroCaribbean origin;  pregnancy testing when pregnancy is possible;  blood grouping (group and save) in patients with a normal preoperative haemoglobin and an anticipated blood loss of 10–15 percent blood volume (blood volume 70 mL/kg in adult), and all patients who have had a previous blood transfusion in case of the presence of antibodies;  blood grouping and cross-matching in all patients with an expected blood loss in excess of 15 percent blood volume (4750 mL in the average adult) or at significant risk of sudden severe haemorrhage. The number of blood units requested for a particular surgical procedure should follow written guidelines drawn up in consultation with the Hospital Transfusion Committee;



Chapter 38 Preparation of the patient for surgery



 electrocardiography in any patient with cardiovascular disease or in asymptomatic patients aged over 60–70 years;  chest x-ray in patients with signs or symptoms of cardiac, respiratory or multisystem disease referable to the chest.



National Institute for Clinical Excellence guidelines on preoperative tests The National Institute for Clinical Excellence (NICE)12 produced its recommendations for the use of routine preoperative tests for elective surgery in June 2003. These are guidelines arising from expert opinion using a consensus development process and the clinical experience of the guideline development group. Specific recommendations are made according to:  grade of surgery 1–4 (minor, intermediate, major, major1);  ASA status 1–3 (Table 38.5);  cardiovascular, respiratory or renal comorbidity;  age of patient.



Table 38.5 Physical status grading of American Society of Anesthesiologists. Grade Grade 1 Grade 2 Grade 3 Grade 4 Grade 5



Physical status Normal healthy patient without any clinically important co-morbidity Patient with a mild systemic disease Patient with one (or more) severe systemic disease which does not present a constant threat to life Patient with systemic disease processes which present a constant threat to life Patient not expected to survive more than 24 hours



Recommendations are made for chest x-ray, ECG, full blood count, haemostasis, renal function, random glucose, urine analysis, blood gases, lung function, sickle cell and pregnancy testing. A wall-chart or booklet shows individual recommendations as green (test recommended), red (test not recommended) and yellow (consider this test). Table 38.6 refers to a patient undergoing intermediate or grade 2 surgery (e.g. tonsillectomy) who is ASA 2 with co-morbidity from cardiovascular disease (perhaps well-controlled, mild hypertension). Table 38.7 refers to a patient undergoing major or grade 3 surgery (e.g. thyroidectomy) who is classed as ASA 3 with co-morbidity from cardiovascular disease. Unfortunately, many recommendations are ‘yellow’ and individual hospitals or departments are still required to produce their own guidelines. Without these, unnecessary tests will continue to be ordered ‘just in case’ they are required by the anaesthetist or surgeon. Staff undertaking clinical preoperative assessments should discuss with patients which tests are recommended, what they involve and the possible implications of an abnormal result. Doctors or nurses ordering the tests should write in the notes that they have discussed the recommended tests and their implications with the patient, who should be informed of the results of the tests.



EXPLANATION OF, AND WRITTEN CONSENT FOR, PLANNED SURGERY Valid consent to treatment is central in all forms of health care, from providing personal care to undertaking major surgery. In the UK, in accordance with the NHS Plan, the Department of Health instituted substantial changes in consent in 2001. The seminal document is ‘Good practice in consent implementation guide’ available on www.doh.gov.uk/consent. Consent is a patient’s agreement for a health professional to provide care. Seeking consent indicates the



Table 38.6 Tests recommended by NICE for intermediate surgery in ASA 2 patient with cardiovascular co-morbidity. Test



Chest x-ray ECG Full blood count Haemostasis Renal function Random glucose Urine analysis Blood gases Lung function ?, local decision required.



] 461



Age (years) Z16 to o40



Z40 to o60



Z60 to o80



Z80



? Yes ? No ? No ? No No



? Yes ? No ? No ? No No



? Yes ? No Yes No ? No No



? Yes ? No Yes No ? No No



462 ] PART 8 PERIOPERATIVE MANAGEMENT Table 38.7 morbidity.



Tests recommended by NICE for major surgery in ASA 3 patient with cardiovascular co-



Test



Chest x-ray ECG Full blood count Haemostasis Renal function Random glucose Urine analysis Blood gases Lung function



Age (years) Z16 to o40



Z40 to o60



Z60 to o80



Z80



? Yes Yes ? Yes No ? ? No



? Yes Yes ? Yes No ? ? No



? Yes Yes ? Yes No ? ? No



? Yes Yes ? Yes No ? ? No



?, local decision required.



whole process of information provision, discussion and decision-making. When a patient formally gives their consent to a particular intervention, this is only the end point of the consent process. Patients may indicate consent nonverbally (for example by presenting their ear for examination in outpatients), orally or in writing. For the consent to be valid, the patient must:  be competent to take the particular decision:  have received sufficient information to take it;  not be acting under duress. When an adult patient lacks the mental capacity (either temporarily or permanently) to give or withhold consent for themselves, no one else can give consent on their behalf. It is good practice to gain written consent if:  the treatment or procedure is complex or involves significant risks;  the procedure involves general/regional anaesthesia or sedation;  providing clinical care is not the primary purpose of the procedure;  there may be significant consequences for the patient’s employment, social or personal life;  the treatment is part of an approved project or programme of research. In the UK, standard forms are used in all NHS hospitals. Four differing forms are available to cover: 1. adult patients undergoing surgery who are competent to consent; 2. parental agreement for a child or young person; 3. adult patients undergoing treatment or procedures where the patient is expected to remain alert and an anaesthetist is not involved 4. when an adult patient is unable to consent.



For an adult patient, the consent process begins with the provision of information to the patient and discussion of treatment options and oral agreement may be reached that particular surgery is appropriate. When written consent is appropriate, the patient should be familiar with the contents of their consent form before they arrive for the procedure and should have received a copy of the page documenting the decision-making process. They may be invited to sign the form at any appropriate time before the procedure – in outpatients, preadmission clinic or when they arrive for treatment. A member of the health care team must check immediately before treatment that the patient has no concerns and that their condition has not changed. Patients must not be under duress to give valid consent and patients should not be expected to give consent when already changed for theatre or, indeed, in the anaesthetic room.



EXPLANATION OF, AND CONSENT FOR, ANAESTHESIA It is the duty of the anaesthetist to gain consent for anaesthesia, but specific written consent for anaesthesia is not required in the UK. Each patient should receive a general leaflet about anaesthesia in outpatients. This allows the patient time to consider the issues surrounding provision of anaesthesia. Discussion between the anaesthetist and patient addresses (as relevant):  previous anaesthetic problems;  likely difficulties or risks in provision of anaesthesia for the planned surgery;  local versus general anaesthesia;  general conduct of general anaesthesia and recovery;  intravenous fluids;  blood transfusion;  pain relief;  control of nausea/vomiting;



Chapter 38 Preparation of the patient for surgery



] 463



 urinary catheterization;  nasogastric tube insertion;  placement of arterial and central venous lines.



used via the epidural route and the appropriate level of postoperative care.



Patient anxieties are generally to do with death, brain damage, awareness, loss of control, memory loss, pain control, nausea or vomiting and needle insertion.13 Premedication with either an anxiolytic or analgesic is discussed and written as required. In the UK it is possible that the preoperative anaesthetic visit will occur only minutes before planned surgery but a better arrangement exists in France. Here, all patients must be seen in an outpatient anaesthetic clinic before surgery. This gives time for proper preoperative assessment, preoperative preparation and discussion of anaesthetic options.



CONSIDERATION OF VENOUS THROMBOEMBOLISM PROPHYLAXIS



Acute pain management plans Individual plans for postoperative pain management should be formulated with the patient in the preoperative period. In the intraoperative period virtually all patients undergoing head and neck surgery receive intravenous opioids and, where possible, incisions should be infiltrated with a long-acting local anaesthetic such as bupivacaine 0.5 percent (maximum dosage 2 mg/kg) before the end of surgery. Early postoperative pain may be treated by intravenous opioid, often morphine in 2 mg increments every five minutes to a total of 10 mg, administered by the recovery nurse or doctor. Simple oral analgesia in adults, such as paracetamol 1 g six hourly and ibuprofen 400 mg six hourly, should be prescribed regularly. If oral administration is not possible, drugs may be given parenterally, by suppository or via a nasogastric tube. Initiation of analgesia by suppository (such as diclofenac 100 mg) towards the end of surgery is common, provided that the patient has consented to this. More severe pain may be treated in the ward by regular dihydrocodeine 30 mg six hourly or morphine given by the oral or intramuscular route. If it is expected that the patient will require regular morphine, this is best given by patient-controlled analgesia (PCA) in which the patient receives a small dose of morphine intravenously on pressing a button. Preoperative instruction in the use of a PCA machine may be given by the anaesthetist, ward nurse or member of the acute pain team. The acute pain team consists usually of nurses, doctors and pharmacists and provides the lead in the generation of acute pain guidelines and management. The expertise of the acute pain team should be sought preoperatively in any patient in whom postoperative pain is likely to be difficult to manage, particularly those on preoperative opioids for medical or nonmedical reasons. Epidural analgesia is not suitable for surgery in the head and neck territory, but is useful to cover abdominal surgery when this is required to mobilize the bowel for major reconstructions. Acute pain guidelines usually detail the analgesic solution to be



Venous thromboembolism (VTE) is one complication among hospital inpatients and contributes to longer hospital stays, morbidity and mortality. Recent recommendations arise from the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy14 in 2004. Risk factors and conditions predisposing to VTE are:          







history of previous VTE; prolonged immobility or confinement to bed; lower limb, pelvic or abdominal operations; trauma particularly of pelvis or acute spinal injury; obesity; major medical illnesses, such as MI, ischaemic stroke, cardiac failure and acute respiratory failure; oestrogen use, such as oral contraception or hormone replacement therapy; cancer, especially metastatic adenocarcinoma age 440 years; acquired hypercoaguable states, such as lupus anticoagulant and antiphospholipid antibodies, hyperhomocysteinaemia, dysfibrinogenaemia and myeloproliferative disorders; inherited hypercoaguable states, such as activated protein C resistance, protein C or S deficiency, antithrombin deficiency or prothrombin gene mutation.



All patients undergoing surgery should be placed in a risk stratification for VTE, with particular importance attached to identifying the high- and very high-risk patient. Expert haematological assessment is required when acquired or inherited hypercoaguable states are suspected.  Low risk. Uncomplicated surgery in patients aged o40 years with minimal immobility postoperatively and no risk factors.  Moderate risk. Any surgery in patients aged 40–60 years, major surgery in patients o40 years and no other risk factors, minor surgery with one or more risk factors.  High risk. Major surgery in patients 460 years, major surgery in patients aged 40–60 years with one or more risk factors.  Very high risk. Major surgery in patients 440 years with previous VTE, cancer or known hypercoaguable state. An appropriate strategy for the prevention of VTE should be formulated for each patient and written guidelines should exist in each department. The preventative measures which have been shown to reduce the incidence



464 ] PART 8 PERIOPERATIVE MANAGEMENT of VTE are early mobilization, compression elastic stockings, pneumatic intermittent calf compression which provides rhythmic external compression of the lower limb or calf to a pressure of 35–40 mmHg for ten seconds every minute, low-dose unfractionated heparin (LDUH) and low molecular weight heparin (LMWH). LMWH has the advantages over LDUH of once daily dosage and lower incidence of heparin-induced thrombocytopenia, but is more expensive. Comparative studies of LDUH and LMWH have shown a broadly similar efficacy but comparisons are hampered by the dosage and timing. LDUH is usually given as 5000 units 8–12 hourly and LMWH may be given in low doses (o3400 anti-Xa IU daily) or high doses (43400 antiXa IU daily). High-dose LMWH appears to be associated with more surgical bleeding problems. Unfortunately, little data exist on preventative measures in head and neck patients exclusively and data appropriate to groups may not benefit an individual patient. Surgery itself in the head and neck territory is considered low risk and there is the additional perceived problem of heparin causing minor, but surgically troubling, bleeding. This may cause difficulty during microscopic surgery or surgery near the skull base or brain, or result in an increase in incidence of wound haematoma. A reasonable strategy for VTE prophylaxis in head and neck patients might be:  Low risk patients. No specific prophylaxis other than early ambulation.  Moderate risk patients. Either LMWHo3400 U daily, LDUH 12 hourly, compression elastic stockings or intermittent pneumatic compression. Each alone is better than no prophylaxis.  High risk patients. LDUH eight hourly, or LMWH43400 U daily plus compression stockings, or compression stockings and intermittent calf compression if anticoagulation considered inadvisable.  Very high risk. LDUH eight hourly or LMWH43400 U daily plus compression stockings and intermittent pneumatic compression.



CONSIDERATION OF SPECIAL REQUIREMENTS Patients undergoing surgery in the head and neck region may lose, temporarily or permanently, sensation or motor activity in the distribution of the cranial nerves. Particularly important from a patient’s perspective are the loss of communication arising from impairment of hearing, speech or vision, disabling sensation such as vertigo through vestibular dysfunction or alteration to normal activity such as eating, drinking or breathing. Patients who will be unable to speak postoperatively require a plan for communication, such as a bell to attract



attention, prepared cards indicating common needs, pen and paper or the use of hand-signals. For those patients undergoing laryngectomy, the preoperative period is the appropriate time to discuss voice reconstruction options and for the patient to meet the speech therapist and another patient who has undergone, and recovered from, similar surgery. The disease process or nature of surgery may interfere with nutrition. Failure of adequate nutrition may lead to morbidity and mortality through infection, failure of wound healing and gross catabolism. Nutrition may be provided orally, through a nasogastric tube, via a percutaneous or open gastrostomy, via an open feeding jejunostomy or intravenously. The enteral route is superior in all regards to the intravenous route and should be used whenever possible. Poorly nourished patients require preoperative supplementation and there is some evidence that a week of enteral (but not parenteral) nutrition improves outcome. A feeding gastrostomy or jejunostomy can be placed in a planned manner at the time of extensive surgery to cover postoperative feeding. Patients in whom the small or large bowel is required for surgical reconstruction of the upper GI tract may require bowel preparation with a low residue diet or intravenous fluids and laxatives.



PLANNING AND SCHEDULING OF THEATRE TIME AND POSTOPERATIVE CARE The Confidential Enquiry into Postoperative Deaths (CEPOD)15 process identifies patients as being elective when the surgery can be planned for the convenience of the patient and surgeon, scheduled when admission will be prioritized within a few weeks of initial referral, urgent indicating an unplanned admission where resuscitation can be achieved before theatre and emergency in which surgery is required coexistent with resuscitation. A better scheme starts by indicating whether the patient is listed or unlisted, since the classification of CEPOD urgent or emergency is not always understood or useful. The operating theatre list should be presented to theatre reception in the manner required within that hospital. Generally, a theatre list should contain the patients’ name, hospital number, date of birth or age and planned surgery. The side of surgery should be recorded as left or right. Some hospitals use the Office for Population Censuses and Surveys (OPCS) or Read codes. The first patients on a list are generally children in ascending age, or shorter operations before longer ones or patients with diseases such as diabetes. Patients with latex allergy must be scheduled first to allow correct preparation of the operating theatre. Scheduling problems should be resolved by discussion between the surgeon, anaesthetist and theatre sister. Some patients will require a planned extended stay in Recovery, or admission to HDU/ITU. Generally, this will be for longer, major



Chapter 38 Preparation of the patient for surgery



procedures, when ventilation is required postoperatively, when the airway is at risk or if extended patient monitoring is required for cardiorespiratory disease. The bed should be booked as soon as surgery is planned for a particular date. Patients should know about the plans for the immediate recovery period (which may change) and may benefit from seeing the HDU/ITU preoperatively or talking to a nurse from the unit.



PREPARATION ON THE DAY OF SURGERY It should be possible to estimate a provisional time of surgery and preparation aims to make certain that the patient is ready when the theatre calls for the patient. Most units use a form with tick boxes to cover the areas of baseline blood pressure, heart rate, weight, allergies, false or capped teeth, patient’s wrist band present, consent form signed, notes and x-rays present, investigations present, blood cross-matched, time of last oral intake and details of premedication. The side and site of surgery should be indicated with an indelible marker by the surgeon who reviews the patient on the day.



Oral intake Patients should not eat or drink prior to anaesthesia to reduce the risk of aspiration of gastric contents. The Association of Anaesthetists of Great Britain and Ireland recommends the following fasting periods which are now generally accepted:16  six hours for solid food, infant formula or other milk;  four hours for breast milk;  two hours for clear nonparticulate and noncarbonated fluids. There is no evidence that safety is improved by extending these fasting times. Some patients should not be left for long periods without fluids and consideration should be given to intravenous fluid administration. These include elderly or sick patients, children, those undergoing bowel preparation and breastfeeding mothers.



Medication Generally, all regular medication should be given on the day of surgery unless specifically crossed off by the anaesthetist or surgeon. It is particularly important to give cardiac medication for angina, hypertension or arrythmia. Diuretics are often omitted. Diabetic medication needs to be specifically addressed by the anaesthetist but generally tablets are not given on the morning of surgery and insulin given in either a reduced dosage subcutaneously with a covering intravenous infusion of 5 percent dextrose, or as a sliding scale. In a sliding scale the



] 465



patient receives a constant dextrose infusion such as dextrose 4 percent with saline 0.18 percent (with potassium 20 mmol in 1000 mL bag) at 100 mL/h and an insulin infusion adjusted one to two hourly according to blood glucose. There is some evidence that tight control of glucose to a range of 4–6 mmol/L is beneficial. Some oral medication, for example nitrates, may be prescribed as a transdermal preparation. Patients on oral steroids require an additional dose parenterally, either 50 mg hydrocortisone intravenously 8–12 hourly or an infusion of 150–200 mg/24 h.



KEY POINTS  Consider day surgery where possible.  Assess the risk of operation/anaesthesia.  Moderate or severe cardiorespiratory disease requires medical/anaesthetic review.  Routine preoperative tests are expensive and unnecessary.  Valid, written consent required for most surgery.  Consider venous thromboembolism prevention.  Book HDU/ITU bed (if required) as soon as date of surgery known.  Patients presenting for ENT surgery may be of either gender and of any age, with general health at any point between fit and moribund, for elective, scheduled or emergency surgery lasting between a few minutes and many hours. Perioperative care may need to be appropriate for minor surgery such as insertion of grommets as a day case or extensive head and neck surgery requiring intensive care.



Best clinical practice [ The patient should be directed into the correct pathway as soon as surgery is contemplated, and sufficient oral and written information given to inform the patient and initiate the consent process. [ Attendance at a preadmission clinic a few weeks before surgery allows a full history and examination to be taken, leading to the identification of any medical, anaesthetic or personal problems. This is particularly important when the interval between scheduling and admission for surgery is more than a few months. [ The requirement for specific preoperative blood tests, investigations and blood cross-matching should be agreed by the whole team and produced in a written



466 ] PART 8 PERIOPERATIVE MANAGEMENT form which is available to doctors and nurses involved in preoperative preparation. [ Preoperative assessment of the problem patient by an anaesthetist, preferably the one who will be involved, is valuable and should be easy to arrange. [ Preoperative tests taken at preadmission should be reviewed before the admission for surgery so that abnormalities can either be treated in time or the patient rescheduled. [ Notify the acute pain team of any patient in whom postoperative pain control is likely to be difficult, particularly those with opioid dependency or on a drug withdrawal programme.







 



REFERENCES 1. www.archive.modern.nhs.uk/scripts/ default.asp?site_id=28&id=8100 2. BADS. BADS Directory of procedures. London: BADS, 2006. 3. www.healthcarecommission.org.uk/ serviceproviderinformation/reviewsandinspections/ acutehospitalportfoliohomepage.cfm 4. Prause G, Ratzenhofer-Komenda B, Smolle-Juettner F, Krenn H, Pojer H, Toller W et al. Operations on patients deemed ‘‘unfit for operation and anaesthesia’’: what are the consequences? Acta Anaesthesiology Scandinavica. 1998; 42: 316–22. 5. Ausset S, Bouaziz H, Brosseau M, Kinirons B, Benhamou D. Improvement of information gained from the preanaesthetic visit through a quality assurance programme. British Journal of Anaesthesia. 2002; 88: 280–3.







6. ACC/AHA guideline update for perioperative cardiovascular evaluation for non-cardiac surgery – executive summary. Anesthesia Analgesia. 2002; 94: 1052–64. 7. Goldman L, Caldera DL, Nussbaum SR. Multifactorial index of cardiac risk in non-cardiac surgical procedures. New England Journal Medicine. 1977; 297: 845–50. 8. Jin F, Chung F. Minimizing perioperative adverse events in the elderly. British Journal of Anaesthesia. 2001; 87: 608–24. 9. Mangano DT. Assessment of the patient with cardiac disease. An anesthesiologist’s paradigm. Anesthesiology. 1999; 91: 1521–6. 10. American Society of Anesthesiologists Task Force on Preanesthesia Evaluation. Practice advisory for preanesthesia evaluation. Anesthesiology. 2002; 96: 485–96. 11. Preoperative assessment – the role of the anaesthetist. The Association of Anaesthetists of Great Britain and Ireland, November 2001. www.aagbi.org 12. National Institute for Clinical Excellence, MidCity Place, 71 High Holborn, London WC1 V 6NA or www.nice.org.uk 13. Matthey P, Finucane BT, Finegan BA. The attitude of the general public towards preoperative assessment and risks associated with general anaesthesia. Canadian Journal of Anaesthesia. 2001; 48: 333–9. 14. Geerts WH, Pineo GF, Heit JA, Bergqvist D, Lassen MR, Colwell CW et al. Prevention of venous thromboembolism: the seventh ACCP conference on antithrombotic and thrombolytic therapy. Chest. 2004; 126: 338–400. 15. www.ncepod.org.uk 16. Soreide E, Eriksson LI, Hirlekar G, Eriksson H, Henneberg SW, Sandin R et al. Preoperative fasting guidelines: an update. Acta Anaesthesiologica Scandinavica. 2005; 49: 1041–7.



39 Recognition and management of the difficult airway ADRIAN PEARCE



Definitions Prevalence Evaluation Prediction of difficulty Strategy Alternative techniques for tracheal intubation Role of the classic LM Failed ventilation and emergency cricothyrotomy



467 468 468 471 472 473 475 476



Obstructed airway Extubation and recovery Follow-up Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



477 483 484 485 485 485 486



SEARCH STRATEGY The data in this chapter are supported by a PubMed search using the key words difficult intubation, difficult airway, obstructed airway.



DEFINITIONS The difficult airway is the clinical situation in which a practitioner experiences difficulty with adequate maintenance and/or protection of the airway. Three airway devices are in common use – the face mask, laryngeal mask and tracheal tube – and each offers a different level of airway protection and maintenance. A cuffed tracheal tube offers the highest level and for this reason is prominent in airway management plans in head and neck surgery. Two broad airway problems can be defined, difficult ventilation and difficult intubation. Difficult mask ventilation was defined by the American Society of Anesthesiologists (ASA)1 as the inability to maintain the oxygen saturations above 90 percent by face mask inflation with 100 percent inspired oxygen or to reverse signs of inadequate ventilation. Oxygen stores in the body are exhausted within a few minutes and difficult or failed ventilation will rapidly result in morbidity or mortality



from hypoxaemia. Difficult intubation was defined by the ASA as the inability to complete tracheal intubation within three attempts at direct laryngoscopy or within ten minutes. Provided that face mask ventilation is possible, failed tracheal intubation by itself should not result in hypoxaemia unless the failure of airway protection leads to gross airway soiling from gastric contents or blood. Unfortunately, failed tracheal intubation is, in clinical practice, associated with the problems of unrecognized oesophageal intubation, damage to the airway and hypoxaemia. Difficult intubation may be defined by the number of attempts at direct laryngoscopy or time to achieve intubation, but also by the view of the laryngeal structures seen at direct laryngoscopy. Cormack and Lehane2 described the commonly used classification of the best view of laryngeal structures seen at direct laryngoscopy. Grade 1 is visualization of the entire laryngeal aperture, grade II is visualization of the posterior portion of the laryngeal aperture, grade III is visualization of the



468 ] PART 8 PERIOPERATIVE MANAGEMENT epiglottis only and grade IV is no view of any laryngeal structures. Difficult laryngoscopy (and therefore difficult intubation) indicates that it is not possible to see any portion of the vocal cords (grade III/IV) after multiple attempts at conventional direct laryngoscopy.3 A final method of defining difficult intubation is through the need for specialized equipment, often taken as requiring an intubation device other than the standard Macintosh or straight-blade laryngoscope. This has some practical significance because when an anaesthetist fails to intubate by direct laryngoscopy (equipment present in each operating theatre) it is often necessary to use equipment from a difficult intubation trolley located centrally in a theatre complex. Difficult direct laryngoscopy is only one cause of difficulty with intubation. It may be easy to visualize the larynx but intubation is unsuccessful because the larynx, subglottis or trachea are abnormally narrowed or distorted.



PREVALENCE Both difficult intubation and difficult ventilation are uncommon. The prevalence of Cormack and Lehane laryngoscopic grade III is 1.5 percent in the general population. Requiring more than three attempts at direct laryngoscopy occurs in approximately 0.4 percent patients and the average anaesthetist will abandon intubation in approximately 1:2500 general surgical patients. In 1200 consecutive ENT and general surgical patients the overall prevalence of difficult intubation (defined as requiring specialist equipment) was 4.2 percent.4 The highest prevalence was 12.3 percent in ENT cancer surgery, 3.5 percent in ENT noncancer surgery and 2.0 percent in general surgical patients. This confirms the clinical impression that difficult intubation is more common in patients undergoing head and neck surgery, particularly in those patients following extensive surgery, flap reconstruction and postoperative radiotherapy or with an obstructed airway. [**] It is difficult to know the precise prevalence of difficult face mask ventilation. Catastrophic failure leading to serious morbidity or mortality is generally quoted as 1:10,000–1:100,00. However, in any large series, a number of problem patients are identified preoperatively and do not receive a general anaesthetic. A North American study involved 18,500 patients of whom 18,200 were intubated under general anaesthesia with 1.8 percent requiring more than two attempts at direct laryngoscopy and no patient being impossible to mask ventilate.5 Approximately 300 patients underwent awake intubation and it is this group of patients that is likely to contain those who would have proved difficult to ventilate if anaesthetized. A study of 1502 patients determined a prevalence of difficult mask ventilation of 5 percent but the definition used was that the anaesthetist considered the difficulty was clinically



relevant and could have led to potential problems if mask ventilation had to be maintained for a longer period.6



EVALUATION The aim of airway management is to adequately maintain and protect the airway by use of the face mask, laryngeal mask or tracheal intubation, and preoperative evaluation seeks to initially determine which airway device is required and whether there will be any difficulties in the use or insertion of it. The face mask provides no airway protection and is little used in anaesthesia for head and neck surgery, except for operations such as insertion of grommets. The decision as to whether to use the laryngeal mask or tracheal intubation is taken after considering such factors as the length of surgery, surgical access, requirement for positive pressure ventilation and risk of airway soiling from either blood, pus, cerebrospinal fluid or gastric contents. The airway may be evaluated according to the scheme outlined in Table 39.1.



History The anaesthetic or hospital notes may indicate previously encountered difficulty with airway management. The patient may pass on verbal or written information from a previous anaesthetist that they are difficult, or difficulty may be inferred from a history of displaced front teeth, bruised lips, excessive sore throat or an unexpected stay in ITU. Past surgery or radiotherapy, or the current surgical condition, may be relevant if it affects the head, neck or mediastinum. A number of medical conditions, such as rheumatoid arthritis, obstructive sleep apnoea and acromegaly, have some association with difficult airway management. In a prospective study of 128 patients with acromegaly, laryngoscopy was difficult (laryngoscopic view grade III) in 10 percent.7 This indicates that the prevalence is six to eight times higher than in normal patients but 90 percent of acromegalics are still easy to intubate. [**] There are a number of congenital Table 39.1



Scheme for evaluation of the airway.



Evaluation History



Examination Investigations



Previous airway difficulty Previous surgery Current surgical condition Current medical condition General Specific predictive tests MR imaging CT imaging Flow-volume loop Flexible nasendoscopy



Chapter 39 Recognition and management of the difficult airway



conditions, such as Treacher-Collins and Pierre-Robin, in which airway management, particularly intubation, is often difficult. Figure 39.1 illustrates a patient with Hunter’s syndrome in which abnormal mucopolysaccharide is deposited in the tissues. Characteristically, he was difficult to intubate for a tonsillectomy to alleviate obstructive sleep apnoea and required an emergency tracheostomy in the recovery period. He is pictured in his late teens when his original standard tracheostomy tube had been replaced by one designed to circumvent lower tracheal and carinal deposits. This T-Y silastic stent passes from just below the vocal cords into each main bronchus (an inverted Y shape) with a limb passing out through the tracheostomy (the T component).



] 469



cricothyroid membrane, limited mouth opening, limited head/neck movements;  voice change, shortness of breath, stridor, inability to lie down. A number of these factors, such as the appreciation of a short neck or receding jaw (Figure 39.2), are subjective. This does not diminish their importance since professional judgements may often be subjective. However, there is a vast literature on prediction of difficult intubation by specific or objective tests. A number have been introduced and the five tests most commonly used are gape, jaw slide, thyromental distance, Mallampati and atlanto-occipital movement. GAPE



Examination General examination of the patient looks for the features in the following list and the practitioner may be alerted to possible difficulties by various findings:  trauma, burn, swelling, infection, scarring, haematoma of the mouth, tongue, larynx, trachea or neck;  large tongue, receding jaw, high-arched palate, prominent upper incisors, short thick neck, large breasts, microstomia, fixed larynx, impalpable



Figure 39.1



Mucopolysaccharidosis (Hunter’s syndrome).



Gape is the measurement of maximal mouth opening and is usually expressed as interincisor distance in fingerbreadths or centimetres. Normal values are 3 fb or 5 cm. A mouth opening of 2 fb is limited and 1 fb is severely limited making direct laryngoscopy very difficult. It is difficult to insert a laryngeal mask when the gape is less than 1–1.5 cm (Figure 39.3). JAW SLIDE OR MANDIBULAR PROTRUSION



Functions are graded as follows: class A if the lower jaw can be protruded beyond the top teeth; class B if the lower



Figure 39.2 Receding jaw.



470 ] PART 8 PERIOPERATIVE MANAGEMENT



Figure 39.4 A normal thyromental distance.



Young arbitrarily introduced a subdivision of class 3 (tongue against soft palate) and class 4 (tongue meets hard palate). ATLANTO-OCCIPITAL MOVEMENT Figure 39.3



Limited mouth opening due to dental abscess.



teeth can only reach the top teeth; and class C if the lower teeth will not reach the top teeth. The value of testing this function is that, in intubation by direct laryngoscopy, the lower jaw must slide forward. In some scoring systems, mouth opening and jaw slide are combined, the greatest difficulty indicated by a gape of o3.5 cm and class C jaw slide. Another method of testing mandibular protrusion is the upper lip bite test8 in which the patient demonstrates how much of the upper lip may be covered by the lower incisors. Class 1 indicates ‘biting’ above the vermilion line, class 2 below the vermilion line and class 3 inability to bite the upper lip. THYROMENTAL DISTANCE



This measurement, described by Patil, is from the mentum to thyroid notch in full neck extension. The normal measurement is 6–7 cm or 3 fb (Figure 39.4). A short distance (2 fb) indicates a ‘high’ larynx and difficult direct laryngoscopy.



MALLAMPATI



Mallampati’s contribution was a test to assess oropharyngeal space. The test asks a seated patient to open their mouth fully and extend the tongue maximally. The practitioner notes which posterior pharyngeal structures are visible. Mallampati described only three grades: class 1 indicates that the posterior pharyngeal wall, fauces and uvula are visible; class 2 indicates that only a part of the fauces, posterior wall and uvula are visible; and class 3 indicates that the tongue meets the palate. Samsoon and



Optimal head and neck positioning for direct laryngoscopy requires cervical spine flexion and almost maximal extension of the head on the spine. A simple clinical test to look at atlanto-occipital movement is for the clinician to ask the patient to flex their neck maximally and then nod; the observer’s hand placed posteriorly on the neck makes certain that the nodding motion is at the level of the cranium on the upper cervical vertebrae. Atlantooccipital extension may be clinically graded as normal or reduced or measured and the normal value is 351.



Investigations Plain x-rays may demonstrate abnormalities, such as enlargement of the retropharyngeal space (Figure 39.5), a swollen epiglottis, tracheal deviation or narrowing, a radio-opaque foreign body or obstructive emphysema suggesting a ball-valve obstruction in the relevant bronchus. However, imaging of the whole airway by CT or MR scan is better and will show narrowing or distortion and allow planning of airway instrumentation. Flexible nasendoscopy under topical anaesthesia is extremely useful in delineating supra- or glottic pathology, and a longer flexible fibrescope can inspect the whole respiratory tract although generally this requires sedation. Another useful test is the flow-volume loop. This expresses flow during expiration and inspiration as a function of lung volume. Airflow is measured during inhalation from residual volume to total lung capacity and exhalation back to residual volume. Extrathoracic obstruction causes limitation of inspiratory flow whilst intrathoracic obstruction causes limitation in expiratory flow. Limitation in expiratory flow is particularly noticeable because the highest flow-rates are usually present in peak expiration.



Chapter 39 Recognition and management of the difficult airway



Figure 39.5



Retropharyngeal abscess.



PREDICTION OF DIFFICULTY The practitioner forms a professional judgement as to whether airway management plans need to be altered from that carried out normally. It is easy to be definite or accurate in prediction when there are ‘barn-door’ abnormalities in the past anaesthetic history, past surgical history, examination and current disease process or from imaging. Examples would be a known history of failed intubation, presentation with breathing difficulty or stridor, absent mouth opening or previous head and neck reconstructive surgery. In a study of 181 patients with pharyngolaryngeal disease,9 the single most predictive factor of difficult intubation was a tumour in the supraglottic region. In a study of 320 patients with a goitre undergoing thyroid surgery,10 the presence of tracheal compression, presence of dyspnoea or a cancerous goitre were the three major predictive factors. [**] However, when there are no abnormalities in the anaesthetic, medical or surgical history, the presenting disease process does not affect the head, neck or mediastinum and the patient does not ‘look’ difficult, then it is not possible to predict difficulty accurately. Most attention has been on predicting difficult direct laryngoscopy using various specific or predictive tests



] 471



(described above), combination of tests and scoring systems. All are imperfect and the reason is partly the low prevalence of difficult intubation. Test sensitivity indicates the ability of the test to label a difficult patient as difficult, test specificity the ability to label a normal patient as normal and the positive predictive value (PPV) is the proportion of patients found to be difficult out of all patients predicted by that test to be difficult. Table 39.2 shows values of test sensitivity, specificity and PPV for various tests. It can be seen that an individual test, such as Mallampati, has a low PPV indicating that most patients predicted to be difficult will, in fact, be normal. A study comparing the upper lip bite test with the Mallampati in 1425 patients concluded that ‘both tests are poor predictors as single screening tests’.11 The more tests that are abnormal increases the likelihood that the patient will be difficult to intubate. Another approach is to produce a score from consideration of various predictive tests, with appropriate weighting. In Wilson et al.’s risk sum,12 five aspects of examination are used – weight, head and neck movement, jaw movement, receding mandible and buck teeth. Each factor is allocated a score of 0, 1 or 2 depending on severity. Wilson’s group suggested that a total score of 2 or more would provide a sensitivity of 75 percent and specificity of 85 percent. These figures may appear to be good but one-quarter of difficult patients will be missed and there will be 1500 false alarms per 10,000 patients. A more recent French study4 produced a scoring system involving seven factors – previous difficult intubation, pathologies associated with difficult intubation, clinical symptoms of airway pathology, interincisor gap and mandibular luxation, thyromental distance, maximum range of head and neck movement and Mallampati. The maximum score is 48 and a score of 11 provided the best level of sensitivity and specificity. The subject of prediction of difficulty is fraught with difficulties arising from studies with small numbers of patients, definitions, curious mathematics and inappropriate conclusions. The topic has been elegantly examined in an editorial which provides a review of the mathematics and an extensive list of references.13 A recent meta-analysis of bedside screening tests for predicting difficult intubation in apparently normal patients concludes that they have limited value.14 [**]



Table 39.2 Test sensitivity, specificity and positive predictive value from the literature and reference 13.13 Test Thyromental Mallampati Wilson risk sum Mouth opening Neck movement



Sensitivity % 65–91 42–56 42–55 26–47 10–17



Specificity % 81–82 81–84 86–92 94–95 98



PPV % 8–15 4–21 6–9 7–25 8–30



472 ] PART 8 PERIOPERATIVE MANAGEMENT It must be remembered that the objective tests aim only to predict difficult direct laryngoscopy when the tongue has a normal compliance and the respiratory tract is normal. Fixation or restricted movement of the tongue was thought to be the direct cause of failed conventional intubation in a series of five patients15 and an abnormality, such as a vallecular cyst, may be symptomless and will not be predicted but may cause great problems with airway management.16 A number of patients who had been found to be unexpectedly difficult to intubate were reviewed and it was discovered that they all had lingual tonsillar hypertrophy.17 Airway evaluation is an essential part of preoperative assessment. It may be rewarded by the detection of severe or obvious problems that necessitate a change from ‘normal’ management. When there are no obvious problems, evaluation is imperfect and safe airway management in all circumstances depends on the adoption of an airway strategy that is able to respond to unexpected difficulty with intubation or oxygenation.



STRATEGY The recommendations of the ASA on management of the airway1, 3 promoted the five-step linear model of



evaluation of the airway, preparation for difficulty, strategy at intubation, strategy at extubation and followup. The ASA difficult airway algorithm presents an overall scheme of planning airway management and a recent version18 is shown in Figure 39.6. The algorithm indicates that airway difficulty may be predicted or unexpected and, in the unexpected limb, attention must first address oxygenation before employing alternative means of intubation. Strategy indicates a combination of plans, also known as plan A/plan B at both initial instrumentation of the airway (induction or intubation) and at the end of surgery (eduction or extubation). Each plan at the start of anaesthesia addresses ventilation of the patient, the abolition of laryngeal reflexes (if required), the airway device and method of insertion, and the abolition of patient distress. The default strategy when intubation is required is shown in Table 39.3. During intubation, there is no ventilation and the patient is oxygenated by intermittent face mask ventilation. Laryngeal reflexes are abolished by muscle relaxation, patient distress is overcome by general anaesthesia and intubation is by optimal direct laryngoscopy. The components of optimal direct laryngoscopy are optimal head and neck position, muscle relaxation, external laryngeal manipulation, laryngoscope blade length/design and use of the gum-elastic bougie. The bougie is the most useful



Difficult airway Recognized Unrecognized Un



co op pa er at tie ive nt



Proper preparation



General anaesthesia +/− paralysis



LMA, Combitube, TTJV



No



Mask ventilation



Emergency pathway Awaken



Surgical airway



Yes nonemergency pathway



Awake intubation choices



Intubation choices Fail



Intubation choices



Regional anaesthesia



Succeed Succeed



Surgical airway



Cancel case, regroup



Fail Confirm Awaken



Surgical airway



Anaesthesia w/ mask ventilation



Figure 39.6



ASA difficult airway algorithm. Redrawn from ref. 18, with permission.



Extubate over jet stylet



Chapter 39 Recognition and management of the difficult airway Table 39.3



Figure 39.7 intubation.



Plan A



Plan B



ALTERNATIVE TECHNIQUES FOR TRACHEAL INTUBATION



Face mask inflation Muscle relaxation



Failed ventilation Muscle relaxation



When intubation by direct laryngoscopy fails, a number of alternative techniques should be considered.



Default airway strategy for intubation.



Means of ventilation Abolition of laryngeal reflexes Device/technique for intubation Control of patient distress



] 473



Optimal direct Failed intubation laryngoscopy General anaesthesia General anaesthesia



The gum-elastic bougie is a very effective aid to



‘low-tech’ device (Figure 39.7) and it proved successful in 80 percent of the unexpected difficult intubations in one series of 11,257 intubations.19 Its correct use during attempted direct laryngoscopy is by placement through the glottis, either blindly or by educated guess, confirming placement by feeling the tracheal rings and hold-up at 25 cm when the bougie contacts the carina, followed by railroading the tracheal tube with the laryngoscope still in situ. The components of optimal direct laryngoscopy indicate that a limited number of attempts are possible, perhaps to change the length of blade, use the bougie or apply external laryngeal pressure. However, generally, no more than three attempts should be made and recent evidence in 2833 critically ill patients indicates increased morbidity with more than two attempts.20 The correct decision is to acknowledge ‘failed intubation’ and abandon plan A. Within default strategy, plan B addresses both failed ventilation and failed intubation by using alternative techniques. In response to the request by the Royal College of Anaesthetists for the display in each hospital of strategies to combat unexpected failed intubation and/or failed ventilation, the Difficult Airway Society (www.das.uk.com) has produced guidelines and flowcharts for the UK.21 [*]



Other laryngoscope blade design The Macintosh curved design is the blade most commonly used by anaesthetists. However, it has a lower success rate than straight blade direct laryngoscopy by the right paraglossal approach. Recently, a British enthusiast, John Henderson, has designed his own straight blade and makes a good case for the reintroduction of the straight blade into mainstream anaesthesia.22 It is a similar technique to that employed by head and neck surgeons, and there seems to be little reason that anaesthetists and surgeons should view the larynx by differing techniques. Another blade, originally invented by McCoy, employs a lever to increase markedly the angulation of the tip (Figure 39.8). A number of direct laryngoscopy blades have been introduced over the years incorporating prisms or mirrors to try and provide an indirect view of the larynx but these have not become mainstream devices.



Rigid fibrescopes Two notable rigid fibrescopes, Upsherscope (Figure 39.9) and Bullard, have been introduced. A fibreoptic viewing bundle is incorporated into a rigid curved blade. The tracheal tube is loaded into or onto the device, the blade is introduced into the pharynx and is then positioned under fibreoptic vision. When the tip of the scope has been positioned directly in front of the larynx, the tracheal tube is passed out of its guiding channel or over an introducer into the trachea. The Bullard appears to be the more popular device and has the support of a Canadian consensus group.23 The device is expensive, training is required and it has not become widely popular in the UK.



Flexible fibrescopes Tracheal intubation using the flexible fibrescope was first described in 1967 by Peter Murphy,24 an anaesthetic senior registrar working at the National Hospital, Queen Square, London, and flexible intubating fibrescopes have been commercially available for over 20 years. Several textbooks25, 26 are concerned solely with the technique. A standard adult intubating fibrescope has a length of 60 cm and a nominal external diameter of 4 mm. Its narrow diameter allows it to pass through the nose or mouth, its flexibility allows it to conform to the anatomy of the patient, the working channel can be used to instil local anaesthesia, oxygen or to pass wires in the antero- or



474 ] PART 8 PERIOPERATIVE MANAGEMENT



Figure 39.9 Using the Upsherscope. Figure 39.8 McCoy blade with a hinged tip.



retrograde direction to aid intubation, the technique is visual and can be used to confirm correct positioning of the tube in the trachea. It is fairly easy to use anaesthetic breathing system connectors or attachments which allow concurrent ventilation of the patient during intubation, or specially adapted airways to make oral intubation easier. It is not surprising that such an instrument has become the safest and most successful technique of intubation. A Swiss study27 examines the results of 13,248 intubations from one hospital in which the fibrescope is used widely as the primary or back-up intubation technique. The failed intubation rate was 0.045 percent. [**] Ovassapian28 gives figures of 98.4 percent success rate in over 2,000 fibreoptic intubations. The complication rate is extremely low. Fibre-endoscopic skills appear in the core competencies for trainee anaesthetists and include visual inspection of the respiratory tract for diagnostic purposes and placement of double lumen tracheal tubes. The problems associated with the intubating fibrescope are that some skill is necessary, the device is expensive and repays careful handling, and disinfection requires chemical agents. There has been particular concern over the inability of cold sterilizing agents to destroy prions. The recommendation from the Department of Health is that a



register should be maintained such that all patients treated with an individual fibrescope can be traced easily.



Intubating laryngeal mask airway The intubating laryngeal mask was devised by Brain and introduced in 1999. The intubating laryngeal mask airway (ILMA) kit (Figure 39.10) differs from the classic laryngeal mask (LM) in several ways. The stem is a rigid highly curved metal tube with a handle, there is an epiglottic elevator bar, a ramp at the junction of the stem and bowl directs the tube appropriately and it is supplied with a special wire-spiral tracheal tube with novel bevel. The described technique of intubation in the anaesthetized, paralyzed patient is for the mask to be placed and the patient ventilated through it. Mask placement appears to be easy provided the mouth opening is more than 2 cm. The tracheal tube is inserted through the stem and advanced slowly without force. As the tip of the tube emerges from the stem, it lifts the epiglottic elevator bar and the route is now clear for the tube to be advanced into the trachea. This blind intubation has a success rate of 95 percent or so if two to three manipulations of mask position and tube advancement are allowed. A fibreoptic modification can be used which allows a visually guided technique and would seem to be preferable. In the largest



Chapter 39 Recognition and management of the difficult airway



] 475



Table 39.4 Role of the classic LM in difficult airway management. Role of the classic LM



Figure 39.10 airway.



Components of the intubating laryngeal mask



series of difficult intubations published so far,29 the ILMA was used in 254 patients with a known difficult airway, including 50 patients with airway distortion due to tumour, surgery or radiotherapy. The successful intubation rate was 96.5 percent with a blind technique and 100 percent when used with the fibrescope. [**] Another North American study30 compared intubation by the ILMA in anaesthetized patients with awake fibre-optic intubation in 38 patients with suspected difficult intubation and found them both to be 100 percent successful.



Lighted stylets The technique of transillumination of the neck to guide oro- or nasotracheal intubation was described first in 1959. A lighted stylet31 uses the principle of transillumination and takes advantage of the anterior or superficial location of the trachea. A number of commercially available devices have been produced over the years and the most recent, the Trachlight, appears to be the most successful. The tracheal tube is loaded onto the stylet which has a distal bulb, and the stylet is shaped into a hockey-stick. The lighted stylet is introduced into the oropharynx from the side and brought into the midline. The tip of the lightwand is passed around the tongue and a bright, well-circumscribed circle of light seen externally at the level of the hyoid indicates that the tip lies in the vallecula. The tube is advanced into the trachea without resistance. In one large series of anticipated difficult intubation, successful Trachlight use by the inventor resulted in intubation in 99 percent patients with a mean time of 26 seconds.



ROLE OF THE CLASSIC LM The classic LM is the most successful of all supraglottic airway devices for maintenance of the airway. In one study of over 11,000 patients, the incidence of failed placement was 1:600 patients. It is, of course, designed for maintenance of the airway in planned, elective surgery. It



As the desired airway device Instead of a tracheal tube Rescue device in failed ventilation Conduit during emergence Conduit for fibre-endoscopy of the airway Conduit for intubation:



Blind Bougie Fibreoptic-guided bougie Fibreoptic Aintree catheter



has the advantages, when compared with tracheal intubation, of easier placement, no requirement for muscle relaxation and is tolerated in situ by the awakening patient. The major disadvantages are that it does not offer the same level of airway protection as a cuffed tracheal tube against gross gastric regurgitation and it does not traverse the larynx, so is no protection against airway occlusion by glottic or infraglottic pathology. It has, however, proved to be a very useful device in difficult airway management and appears in several places in airway algorithms (Table 39.4).



Intubation via the LM In normal use, the LM should be seated in close proximity to the vocal cords and it is not surprising that it can provide a route for passage of a tracheal tube. There are five methods for this (Table 39.4). A size 6.0 mm will pass through the stem/connector of a size three or four LM and a size 7.0 mm tube will pass through the size five LM. Blind placement, in which the tube is lubricated and advanced blindly, has a success rate of only 50–90 percent and passing a bougie through the LM first has an even lower success rate. Techniques under vision have appreciably higher success rates and flexible fibreoptic techniques are particular useful. The fibrescope may be used to guide placement of a bougie into the trachea with intubation occurring over the bougie. Fibreoptic-assisted techniques are useful because they are visually guided (Figure 39.11). The core technique is to insert the LM, load a 6.0 mm tube onto the fibrescope, pass the fibrescope into the trachea through the stem of the LM and slide the well-lubricated tube into the trachea. It is also possible to ventilate an anaesthetized patient through the LM whilst intubation is in progress. This introduces the concept of the laryngeal mask as a dedicated airway,32 ‘a device used for maintenance of



476 ] PART 8 PERIOPERATIVE MANAGEMENT always be considered. If oxygenation cannot be achieved by face mask or LM, it may be worthwhile attempting intubation by direct laryngoscopy. This must be a brief attempt only and must not delay oxygenation by the next step. Airway deaths, unfortunately, often involve prolonged fruitless attempts to intubate when oxygenation is the immediate necessity. When the anaesthetized patient cannot be oxygenated by face mask or LM, and tracheal intubation is not possible, emergency oxygenation should be attempted directly into the respiratory tract below the level of the vocal cords. Anaesthetists generally use the cricothyroid membrane (CTM) which has a number of desirable properties: Figure 39.11 The laryngeal mask is a good conduit for fibreoptic intubation.



        



superficial; easy landmarks to locate; present in most patients; rarely calcifies; relatively avascular; wide enough to accept 6.0 mm tube; inferior to vocal cords; cricoid ring holds airway open; posterior lamina protects back wall.



There are three types of cricothyrotomy as detailed in the following sections.34



Needle or small cannula cricothyrotomy



Figure 39.12 Aintree catheter – a hollow bougie inserted over a fibrescope.



the airway whilst other airway interventions (e.g. intubation) are in progress’. Development of this concept led to the design of the Aintree catheter,33 a hollow bougie which may be placed over the fibrescope and inserted through the LM into the trachea (Figure 39.12). Effectively, the technique places a hollow bougie under vision into the airway, over which a tracheal tube is railroaded.



FAILED VENTILATION AND EMERGENCY CRICOTHYROTOMY Failed ventilation refers to the situation where a patient has been anaesthetized and muscle relaxants administered but it is not possible to provide positive pressure ventilation by use of the face mask and oral airway. A prepared sequence of steps should commence to provide rescue oxygenation as quickly as possible. Bag/mask ventilation in which two hands are used to try to maintain airway patency and another hand squeezes the anaesthetic reservoir bag should be followed by insertion of a classic LM. This may prove life-saving and must



A narrow-calibre rigid needle or flexible cannula is inserted through the CTM in a caudad direction. In adults, an appropriate size is 14 G with an internal diameter of 2 mm (Figure 39.13). The resistance to flow through such a calibre is high and this has implications for inspiration and expiration; in inspiration, adequate gas flows cannot be obtained by the pressures generated within a standard anaesthetic breathing system and exhalation of 500 mL takes >30 seconds. Inspiratory gas flows of 500 mL/s require oxygen at a pressure of 2–4 bar supplied by a Sanders injector (Figure 39.14). The Sanders injector attaches to the 4 bar (400 kPa, 4 atmospheres) oxygen pipeline and has a hand-operated lever to control gas flow during inspiration. Exhalation occurs through the upper airway and particular attention must be taken to ensure that this happens, otherwise airway pressures rise and pulmonary barotrauma develops. A needle cricothyrotomy is a temporary measure and consideration must be given to creation of a tracheostomy in relevant circumstances. It is possible for ventilation via needle cricothyrotomy to be used in the planned elective or urgent case. Figure 39.15 shows a patient whose partial denture had fallen into the pharynx in such a position that the larynx was obscured. The cricothyrotomy needle was placed in the awake patient and its correct position in the trachea confirmed by aspiration of free air and attachment of the capnograph to show an appropriate trace. Following intravenous anaesthesia and



Chapter 39 Recognition and management of the difficult airway



] 477



Figure 39.13 Ravussin style cricothyrotomy cannula. Figure 39.15 Planned use of transtracheal jet ventilation.



Surgical cricothyrotomy



Figure 39.14 Sanders injector.



muscle relaxation, the Sanders technique of intermittent application of high pressure oxygen provided adequate oxygenation and ventilation. Exhalation through the upper airway was unobstructed.



The technique is to make a 3 cm midline incision in the skin followed by a horizontal stab incision in the inferior part of the cricothyroid membrane. The incision is spread horizontally and vertically and a 6 mm tube is inserted. The complication rate was 40 percent in one series of 38 emergency surgical cricothyrotomies. Misplacement and bleeding requiring ligation of vessels were the most common problem. A more rapid four-step technique has been described35 in which a single horizontal incision is made through the skin and cricothyroid membrane together. In a cadaver study, emergency physicians compared the standard and rapid technique. The rapid technique was faster than the standard 43 versus 134 seconds, but the complication rate was higher. The horizontal incison through the skin may cause more haemorrhage. A cricothyrotomy tube should not be left in place for more than a few days and conversion to tracheostomy prevents the complications of dysphonia and subglottic stenosis.



Large cannula cricothyrotomy A purpose-built cannula with an internal diameter of >4 mm allows adequate inspiratory gas flows with the pressures generated by the standard breathing system and exhalation of 500 mL takes approximately 5–6 seconds. This is an advantage because there is no requirement for high pressure oxygen. It is easy to attach the capnograph to the circuit for confirmation of correct positioning within the trachea, and to suction the respiratory tract. Exhalation occurs through the cannula, even in the presence of complete upper airway obstruction. The cannula may be a cannula-over-needle or a Seldinger-type dilatational device. Cannula-over-needle devices place an uncuffed tube whereas the Melker Seldinger device allows positioning of a 6.0 mm cuffed tube. There are advantages in placing a cuffed tube in allowing controlled positive pressure ventilation and protecting the airway.



OBSTRUCTED AIRWAY The obstructed airway36 is one in which the primary symptoms or signs are due to narrowing or distortion of the airway. There are two broad clinical presentations. In acute obstruction (Figure 39.16), a previously normal person develops problems over a matter of minutes or hours. The aetiology is usually one of inhaled foreign body or abnormal fluid accumulation, such as blood, pus or oedema. Typical clinical scenarios are infections in the head and neck, postoperative haematoma, and airway swelling secondary to anaphylaxis or angiotensin-converting enzyme inhibitors. The rapidity of onset produces prominent signs of difficulty with breathing and the patients may present in extremis. Imaging of the airway by x-ray, CT or MR scan is often inappropriate because



478 ] PART 8 PERIOPERATIVE MANAGEMENT



Figure 39.16 Patient with Ludwig’s angina.



the patient may be unable to adopt the necessary position to complete the procedure and the radiology suite is not an appropriate location for a patient with a deteriorating airway. The management of acute obstruction includes 100 percent oxygen by face mask, trial of administration of heliox, nebulized adrenaline (epinephrine) if oedema is prominent, drainage of pus, tracheal intubation or emergency cricothyrotomy or tracheostomy. Heliox is a combination of oxygen 21 percent and helium 79 percent and is three times less dense than air because of the low atomic weight of helium (4) compared with nitrogen (14). In airway obstruction, turbulent gas flow is inversely proportional to the square root of density and the use of heliox improves gas flow in turbulent conditions and also promotes laminar flow. It is suggested also that carbon dioxide molecules diffuse four to five times faster through heliox than an equivalent oxygen–nitrogen mixture and this may enhance carbon dioxide excretion in the airway. In chronic obstruction, the airway pathology has developed over a period of weeks or months and is usually due to growth of tissue or to scarring. The slow onset allows development of enlarged intercostal muscle mass and patients can tolerate a very significantly narrowed airway without symptoms. Generally, a patient will not be dyspnoeic at rest until the airway is narrowed



to o5 mm diameter, although they are likely to be short of breath on exercise. The slower time course of obstruction allows complete imaging of the airway and controlled intervention. Appropriate imaging in the stable, chronic condition is through flexible nasendoscopy and CT or MR imaging of the entire airway. Flow-volume loops may be helpful in detecting that a patient’s problem is large airway narrowing and not truly pulmonary, or assessing the degree of tracheal narrowing before planning surgery. Evaluation of the obstructed airway seeks to define the degree of obstruction, likely site, rapidity of onset and likely time course of deterioration by history, examination and special investigations. The history from the relative or patient usually indicates whether this is acute, chronic or acute-on-chronic. Specific symptoms/signs in airway obstruction are degree of difficulty with breathing, stridor (noisy breathing) and phase of respiration of any stridor. Stridor indicates that the airway is narrowed to o50 percent normal diameter, but this degree of airway narrowing is not always accompanied by stridor. Expiratory stridor indicates an infraglottic problem and inspiratory stridor a supra or glottic aetiology. There may be a positional aspect to the difficulty with breathing and patients may be more comfortable in the sitting or lateral position. Voice change and difficulty with swallowing may indicate the site or extent of disease, and pyrexia or sepsis indicates an infective component. Unilateral reduction in breath sounds in the chest indicate specific bronchial obstruction. With progressive obstruction, signs of ineffective ventilation and poor gas exchange are present. These include agitation, anxiety, confusion, restlessness and depressed level of consciousness. Increased work of breathing is indicated by a high respiratory rate, use of accessory muscles of respiration, flaring nostrils, sweating and tachycardia. Oximetry is not a good monitor of work of breathing and normal oxygen saturation does not indicate that all is well. In severe acute obstruction with untrained respiratory muscles, exhaustion occurs relatively quickly with a resulting decrease in effective minute ventilation, hypoxaemia, bradycardia and death.



Management strategy in the obstructed airway The Confidential Enquiry into Perioperative Deaths (CEPOD) report37 1996/97 examined 30 patients who had died within 30 days of surgery for management of lesions in the upper airway or who presented with stridor. In most cases death was due to the underlying disease process but in some cases difficulties with maintenance of the airway led to death in the operating theatre or recovery unit. Brief case scenarios are given in the report and one describes a 76 year-old patient with fixed cervical spine flexion (due to rheumatoid arthritis) and a retrosternal goitre. After induction of anaesthesia with thiopentone and muscle relaxation with suxamethonium,



Chapter 39 Recognition and management of the difficult airway



the patient proved impossible to intubate and mask ventilate adequately. An emergency tracheostomy proved difficult and cardiac arrest occurred in theatre. The surgical and anaesthetic assessors made several good recommendations. The patients should be seen by the consultant surgeon and anaesthetist who consider carefully the plan for securing the airway. Imaging of the airway should be obtained, if possible, to allow delineation of whether the airway is narrowed at supraglottic, glottic or infraglottic level. Airway strategy should include a primary and back-up plan with coherence between the individual plans. If a long-acting muscle relaxant is used in plan A, plan B cannot rely on a patient ‘waking-up’. If a surgical tracheostomy is the back-up plan, anaesthesia should start in the operating theatre with the equipment, surgeon and theatre team immediately ready – it is foolish to start anaesthesia in the anaesthetic room and scramble into theatre when plan A fails. [*] There has been much discussion in the anaesthetic literature about provision of anaesthesia in the presence of the abnormal airway but general agreement that classification according to the site of obstruction is helpful. Currently, the terms used by anaesthetists are not as precise as the ones used by surgeons to localize tumours. ‘SUPRAGLOTTIC’ OBSTRUCTION



In ‘supraglottic’ obstruction the obstruction is above or superior to the glottis and the glottic aperture is of normal dimensions. An example would be a base of tongue tumour (Figure 39.17), although clearly this is not truly a supraglottic structure in disease terminology. Direct laryngoscopy is likely to prove difficult and may cause bleeding if the blade contacts the tumour. It may also prove difficult to ventilate by face mask following induction of general anaesthesia. In this circumstance it is helpful to retain spontaneous respiration and to use an alternative means of intubation. Awake fibreoptic intubation (Figure 39.18) by the nasal or oral route has much



Figure 39.17 Magnetic resonance image of tongue base tumour.



] 479



to recommend it38 and was used successfully in a series of 26 adult patients with deep neck infections including classical Ludwig’s angina.39 [**] Awake intubation is a misnomer because it is very difficult to intubate a truly awake patient. The more correct term is tracheal intubation under topical anaesthesia with conscious sedation. It has a very good record of safety in airway management because the patient maintains their own airway and continues with spontaneous respiration until the airway is secure. It is possible for the patient to adopt a change in position, such as sitting up, and to aid intubation by protruding their tongue, vocalizing or taking deep breaths. There are a number of intubation techniques in an awake patient, such as oral direct laryngoscopy, blind intubation through the nose, through the classic or intubating laryngeal masks or a retrograde wire technique. Awake fibreoptic intubation appears to be the best technique, combining awake intubation with a visually guided method of both inspecting and intubating the airway, applying local anaesthesia to the airway and confirming correct positioning of the tube within the trachea. Fibreoptic intubation is generally easier in the awake rather than anaesthetized, paralysed patient because airway patency is maintained, the airway opens and closes with respiration and the flow of gas indicates the route to the larynx. There are a number of specific practical steps and attention to detail is required. Premedication may be employed if there is no airway embarrassment, but should be avoided if there are symptoms or signs of airway obstruction. An antisialogogue is important and absence of secretions allows earlier and more profound topical anaesthesia, and easier fibre-endoscopy. Preoperative



Figure 39.18 Awake fibreoptic intubation.



480 ] PART 8 PERIOPERATIVE MANAGEMENT intramuscular atropine 0.6 mg, hyoscine 0.4 mg or glycopyrrolate 0.4 mg are suitable but it is common to administer glycopyrrolate 0.2–0.4 mg intravenously as soon as the patient arrives in the operating room. Sedation aims to provide a comfortable patient who maintains spontaneous respiration, airway patency and verbal contact. Small incremental doses of a benzodiazepine and opioid are administered, taking care not to produce oversedation. Typical total doses are midazolam 1–5 mg and fentanyl 25–100 mg. Both drugs have a peak onset of five minutes and a specific antagonist (flumazenil and naloxone, respectively). Benzodiazepines cause relaxation of genioglossus and destabilization of the airway and opioids are associated with central respiratory depression so neither drug is benign. However, appropriate doses produce a compliant patient who is not unduly upset by airway topicalization or instrumentation, and is usually amnesic. Propofol, an anaesthetic agent, provides sedation at subhypnotic concentrations and is particularly useful when given by target-controlled infusion. Target controlled infusion (TCI) devices are sophisticated syringe pumps which incorporate a pharmacokinetic model of the relevant drug. The operator enters the age and weight of the patient and the desired blood level. The TCI pump calculates and delivers the appropriate bolus dose to reach the required blood level and the subsequent infusion required to maintain it. Appropriate starting blood levels for propofol are 0.5–1.0 mg/mL. Remifentanil, an ultrashort-acting opioid, may also be used by TCI and is gaining a reputation as a useful opioid for awake intubation. If the nasal route is chosen for intubation, a topical vasoconstrictor should be applied. This may be xylometazoline or ephedrine 0.5 percent drops or cocaine (3 mL 5 percent), which produces both vasoconstriction and topical anaesthesia. Topical anaesthesia may be provided by:    



nebulization; translaryngeal administration; specific nerve blocks; transendoscopic administration.



Nebulization sounds attractive but use of the technology which provides particles for alveolar deposition of drug (e.g. salbutamol nebulizer), may lead to disappointing results partly because the particles are too small but also because a large amount of the drug escapes to the atmosphere. High-drug concentrations may be effective and nebulized lignocaine 10 percent in a dose of 6 mg/kg has been described as effective. A useful variant is the production of a larger droplet size produced during inspiration. This manual spray-and-inhale technique employs a 22 g Venflon attached to a constant oxygen flowrate of 1–2 L/min. Small increments of lignocaine 4 percent are injected through the port of the cannula and coordinated with inspiration. One spray-and-inhale regime is to use 3 mL lignocaine 4 percent with ephedrine 15 mg to the nose, wait three minutes and apply the



second 3 mL lignocaine 4 percent, asking the patient to inspire deeply and slowly through the nose. The droplets are inhaled and deposited onto the larynx and trachea. After a three minute wait, a further 3 mL lignocaine 4 percent is nebulized during slow forced inspiration. Maximum lignocaine dosage should be 9 mg/kg. Translaryngeal administration has a long history of safe use. A study reviewing 17,500 administrations reported only six noteworthy complications.40 A 22 G cannula or needle is passed through the cricothyroid membrane or trachea and 3–4 mL lignocaine 4 percent is injected preferably at end-expiration. The injection provokes a short period of intense coughing which distributes the drug to the glottis and above. The experimental addition of methylene blue to the local anaesthetic shows staining of the superior aspect of the vocal cords in 95 percent of patients. Appropriate specific nerve blocks are of the superior laryngeal and glossopharyngeal nerves. The internal branch of the superior laryngeal nerve supplies sensation to the under surface of the epiglottis and the superior surface of the vocal cords. It may be blocked on each side as it traverses the thyrohyoid membrane. Extension of the head and neck aids identification of the hyoid and thyroid cartilages. A 22 G needle is placed inferiorly to the greater horn of the hyoid, passed into the membrane and 2 mL lignocaine 2 percent is injected. The glossopharyngeal nerve supplies sensation to the posterior third of the tongue, superior part of the epiglottis, lateral pharyngeal wall and inferior surface of the soft palate. The nerve may be blocked behind the anterior pillar of the tonsillar fossa. With full mouth opening the tongue is grasped and pulled to the contralateral side. A 20 G spinal needle is inserted to a depth of 5 mm into the base of the anterior tonsillar pillar at the level of the reflection onto the tongue, and 2 mL lignocaine 2 percent injected. Bilateral glossopharyngeal nerve blocks will abolish the gag reflex and allow greater manipulation in the oropharynx or direct laryngoscopy, perhaps when placing a large double lumen tube. These specific nerve blocks are not performed routinely. Transendoscopic administration of lignocaine 4 percent through the working channel of the intubating fibrescope is an extremely effective means of applying local anaesthetic to the airway under vision. This sprayas-you-go technique is highly favoured and noninvasive. The intubating fibrescopes have connectors for injection but an easier alternative is to place an epidural catheter, cut to produce one terminal hole, substantially into the working channel (Figure 39.19). Aliquots of lignocaine 4 percent to a maximum dose of 9 mg/kg are administered. It is helpful for the fibrescope to be attached to a CCTV system, particularly for training. An appropriate size tracheal tube is loaded onto the fibrescope and the scope is introduced under vision into the mouth or more patent nostril. The fibrescope is advanced without



Chapter 39 Recognition and management of the difficult airway



Figure 39.19 Injection of lignocaine through an epidural catheter inserted into working channel of fibrescope.



touching the mucosa until the vocal cords are seen. Additional local anaesthetic may be applied before the fibrescope is advanced to the carina. The tube is advanced or railroaded over the fibrescope. This may be difficult because the bevel impinges on the larynx. Use of a small diameter tube and rotation of the tube minimize this problem. Awake fibreoptic intubation may be difficult when the airway anatomy is abnormal and when there is blood or secretions in the airway. An illustrative case scenario indicates some of the pitfalls in management. An adult patient underwent removal of a fishbone impacted in the lower pharynx/upper oesophagus. Three days later the patient was pyrexial, with a sore throat, unable to swallow and with limited mouth opening. A lateral x-ray of the neck showed a retropharyngeal abscess. The patient was seen by two anaesthetic trainees who did not inform the consultant on call. They decided on an awake fibreoptic intubation but administered too much sedation. In a deteriorating situation with a semirousable patient they attempted nasal fibreoptic orotracheal intubation (FOI) with the patient supine. They were unsuccessful and were moved aside by the consultant surgeon who managed to intubate the patient fibreoptically with the patient sitting, leaning forward on the edge of the operating table. Mistakes made here include failure to appreciate the seriousness of the condition and to inform a consultant, failure to realize that awake nasal FOI in the supine position would be difficult in the presence of retropharyngeal swelling and failure to realize that airway patency would be compromised by any sedation.



PERIGLOTTIC/GLOTTIC OBSTRUCTION



In these situations it may be difficult to visualize the vocal cords by direct laryngoscopy and the glottic aperture may be significantly narrowed or distorted. Much lively discussion is evident over the correct anaesthetic technique. When awake fibreoptic intubation is used, it allows visualization of the airway but becomes less useful



] 481



in the patient with stridor due to glottic narrowing. The fibrescope has little rigidity to ‘push’ through a narrow hole and attempts to do this may precipitate bleeding and oedema. If the scope is passed through a small hole, the airway is completely obstructed for a short time and patients feel uncomfortable at this stage. There have been a few reports41, 42 of destabilization of the airway by applanaesthetic agent to the airway and this is a reminder of the need to work at all times in the correct environment for immediate activation of the preformulated back-up plan. Awake fibreoptic intubation is, therefore, a technique which should always be considered but may not be suitable. It is not appropriate when adequate operator skill is not present, in children and uncooperative adults and in the opinion of some anaesthetists when stridor is present. Mason and Fielder43 argue that the correct anaesthetic technique in the presence of stridor due to periglottic/ glottic pathology is inhalational induction of general anaesthesia. [*] This permits a gradual onset of anaesthesia and maintains spontaneous respiration, even at a depth of anaesthesia appropriate for direct laryngoscopy and intubation. The maintenance of spontaneous respiration is viewed as highly desirable with supraglottic airway obstruction. A typical case scenario would be a child with epiglottitis and, in a survey of college tutors in the UK, 98 percent of anaesthetists would choose this form of anaesthesia. The face mask is applied to the patient who is in the most comfortable position (sitting if necessary) and 100 percent oxygen administered for a few minutes. The volatile agent is administered in increasing concentrations until a surgical level of anaesthesia is obtained. The agent commonly used initially is either halothane or sevoflurane and both drugs are nonirritant so do not provoke coughing. The speed of onset of anaesthesia is inversely proportional to the blood–gas solubility and is therefore faster with sevoflurane (0.6) than halothane (2.4). However, it is more difficult to establish sufficient depth of anaesthesia to instrument the airway with sevoflurane and it has a more rapid offset than halothane. Halothane may be associated with increased cardiac rhythm irritability and is now difficult to obtain in the UK. It is usual, therefore, to start with sevoflurane but change over to isoflurane which does permit adequate levels of anaesthesia. It is not an easy anaesthetic to administer in the presence of stridor and requires a sanguine anaesthetist. There may be periods of increasing obstruction due to glottic irritability or change in position. Generally, there should be no change in the position of the anaesthetist’s hands or the face mask and no attempt in light planes of anaesthesia to provide positive pressure ventilation. Insertion of an oral airway is risky but there may be benefit in a nasopharyngeal airway, although it is perhaps useful to have applied a vasoconstrictor to the nasal mucosa first. Glottic irritability is confined to light planes of anaesthesia and should resolve, although induction



482 ] PART 8 PERIOPERATIVE MANAGEMENT may take much longer than normal because of the reduced alveolar ventilation. When an adequate depth of anaesthesia is reached, which may take 20 minutes, direct laryngoscopy is undertaken. The view may be quite abnormal and it may be necessary to press on the chest and observe the egress of bubbles to detect the glottis. A small tube will be needed and the use of a bougie should be considered. If intubation is not possible, the face mask is reapplied and a tracheostomy undertaken. Plan B must be formulated and ready so that anaesthesia is induced in the operating theatre with the surgeon scrubbed and ready to undertake tracheostomy. It is apparent that the safety of any approach is the combination of plans and the close cooperative working of the surgeon, anaesthetist and theatre team. In the common scenario of a known obstructing glottic tumour which requires initial histology and debulking, it is possible to construct a primary plan of (in the operating theatre) preoxygenation, followed by intravenous induction and rapid muscle relaxation. Direct laryngoscopy using a bougie and size 5.0 mm microlaryngeal tube is attempted. If it is unsuccessful, the surgeon is in the best situation to undertake tracheostomy – the patient is as well oxygenated as possible, unconscious and remains still. This technique is logical and popular but appears ‘heretical’ since it abolishes spontaneous respiration. However, it illustrates that safety lies in the combination of plans rather than any particular plan A, that the site of obstruction requires plans which are specific for that level of obstruction and the safety which arises from the close working of experienced surgeon and anaesthetist. There is an increasing evidence-base to the practice of planned prior placement of a transtracheal ventilation catheter under local anaesthesia and using this as a route for oxygenation and ventilation during intubation attempts under general anaesthesia. The technique was used in 11 patients over 22 months in one institution with great success.44 [**]



TRACHEOSTOMY UNDER LOCAL ANAESTHESIA



This should be considered in any patient with an obstructed airway. It is particularly appropriate as the primary plan when the disease process is a large friable mass or abscess in the supraglottis or glottis and intubation attempts may destabilize or compromise the airway. There are differences between countries and between surgeon–anaesthetist pairs as to when the patient undergoes traditional intubation in the awake or anaesthetized state or awake tracheostomy. Elements within the decision making are the availability of a skilled fibre-endoscopist, the ability of the patient to cooperate with the procedure and adopt a suitable position, the pretracheal anatomy and the likely time for resolution of the disease process. It may be a very difficult procedure in



patients with short stocky necks, a previous tracheostomy or post-radiotherapy with respiratory distress. Tracheostomy under local anaesthesia is undertaken in the operating theatre with the patient breathing oxygen or heliox, monitored by noninvasive blood pressure, ECG and pulse oximetry with intravenous access. A semisitting position with a roll under the shoulders and neck extension is ideal. Generally, reassurance is given to the patient but intravenous sedation is not required and should be used cautiously. Restlessness during the procedure may be due to hypoxia, hypercarbia or an inability to breathe in that position. Sedation may destabilize the airway leading to sudden hypoxia and loss of consciousness. The anaesthetist must be prepared for the back-up plan if the patient deteriorates. Placement of a cricothyrotomy needle at the outset may be useful when it is known that the pathology is supraglottic. The cricothyrotomy needle does not interfere with a surgical tracheostomy and can be used to provide oxygenation. If the patient becomes so restless that the surgeon is unable to operate, consideration should be given to providing general anaesthesia and oxygenation through the needle. Another option in the distressed patient is to provide sedation/anaesthesia by addition of a volatile anaesthetic agent, such as sevoflurane, to the breathing system with 100 percent oxygen. At best, the airway proves to be adequate enough to allow a surgical plane of anaesthesia to be reached and the tracheostomy is undertaken in a relatively unhurried fashion on 100 percent oxygen/ sevoflurane by face mask. At worst, the airway deteriorates with the onset of sedation/anaesthesia but the patient stops moving and a rapid emergency tracheostomy can be undertaken. In a rapid tracheostomy, the surgeon enters the airway with one or two incisions and a small cuffed microlaryngeal or armoured tube size 5.0–6.0 mm is inserted. Once the patient has been stabilized, a more measured exploration of the neck and fashioning of a formal tracheostomy may be undertaken. It is important to verify that the tube is within the trachea before ventilation starts otherwise gas may be forced into the mediastinal tissues. Signs of correct placement when undertaking a tracheostomy under local anaesthesia (with the patient breathing spontaneously) are firstly regular respiratory movement of the reservoir bag of the anaesthetic breathing system connected to the tracheostomy tube, and the presence of six successive breath-related carbon dioxide traces on the capnograph. If no carbon dioxide is detected in the breathing system, inflation of the cuff of the tracheal tube or tracheostomy and connection of the capnograph to the breathing system should be checked first. If both are correct, failure to detect carbon dioxide indicates that the tube is not in the airway. When an emergency tracheostomy is undertaken in an apnoeic patient, the confirmatory signs of anaesthetic bag movement and capnography can be obtained only by applying a number of positive pressure breaths. If the tube is not within the trachea, these



Chapter 39 Recognition and management of the difficult airway



positive pressure breaths into the mediastinum may prove deleterious. An alternative confirmatory device in these circumstances, although not widely used, is the oesophageal detector device. The principle is simple and takes advantage of the structural differences between the oesophagus and trachea. In the original version described by Wee in 1988, an empty 60 mL syringe is attached to the 15 mm connector of the inserted tracheal tube and aspiration attempted. Aspiration of air is not possible if the tube is in the oesophagus because the mucosa is ‘sucked’ over the end of the tube, whereas the more rigid cartilaginous structure of the trachea allows free aspiration of air. The syringe can be replaced by a self-inflating bulb with a volume of approximately 75 mL and in this version the bulb is squeezed flat before being attached to the inserted tube. If the bulb reinflates immediately, the tube is in the trachea and if the bulb does not reinflate the tube is in a false passage or the oesophagus. SUBGLOTTIC AND MIDTRACHEAL OBSTRUCTION



Imaging is particularly useful in delineating the length of narrowing, the diameter of the airway at its narrowest and that sufficient distance is present inferiorly to the obstruction to permit the cuff of a tracheal tube to be positioned above the carina. The obstruction may arise from external pressure, such as a retrosternal goitre or other mediastinal mass, from a mass arising from the trachea, from an inflammatory condition such as Wegener’s granulomatosis, from previous surgery or from damage due to prolonged intubation. The type of narrowing may range from a short subglottic stenosis due to previous prolonged intubation to a narrowing of several centimetres in the midtrachea due to tumour. In the presence of stridor, the principles of management vary according to whether it can be bypassed by tracheostomy. This will be true for subglottic disease but the CEPOD assessors noted that in two patients it had been difficult to bypass a mid/low



] 483



tracheal lesion with a standard length tracheostomy tube. Awake fibreoptic intubation has a role in management of trachea narrowing, allowing inspection of the airway and confirmation that the tip of the tracheal tube has passed beyond the obstruction. Rigid bronchoscopy is an extremely effective means of managing these patients.



LOW TRACHEAL OBSTRUCTION



Narrowing of the lower trachea or carina presents great difficulty. The anaesthetic literature contains case reports of failed airway maintenance leading to death.45 This characteristically occurs after induction of general anaesthesia or muscle relaxation when, presumably due to loss of muscle tone, airway patency is lost. Tracheal intubation may not provide an adequate airway because the obstruction is beyond the tip of the tube. Occasionally, the presence of carinal obstruction is not known and anaesthetic induction, intubation or indeed extubation may result in unexpected disaster. When imaging has provided good preoperative localization of obstruction, a number of options may be used. Rigid bronchoscopy is invaluable and will often provide a route for ventilation (Figure 39.20). The rigid bronchoscope may also act as a guide to therapy, such as lasering of a tumour or introduction of a tracheobronchial stent. Surgical resection of carinal lesions requires specialist anaesthetic techniques including jet ventilation and undertaking surgery during cardiopulmonary bypass.



EXTUBATION AND RECOVERY At the end of surgery a decision is made as to where and when the airway device should be removed. There is little problem with removal of a laryngeal mask. This is usually tolerated well by a patient until they are awake. Tracheal intubation is common in head and neck surgery due to



Figure 39.20 Rigid bronchoscopy used for ventilation and conduit for tracheal stent.



484 ] PART 8 PERIOPERATIVE MANAGEMENT the constraints of providing clear operative fields for the surgeon and maintaining a secure airway to the distal trachea. Extubation requires as much thought, and gives rise to as much difficulty, as intubation. Transient difficulties with oxygen saturations are common due to coughing, breath-holding and laryngospasm, particularly in children. Extubation problems may arise in those patients who were difficult to intubate and those who were not difficult to intubate but in whom surgery has affected the airway. Default strategy at extubation when no difficulty is expected is for the anaesthetist to remove any pharyngeal packs, suction the pharynx under direct vision, administer 100 percent oxygen, antagonize residual neuromuscular blockade and consider whether to remove the tube in the anaesthetized or awake state. After extubation, 100 percent oxygen is administered by face mask and the patient observed by the anaesthetist until it is clear that the patient is safe to go to the post-anaesthetic care unit (Recovery). In the patient with a normal airway who was difficult to intubate, it is prudent to make certain that oxygen stores are maximal and to extubate in the awake state. Lung oxygen stores can be considered maximal when the end-tidal (i.e. alveolar) oxygen is 91 percent. This may take at least five minutes of breathing 100 percent oxygen or longer if nitrous oxide has been used. Awake extubation refers to removal of the tracheal tube when the person has opened their eyes and is able to obey commands. An additional option is to assess the leak around the tube before removal. This may be carried out by applying positive pressure to the tube, deflating the cuff and listening for egress of gas around the tube, or alternatively occluding the tube in spontaneous respiration and making certain that inspiration can occur around the tube. Failure of a leak test indicates that the tube is a very tight fit within the airway and an inadequate air passage may be left after extubation. In the patient with an abnormal airway, either present preoperatively or due to surgery, consideration should be given to keeping the tracheal tube in situ for 24–48 hours until any airway oedema subsides. The patient should be nursed in a high-dependency unit with an appropriate level of sedation to avoid inadvertent removal of the tube. In some circumstances it is appropriate to perform a tracheostomy to provide a secure airway in the first few postoperative days. Extubation of the high-risk airway requires a strategy. Awake extubation after maximal oxygenation in the presence of an anaesthetist in a well-equipped environment may be a good plan A, but what happens if it fails? The situation rapidly becomes critical with a struggling, hypoxic patient possibly with blood in the oropharynx. One possibility is to extubate over a thin bougie and to leave the bougie in the airway for a period until it is certain that the patient is coping satisfactorily. The bougie acts as a guide if reintubation is required and, if it is hollow, may be used for emergency oxygenation.



Problems may arise in the Recovery unit or postoperatively on the ward. Of most concern is postoperative bleeding following carotid endarterectomy or removal of a parapharyngeal mass. The physical presence of a mass of blood may compress the airway itself but also induces mucosal oedema, perhaps by impairment of lymphatic drainage. The deterioration of the airway may be very dramatic and necessitate emergency cricothyrotomy or tracheostomy as part of resuscitation. In a dire situation it is always worth fitting an LM. It is helpful to open the wound and evacuate the blood clot and this may provide temporary improvement. The patient is returned to theatre for surgical exploration. Intubation should take place with the patient breathing spontaneously, if possible. This may be by awake intubation or with inhalational anaesthesia. Blood in the pharynx may impair the view and the first response is to try suctioning. An LM may cover the larynx and provide some respite before being used as a conduit for intubation. Emergency tracheostomy may be required (Figure 39.21). In this patient, drainage of a peritonsillar abscess had been undertaken with intubation by awake fibreoptic intubation. The patient had been extubated and returned to Recovery. Approximately 45 minutes later, the patient developed severe breathing difficulties and was returned immediately to the operating theatre. Awake intubation was not possible due to soiling of the airway, the patient was too restless to adopt a position suitable for formal tracheostomy and, in a deteriorating situation, the surgeon managed to carry out a rapid emergency tracheostomy.



FOLLOW-UP Following difficulties with airway management, a certain scheme should be followed. An account of the problem and management should be written in the anaesthetic record and in the hospital notes. The patient needs to be reviewed clinically to detect and treat any morbidity, an explanation is required for the patient with details of the problem encountered and management, and a written



Figure 39.21



Post-extubation emergency tracheostomy.



Chapter 39 Recognition and management of the difficult airway



account should be sent to the patient with a copy to their general practitioner.46 If the problem with airway management is likely to be recurrent with subsequent anaesthetics, consideration should be given to the patient registering with Medic Alert and wearing a bracelet or to registering the patient with the difficult airway database supported by the Difficult Airway Society. [*] Immediate morbidity or mortality from difficult airway management arise from the effects of severe hypoxia, hypercarbia or cardiovascular responses, from failure to adequately protect the airway leading to aspiration and from physical trauma to the airway during attempts at intubation or resuscitation. Airway damage may occur even when airway management has not been notably difficult. Valuable information can be obtained from detailed analysis of the medical information contained in insurance reports, once claims for negligence have been settled or closed. In an analysis of such closed claims in North America,47 6 percent of 4460 claims were for airway injury. The most frequent sites of injury were the larynx (33 percent), pharynx (19 percent) and oesophagus (18 percent). Approximately 20 percent of laryngeal injuries were associated with difficult intubation and included granuloma formation, arytenoid dislocation and hoarseness. Injuries to the pharynx and oesophagus had a much stronger association with difficult airway management. Half of all pharyngeal injuries and 68 percent of pharyngeal perforations were associated with difficult intubation. There were five deaths in the pharyngeal injury claims and all involved perforation and the development of mediastinitis. The oesophageal injuries involved a significantly greater proportion of females and patients older than 60 years than the other sites and oesophageal perforation involved difficult intubation in 67 percent of claims. Oesophageal injuries were the most severe and were associated with a poor outcome with 19 percent mortality. Pharyngo-oesophageal perforation is a serious condition (overall mortality 25 percent) and risk factors include difficult intubation, emergency intubation and intubation by inexperienced personnel. Perforation may also be caused by passage of a nasogastric tube. The triad of surgical emphysema, chest pain and pyrexia should be sought and treatment with antibiotics, limitation of oral intake and surgical review initiated as soon as possible. In the closed claims study, surgical emphysema was only evident in 56 percent of patients and the diagnosis was sometimes delayed. It is suggested that treatment within 24 hours improves outcome. [*]



KEY POINTS  The difficult airway is an important feature in head and neck surgery.  Difficult intubation and difficult mask ventilation are different entities.



] 485



 Evaluation of the airway is imperfect and may fail to predict difficulty.  Airway obstruction may be acute or chronic.  Imaging of the airway is essential, when possible, to delineate the level of obstruction.  Airway strategy means a primary plan A and back-up plan B.  The anaesthetic room is an inappropriate location for plan A if plan B is a surgical tracheostomy.  Pharyngo-oesophageal perforation due to intubation attempts has a high mortality and needs early detection and treatment.



Best clinical practice [ All patients should undergo airway evaluation as part of preoperative assessment.



[ Strategy must cover unexpected failed intubation and failed ventilation.



[ The LM is a versatile airway device and should always be available.



[ Decisions about management of the obstructed [ [ [ [ [ [



airway are made by senior anaesthetists and surgeons. Awake fibreoptic intubation and tracheostomy under local anaesthesia should be considered in the obstructed airway. Maintenance of spontaneous respiration is recommended when general anaesthesia is employed in the presence of upper airway obstruction. Consider placement of a transtracheal ventilation catheter prior to inducing general anaesthesia in the difficult upper airway. Always confirm correct placement of the tube in the trachea. A strategy is required for extubation. Follow-up is important to detect and treat morbidity caused by airway management.



Deficiencies in current knowledge and areas for future research



$ $ $



Unified surgical and anaesthetic terms describing the location of disease pathology in the airway. Randomized comparative studies of anaesthetic techniques in management of the obstructed airway. National collection of serious adverse incidents resulting from airway management in head and neck disease.



486 ] PART 8 PERIOPERATIVE MANAGEMENT



$ $



National audit of attempts at emergency cricothyrotomy. Annual publication of circumstances of death within 28 days of surgery for the obstructed airway.



 14.



15.



REFERENCES







1. Caplan RA, Benumof JL, Berry FA, Blitt CD, Bode RH, Cheney FW et al. A practice guideline for management of the difficult airway. Anesthesiology. 1993; 78: 597–602. 2. Cormack RS, Lehane J. Difficult tracheal intubation in obstetrics. Anaesthesia. 1984; 39: 1105–11. 3. Practice guidelines for management of the difficult airway. An updated report by the American Society of Anesthesiologists Task Force on management of the difficult airway. Anesthesiology. 2003; 98: 1269–77. Definitions and national evidence-based practice guidelines from USA. 4. Arne J, Descoins P, Fusciardi J, Ingrand P, Ferrier B, Boudigues D et al. Preoperative assessment for difficult intubation in general and ENT surgery: predictive value of a clinical multivariate risk index. British Journal of Anaesthesia. 1998; 80: 140–6. 5. Rose DK, Cohen MM. The airway: problems and prediction in 18,500 patients. Canadian Journal of Anaesthesiology. 1994; 41: 372–83. 6. Langeron O, Masso E, Huraux C, Guggiari M, Bianchi A, Coriat P et al. Prediction of difficult mask ventilation. Anesthesiology. 2000; 92: 1229–36. 7. Schmitt H, Buchfelder M, Radespiel-Troger M, Fahlbusch R. Difficult intubation in acromegalic patients. Anesthesiology. 2000; 93: 110–4. 8. Khan ZH, Kashfi A, Ebrahimkhani E. A comparison of the upper lip bite test (a simple new technique) with modified Mallampati classification in predicting difficulty in endotracheal intubation: a prospective blinded study. Anesthesia and Analgesia. 2003; 96: 595–9. 9. Ayuso MA, Sala X, Luis M, Carbo JM. Predicting difficult orotracheal intubation in pharyngo-laryngeal disease: preliminary results of a composite index. Canadian Journal of Anaesthesia. 2003; 50: 81–5. 10. Bouaggad A, Nejmi SE, Bouderka MA, Abbassi O. Prediction of difficult tracheal intubation in thyroid surgery. Anesthesia and Analgesia. 2004; 99: 603–6. 11. Eberhart LH, Arndt C, Cierpka T, Schwanekamp J, Wulf H, Putzke C. The reliability and validity of the upper lip bite test compared with the Mallampati classification to predict difficult laryngoscopy: an external prospective evaluation. Anesthesia and Analgesia. 2005; 101: 284–9. 12. Wilson ME, Spiegelhalter D, Robertson JA, Lesser P. Predicting difficult intubation. British Journal of Anaesthesia. 1988; 61: 211–6. 13. Yentis SM. Predicting difficult intubation – worthwhile exercise or pointless ritual. Anaesthesia. 2002; 57: 105–9.



16.



17.



18. 19.



20.



 21.



22.



23.



24. 25. 26. 27.



28. 29.



30.



Shiga T, Wajima Z, Inoue T, Sakamato A. Predicting difficult intubation in apparently normal patients: a metaanalysis of bedside screening test performance. Anesthesiology. 2005; 103: 429–37. Excellent reference list and analysis of predictive power of individual tests. Rosenstock C, Kristensen MS. Decreased tongue mobility – an explanation for difficult endotracheal intubation? Acta Anaesthesiologica Scandinavica. 2005; 49: 92–945. Kamble VA, Lilly RB, Gross JB. Unanticipated difficult intubation as a result of an asymptomatic vallecular cyst. Anesthesiology. 1999; 91: 872–3. Ovassapian A, Glassenberg R, Rendel GI, Klock A, Mesnick PS, Klafta JM. The unexpected difficult airway and lingual tonsillar hyperplasia: a case series and review of the literature. Anesthesiology. 2002; 97: 124–32. Benumof JL. Laryngeal mask airway and the ASA difficult airway algorithm. Anesthesiology. 1996; 84: 686–99. Combes X, Le Roux B, Suen P, Dumerat M, Motamed C, Sauvat S et al. Unanticipated difficult airway in anesthetized patients. Prospective validation of a management algorithm. Anesthesiology. 2004; 100: 1146–50. Mort TC. Emergency tracheal intubation: complications associated with repeated laryngoscopic attempts. Anesthesia and Analgesia. 2004; 99: 607–13. Henderson JJ, Popat MT, Latto IP, Pearce AC. Difficult Airway Society guidelines for management of the unanticipated difficult intubation. Anaesthesia. 2004; 59: 675–94. Excellent reference list and national airway management guidelines from the UK. Henderson JJ. The use of paraglossal straight blade laryngoscopy in difficult tracheal intubation. Anaesthesia. 1997; 52: 552–60. Crosby ET, Cooper RM, Douglas MJ, Doyle DJ, Hung OR, Labrecque P et al. The unanticipated difficult airway with recommendations for management. Canadian Journal of Anaesthesiology. 1998; 45: 757–76. Murphy P. A fibre-optic endoscope used for nasal intubation. Anaesthesia. 1967; 22: 489–91. Popat M. Practical fibreoptic intubation. Oxford: Butterworth Heinemann, 2001 ISBN 0 7506 4496 6. Hawkins N. Fibreoptic intubation. London: Greenwich Medical Media, 2000 ISBN 1 84110 060 9. Heidegger T, Gerig HJ, Ulrich B, Kreienbuhl G. Validation of a simple algorithm for tracheal intubation: daily practice is the key to success in emergencies - an analysis of 13,248 intubations. Anesthesia and Analgesia. 2001; 92: 517–22. Ovassapian A. Fiberoptic endoscopy and the difficult airway. Lippincott-Raven, 1996 ISBN 0 7817 0272 0. Ferson DZ, Rosenblatt WH, Johansen MJ, Osborn I, Ovassapian A. use of the intubating LMA-Fastrach in 254 patients with difficult-to-manage airways. Anesthesiology. 2001; 95: 1175–81. Joo HS, Kapoor S, Rose K, Naik VN. The intubating laryngeal mask airway after induction of general anesthesia versus awake fiberoptic intubation in patients with difficult airways. Anesthesia and Analgesia. 2001; 92: 1342–6.



Chapter 39 Recognition and management of the difficult airway 31. Davis L, Cook-Sather SD, Schreiner MS. Lighted stylet tracheal intubation: a review. Anesthesia and Analgesia. 2000; 90: 745–56. 32. Charters P, O’Sullivan E. The dedicated airway: a review of the concept and an update of current practice. Anaesthesia. 1999; 54: 778–86. 33. Atherton DP, O’Sullivan E, Lowe D, Charters P. A ventilation-exchange bougie for fibreoptic intubations with the laryngeal mask airway. Anaesthesia. 1996; 51: 1123–6. 34. Vanner R. Emergency cricothyrotomy. Current Anaesthesia and Critical Care. 2001; 12: 238–43. 35. Brofeldt BT, Panacek EA, Richards JR. An easy cricothyrotomy approach: the rapid four-step technique. Academic Emergency Medicine. 1996; 3: 1060–3. 36. Popat M, Dudnikov S. Management of the obstructed upper airway. Current Anaesthesia and Critical Care. 2001; 12: 225–30. 37. National Confidential Enquiry into Patient Outcome and Death. 1996/97 Report. London: NCEPOD, cited April 07. Available from: http://www.ncepod.org.uk/sum96.htm#32 38. Woodall N. Awake intubation. Current Anaesthesia and Critical Care. 2001; 12: 218–24. 39. Ovassapian A, Tuncbilek M, Weitzel EK, Joshi CW. Airway management in adult patients with deep neck infections:



40. 41.



42.



43. 44.



45. 46.



 47.



] 487



a case series and review of the literature. Anesthesia and Analgesia. 2005; 100: 585–9. Gold MI, Buechel DR. Translaryngeal anesthesia: a review. Anesthesiology. 1959; 20: 181–5. Shaw IC, Welchew EA, Harrison BJ, Michael S. Complete airway obstruction during awake fibreoptic intubation. Anaesthesia. 1997; 52: 582–5. Ho AM, Chung DC, To EW, Karmakar MK. Total airway obstruction during local anesthesia in a non-sedated patient with a compromised airway. Canadian Journal of Anaesthesia. 2004; 51: 838–41. Mason RA, Fielder CP. The obstructed airway in head and neck surgery. Anaesthesia. 1999; 54: 625–8. Gerig HJ, Schnider T, Heidegger T. Prophylactic percutaneous transtracheal catheterisation in the management of patients with anticipated difficult airways: a case series. Anaesthesia. 2005; 60: 801–5. Goh MH, Liu XY, Goh YS. Anterior mediastinal masses: an anaesthetic challenge. Anaesthesia. 1999; 54: 670–4. Barron FA, Ball DR, Jefferson P, Norrie J. Airway alerts. How UK anaesthetists organise, document and communicate difficult airway management. Anaesthesia. 2003; 58: 73–7. Domino KB, Posner KL, Caplan RA, Cheney FW. Airway injury during anesthesia. Anesthesiology. 1999; 91: 1703–11. Analysis of closed claims in USA.



40 Adult anaesthesia ANDREW D FARMERY AND JAIDEEP J PANDIT



Principles of general anaesthesia Premedication Key points Anaesthetic agents Key points Inhalational agents Key points Devices used in airway management Key points Conduct of anaesthesia: principles guiding induction of anaesthesia Monitoring in anaesthesia



488 489 490 490 491 491 491 492 495 495 496



Anaesthetic technique and principles for specific ENT operations Salivary gland surgery Nasal and sinus surgery Laryngeal microsurgery and laryngoscopy Laser surgery Key points Best clinical practice Deficiencies in current practice and areas for future research References



500 503 503 503 503 505 505 505 506



SEARCH STRATEGY AND EVIDENCE-BASE The search strategy for the evidence in the chapter was by Medline search of the key words anaesthesia, ear, nose, throat and maxillofacial, and combinations of the words. Reference lists of retrieved papers were searched manually as were reference lists of standard text (listed in the reference list at the end of the chapter). As is the case with anaesthesia in general the evidence levels are largely from observational, nonexperimental, noncontrolled studies or expert opinion (i.e. levels 3 and 4), or in the case of clinical evidence from low-quality casecontrolled/cohort studies or clinical series and expert opinion (i.e. grades C and D). This arises purely due to the paucity of randomized controlled trials or metaanalyses of the common questions covered in this chapter.



PRINCIPLES OF GENERAL ANAESTHESIA Anaesthesia is a relatively modern specialty. Before the mid-nineteenth century, the scope of operative surgery was small. The advent of anaesthesia, however, facilitated a revolution in surgical practice. Although anaesthetic techniques have changed considerably since this time, the basic principles that underpin the practice have changed little. Modern anaesthetic practice encompasses many things, including perioperative system support, monitoring, postoperative intensive care and pain relief. However, the roots of anaesthetic practice lie in the simple



requirement to facilitate successful surgery. The main aims to this end are summarized in the triad of surgical anaesthesia below: 1. to abolish consciousness, implicit and explicit memory ( = hypnosis); 2. to prevent movement ( = akinesia); 3. to obtund the subconscious response to pain, stress and trauma ( = analgesia). These are both functional and humanitarian aims. The principal benefit to the pioneering surgeon was the abolition of movement in their patients, so now for



Chapter 40 Adult anaesthesia



the first time, subtle and delicate procedures could successfully be undertaken. The abolition of consciousness and recall is a clear humanitarian benefit to the patient, and the obtunding of the response to pain produces better operating conditions as well as having other advantages. Early anaesthesia, to some extent, achieved all three of these aims with a single agent, namely ether. Given in sufficient dosage, ether could produce surgical anaesthesia achieving hypnosis, akinesia and, to some extent, analgesia. The drawback with this kind of ‘single-agent anaesthesia’ was that the doses of ether required to achieve all these was often perilously high. In contrast, modern anaesthesia tackles the triad of anaesthesia using a combination of drugs or agents to satisfy each of the aims separately. This has become known as ‘balanced anaesthesia’ and is depicted in Figure 40.1. In balanced anaesthesia, the ‘anaesthetic vapour’ (such as halothane or isoflurane) is used merely to depress cognitive function and provide the element of ‘hypnosis’. This is achieved at relatively low concentrations; far below those required to achieve akinesia and analgesia and far below concentrations associated with respiratory depression and cardiovascular collapse. The element of analgesia is provided by specific drugs (such as the opioid fentanyl or the opiate morphine), or by supplemental use of regional or local anaesthetic blocks. This combination of hypnotic and analgesic drugs may be sufficient also to provide akinesia, i.e. to abolish movement in response to stimulus, and yet the patient may still breath spontaneously. However, there are occasions when it is desirable to abolish all movement and to relax the tone in all skeletal muscle, for example to facilitate endotracheal intubation and permit mechanical



] 489



ventilation. This can be achieved by specific drugs which block the neuromuscular junction and paralyse skeletal muscle such as atracurium or vecuronium. When the element of akinesia is provided in this way, it is possible to reduce the doses of hynotic and analgesic drugs yet further. So by judicious use of hypnotics, analgesics and muscle relaxation, a state of balanced anaesthesia can be achieved in which the dose of anaesthetic agent is a fraction of that which would be required if it were used as a sole agent.



PREMEDICATION Until relatively recently, premedicant drugs were used religiously and routinely as an accompaniment to general anaesthesia. This habit probably stemmed from the necessity of using such drugs when ether inhalation was standard practice.



The need for anxiolysis Induction of anaesthesia by inhalation of ether and chloroform was a lengthy process, and many patients still have unpleasant memories of it. As these experiences entered the folklore, patients often became more anxious about their impending anaesthetic than their surgery.



The need for an antisialogogue In addition, inhalation of the early anaesthetic agents produced copious secretions which was not only undesirable to the ENT surgeon, but also made spontaneous respiration via a face mask a challenge for the anaesthetist.



The need for an antiemetic Ether is a particularly potent emetic. Even modern volatile agents can contribute to postoperative nausea and vomiting.



Hypnosis



The need for an opiate



Analgesia



Figure 40.1



Balanced anaesthesia.



Akinesia



In modern anaesthesia, rapidly acting, lipid soluble synthetic opioids can be used to produce desired levels of analgesia with precision. By contrast, the traditional opiates (morphine, omnopon, etc.) used in early anaesthesia have a slow onset of action and so were best given an hour or so before the onset of surgery to ensure adequate analgesia. Such opiates were not usually given intraoperatively in any significant dose because, since most patients breathed spontaneously via face masks, any depression of respiration was considered undesirable.



490 ] PART 8 PERIOPERATIVE MANAGEMENT Table 40.1



Modern premedicant drugs. Analgesics



NSAIDs NSAIDs are widely used in the perioperative period. They are remarkably effective and relatively safe when used in healthy patients for a short duration. Although many preparations can be given rectally, they are well absorbed from the stomach and best, and most economically, given orally if possible



Gastric alkalinizers



Prokinetics/antiemetics



Anxiolytics



Aspiration of even small volumes of acid gastric fluid can produce severe respiratory complications. All patients with a history of gastrooesophageal reflux should receive prophylactic treatment to reduce gastric acidity. Many anaesthetists would routinely do this for all patients H2 blockers, e.g. ranitidine Proton pump inhibitors, e.g. omeprazole Sodium citrate. This is a ‘buffer’ which mops up acid and is given, if indicated, immediately preinduction.



Metoclopramide is a dopamine antagonist, which increases gastric motility and emptying and has antiemetic properties. The evidence is not clear whether the benefits of such drugs justify their prophylactic use



The best anxiolysis is provided by good communication between doctor and patient. Wellinformed and reassured patients will seldom require pharmacological anxiolysis. Where indicated, it is usually provided by oral drugs, e.g. benzodiazepines such as temazepam



Paracetamol This is a safe and effective analgesic which can be given orally with an NSAID one hour preoperatively



The combination of intramuscular omnopon and scopolamine (‘om and scop’) was for many years the ‘pre-med’ of choice for adult anaesthesia, since it combined analgesia, anxiolysis, antiemesis and the drying of secretions. Modern anaesthestic drugs and techniques are now so finely honed that the traditional ‘pre-med’ is rarely required and is seldom used. However, in some cases premedication with other drugs is indicated (see Table 40.1).1



KEY POINTS  Sedative premedication is seldom used in modern anaesthetic practice, especially as day-case surgery is increasing.  Pre-emptive analgesia (from simple agents such as nonsteroidal antiinflammatory drugs (NSAIDs) and paracetamol) are usefully given preoperatively to provide early postoperative pain relief.



 Antiemetic and antacid prophylaxis can usefully be given preoperatively to high-risk patients.



ANAESTHETIC AGENTS Intravenous agents Intravenous (i.v.) anaesthetic agents can be used to induce and/or to maintain anaesthesia, although they are more commonly used for induction alone. For many years, the barbiturate thiopentone dominated this class of drugs. However, by far the most commonly used i.v. induction agent in the western world is propofol, which is a derivative of phenol. Most induction agents are extremely lipid soluble, and this accounts for their rapid onset of action, since they are delivered to the brain with a high blood flow and diffuse into it (the brain being principally lipid) with ease. Plasma levels then fall rapidly. This is not due to rapid metabolism, but rather because the drug



Chapter 40 Adult anaesthesia



redistributes (via the large cerebral blood flow) to other lipid-rich tissues with lower blood flows. Metabolic clearance accounts for a much slower ‘background’ elimination of drug and reduction in the brain-drug levels. Propofol is so lipid soluble that it cannot easily be dissolved in aqueous media. For this reason it is prepared as a lipid emulsion and this accounts for its milky appearance. As well as being used for induction, propofol can be infused intravenously to maintain anaesthesia. This is an increasingly popular technique and can be associated with more rapid clear-headed recovery and a reduction in postoperative nausea. Like most anaesthetic agents, thiopentone and propofol are associated with a fall in blood pressure mediated by a number of mechanisms, including reduction in central sympathetic vasomotor tone and reduction in cardiac contractility. These features are particularly marked in the elderly and patients who are hypovolaemic and dehydrated.



KEY POINTS  Intravenous induction agents owe their rapid onset and offset of action to rapid diffusion across the blood–brain barrier due to their high lipid solubility.  Hepatic clearance is moderately slow, and does not play a significant part in the return of consciousness after a bolus dose of intravenous agent.



INHALATIONAL AGENTS Inhalational or ‘volatile’ agents may be used for induction and/or maintenance of anaesthesia, although they are more commonly used for maintenance alone. The historical archetype is ‘ether’ (actually diethyl ether) but the more commonly used modern agents such as isoflurane and sevoflurane are also ethers (halogenated methyl-ethyl ethers). Like the intravenous agents, the volatile agents are lipid soluble: the higher the lipid solubility, the higher the potency. Modern agents are extremely lipid soluble and so their potency is sufficiently high to require only 1 or 2 percent concentration in the inspired gas to induce anaesthesia. The depth of anaesthesia is determined by the partial pressure (i.e. the effective concentration) of the volatile agent in the brain. Since the brain is in equilibrium with the arterial blood, and the latter is in equilibrium with the alveolar gas, then the depth of anaesthesia should be determined by the partial pressure (or concentration) of volatile agent in the alveoli. Given that we can easily measure the



] 491



concentration of anaesthetic vapours in the alveoli (i.e. in the expired breath), we can therefore reliably determine and estimate the anaesthetic depth with precision and confidence. In contrast, we have no means of doing this for i.v. agents. Volatile agents can also be used to induce anaesthesia, and some do so more quickly than others. We have already stated that the depth of anaesthesia is determined by the partial pressure of the agent in the alveoli/blood/ brain. In the blood, only molecules that are not in solution contribute to the partial pressure. As soon as a molecule of gas or vapour enters solution in the blood, it is effectively ‘hidden’ and no longer contributes to the partial pressure. This is analogous to the absorbency of a sponge. When water is absorbed into the sponge it no longer contributes to its external ‘wetness’. A large volume of water will need to be added to a highly absorbent sponge before it contributes to its external wetness, whereas only a small volume needs to be added to a poorly absorbent sponge before it appears wet. So, for agents which are highly soluble in blood (such as traditional ether or halothane), the partial pressure in the blood will be slow to rise because almost as fast as the molecules of gas diffuse across the lung into the blood, they enter solution and are effectively hidden, thus failing to contribute to the partial pressure. Conversely, modern agents, such as sevoflurane, have a low blood solubility. As these molecules enter the bloodstream, only a few enter solution, the remainder do not, and therefore contribute to the partial pressure in the blood. Consequently, the arterial partial pressure rises and equilibrates rapidly with the alveolar gas. In turn, the high partial pressure in the arterial blood equilibrates with the brain and induction occurs rapidly. In summary, modern agents have a high lipid solubility (and hence high potency) and low blood solubility (and hence allow rapid induction).2 The sections above outline the drugs that can be used to induce and maintain anaesthesia. The following section outlines how anaesthetic technique is planned for some common or important ENT operations, with a particular emphasis on management of the patient’s airway. One of the hallmarks of anaesthesia for ENT surgery is the concept of the ‘shared airway’: the surgeon operates in the same anatomical region as the anaesthesia airway devices are situated. Therefore, the detailed properties of the airway device selected and precise conduct of anaesthesia are particularly important to ENT surgery as compared with many other types of surgery.



KEY POINTS  Older agents (ether, halothane) have high solubility in blood and therefore a slow onset and offset of action.



492 ] PART 8 PERIOPERATIVE MANAGEMENT  Modern agents (sevoflurane, desflurane) have low solubility in blood and therefore rapid onset and offset of action.  Because of our ability to measure vapour concentrations in the expired breath, adequate depth of anaesthesia can be determined with confidence. Figure 40.2 Laryngeal mask airway.



DEVICES USED IN AIRWAY MANAGEMENT The simple face mask Induction of anaesthesia invariably results in a degree of collapse of the upper airway resulting in airway obstruction. This is largely due to the reduction in tone of the pharyngeal dilator muscles and genioglossus. These muscles are innervated by the efferent outflow of the ‘respiratory centre’, and tone varies cyclically in much the same way as does the diaphragm under the command of the phrenic nerve. Drugs that depress the respiratory centre and have a tendency to produce a ‘central’ apnoea, also by this mechanism, have a tendency to produce an ‘obstructive’ apnoea. Central and obstructive apnoeas are therefore inextricably linked, and cannot adequately be categorized as separate entities.3 If neuromuscular blocking drugs are also used, then spontaneous ventilation is abolished completely. All these factors require the use of devices that give both anatomical support to the upper airway and also facilitate ventilation. Perhaps the simplest means of achieving these aims is a face mask which, when connected to a suitable supply of oxygen and/or anaesthetic gas, may be used to support the airway manually or (if spontaneous ventilation is absent) allow hand-ventilation of the patient’s lungs. Problems with the face mask technique are that it does not allow good surgical access to the face, head or airway; it requires that one or both of the anaesthetist’s hands are occupied in airway maintenance and active ventilation of the patient’s lungs can be often inefficient (i.e. not all the tidal volume administered to the patient enters the lungs, and some is lost as a ‘leak’).



Supraglottic airway devices A range of supraglottic airway devices (SADs) have been developed which reduce some of these problems. In general, these are tubes (often with inflatable distal components to ‘hold’ the airway in place) that lie in the oropharynx, above the level of the glottis. The device most commonly used is the laryngeal mask airway and others include the cuffed oropharyngeal airway and the ProsealTM laryngeal mask (Figure 40.2). SADs allow



spontaneous breathing and, similar to the face mask technique, facilitate active ventilation by hand. However, since SADs lie above the glottic opening and do not seal it, there is a risk that prolonged positive pressure ventilation of the lungs (for example, if the SAD is connected to a mechanical ventilator) may cause air to enter the oesophagus and stomach. If gastric fluid should then regurgitate, SADs will not completely protect the airway from lung soiling. There is currently a debate concerning the role of some SAD devices in positive pressure ventilation, and some authorities argue that their use during positive pressure ventilation is, in fact, justified.4



Tracheal tubes There is a range of tracheal tubes available (Figure 40.3). These tubes lie in the trachea and (usually) have a small cuff at the distal end, whose purpose it is to protect the airway from soiling (for example by blood or regurgitant gastric contents) and to improve the efficiency of ventilation by minimizing leaks. Tracheal tubes vary in shape, size and the material from which they are made, and these factors are important in relation to their specific use.



CUFFED VERSUS UNCUFFED TUBES



Generally speaking, cuffed tubes are used in adult anaesthesia, and ‘plain’ or uncuffed tubes are used in prepubertal children for the following reason. In adults, the narrowest part of the larynx is at the level of the vocal cords, and so if a tube can be passed between the cords, one can be confident that at any point distal to the vocal cords, the tube will be sitting loosely and will not be wedged-in, causing epithelial damage. The high-volume, low-pressure cuff can be inflated to provide a loose seal at this level. In children however, the narrowest part of the larynx is at the level of the cricoid cartilage and not at the vocal cords. It is, therefore, quite possible to pass a tracheal tube which is small enough to pass through the cords with ease, but which may be wedged-in at the level of the cricoid, and this may cause epithelial oedema or necrosis at this level. This is particularly problematic in small airways because any further reduction in calibre in an already small airway produces a disproportionately



Chapter 40 Adult anaesthesia



] 493



Figure 40.3 Tracheal tubes. (a) A cuffed RAE tube. The tube is preformed in a ‘south-facing’ curve so that the anaesthetic breathing system passes out of the way of the surgical field; (b) a ‘north-facing’ RAE tracheal tube; useful if the anaesthetist and the breathing system are ‘north’ of the patient’s head and the surgeon is working from the ‘south’; (c) a microlaryngeal tube. This has an internal diameter of 5 mm and is sufficiently narrow to allow the vocal cords to be inspected during panendoscopies; (d) a reinforced tracheal tube. Note the metallic spiral in the wall of the tube that allows it to be bent in any direction without kinking the lumen.



greater increase in airflow resistance, since this is inversely related to the fourth power of the airway radius. In order to confirm that the tracheal tube is sitting loosely at all points in the larynx, plain tubes are used and one deliberately seeks to allow a small leak. This should be audible on inflation of the chest under positive pressure and excludes the possibility that the tube is exerting any circumferential pressure on the laryngeal epithelium at its narrowest point. The disadvantages of having this small leak is that ventilation may not be as efficient and the lower airway is not definitively secured and protected from soiling from above. For this reason, if the type of surgery permits, the laryngeal inlet can be packed with damp ribbon gauze. Since tracheal tubes form a sealed conduit into the trachea, they are primarily used when the patient is paralysed and requires positive pressure ventilation (there is no risk of blowing air into the stomach). It is indeed possible to allow the patient to breathe spontaneously through a tracheal tube, but since the glottis and trachea are richly innervated, deep anaesthesia is required to prevent the patient from reflex coughing. The doses required to achieve this may lead to dose-related side effects (as described above). Whereas most SAD devices can be used effectively after relatively short periods of training, the techniques used by the anaesthetist to insert a tracheal tube are more specialized and are described in more detail in Chapter 39, Recognition and management of the difficult airway. Tracheal tubes may be passed either orally or nasally, and these routes are important and can depend upon the type of surgery performed (for example, a nasotracheal tube facilitates better surgical access for dental surgery, but is clearly a hindrance during operations on the nose). Regardless of which airway device is selected, it is vital that there is constant communication between surgeon



and anaesthetist regarding the other’s needs during surgery (and also at the stage of planning an operation). The airway device must not hinder the surgical process because of its size or shape; it must be properly secured in place to prevent dislodgement during head movement by the surgeon; and the direction of the airway device shaft must be appropriate to the operation. Use of surgical gags and props may compress the airway device, so the anaesthetist must be vigilant about the patency of the airway. Airway obstruction may occur not only during induction of anaesthesia, but also after extubation of the trachea or recovery and it is often especially important that the surgeon is present in the operating room at extubation.



Factors influencing choice of device in airway management In general terms, patient factors and surgical factors dictate the choice of which of the above devices is used in anaesthetic airway management.



PATIENT FACTORS INFLUENCING CHOICE OF AIRWAY DEVICE



There are a number of factors or signs in the patient which would suggest to most anaesthetists that a tracheal tube (with the patient consequently artificially ventilated), rather than a SAD device (with the patient breathing spontaneously), should best be used. Gastro-oesophageal reflux Hiatus hernia, peptic ulcer disease, symptomatic reflux, pregnancy, recent ingestion of solids or particulate liquids,



494 ] PART 8 PERIOPERATIVE MANAGEMENT abdominal sepsis or injury all predispose to regurgitaton of gastric contents during induction of anaesthesia. Tracheal intubation (as part of a ‘rapid sequence’ induction of anaesthesia) would minimize the risk of lung soiling.



Obese patients do not always find it easy to breathe when lying supine, even when awake. This difficulty is exacerbated when anaesthetized, and artificial positive pressure ventilation (necessitating tracheal intubation) is therefore often necessary.



neuromuscular blocking drugs. Since these also paralyse spontaneous ventilation, artificial mechanical ventilation is required; an endotracheal tube is most commonly used to facilitate this. While strict muscle relaxation of the degree needed in abdominal surgery is rarely required in ENT surgery, there are some operations in which it is desirable to minimize risk of the patient moving, coughing or swallowing (for example skull base surgery, any ENT operation in combination with neurosurgery or pharyngeal surgery). This is most reliably achieved by using neuromuscular blocking drugs.



Patients with known or anticipated difficult airway



Neuromuscular blockade and facial nerve monitoring



In any patient whose trachea is known or anticipated to be difficult to intubate, it is reasonable to formulate a plan to intubate the trachea during induction of anaesthesia and so achieve a definitive, ‘secured airway’ in the patient before surgery begins. While it is also acceptable (and sometimes necessary should tracheal intubation fail) to use a SAD device, the problem is that should tracheal intubation subsequently be necessary urgently during the course of surgery, this might be difficult or impossible to achieve in a short space of time.



For certain ENT operations, such as vestibular schwannoma resection, middle ear, mastoid and parotid surgery, it is necessary that the surgeon is able to monitor the integrity of the facial nerve and minimize the risk of damaging it during surgery. For this purpose, needle electrodes are placed in the orbicularis oculi and oris muscles that detect EMG potentials generated in these muscles as a result of surgical or electrical stimulation of the facial nerve. The electrical activity evoked in these muscles is amplified and converted into an audible signal by the monitor. By this means, the surgeon is warned of the impending proximity of the nerve should he be unaware of its precise position, and can use the sound signal to moderate his handling of the nerve in those situations where this is necessary. At the completion of surgery the electrical integrity of the nerve can be assessed by proximal stimulation. In addition, it is possible to identify the site of a neuropraxia and, by varying the intensity of the stimulating current, assess its severity. A block that can be overcome by increasing the stimulation current is probably not as bad as one that cannot. Obviously, it is impossible to monitor facial nerve activity in the presence of complete neuromuscular blockade. Partial blockade can be overcome by decreasing the stimulation frequency from 30 to 3 Hz, but information derived in this state may be unreliable. From the anaesthetic standpoint, if direct nerve monitoring is necessary, it is advisable to avoid longacting neuromuscular blocking drugs altogether during surgery. If tracheal intubation is required, it is possible to use a short-acting muscle relaxant to facilitate intubation on the premise that the effect will have worn off by the time monitoring is required by the surgeon. Offset of muscle relaxation should be confirmed by using a peripheral nerve stimulator to stimulate the ulnar nerve (Figure 40.4) and feeling forceful contraction of adductor pollicis. Alternatively, intubation may be accomplished without muscle relaxants by administration of high doses of opioids or volatile anaesthetic agents. An infusion of remifentanil, a very potent and ultra-short acting opioid, often produces good conditions for intubation and allows stable levels of anaesthesia during surgery. In some cases it will be possible to avoid intubation by the use of a supraglottic airway device. It is good practice to test the correct assembly and functioning of the facial nerve



Obesity



SURGICAL FACTORS INFLUENCING CHOICE OF AIRWAY DEVICE



A number of surgical requirements reasonably influence the choice of airway device. Factors related to the ‘shared airway’ For certain operations, it is necessary for the surgeon to have an unobstructed view of the relevant anatomy. Most SAD devices (as a result of the large distal cuff) do not permit a good view of structures distal to the oropharynx. Thus, for almost all periglottic, laryngeal and subglottic operations, a tracheal tube is more suitable or necessary. SAD devices may, however, be used for operations related to the tonsils, anterior tongue, nose, teeth and ears without obstructing the surgical field. Laser surgery Standard (polyvinyl chloride) tracheal tubes are not laserresistant and may ignite if struck by the laser beam. A number of specialized tracheal tubes have been developed which are more laser-resistant. The materials used include various metals, Teflon and ceramics. Some recent work suggests that the flexible laryngeal mask airway is also suitably laser-resistant and, if used in the presence of laser, its distal cuff should be filled with saline (or methylene blue dye so that rupture can be easily detected). Anaesthesia for laser surgery is further discussed below under Laser surgery. Requirement for neuromuscular blockade For certain operations (for example abdominal surgery) it is necessary that the patient’s muscles are fully flaccid to facilitate surgery. This is best achieved by use of



Chapter 40 Adult anaesthesia



Figure 40.4 A peripheral nerve stimulator. This passes a supramaximal (450 mA) stimulus transcutaneously over the desired motor nerve, typically the ulnar nerve at the wrist, in order to assess the degree of neuromuscular block.



monitoring system before surgery by using the peripheral nerve stimulator. A short burst of tetanic stimulation applied over the facial nerve in front of the ear should produce visible contractions of the muscles and both visual and audible alerts from the monitor. Postoperative plan After some major operations of the head and neck (for example for tumour) involving free flap transfer, it is conventional for the postoperative plan to include admission to an intensive care unit for a period of artificial ventilation, which itself would require the presence of a tracheal tube. This plan might also be necessary if the patient has certain medical conditions (for example poor lung function).



KEY POINTS  Anaesthesia and loss of consciousness result in reduced tone in the pharyngeal dilators and posterior displacement of the tongue.  The oropharnyx behaves as a Starling resistor: airflow is critically dependent on pharyngeal tone and transmural pressure.  Maintaining the airway manually with a simple bag and mask is an important clinical skill.



CONDUCT OF ANAESTHESIA: PRINCIPLES GUIDING INDUCTION OF ANAESTHESIA Preoxygenation It is usual to allow a patient to breathe 100 percent oxygen via a face mask for a period of approximately two



] 495



to three minutes before the induction of anaesthesia. This process in known as preoxygenation and has a number of theoretical and real advantages. Perhaps counter-intuitively, breathing 100 percent oxygen in this way does not increase the oxygen content of the blood at all. This is because, for most patients, the arterial blood is almost fully saturated with oxygen even when breathing room air, so breathing oxygen cannot really improve on this. The purpose of this manoeuvre is rather to ‘de-nitrogenate’ the gas within the lungs. This increases the mass of oxygen within the alveolar compartment, which serves as a reservoir during a subsequent apnoea, and markedly delays the rate of desaturation during this period. An alternative technique for preoxygenation is to allow the patient to take three successive vital capacity breaths of 100 percent oxygen. For a resting lung volume of 2 L, and tidal breaths of 2 L, the alveolar oxygen concentration should rise from 16 percent on breath zero, to approximately 60 percent on breath one, 80 percent on breath two and 90 percent on breath three.5, 6



Intravenous induction Perhaps the commonest mode of induction of anaesthesia is by i.v. injection (using one of the drugs listed in the above section, but most commonly probably propofol). This route rapidly induces anaesthesia. The advantage of this is that there is little or no ‘stage of hyperexcitability’. This is a stage just before deep anaesthesia is attained, in which the patient may be paradoxically excitable and there is tongue-biting, vomiting and laryngeal spasm (see below under Inhalational induction). The aim is to titrate the intravenous administration of drug, slowly, according to the observed effect. Injudicious dosing of intravenous induction agents invariably causes loss of spontaneous ventilation, so at least for a period of time after induction (and for longer if neuromuscular blocking drugs are used) the anaesthetist must be confident of maintaining ventilation, using any or all of the means described above. If, therefore, there is any doubt on the part of the anaesthetist that s/he is able to maintain ventilation, then even the most carefully administered i.v. induction is probably not the safest means of inducing anaesthesia. The details and nature of these doubts (i.e. the ability to predict a patient who is difficult to intubate or ventilate) are discussed in Chapter 39, Recognition and management of the difficult airway. If the use of neuromuscular blocking drugs is planned, it is important that the anaesthetist is satisfied that s/he can, if necessary, ventilate the patient with a bag and mask before such drugs are given. The precaution ensures that if for some reason the trachea cannot be intubated, then ventilation can at least be achieved with a bag and mask until spontaneous ventilation resumes.



496 ] PART 8 PERIOPERATIVE MANAGEMENT



Rapid sequence induction This is a special form of i.v. induction that is employed to secure the airway with a tracheal tube, and so protect it from soiling, as quickly as possible. It is particularly indicated in emergency surgery where a patient may have a full stomach or a hiatus hernia with active reflux. It may also be considered where there is bleeding (as in posttonsillectomy bleeding). In a rapid sequence induction we appear to break all the rules detailed above. Here, the induction is not slow, nor the dose titrated to effect so as to minimize the risk of apnoea, airway collapse and obstruction, but rather, a predetermined dose of i.v. agent is given rapidly as a bolus. The muscle relaxant rule is also broken. In the rapid sequence induction we do not test our ability to ventilate the patient with a bag and mask before given neuromuscular blockers, but rather, give a dose of suxamethonium (or other rapidly acting drug) immediately following injection of the induction agent. This adds an extra burden of responsibility on the anaesthetist to be able to intubate the trachea, since we have chosen to neglect our ‘escape plan’. As soon as consciousness is seen to fall, the anaesthetic assistant applies ‘cricoid pressure’ to the larynx. Empirically, it is found that if a force of 40 N (4 kg weight) is applied to the cricoid cartilage using the finger and thumb, the posterior part of the cricoid cartilage will compress the oesophagus posteriorly against the C6 cervical vertebral body and so prevent passive spillage of gastric contents. Laryngoscopy and intubation are now undertaken in the usual manner.



induction is slow, there is a danger that the stage of excitability is prolonged. Second, should collapse of the upper airway occur during inhalational induction then not only anaesthetic vapour, but also oxygen is prevented from reaching the lungs. Thus, inhalational induction may not be as safe or controlled as may be claimed.



Induction after securing the airway Certain patients, especially those whose tracheas are predicted to be difficult to intubate, are subjected to techniques that achieve tracheal intubation with the patient awake or sedated, with anaesthetic induction occurring only after the airway is secured. These methods are discussed further in Chapter 39, Recognition and management of the difficult airway. The methods may also include planned awake tracheostomy under local anaesthesia.



MONITORING IN ANAESTHESIA Derived from the Latin monere – to warn, monitoring is used to describe measurements whose prime purpose is to ‘warn’ of imminent (possibly injurious) events, and allow action to be taken to avoid them, or moderate their effect. The Association of Anaesthetists of Great Britain and Ireland publishes guidelines for minimum monitoring standards, and these are reviewed from time to time. Undoubtedly, the most important ‘monitor’ is the very presence of an anaesthetist throughout the duration of anaesthesia who can synthesize information derived from clinical observation, and specialized devices.



Inhalational induction Basic intraoperative monitoring An alternative means of inducing anaesthesia is by inhalational induction, using one of the vapours discussed above (most commonly either sevoflurane or halothane). Anaesthesia in children, who may be frightened of intravenous cannulation, is often induced in this manner. Historically, this was the first method of induction. One advantage of the technique is that it is theoretically a ‘controlled’ means of induction. If, during the course of inhalational induction, the patient’s upper airway collapses and the patient’s breathing becomes obstructed, then no further anaesthetic vapour can enter the lungs. The patient begins to wake up and thereby the upper airway tone and breathing are restored. For this reason, it has been advocated as the technique of choice in a case of upper airway obstruction and stridor (for example due to supraglottic tumour). Unlike an i.v. induction, therefore, inhalational induction may be used even when the anaesthetist has doubts about the ability to intubate the trachea or maintain ventilation. However, there are some practical disadvantages. First, because



ECG



Continuous single-lead ECG recording is universally used. Electrical activity of the heart gives no information about pump function and circulation. Its purpose is to detect the development of dysrhythmias and/or myocardial ischaemia. Lead II configuration is best for rhythm disturbances (AF, heart block, asystole, VT, VF) since this lead shows the presence (or absence) of P waves best. The CM5 configuration is best for detecting myocardial ischaemia, (ST depression, T wave inversion) (see Figure 40.5).



PULSE OXIMETRY



These devices shine red light through an extremity, usually a finger, and measure the absorbance of this transmitted light by substances in its path. Clearly, there will be a number of substances which absorb and/or



Chapter 40 Adult anaesthesia



RA



] 497



this pulsatile absorbance is due to the added volume of arterial blood which enters during (or shortly after) systole. If this pulsatile signal is subtracted from the background signal we achieve an arterial absorbance signal that is wholly accounted for by the composition of arterial blood. The red light comprises two different wavelengths and these are absorbed by deoxygenated and oxygenated haemoglobin to different degrees. The relative absorbances of these wavelengths are used to calculate the proportions of oxygenated and deoxygenated haemoglobin, yielding a percentage SpO2. The pulse oximeter monitors oxygenation but not ventilation. This is an extremely important concept when considering high-risk patients nursed on the ward and recovery area. Patients breathing supplemental oxygen may still be pink, despite being virtually apnoeic (i.e. imminent respiratory arrest). Pulse oximeters must not be relied upon to monitor the adequacy of ventilation in patients who are at risk of respiratory depression or obstruction. In addition, pulse plethysmography (the pulsatile waveform display shown on the pulse oximeter) gives a beat-to-beat indication of circulation, which is the electronic equivalent of keeping one’s finger on the pulse at all times. This has become an invaluable tool.



LA



II



LL



(a)



RA



BLOOD PRESSURE



Automatic noninvasive blood pressure LA



LL



(b)



Figure 40.5 (a) For standard lead connection, the right arm lead is placed on the right shoulder, the left arm lead on the left shoulder, and the left leg lead is placed centrally to form an equilateral triangle. Selection of ‘lead II’ on the monitor will connect leads RA to the negative terminal and LL to the positive terminal. Lead LA acts as an earth lead. (b) For CM5 configuration, lead RA is placed on the central manubrium and lead LA at the V5 position in the anterior axillary line. The monitor is set to display ‘lead I’, which connects RA to the negative terminal and LA to the positive terminal. Lead LL is earth/ground/indifferent and its placement is not critical.



scatter this light energy, including arterial haemoglobin, venous haemoglobin, skin, bone and nail-bed. However, in addition to the background absorbance, there will be a component that varies throughout the cardiac cycle, and



Automatic noninvasive blood pressure (NIBP) uses the principle of oscillotonometry applied to an inflatable cuff around the arm. This principle differs from the more familiar auscultatory method. The mean blood pressure measured by oscillotonometry is reasonably accurate. Systolic and particularly the diastolic values are less so. These automatic NIBP devices have gained widespread acceptance because they are safe and easy to use. They can be set to take measurements every few minutes and produce a printed chart, which is useful for detecting trends in theatre or the high dependency area. NIBP measurement has a number of disadvantages. Firstly, its readings, even if taken frequently, are ‘intermittent’ and so do not allow one to detect sudden circulatory collapse. NIBP measurements tends to overestimate low pressures, and underestimate high pressures. It is often impossible to get any reading from patients who are ‘shocked’ and peripherally shut down. Invasive monitoring of blood pressure Invasive monitoring of blood pressure (IBP) can be measured by siting a cannula in a suitable artery, usually the radial. This has the advantage of allowing beat-tobeat blood pressure measurement, which allows extremely rapid changes to be detected almost in real time. The IBP monitor allows the arterial ‘waveform’ to be displayed on screen. The morphology of this waveform



498 ] PART 8 PERIOPERATIVE MANAGEMENT (e.g. the rate of upstroke, pulse pressure, presence ‘pulsus paradoxus’ with respiration) provides useful information on circulatory and vasotonic status. Having an arterial cannula in situ also permits easy and regular blood gas sampling. IBP measurements are reliable and accurate in both high and low pressure states. Disadvantages are few, but include the fact that siting such cannulae can be fiddly, and the disposables are relatively expensive. There is a very small risk of thrombosis, distal ischaemia and infection.



GAS ANALYSIS



Inspired oxygen fraction (FIO2) It is axiomatic that oxygen is vital for safe anaesthesia. However, since it is odourless and colourless it is not straightforward to know how much is being given to a patient, or more worryingly, whether any is being given at all! Deaths still occur due to administration of gas devoid of oxygen, either by inadvertently giving pure nitrous oxide or carbon dioxide, or as a result of an error in gas pipeline connections. It is therefore vital that the labelled identity of the gas in a cylinder or pipeline is not relied upon, but that the composition of gases as they leave the anaesthetic machine and enter the patient’s airway is monitored. This is usually carried out by means of continuous aspiration of a sample of gas from the airway into a rapid gas analyser which can display the concentration of inspired and expired oxygen breathby-breath. Most analysers are also capable of monitoring other gas, such as CO2 and anaesthetic gases and vapours. Inspired/end-tidal anaesthetic agent concentration By sampling the concentration (or partial pressure) of anaesthetic vapours in the expired breath, one can estimate the anaesthetic partial pressure in the alveolar gas and hence the arterial blood and brain. This is a very reliable way of monitoring anaesthetic depth because the dose-response curves (or more accurately the partial pressure-response curves) for these agents show very little interindividual variation. So, for example, if the end-tidal concentration of isoflurane is 1 percent, we can be extremely confident that for all patients, anaesthetic depth is adequate and there will be no danger of ‘conscious awareness’. Inspired and expired PCO2 Inspired CO2 should be zero or near zero. Rising inspired CO2 concentrations indicate that the patient is rebreathing exhaled breath and this usually results from a faulty breathing circuit or inadequate gas supply from the anaesthetic machine. Expired PCO2, and its waveform, is one of the most important monitors in anaesthesia. The term ‘end-tidal’ refers to the concentration at the very end of expiration



and this is taken to represent alveolar gas. The ‘capnogram’ (a graphical representation of expired PCO2 versus time) provides an enormous amount of information on cardiorespiratory function. It is one of the earliest and most robust indicators of whether the trachea or oesophagus has been intubated. The presence of expired CO2 and a normal ‘alveolar waveform’ definitively confirms endotracheal placement. Likewise, the absence of an alveolar waveform is strongly suggestive of oesophageal intubation. The end-tidal PCO2 is useful to confirm that alveolar ventilation is adequate. Hypoventilation and hyperventilation result in hypercapnia and hypocapnia, respectively. Clues to a number of other cardiopulmonary abnormalities can be gained from capnography, such as falling cardiac output, pulmonary embolus, V/Q maldistribution, the patient ‘fighting’ the ventilator. See example capnograms in Figure 40.6.



AIRWAY PRESSURE, TIDAL VOLUME



If patients are ventilated mechanically, the ventilator provides information on airway pressure and expired volume. Pressure is required to pass a volume of gas into the lungs. Two separate processes contribute to this pressure: 1. the compliance (or elastance) of the lung (the springiness of the spring); 2. the resistance to airflow (the narrowness of the tubes). The compliance of the lungs is not usually a problem in most theatre cases (cf. ITU cases). Sudden changes in airway pressure usually indicate sudden changes in resistance (for example, sudden onset of bronchospasm in anaphylaxis, or kinking of the endotracheal tube by a Boyle Davis gag).



TEMPERATURE



Patients tend to get cold during surgery. This is because of:  heat at loss to the environment (cold theatre, naked body, exposed body cavities);  altered homeothermic mechanisms under anaesthesia. This is bad for a number of practical and theoretical reasons. Hypothermia delays wound healing, depresses immunity, adversely affects skin integrity (so prone to pressure sores) and adversely affects coagulation. Temperature is usually measured with an oro- or nasopharyngeal thermister. Rectal and intravascular varieties also exist. Our aim is to maintain normothermia by means of warm air convection blankets (e.g. BairHuggerTM) and by using an intravenous fluid warmer if the need for large volumes of fluid is anticipated.



Chapter 40 Adult anaesthesia



] 499



PCO2



5



PCO2



5



PCO2



(a)



5



(f)



PCO2



5



PCO2



(b)



(g)



PCO2



5



PCO2



(c)



(h)



(d)



(e)



PCO2



PCO2



5



(i) Time



Figure 40.6 Example capnograms: (a) Normal capnogram series. Note: regularity of ventilatory cycle, flat alveolar plateaux, normal PACO2 value, zero inspired PCO2. (b) Hypoventilation. Shape of capnogram is normal, and alveolar plateau preserved, so airway flow must be normal. The likely cause is bradypnoea due to opiates. (c) Hyperventilation. Note that respiratory rate is normal so it is likely that the patient is being overventilated (tidal volume too great). Another rarer cause would be systemic shock with poor venous return. (d) Sudden fall in PCO2 despite no change in ventilation. Causes: Pulmonary embolus (clot or air); circulatory arrest. (e) Loss of normal morphology. Small blips indicate insufficient alveolar gas is making it as far as the sample probe (at the lips). This is either because the tidal volume is very small (only just exceeding the dead space), or there is some upper airway obstruction which is physically hindering gas efflux on the final part of its journey from larynx to lips. In the case of ventilation via a face mask, this could indicate air leak from a poor seal, i.e. ventilation may be normal, but alveolar gas is not sampled properly because of the poor face mask seal. (f) Tachypnia. Rapid respiration is normal in babies. In adults, it may indicate inadequate anaesthesia or analgesia. In this example, the ‘response time’ of the analyser is not rapid enough to measure the true capnogram at this respiratory rate, so it appears as a series of rapid blips, the peaks of which do not represent true capnogram (dotted). (g) V/Q mismatch/lower airway obstruction. Note the sloping alveolar ‘plateau’. Alveolar gas is not homogeneous. Low-resistance units, by definition, are well ventilated and have a low PCO2, and because they are low-resistance, they empty first. High resistance units are poorly ventilated and have a high PCO2. Because they are slow to empty, they dominate the latter part of expiration. This is classically seen in COAD and asthma. (h) ‘Curare notch’. The notch in the alveolar plateau indicates that a patient may be trying to breathe in at this point (i.e. to fight against the ventilator). It usually means that the muscle relaxant is starting to wear off, this should be checked with the peripheral nerve stimulator before giving further doses. (i) Rebreathing/inadequate fresh gas supply. Here, the inspired PCO2 is not zero. Some degree of ‘rebreathing’ is occurring as breathing in and out of a paper bag. The usual reason is failure of the gas supply from the anaesthetic machine to the breathing circuit. The patient’s expired gas is not washed away by fresh gas, and so it ends up being rebreathed at the next inspiration.



500 ] PART 8 PERIOPERATIVE MANAGEMENT CENTRAL VENOUS PRESSURE/PULMONARY ARTERY WEDGE PRESSURE/CARDIAC OUTPUT/TRANSOESOPHAGEAL ECHOCARDIOGRAPHY



This is not routine monitoring, but is used in situations where indicated either by:  the patient’s preoperative medical condition, e.g. left ventricular failure, sepsis;  the surgery is likely to produce significant deviations in these parameters, e.g. large blood loss or compression of the mediastinum and great vessels.



for clinical and medico–legal purposes. For patients breathing spontaneously, i.e. not paralysed, ‘awareness’ under anaesthesia is very unlikely because if such patients became ‘light’, they would indicate this by moving or coughing, long before awareness supervened. In paralysed patients however, such early warning signs are absent. This is not routine monitoring.



ANAESTHETIC TECHNIQUE AND PRINCIPLES FOR SPECIFIC ENT OPERATIONS



The aim is to detect and correct changes in left and/or right ventricular preload, afterload or contractility indicated by changes in venous and pulmonary pressures, arterial pressure and cardiac output. Transoesophageal echocardiography (TOE) is particularly useful as it provides information on ventricular volume rather than pressure.



The above sections indicate the broad principles used to determine choice of anaesthetic drugs, airway device and mode of induction. We now turn to some specific operations in ENT surgery and outline the relevant anaesthetic considerations.



MONITORING OF NEUROMUSCULAR FUNCTION



Tonsillectomy and adenoidectomy



The problems of neuromuscular blockade in the context of facial nerve monitoring in middle ear and parotid surgery has been discussed. In general, the anaesthetist monitors the adequacy or completeness of neuromuscular blockade by means of a peripheral nerve stimulator. This comprises two adhesive electrodes placed over a convenient peripheral nerve (usually the ulnar) by which current pulses of around 50 mA are passed. These pulses provoke a visible, palpable and unfatiguable twitch in the relevant muscles in the unparalysed patient. In the completely paralysed patient, no twitches are palpable. For degrees of paralysis between these extremes, a reduction in twitch amplitude (relative to the unparalysed twitch) is observed and, more particularly, a diminution in twitch amplitude with each successive impulse is seen, which is characteristic of the nondepolarizing (curare-like) neuromuscular blockers. It is important to remember that the ability to elicit palpable twitches is a crude test of neuromuscular function, in contrast to the facial nerve monitor used by the ENT surgeon, which detects microvolts of EMG activity, rather than gross movement. It is possible for patients to be moderately but adequately blocked from the anaesthetic perspective, but satisfactorily unblocked from the perspective of facial nerve monitoring. The important point here is that if this approach is used, the surgeon, as well as the anaesthetist, needs to be aware of it. INDICES OF ANAESTHETIC DEPTH: EEG/AEP/BIS



A number of devices that measure and process EEG signals are available for estimating anaesthetic depth during surgery. In addition, auditory evoked potentials can be used since increasing anaesthetic depth increases the latency and reduces the amplitude of the early cortical responses. Knowledge of anaesthetic depth is important



The most common reason for this operation is chronic or recurrent infection, and it is usually carried out in children. Thus, the incidence of upper respiratory tract infection on day of surgery is common, often requiring postponement of the operation. Extremely anxious children may need sedation on the ward using oral midazolam 0.5–1 mg/kg, and may need inhalational induction of anaesthesia, but this is rare. Most children, however, accept local anaesthetic cream (EMLA or Ametop) which facilitates i.v. cannulation and induction. It is usual for parents to be present during induction. Other reasons for the operation include: as part of treatment for sleep apnoea syndrome or snoring; excision biopsy for suspected malignancy; or peritonsillar abscess. These operations are usually performed in adults.



AIRWAY MANAGEMENT



Most anaesthetists would probably choose to intubate the trachea because this secures the airway and facilitates ventilation more definitively. Commonly a ‘preformed’ tracheal tube is used, such as an RAE tube (Figure 40.3). This tube is designed such that its outside end (which connects to the breathing circuit) points ‘south’ naturally (i.e. sits under the Boyle Davies gag) without having to be bent. The act of bending tubes usually results in kinking and obstruction. An alternative tube is a reinforced tube, whose shaft is flexible and kink resistant. An integral metallic spiral provides this protection from external compression and kinking (Figure 40.3). Since tonsillectomy is mostly carried out in children, uncuffed or ‘plain’ tracheal tubes are mostly used. This means that the lower airway is not definitively sealed and protected from soiling by blood. In addition, the laryngeal inlet cannot easily be sealed with a throat-pack (as would be done in



Chapter 40 Adult anaesthesia



other types of surgery) because this would most likely obscure the surgical field. The anaesthetist therefore relies on the surgeon to prevent blood from entering the larynx by careful haemostasis and use of suction. It is now relatively common for the flexible, reinforced version of the LMA to be used, as it is suggested that the inflated cuff, which sits around the outside of the laryngeal inlet, acts as an effective barrier to lung soiling. The reinforced LMA has a narrower tube portion which is not only kink resistant and sits under the Boyle Davies gag, but also obscures the surgical view less than the standard LMA. Whichever airway device is used, care should be taken that it remains patent during surgery, especially at the point where the Boyle Davis gag is placed (Figure 40.7).



MAINTENANCE



The traditional technique, with a number of advantages, is to use a volatile anaesthetic agent and to allow the patient to breathe spontaneously. In this technique, after induction of anaesthesia, a short-acting muscle relaxant such as suxamethonium or mivacurium is used to facilitate tracheal intubation. Spontaneous respiration will soon resume and the patient allowed to breathe a mixture of oxygen and nitrous oxide with the addition of a volatile agent such as isoflurane or sevoflurane. As discussed above, a relatively deep plane of anaesthesia will be required for the patient to tolerate the tracheal tube without coughing. Analgesia may be provided by an opioid, and titrated such that respiration is not depressed.



] 501



positioned in the lateral position, with a pillow placed under the upper chest and the head tilted down so that any blood or secretions will drain without soiling the airway. The patient, still breathing spontaneously, is then extubated at the peak of inspiration, still under a ‘deep’ plane of anaesthesia. The subsequent expiration should clear any small quantity of blood or secretions from the laryngeal inlet, and respiration should continue unimpeded and without coughing. The patient is then transferred to the recovery ward where, with modern fast onset/offset agents, the volatile agent will soon be washedout, and consciousness return. THE POSTOPERATIVE BLEEDING TONSIL



This is a serious complication. Anaesthesia and surgery may be necessary after appropriate fluid and volume resuscitation. The patient (especially a child) may have swallowed most of the blood and must be treated as a high risk for regurgitation and aspiration. At the same time, the trachea may be difficult to intubate due to the presence of blood in the oropharynx and oedema from the recent surgery. The options for anaesthetic induction are either a standard ‘rapid sequence’ intravenous technique (probably favoured by most anaesthetists), or a careful inhalational induction with the patient in the head down, lateral position. There must be good communication between the teams: the surgeon must be prepared to establish an emergency surgical airway or tracheostomy should tracheal intubation fail.



Ear surgery EXTUBATION/RECOVERY



At the end of the case, the laryngeal inlet is examined with the aid of a Macintosh laryngoscope and gently suctioned to confirm that haemostasis is secure. The patient is



Surgery on the external ear is performed for a variety of reconstructive or cosmetic reasons. Surgery on the middle ear is performed to restore hearing, eliminate infection, treat cholesteatoma or, rarely, for neoplasm.



Figure 40.7 The relative positions of: (a) the tracheal tube; and (b) the tonsillectomy (Boyle Davis) gag. Note how the blade retracts the tongue and the tube upwards.



502 ] PART 8 PERIOPERATIVE MANAGEMENT The main anaesthetic implication is that the head must be turned so that the ear on which the operation occurs is turned upward. It is one instance in ENT surgery when, strictly speaking, the airway is not ‘shared’ and so anaesthetic considerations, rather than surgical considerations, predominate in choice of airway or anaesthetic drugs. However, where tympanoplasty, tympanomeatal flaps, stepedotomy or stapedectomy is performed, it is advisable to avoid nitrous oxide. Nitrous oxide (due to its relative solubility) enters the middle ear cavity at a rate faster than nitrogen (the ambient gas in the air) leaves the space: consequently, there is a rise in middle ear pressure which can displace the structures being operated on. Myringotomy (grommet) surgery, is usually a very short operation in children and any suitable anaesthetic technique may be employed. More invasive operations of the ear include exploratory middle ear surgery and mastoidectomy. The surgical approach is often by a post-auricular approach, through the mastoid. The main consideration is that the facial nerve needs to be monitored during surgery (see above under Monitoring of neuromuscular function).



HYPOTENSIVE ANAESTHESIA



In certain operations, notably major ear surgery, special techniques are employed to improve the view of the surgical field by minimizing bleeding. These techniques have become collectively known as ‘hypotensive anaesthesia’ although they comprise a number of other approaches in addition to hypotension. Blood in the surgical field may arise from arterial, capillary or venous bleeding. Arterial bleeding is directly proportional to mean arterial blood pressure (MABP), which is governed by the following relationship: MABP ¼ cardiac output  peripheral vascular resistance ¼ ðstroke volume  heart rateÞ  peripheral vascular resistance



A reduction in stroke volume (via a reduction in contractility) and a reduction in heart rate can be achieved by b blockade. A reduction in peripheral



vascular resistance can be achieved (via vasodilatation) with alpha blocking drugs (such as phentolamine) or increased concentrations of volatile anaesthetic agents. Capillary bleeding depends on local flow in the capillary bed, which in turn is dependent on upstream arterial pressure and downstream venous tone, in addition to local metabolic factors. Capillary bleeding can be moderated by the use of locally infiltrated adrenaline (although the systemic effects of this will transiently confound the anaesthetist’s attempts to control blood pressure) and by reduction in arterial and venous PCO2 by hyperventilation. A reduction in venous tone and backpressure will reduce both capillary and venous bleeding. Venous tone can be moderated by the use of intravenous nitrates, and positioning can reduce the hydrostatic venous pressure, for example a 251 head up tilt can be employed in middle ear surgery. In young healthy patients, blood flow to vital organs such as the brain, kidney and heart is regulated such that it remains constant despite variations in perfusion pressure. This autoregulation occurs within a mean arterial pressure range of 60–160 mmHg. At pressures below 60 mmHg mean, blood flow becomes ‘pressure dependent’ and will diminish with diminishing pressure, (see Figure 40.8). For such patients, MABP may safely be lowered to 60 mmHg without diminution of vital organ blood flow. Even at pressures lower still, a modest reduction in organ blood flow is unlikely to be clinically significant since there is little metabolic demand under these basal conditions. However, in elderly patients, and particularly those with systemic hypertension or coronary artery disease, the imposition of deliberate hypotension should not be undertaken lightly. In systemic hypertension, the normal autoregulation curve (shown in Figure 40.8) is shifted to the right, and so even moderate degrees of induced hypotension may produce pressure-dependent flow and reduce vital organ perfusion. The ‘re-setting’ of these autoregulatory mechanisms is time-dependent and may take several weeks to normalize after a hypertensive



Organ blood flow



Normal patient Hypertensive patient 60



160 MABP (mmHg)



Figure 40.8 Autoregulation curve.



Chapter 40 Adult anaesthesia



patient is started on antihypertensive therapy. It is for this reason that it is unwise to embark on anaesthesia in hypertensive patients, even though we can quite easily and quickly achieve systolic and diastolic blood-pressure values that are in the normal range. The second group of patients in whom induced hypotension should be effected with caution are those with coronary or cerebrovascular disease. The arterial pressure just downstream of a stenosis may be 20 or more mmHg lower than the upstream (or systemic) arterial pressure. So, whilst a mean systemic pressure of 60 mmHg may be adequate for the majority of an arteriopathic patient’s organ systems, such a pressure may only generate 40 mmHg of pressure downstream of a fixed stenosis, thus producing a diminished and pressure-dependent flow with a risk of myocardial ischaemia or cerebral thrombosis.7



SALIVARY GLAND SURGERY Parotidectomy and removal of submandibular gland may be performed for tumour, stone or chronic infection. Any suitable anaesthetic technique may be used, but the facial nerve may need to be monitored.



NASAL AND SINUS SURGERY Nasal surgery is performed for cosmetic or functional restoration of the nasal airway, and includes operations such as septoplasty (operations to the nasal septum), rhinoplasty (operations to remodel the nasal contour), turbinectomy or these combined. Sinus surgery is performed to eliminate infection, polyps or neoplastic conditions of the sinuses, and the goal is to provide aeration of the sinuses so that secretion can drain adequately into the nasopharynx. The operations are usually performed in young adults. Anaesthetic considerations for all these operations are similar. An oral (as opposed to nasal) airway device is necessary (usually a tracheal tube in combination with a gauze throat pack to collect any blood which trickles from the nose). Nasal bleeding is often minimized by surgical application of cocaine mixtures to the nose before the operation begins. At the end of surgery, the patient’s nose is usually packed, so a good oral airway should be established to allow the patient to breathe spontaneously.



LARYNGEAL MICROSURGERY AND LARYNGOSCOPY These operations are performed for diagnostic or therapeutic purposes and may be combined with an examination of the pharynx, oesophagus or bronchial tree. Often, lasers are used to treat isolated lesions of the larynx.



] 503



Because of the rigid nature of the laryngoscope, a small size (6 mm internal diameter or less) tracheal tube (a microlaryngeal tube) is commonly employed, and if lasers are to be used, an appropriate laser-resistant type is needed. This type of tracheal tube is sufficiently flexible and long that it does not interfere with the surgical field, although good communication is needed between surgeon and anaesthetist throughout the procedure. The operation is usually relatively short, but laryngospasm or postoperative airway obstruction due to oedema of the vocal cords may occur. The risk of this may be minimized if steroids (for example dexamethasone 8 mg intravenously) are used peroperatively.



LASER SURGERY The use of laser, especially for laryngeal microsurgery, has become widespread. Lasers are also used for stapedectomy, tympanoplasty, turbinate surgery and oropharyngeal surgery. There are various types of laser (CO2, Nd:Yag, KTP; all named after the source of substrate for the laser beam). It is important to know the general properties of the laser beam that is being used. The Nd:Yag laser (yttrium-aluminium-garnet) was introduced in the 1980s, and its main advantage over CO2 laser is that it can be delivered by means of an easily handled flexible fibreoptic light cable. It has more coagulative but less cutting ability than a CO2 laser. The KTP laser is a variant of the Nd:Yag, in which the laser beam is passed through a potassium titanyl phosphate crystal. Lasers can produce either a parallel or divergent beam. Parallel beams impart much greater power densities to the tissues they strike than divergent beams. The greatest concern during use of lasers is the possibility of uncontrolled combustion of either tissues or extraneous materials (including airway devices) if accidentally struck for prolonged periods by the laser beam. A fire in the patient’s airway is serious and is often fatal. The steps shown under Best clinical practice are reasonable precautions to minimize this risk.8



Bronchoscopy Rigid bronchoscopy is performed for a variety of reasons, including inhaled foreign body, diagnosis of endobronchial lesions, and staging of disease prior to pulmonary resection. The type of patient who presents for this procedure can therefore be extremely variable, from a small child to an elderly smoker with severe lung disease. An important anaesthetic consideration is that the large rigid bronchoscope competes for space with anaesthetic devices within the trachea itself (Figure 40.9). Pulmonary ventilation needs to continue during the procedure and this can be achieved in the following ways:  Using a small tracheal tube, though the risk is that the bronchoscope may dislodge this.



504 ] PART 8 PERIOPERATIVE MANAGEMENT



Figure 40.9 A Sanders injector and rigid bronchoscope. High-pressure oxygen is fed from the regulator to a jet orifice, which clips or screws loosely onto a sideport of the rigid bronchoscope. The jet is controlled by the hand trigger. Note that the connection of the jet orifice to the sideport is not sealed and there is considerable leak so the patient is not exposed to the full pressure of the injector, but rather room air is entrained into the scope by the jetting oxygen. Note also how the conventional anaesthetic circuit can also be attached (without leak) to the sideport to allow conventional ventilation (although the leak at the level of the bronchoscope and the vocal cords may make ventilation in this way difficult).



 Using the technique of ‘apnoeic oxygenation’. This involves oxygenating the patient’s lungs with 100 percent oxygen until the lung volume is depleted of nitrogen and effectively full of oxygen and no other gas. All airway devices are removed save for a small catheter, which insufflates oxygen into the oropharynx. The surgeon is then allowed to proceed with bronchoscopy with the patient apnoeic. Because of the relatively low rate of basal oxygen consumption, and the continued minimal delivery of oxygen via the catheter, it is theoretically possible to continue oxygenation indefinitely. However, the duration of the procedure is limited by the accumulation of carbon dioxide (active pulmonary ventilation is necessary for its removal). This accumulates in the patient’s blood at a rate of approximately 0.5 kPa per minute, which limits the safe time of apnoea to approximately ten minutes.  Employing a method known as ‘jet ventilation’ using a Sanders injector (Figure 40.9). This injector is essentially a pressure relief valve and tubing, one end of which is attached directly to the high pressure (4 bar) oxygen pressure supply on the anaesthetic machine. The other end of the tubing is attached to the rigid bronchoscope. Opening the injector blows oxygen at high pressure into the lungs and inflates them. Because of the high pressures involved, such injectors can be extremely dangerous to use without training. After injection of oxygen, sufficient time must then be allowed for air to leave the lungs (passive expiration).



 Some bronchoscopes have a side-port (Racine adaptor) which allows the standard anaesthetic/ oxygen tubing to be attached: the anaesthetist can hand-ventilate the patient’s lungs in a conventional manner using this attachment during bronchoscopy (Figure 40.9). Since the anaesthetic management for rigid bronchoscopy is varied and relatively specialized, good communication between anaesthetist and surgeon is essential.



Laryngectomy This is normally performed for tumour of the larynx. A total laryngectomy involves removal of all tissues from the valecula (and sometimes base of tongue) to the second or third tracheal rings. Occasionally, the thyroid gland also needs to be removed. The pharynx is closed in a T-shape, and the trachea is brought out to the skin as an endtracheostomy (so no special tube is required long term to maintain airway patency). There are variations of this operation. A supraglottic laryngectomy involves resection of all tissue from the base of tongue to the vocal cords (which are left intact). The strap muscles are also preserved. A temporary tracheostomy is required postoperatively. A hemilaryngectomy involves removal of the epiglottis and just one half of the larynx (including one true and false vocal cord). The main anaesthetic considerations are usually:  the preoperative state of the patient;  tracheal intubation during anaesthetic induction;  changes of airway device during surgery.



Chapter 40 Adult anaesthesia



PREOPERATIVE PATIENT STATE



Laryngeal tumours commonly occur in elderly smokers, so there may be considerable coexistent cardiovascular and respiratory disease which the anaesthetist needs to manage during surgery. There may need to be a preoperative plan to optimize any coexistent medical conditions and additional monitoring (for example invasive arterial pressures and central venous monitoring) may be necessary during surgery. If lung function is particularly poor, the plan may include a period of elective postoperative ventilation in intensive care.



] 505



 The return to theatre of a bleeding tonsil poses special problems for the anaesthetist due to the probability of a full stomach and obscured laryngoscopy. Ear surgery  Moderate hypotension may help to reduce bleeding in the surgical field.  It requires a multimodal approach rather than the use of a single agent or technique.  It should be used with caution in patients with uncontrolled hypertension, cerebrovascular or ischaemic heart disease.



TRACHEAL INTUBATION



The presence of a laryngeal tumour may make conventional tracheal intubation difficult, and specialized techniques may be necessary (for example awake intubation) during induction of anaesthesia. These are further discussed in Chapter 39, Recognition and management of the difficult airway.



AIRWAY DURING SURGERY



The operation might begin with the patient’s trachea intubated with a standard (Magill) orotracheal tube. As surgery proceeds, this may (for reasons of surgical access) need to be exchanged for a tracheal tube which is inserted directly percutaneously into the trachea and shaped to lie against the neck, pointing caudally (for example Montandi tube, reinforced flexible tube or RAE tube). Finally, as the permanent end-tracheostomy is fashioned, all tracheal tubes will be removed. Thus, the anaesthetist needs to have ready a number of sterile airway devices, sterile connectors and tubing which often need to be sited directly by the surgeon. Close communication is clearly necessary.9



KEY POINTS  Anaesthesia for ENT surgery epitomizes the quintessentially anaesthetic ‘art’ of airway mastery.  It requires a unique dialogue between surgeon and anaesthetist who share the same territory. As such it requires, perhaps more than any other branch of surgery, an understanding of, and respect for each others contribution towards a common end. Tonsillectomy and adenoidectomy  The airway may be maintained by either insertion of a tracheal tube, or with a laryngeal mask. The choice is down to operator preference.



Best clinical practice Laser surgery [ The surgeon should exercise caution and avoid any direct strikes to the airway device. This risk might be minimized during oral surgery if a Boyle Davis gag without the Doughty modification is used, thus protecting the shaft of most airway devices. If the Doughty modification is used, then the shaft of the airway device might be protected using damp gauzes or swabs. [ The minimum power output of the laser should be used and the duration of laser strike limited. It has been found that the risk of combustion increases with duration of strike at the same power. [ Both oxygen and nitrous oxide support combustion. Ideally, a mixture of oxygen (perhaps limited to no greater than 30 percent) mixed with air should be used to ventilate the patient’s lungs during anaesthesia, to minimize the risk of combustion. [ Cuffs of tracheal and SAD devices are more vulnerable to laser strike than are the shafts of these devices. Therefore, the cuffs should be filled with saline (or coloured water) rather than air. This protects them, to some extent, from combustion with laser strike; it slows any deflation of the cuff after strike; it aids, to some extent, in preventing spread of any fire; and the leak of fluid allows the surgeon to see that the cuff has been ruptured.



Deficiencies in current practice and areas for future research The scope for research in anaesthesia is huge. At the most fundamental level, it is not known how anaesthetic drugs induce unconsciousness and research in this field will span neurophysiology as well as, perhaps,



506 ] PART 8 PERIOPERATIVE MANAGEMENT psychology. Of more immediate relevance to anaesthesia in the context of head and neck surgery are the following areas: Drug developments – In particular, the search continues for anaesthetic drugs that have minimal effects on the cardiovascular and respiratory systems. It is expected that such drugs will possess a better safety profile, with fewer adverse cardiorespiratory side effects. – The search for drugs that minimize – or treat – postoperative nausea and vomiting is also important. While not addressing life-saving issues, this area of research certainly aims to improve quality of life, and also costs, since delayed discharge from hospital due to nausea/vomiting is an important cause of additional costs to a healthcare organization. Airway management. Current techniques for airway management in those patients who present particular difficulties (especially common in head and neck surgery) involve relatively expensive items of equipment such as fibreoptic endoscopes, and also require additional training in their use by experts who are proficient in managing these difficult cases. It would be desirable to introduce a technique – or equipment – which is both low-cost and which requires minimal expertise in its use to maintain an airway. However, developing such devices or techniques will require far more understanding of the reasons underlying the concept of the ‘difficult airway’ than we have at present. Reducing blood loss and blood replacement. Anaesthetic techniques that help minimize blood loss during surgery (such as ‘hypotensive anaesthesia’ in which the patient’s blood pressure is held at low levels) require further research. This research also involves developments in more precise monitoring of



the circulation. Related to this are areas of research in artificial haemoglobins or haemoglobin substitutes designed to minimize the transfusion of blood and so reduce the attendant risks.



$



$



REFERENCES







 



$







1. Baker L. Premedicant drugs. In: Wilson I, Allman KG (eds). Oxford Handbook of Anaesthesia, 2nd edn. Oxford: Oxford University Press, 2006. 2. Smith G. Inhalational anaesthetic agents. In: Aitkenhead AR, Rowbotham DJ, Smith G (eds). Textbook of Anaesthesia, 4th edn. London: Churchill Livingstone, 2001. 3. Farmery AD. Physics and physiology. In: Calder I, Pearce A (eds). Core Topics in Airway Management. Cambridge: Cambridge University Press, 2005. 4. Cook TM. Supraglottic airway devices. In: Calder I, Pearce A (eds). Core Topics in Airway Management. Cambridge: Cambridge University Press, 2005. 5. Farmery AD. Physiology of apnoea and hypoxia. In: Calder I, Pearce A (eds). Core Topics in Airway Management. Cambridge: Cambridge University Press, 2005: Chapter 2. 6. Pandit JJ, Duncan T, Robbin PA. Total oxygen uptake with two maximal breathing techniques: a physiologic study of preoxygenation. Anesthesiology. 2003; 99: 841–6. 7. Dodds C. Hypotensive anaesthesia. In: Aitkenhead AR, Rowbotham DJ, Smith G (eds). Textbook of Anaesthesia, 4th edn. London: Churchill Livingstone, 2001. 8. Pandit JJ, Chambers P, O’Malley S. KTP laser-resistant properties of the reinforced laryngeal mask airway. British Journal of Anaesthesia. 1997; 78: 594–600. 9. Roberts F. ENT surgery. In: Wilson I, Allman KG (eds). Oxford Handbook of Anaesthesia, 2nd edn. Oxford: Oxford University Press, 2006.



41 Paediatric anaesthesia ALISTAIR CRANSTON



Introduction Preoperative assessment Equipment and techniques Specific surgical procedures



507 508 510 513



INTRODUCTION This chapter is intended to provide an overview of the principles of anaesthetic care for children in general, with specific reference to ear, nose and throat (ENT) surgery. The introductory section briefly deals with the organization of services for paediatric surgery and anaesthesia. This is followed by a consideration of preoperative factors, including physical assessment, preoperative fasting and premedication. The third section of the chapter addresses the equipment and skills necessary for the safe conduct of anaesthesia in children, with a particular emphasis on management of the airway. Induction of anaesthesia and physiological monitoring during anaesthesia are also covered in this section. The final section comprises an account of anaesthetic considerations and conduct for specific operative procedures commonly encountered in paediatric ENT surgery. Ear, nose and throat surgery accounts for a significant percentage of all paediatric surgical procedures and provides the basis of many anaesthetists’ experience with children. Many of these procedures are undertaken in centres where little or no other paediatric surgery is encountered. Whilst the majority of such operations are minor and of short duration, the nature of some ENT surgical procedures, particularly those involving the airway, require considerable expertise and experience. The development of cooperation and understanding between surgeon and anaesthetist is of enormous importance when undertaking such procedures.



Key points References Further reading



522 522 525



Fundamental to the delivery of a safe, high-quality anaesthetic service for children is a suitably trained anaesthetist with continuing paediatric experience. Such an individual, however, must work in an environment with appropriate facilities and support if the reasonable expectations of children and their carers are to be met. The UK Department of Health defined the requirements for such a service in The Welfare of Children and Young People in Hospital.1 There has been debate concerning who, where and what constitutes best practice in the delivery of surgical services for children, addressing such issues as training, ongoing experience and the merits of caring for children in a specialist paediatric centre as opposed to a general hospital environment.2, 3, 4 The majority of children presenting for surgical treatment in the UK have been cared for in the district hospital environment and there is no reason why most children should not be treated outside specialist centres if standards can be maintained. The 1989 National Confidential Enquiry into Perioperative Deaths reviewed the surgical and anaesthetic care of children in hospital.5 The report highlighted the problems of ‘occasional paediatric practice’ and that the outcome of surgery and anaesthesia for children is related to experience. Numerous changes, which have required considerable reorganization in service delivery for children, have been recommended. It is increasingly the case that surgeons, anaesthetists and others are now working to defined values and standards and are organizing and auditing paediatric surgical services in a way that has led to a better defined and improved service for children.



508 ] PART 8 PERIOPERATIVE MANAGEMENT



PREOPERATIVE ASSESSMENT Preoperative assessment for surgery involves an exchange of information between anaesthetist, child and carers that addresses three important issues vital to the conduct of a safe and satisfactory procedure. The first objective is to establish the nature of the procedure and to determine that the child is fit for surgery. Additionally, anatomical or physiological factors that may be specific to the condition to be treated or represent an unrelated disease process should be identified. Information obtained here will inform the development of the optimal anaesthetic care plan and may prompt further investigation or treatment before the procedure is undertaken. Thirdly, establishment of a rapport with children and their carers based on honesty and trust is vital to allay anxiety.



Psychological preparation for anaesthesia and surgery For many children and their carers, the psychological disruption attendant upon a surgical procedure is at least as disturbing as the effects of the operation itself. Anaesthesia is particularly likely to contribute to emotional disturbance and so the conduct of pre-anaesthetic preparation is of great importance. Care should be taken to understand and address the concerns of both the child and their parents simply but truthfully. An understanding of age-specific responses to stress and the provision of a dedicated, child-friendly environment are important to consider when preparing children for surgery.



Day case surgery The majority of minor ENT surgical procedures are undertaken on a day-stay basis, which affords an effective use of resources and has advantages for the child and their family. Pre-admission assessment of children, often nurse-led and employing questionnaire screening, speeds the admission process, limits wastage of operating time and reduces preventable cancellation on the day of surgery.



Physical examination All children receive a full physical examination before anaesthesia and surgery, with emphasis placed on respiratory and cardiovascular systems. A careful assessment of the airway, including mouth opening, loose dentition and nasal obstruction is important to anticipate difficulties in airway management during anaesthesia. Attention should also be paid to coexisting anomalies or disease that may require modifications of the



anaesthetic technique or particular precautions to be taken.



Respiratory assessment Airway and respiratory complications are the most common causes of peri-anaesthetic morbidity in children. Children presenting for ENT surgical procedures frequently exhibit concurrent or associated respiratory disease, emphasizing the importance of thorough preoperative evaluation. Healthy children, in particular infants and toddlers, have a significantly greater risk of respiratory complications than do healthy adults.6 This is predominantly related to maturational aspects of the respiratory system and the effects of anaesthesia on active mechanisms that help to maintain lung volume and functional residual capacity in infants and small children. In addition, stimulation of vagal airway receptors tends to increase the incidence of laryngospasm and apnoea in infants. The inherent predisposition of small children to respiratory complications is compounded by the frequency of upper respiratory tract disturbances in the ENT surgical population and the decision to proceed with or defer surgery in a child with signs of an upper respiratory tract infection (URTI) presents a common dilemma. The features of URTI associated with anaesthetic complications may be difficult to define and the decision to proceed or not is often based on personal experience and a hierarchy of symptoms. Two recent studies7, 8 have examined the incidence of and risk factors for adverse respiratory events in children with URTI. These studies suggest that many children with acute or recent URTI can undergo minor elective procedures such as myringotomy safely. The variables most likely to predict an adverse event are outlined in Table 41.1. Asthma is common in childhood and there is a perception that the prevalence is increasing. Optimization of maintenance medication may be required prior to surgery but, in the absence of significant bronchospasm or intercurrent respiratory infection, most children can be safely managed with attention to the avoidance of anaesthetic agents likely to provoke histamine release. Upper airway obstruction is frequent in paediatric ENT patients and evaluation of the cause and severity is important both in terms of airway management during anaesthesia and in determining the risk of airway difficulties in the postoperative period. Adenotonsillar hypertrophy is the most common cause of upper airway obstruction but others, notably congenital and acquired anomalies of the jaw, nasal and oral airways and larynx as well as craniofacial anomalies, particularly of the midface, are also encountered in this population. In addition, children with other conditions, such as Down’s syndrome or neuromuscular disease, may develop airway



Chapter 41 Paediatric anaesthesia Table 41.1



] 509



Disease and anaesthetic factors likely to predict an adverse event in children with URTI.7, 8



Patient factors Parental report of a ‘cold’ Presence of a productive cough Passive smoke exposure Presence of snoring Presence of nasal congestion



Anaesthetic factors Tracheal intubation versus face mask or LMA Lack of anticholinesterase administration Thiopentone or halothane induction versus propofol or sevoflurane Surgery involving the airway



obstruction in the presence of relatively mild degrees of hypertrophy. Although minor airway obstruction and benign snoring are common in children and do not usually pose particular risks, more severe airway obstruction may cause obstructive sleep apnoea syndrome (OSAS) and may predispose to significant airway compromise in the postoperative period. Preoperative OSAS, characterized by disturbed sleep with snoring, apnoeic pauses and hypoxia, in children undergoing adenotonsillectomy is a risk factor for postoperative respiratory compromise,9, 10 and prediction of those most at risk is the subject of extensive studies.11, 12, 13, 14



Preoperative fasting Preoperative fasting has long been considered essential for elective surgery. Prolonged fasting, however, does not reduce the aspiration risk during anaesthesia and current practice has focused on reduction of starvation times and a greater appreciation of other risk factors for regurgitation and aspiration of gastric contents.15 The current practice in most centres for elective surgery, summarized in Table 41.2, is based on ensuring minimal residual gastric volume at induction of anaesthesia and avoiding patient distress and potential under hydration associated with prolonged starvation. This practice is supported by studies in children and infants16, 17, 18 and appears to improve the preoperative experience of children and parents.19 In children undergoing emergency procedures, however, a full six-hour fasting period may be considered necessary, unless the urgency of the surgery dictates otherwise and anaesthesia is conducted assuming that the child has a full stomach.



Premedication Pharmacological premedication in paediatric practice remains a controversial subject. The routine use of premedicant schedules has largely been abandoned in favour of an individualized approach with specific goals in mind. Although the general trend has been a reduction in premedication and certainly in the use of



Table 41.2 Minimum preoperative fasting times in infants and children for elective surgery.



Children Infants



Clear fluids



Breast milk



Formula or cow’s milk



Solids



2 hours 2 hours



– 4 hours



6 hours 6 hours



6 hours 6 hours



intramuscular agents, topical local anaesthetics, sedative agents and anticholinergics are still widely used. The topical local anaesthetic agents EMLA cream20 and Ametop gel21 are effective in reducing the pain of venepuncture and in facilitating intravenous induction. Although effective, in some children the anticipation of an unpleasant experience may outweigh the benefits afforded. The value of sedative premedication is questionable for many children and is certainly no substitute for sympathetic psychological preparation. Historically, many classes of drugs have been used for this purpose but currently benzodiazepines and particularly Midazolam administered orally or, less frequently, intranasally, are the most popular agents.22 Midazolam affords anxiolysis and amnesia and its elimination half-life of two hours makes it suitable for use in day-case or inpatient procedures. Timing of premedication is important, maximal effect being within 30 minutes of oral administration. Preoperative administration of anticholinergic agents, particularly Atropine, has long been considered best practice in paediatric anaesthesia, especially in infants. Newer anaesthetic agents and techniques have modified this approach and there has been a general decline in the use of anticholinergic premedication,23, 24 although some still advocate its routine use.25 Procedures involving the pharynx or airway and ophthalmic operations are particularly likely to provoke vagal reflexes and many anaesthetists continue to use anticholinergics in these cases. There is an increasing trend in paediatric anaesthesia to prescribe simple analgesic or antiinflammatory drugs, such as paracetamol, preoperatively to children undergoing minor procedures. In appropriate dosage (30–40 mg/kg), this appears to be effective in reducing the use of postoperative opiate analgesia with its attendant problems of nausea and vomiting26 and is safe in terms of residual gastric volume at induction of anaesthesia.27



510 ] PART 8 PERIOPERATIVE MANAGEMENT



EQUIPMENT AND TECHNIQUES Equipment for paediatric anaesthesia Historically, anaesthetists dealing with children were faced with the task of adapting equipment designed for adult use and using their ingenuity to produce results that were at best mediocre by modern standards. Significant advances in technology and the understanding of paediatric physiology now make it unacceptable to consider anaesthesia for children without a full range of equipment designed specifically for that purpose. Most anaesthesia departments now have a designated ‘lead’ for paediatric anaesthesia with particular responsibility for the procurement and maintenance of appropriate equipment.



Paediatric breathing systems The observations of Phillip Ayre28 in 1937 on the effects of using a closed breathing system designed for adults in infants led him to develop the T-piece system with minimal dead space and resistance to breathing and a continuous supply of oxygen close to the airway. The Jackson Rees modification of this system29 described in 1950 remains popular with many anaesthetists despite many subsequent advances (Figure 41.1). The classification of anaesthetic breathing systems and an understanding of their functional characteristics30, 31 has allowed the development of a number of different systems appropriate for paediatric use.32, 33 Other considerations, such as reduction of atmospheric pollution by inhalational agents and the economic use of newer, expensive inhalation agents, have further influenced development.



Airway management equipment Much of the morbidity in anaesthetized children is related to airway difficulties, and safe management of the airway



is, perhaps, the most important aspect of paediatric anaesthesia. This is particularly the case in many ENT procedures where surgery involves a ‘shared airway’ between surgeon and anaesthetist. Notwithstanding the huge range of airway devices available, the anaesthetist’s most important assets are sound basic airway management skills and an understanding of the anatomy and physiology of the paediatric airway. Equipment choices for airway maintenance during anaesthesia are a face mask with or without an oral or nasopharyngeal airway, a laryngeal mask airway, a tracheal tube or a more complex method such as tracheostomy. Face masks are used during spontaneous breathing for short procedures, such as myringotomy and for preoxygenation and inhalational induction of anaesthesia. Ideally, they should be nonthreatening, easy to apply and achieve a good seal with minimal pressure on the face. Clear plastic face masks with cushioned rims are available in appropriate shapes and sizes for children of all ages and fulfil these requirements in all respects. Most of these are latex-free and some are available with ‘flavour’ impregnated rims, which may be helpful during inhalational induction. The laryngeal mask airway (LMA)34 introduced in the mid-1980s has greatly assisted in the management of both the routine and the difficult airway. Although originally developed for adults, LMAs are now available in sizes suitable for use in children of all ages, including neonates (Table 41.3). Laryngeal masks are particularly useful in many ENT procedures, allowing adequate surgical access whilst avoiding intubation of the trachea. The reinforced LMA, which has a kink-resistant flexometallic oropharyngeal tube, is commonly used in head and neck surgery (Figure 41.2). As well as its extensive use in routine anaesthetic practice, the LMA is increasingly being used as an adjunct to fibrescopic examination and intubation of the paediatric airway. Tracheal intubation remains the standard for airway maintenance during many procedures. The anatomical features of the child’s airway influencing tube selection are well known36 and hinge on the nature of the cricoid ring. Generally, a tracheal tube of the largest possible



Table 41.3



LMA sizes for paediatric patients.



Patient weight (kg)



Figure 41.1 ‘T’ piece breathing systems. Original system with metal connector and rubber hoses (above). Modern single use system (below).



Less than 5 5–10 10–20 20–30 Not chosen by weight Not chosen by weight Not chosen by weight



Suggested laryngeal mask size 1 1.5 2 2.5 3 4 5



Adapted from Ref. 35, with permission.



Chapter 41 Paediatric anaesthesia



Figure 41.2 LMAs. Reinforced airway (top) and standard airway (bottom).



internal diameter should be chosen to minimize resistance to gas flow and avoid an excessive leak around the tube. It is important, however, to avoid inserting too large a tube, which may cause mucosal damage with the attendant risks of post-extubation airway difficulties. Uncuffed, disposable polyvinyl chloride (PVC) tubes are available in a wide range of internal diameters and are commonly used in children under eight years of age. Although there is considerable individual variation, it is generally the case that a child’s age, rather than height, weight or other characteristics, correlates best with the optimal tube diameter. Numerous formulae to help choose the correct diameter have been proposed of which the most often quoted (for uncuffed PVC tubes) is:37 Age ðyearsÞ 3 þ 3:75 ðfor children under six years of ageÞ Age ðyearsÞ Internal diameter ðmmÞ ¼ 4 þ 4:5 ðfor children over six years of ageÞ: Internal diameter ðmmÞ ¼



(It should be noted that tubes are sized according to the internal diameter and that there may be some variation in the external diameter depending on the manufacturer and style of tube.) Below the age of two years, this guide is not reliable and the use of any formula must be modified by individual considerations. The length of the tracheal tube is similarly important, the distance between the vocal cords and the carina being age- and height-dependent. Endobronchial intubation is a common hazard and the effect of flexion or extension of the neck in some procedures must be considered. The following formulae have been suggested as a guide to tube length: Length ¼ Age ðyearsÞ=2 þ 12 ðcmÞ for orotracheal intubation Length ¼ Age ðyearsÞ=2 þ 15 ðcmÞ for nasotracheal intubation: As for tube diameter, these formulae are useful guides but correct tube position must be confirmed by careful auscultation, capnography and, if necessary, a chest x-ray.



] 511



In older children approaching puberty, cuffed endotracheal tubes are used, reflecting the anatomical development of the airway. Despite concerns about the potential for mucosal damage, the use of cuffed tubes (with high volume, low pressure cuffs) is increasing in younger children and even newborns.38 Provided that adjustments are made to size calculations and that the reduced margin of safety for endobronchial intubation39 is appreciated, this has been shown to be safe and may afford advantages in terms of gas leak and environmental pollution and may reduce the need for repeat intubation if an incorrect tube size is selected. Endotracheal tubes are available in a variety of materials, although the use of PVC and silicone rubber is now almost universal with the exception of tubes designed for laser procedures, discussed under Laser endoscopy. Similarly, a range of configurations has been developed (Figure 41.3). Preformed tubes, such as the ‘south-facing’ Ring, Adair, Eluyn (RAE) tube,40 ‘north-facing’ nasal tubes or armoured tubes may be of particular benefit in head and neck procedures, allowing security of fixation with less interference in the surgical field than standard tubes. A huge range of laryngoscopes and blades is available. Anatomical considerations and, to some extent personal choice, determine the most appropriate blade to use. In general, the relatively cephalad position of the infant larynx and the long epiglottis makes intubation easier with a straight-bladed instrument and these are often used in children under six months of age. Mackintosh pattern curved blades used in older children are positioned in the vallecula and indirectly lift the epiglottis. Vagal innervation of the posterior epiglottis makes bradycardia a potential hazard of the use of a straightbladed technique. Various blade designs, with different cross sections and shapes, are produced, as well as different blade lengths. Not infrequently, failure to position and stabilize the head correctly can lead to intubation difficulties. Attention to basic details such as this should be assured before resorting to additional equipment. Bougies or introducers over which a tube can be advanced are often helpful but care must be taken to avoid trauma to the larynx or trachea. In situations where anatomical abnormalities present particular problems, the use of a fibreoptic technique is often successful provided the operator is trained and practised in the use of such equipment.



Vascular access Secure access to the circulation is mandatory for all children undergoing anaesthesia. Intravenous access should be obtained before induction of anaesthesia in emergencies, where fluid resuscitation is necessary or when an intravenous induction of anaesthesia is planned.



512 ] PART 8 PERIOPERATIVE MANAGEMENT



Figure 41.3 Endotracheal tube patterns. From above; ‘RAE’ north and south facing, armoured uncuffed, cuffed and uncuffed pvc, Portex ‘Ivory’ for nasal intubation with metal ‘Magill’ connector.



The use of topical local anaesthetic creams (see under Premedication) is increasingly common in paediatric practice and has led to an increase in the use of intravenous induction of anaesthesia in children of all ages. Intravenous cannulae are available in a range of designs and size, generally made from inert materials such as Teflon, polyurethane or silicone. Appropriate sites for cannulation depend on the age of the child.41 Following cannulation, secure fixation should be ensured and attention taken to ensure access to the cannula after positioning and draping for surgery. In some situations, central venous or arterial cannulation may be necessary. Such interventions should not be undertaken lightly and performed only by those trained in the appropriate techniques.



Heat conservation Control of body temperature is an important consideration in paediatric surgical patients. Most children become hypothermic to some degree unless active measures are taken to avoid this. Hypothermia results partly from environmental exposure but more importantly from anaesthesia-induced failure of thermoregulatory control.



Even a modest reduction in body temperature may be undesirable; consequently measurement of body temperature and active measures to conserve heat should be a standard part of paediatric anaesthetic practice. General anaesthesia inhibits normal thermoregulatory mechanisms and superimposed heat loss to the environment predominantly radiant in nature increases the likelihood of hypothermia. Different body tissues have different temperatures and measurements are usually described as ‘core’ or ‘peripheral’. Measurement of core temperature is of greatest clinical importance and thermocouple or thermistor thermometers are widely available. Common sites for core temperature measurement, such as the nasopharynx, distal oesophagus or tympanic membrane, may be inappropriate during ENT surgery! Avoiding unnecessary exposure of the child is a simple, but effective, method of reducing heat loss, the area of skin covered being more important than the material used. In particular, the large surface area of the head in infants should remain covered if possible. Similarly raising the ambient temperature will help, although the temperature required to maintain normothermia in neonates (c. 26 1C) can create uncomfortable working conditions in the operating theatre. Hence, maintenance of the room temperature at a comfortable level and



Chapter 41 Paediatric anaesthesia



warming the microclimate around the child is more usual practice. A number of warming devices are available, but forced warm-air convective blankets appear to be the most effective and are commonly used in paediatric practice.42 When active warming devices are used, temperature monitoring is mandatory to guard against unintentional hyperthermia.



Induction of anaesthesia The transition from an awake to an anaesthetized state creates a number of physiological changes, with potentially hazardous consequences. It is the responsibility of the anaesthetist to conduct a safe induction of anaesthesia whilst maintaining a calm and unhurried manner that will reassure the child and, commonly, accompanying parents and carers. Immediately before induction of anaesthesia, patient identity, preparation and consent for all procedures should be confirmed. The function of any anaesthetic and ancillary equipment should be checked and the availability of appropriate drugs and, importantly, skilled assistance should be assured. Parental presence during the induction of anaesthesia is now the norm in UK paediatric practice and is generally agreed by both anaesthetists and parents to be helpful. The anxiety felt by many parents should not be underestimated and it is important that accompanying their child is a choice rather than compulsory. The changes that accompany induction of anaesthesia necessitate careful monitoring, both by clinical observation and through the use of physiological monitoring devices. It is often possible to apply monitoring equipment prior to the induction of anaesthesia and pulse oximetry is viewed as essential in this respect. ECG and blood pressure monitoring are desirable, but may be deferred until after induction in healthy children. Inhalational induction of anaesthesia has been the mainstay in paediatric anaesthesia and remains in common use. The development of agents such as Sevoflurane, with low blood gas solubility (hence rapid induction), less cardiovascular depression than older agents and a less pungent odour has allowed smoother and more rapid anaesthesia.43 The use of topical anaesthetic creams and the availability of fine gauge cannulae have similarly improved the acceptance of intravenous anaesthesia in children. Frequently, the choice of induction technique is of no great clinical significance and is often influenced by personal preference, either on the part of the anaesthetist or the child. However, clinical considerations may influence the choice of technique, for example in the child with a difficult or partially obstructed airway most anaesthetists regard an inhalational technique as the safest, allowing the maintenance of spontaneous respiration.



] 513



Physiological monitoring during anaesthesia Modern anaesthesia is expected to provide optimal conditions for surgical interventions whilst ensuring patient safety, minimizing morbidity and mortality. Advances in the last 20 years have allowed the development of equipment that provides almost instantaneous detection of physiological changes, particularly respiratory and cardiovascular, which has greatly enhanced the safety of paediatric anaesthesia. Minimum standards for monitoring during anaesthesia44, 45 have been widely adopted, although it must be appreciated that increasing reliance on complex electronic equipment, which may fail, should assist rather than replace basic clinical monitoring and that interpretation of information must be guided by the individual clinical situation. Routine patient monitoring includes ECG, pulse oximetry, blood pressure measurement and capnography. The importance of temperature monitoring, particularly in infants, is mentioned under Heat conservation. Routine monitoring, particularly of respiratory function and oxygenation should be continued during recovery from anaesthesia. Unfortunately, monitoring the effect of anaesthetic agents on the brain is difficult and although methods based on processed EEG signals may be used,46 these may be cumbersome and impractical in routine clinical practice. Monitoring of neuromuscular blockade is advisable, particularly in neonates or during prolonged procedures, primarily to ensure adequate recovery from relaxants at the end of surgery. During some procedures, for example parotid surgery, the surgical team may rely on stimulation of the facial nerve to avoid damage to this structure. In such surgery, monitoring of neuromuscular transmission is of particular relevance to ensure a muscular response to nerve stimulation. In some situations, patient or procedure considerations may merit more complex or invasive monitoring such as central venous pressure, direct arterial pressure, blood gas estimation or cardiac output measurement.



SPECIFIC SURGICAL PROCEDURES The principles of preoperative preparation and conduct of anaesthesia described under Preoperative assessment outline the general approach to the paediatric surgical patient. Whilst these are applicable to all procedures, specific surgical operations may require the use of particular techniques to allow adequate surgical access and to ensure patient safety and comfort.



Operations on the ear MYRINGOTOMY AND GROMMET INSERTION



Ventilation of the middle ear in children is a common procedure usually performed on a day case or outpatient



514 ] PART 8 PERIOPERATIVE MANAGEMENT basis. The majority of patients are satisfactorily managed with a standard inhalation or intravenous induction of anaesthesia and airway maintenance with an LMA provides good operating conditions47 and may have some advantages over a simple face mask technique.48 Despite the potential for nitrous oxide to increase middle ear pressure, its use is widespread for grommet insertion and has not been associated with increased postoperative vomiting.49 Paracetamol or nonsteroidal antiinflammatory agents afford adequate postoperative analgesia in most patients.50, 51 Modification of anaesthetic technique or overnight hospital admission may be necessary in grommet insertion cases in a significant minority of children with underlying conditions (such as palatal clefting, craniofacial or chromosomal anomalies) or with coexisting adenoidal hypertrophy and potential airway difficulties.



CONGENITAL EAR DEFECTS



Congenital defects of the ear may necessitate surgical intervention to reconstruct the external ear or for the implantation of bone-anchored hearing aids. Whilst the surgery is superficial and rarely of prolonged duration, the association of ear defects with other dysmorphic features of the face and mandible (particularly Treacher Collins and hemifacial microsomia syndromes) may pose significant anaesthetic concerns. These may include difficulties in visualizing the larynx and tracheal intubation52, 53 and careful assessment and planning of airway management are essential. Scrutiny of previous anaesthetic records can provide valuable information regarding airway management strategies. Fortunately, unless mouth opening is very severely limited, the LMA often allows adequate airway maintenance for these patients.54 If tracheal intubation is necessary, a full range of equipment to assist in the management of difficult intubation and experienced assistance is essential.



MIDDLE EAR EXPLORATION, MASTOIDECTOMY, MYRINGOPLASTY, COCHLEAR IMPLANTATION



Anaesthesia for middle ear surgery requires an appreciation of the conditions necessary for successful microsurgery including a quiet, bloodless field, the influence that nitrous oxide may have on a tympanic graft and, in some procedures, the importance of preservation of the facial nerve. Although adequate conditions may be achieved with a number of different anaesthetic techniques, common to them all is attention to detail. Careful preparation and judicious use of premedication will aid a smooth induction of anaesthesia avoiding hypertension and tachycardia. Subsequently, the avoidance of airway obstruction and minimizing mean airway pressure will help reduce venous bleeding which is often



more troublesome than arterial bleeding. When positioning the child for surgery, a 15 to 20 1 head-up tilt is helpful to reduce venous congestion, provided that compression of the neck, chest and abdomen is also avoided. The rapidity with which depth of anaesthesia can be altered with modern inhalational or intravenous agents such as Desflurane and Remifentanil means that hypotensive agents are rarely required. Tracheal intubation and controlled ventilation is a common technique for middle ear surgery, although the use of a reinforced LMA may reduce the potential for airway complications at the end of surgery. Clearly, avoidance of coughing, laryngospasm and straining at the end of surgery is equally important as at induction of anaesthesia. The solubility of nitrous oxide allows its diffusion into and out of air-filled spaces more rapidly than air leading to an increase in volume of the space, or an increase in pressure in noncompliant spaces such as the middle ear. Thus, pressure in the intact middle ear increases during nitrous oxide anaesthesia.55 When nitrous oxide is discontinued, pressure falls as nitrous oxide diffuses out faster than air diffuses in. These pressure changes may distort or displace a tympanic graft and so nitrous oxide should be discontinued before the middle ear is closed. The proximity of the facial nerve to the surgical field in some middle ear procedures demands careful monitoring to avoid nerve damage. A nerve stimulator may be used in the surgical field to identify branches of the facial nerve. It is important that the anaesthetist is aware of this and monitors the effect of any neuromuscular relaxant agents used. The anaesthetic considerations for cochlear implantation are essentially the same as for other middle ear surgery although surgery, particularly if bilateral, may be prolonged. Profound sensorineural hearing loss may be associated with other features of relevance to the anaesthetist. For example, the prolongation of the QT interval in the Jervell and Lange-Nielsen syndrome presents a daunting challenge.56



Operations on the nose, nasopharynx and pharynx CHOANAL ATRESIA



Bilateral occlusion of the posterior nares may cause acute respiratory distress in the neonatal period and maintenance of a patent oral airway or, sometimes, endotracheal intubation is necessary until surgical intervention is undertaken. Repeated anaesthesia may be required for nasal dilatation and restenting during early infancy. In some cases, the cytotoxic agent Mitomycin C may be topically applied following nasal dilatation57 and great



Chapter 41 Paediatric anaesthesia



] 515



care should be taken to avoid exposure of surgical, anaesthetic and theatre staff. ADENOIDECTOMY AND TONSILLECTOMY



Adenotonsillectomy remains one of the most common paediatric surgical procedures and comprises much of many anaesthetists’ paediatric practice. Although often considered a routine procedure, a minority of patients with severe upper airway obstruction from adenotonsillar hypertrophy present stern challenges for even the most experienced anaesthetist. Considerable variation in pre-, intra- and postoperative management of children undergoing adenotonsillectomy exists and controversies abound. Preoperative assessment is informed by the indication for adenotonsillectomy and, as indicated above, identification of children with OSAS is of great importance when planning intra and postoperative care. Although largely a clinical diagnosis, sleep studies with pulse oximetry and cardiological assessment may be required in children with severe or longstanding symptoms. Prediction of postoperative apnoea and hypoxia in children with severe OSAS will necessitate careful observation with pulse oximetry, supplemental oxygen and occasionally airway support.11, 12 Particular difficulties may be expected in younger children under three years old13 and those with other congenital or acquired anomalies. Routine preoperative coagulation testing for healthy children is not generally indicated unless a history of bleeding tendency is suspected.58 Factors dictating an appropriate anaesthetic technique include the preoperative and expected postoperative degree of airway compromise, airway security during surgery, postoperative pain relief and control of nausea and vomiting, postoperative bleeding and the proposed duration of hospital stay. In addition, recent concerns regarding the potential transmission of prion disease (variant Creutzfeld Jakob disease) have, in the UK, determined anaesthetic technique to some extent. In children with significant obstructive symptoms induction of anaesthesia may precipitate severe obstruction, which can usually be improved by continuous positive airway pressure (CPAP) and the insertion of an oropharyngeal airway. In such patients, a technique employing endotracheal intubation, ventilation and minimal opiate use affords good operative conditions and an awake child at the end of surgery. Careful postoperative observation with supplemental oxygen therapy and sometimes nasophayngeal airway support or even tracheal intubation may be needed in the most severely obstructed patients. Particular attention should be directed to the care of such children on the first postoperative night. Protection of the airway from blood and tissue debris is essential during surgery. To achieve this, most anaesthetists employ endotracheal intubation or the reinforced LMA. Tracheal intubation, usually with a



Figure 41.4 Insufflation of oxygen and volatile anaesthetic agent via nasopharyngeal tube.



preformed plastic RAE pattern tube40 (see Figure 41.4) has been the mainstay of airway management prior to the introduction of the reinforced LMA and is usually satisfactory for both surgeon and anaesthetist. Potential problems inherent in the fixed length of such tubes are well known and occasionally difficulties may be encountered with compression from the mouth gag, particularly if the gag is short or opened excessively. These difficulties are more likely with small tube sizes (less than 4.5 mm internal diameter) as is the potential for ‘wedging’ of the tube in the gag blade and accidental extubation of the patient when removing the gag.59 The reinforced LMA has been widely used for tonsillectomy60, 61 and may have some advantages over tracheal intubation. Placement of the mask does not require the use of muscle relaxants, creates a seal that protects the larynx from blood and debris and can be left in place to secure the airway until protective reflexes have returned. Use of the LMA requires full surgical cooperation and may necessitate particular care when positioning the mouth gag. Difficulties with surgical access62 may preclude the use of the LMA, particularly in smaller children (less than 15 kg) or when other complicating factors are present. A meta-analysis of the use of LMA in tonsillectomy concludes that this is a safe alternative to intubation provided full cooperation between surgeon and anaesthetist is ensured.63 Analgesia following adenotonsillectomy in children has been achieved by a variety of means, most of which



516 ] PART 8 PERIOPERATIVE MANAGEMENT Table 41.4



Suggested doses of paracetamol and NSAIDs in children.



Oral paracetamol Rectal paracetamol Oral ibuprofen Oral diclofenac Rectal diclofenac



Loading dose (mg/kg)



Maintenance dose (mg/kg)



Dose interval (hours)



Maximum daily dose (mg/kg/24 h)



20 30 5 N/A N/A



15–20 20 4–10 1 1



4–6 6–8 6–8 8–12 8–12



90 90 20 3 150 mg/day



Reproduced from Ref. 64, with permission.



involve a combination of paracetamol and nonsteroidal antiinflammatory drugs (NSAIDs), often with the addition of low-dose opioids such as codeine phosphate. The combination of paracetamol and codeine has long been used in paediatric practice and appears to be effective. Many anaesthetists use intravenous opioid drugs such as fentanyl or morphine intraoperatively, although this may increase postoperative nausea and vomiting and caution is advocated in the presence of severe obstructive symptoms. There has been much debate regarding the efficacy and safety of varying doses and routes of administration of both paracetamol and NSAIDs in children. A suggested dosage schedule is indicated (Table 41.4) based on the Royal College of Paediatrics and Child Health (RCPCH) guidelines.64 Guidelines for the use of NSAIDs in the perioperative period based on a structured literature review have been issued by the Royal College of Anaesthetists.65 There is good evidence for the analgesic efficacy of NSAIDs compared with opioids after adenotonsillectomy, although a risk of increased bleeding has been shown in some studies. The guidelines suggest that NSAIDs (but not paracetamol) be avoided in children with proven asthma, particularly if associated with nasal polyps, severe eczema or atopy, although most anaesthetists believe that it is reasonable to use NSAIDs in asthmatic children in whom previous exposure to these drugs has been uncomplicated. Studies published since this review are generally supportive of the analgesic benefits of NSAIDs although are not conclusive on the issue of increased bleeding risk.66, 67, 68 A recent systematic review concluded that the use of NSAIDs in paediatric tonsillectomy did not cause any increase in bleeding requiring a return to theatre and that there was less nausea and vomiting when NSAIDs were used compared to other analgesics.69 It has been suggested that the surgical technique employed during tonsillectomy may influence postoperative pain although the evidence for this is conflicting.70, 71, 72 Similarly, the infiltration of the tonsillar bed with local anaesthetic agents is advocated by some authors,73, 74 whilst a systematic review of this practice concludes that there is no evidence that postoperative pain control is improved.75 Nausea and vomiting following adenotonsillectomy is a significant problem in terms of patient comfort and



satisfaction and may delay discharge of children undergoing day care. The 5-hydroxytryptamine antagonists ondansetron and tropisetron are effective in reducing postoperative nausea and vomiting either alone or in combination with dexamethasone.76, 77 Primary haemorrhage requiring surgical intervention in the first few hours following adenotonsillectomy is an uncommon (0.5–1 percent) but potentially life-threatening event that requires immediate attention. Blood loss is often difficult to estimate but clinical signs of tachycardia, pallor and reduced capillary refill should prompt intervention. Blood pressure is usually maintained in children until 25 percent or more of circulating volume has been lost and hypotension is a late and very worrying sign. Rapid volume replacement with crystalloid or colloid solutions should be instituted and hypovolaemia corrected whilst a theatre is prepared for surgery. During this time, blood is cross-matched and a full blood count and a coagulation screen performed. The anaesthetic approach to a bleeding post-tonsillectomy patient must address the problems presented by the potential for regurgitation of swallowed blood together with active bleeding obscuring the view at laryngoscopy. Although opinions differ, most anaesthetists would favour a rapid sequence intravenous induction of anaesthesia with preoxygenation and cricoid pressure with the patient in a head down position on the operating table. In addition to the usual anaesthetic equipment, a back-up suction device and tracheal tubes smaller than that used at the original procedure should be to hand. Following the provision of a secure airway, close cardiovascular monitoring and further fluid or blood transfusion are undertaken during surgery. The stomach should be emptied via a wide bore nasogastric tube and extubation should be performed with the child awake in a lateral, head down position. A postnasal pack placed to stem bleeding from the adenoid bed is poorly tolerated by younger children and may necessitate sedation and intubation in an intensive care unit until the pack is removed. Adenotonsillectomy is now frequently performed on a day stay basis, particularly in North America, and appears safe provided that careful selection criteria and meticulous operative techniques are followed. Strict observation in the postoperative period, usually for four to six hours



Chapter 41 Paediatric anaesthesia



postoperatively, is ensured and clear instructions issued for homecare and support following discharge. Despite an increase in day care tonsillectomy in the UK, many patients remain in hospital overnight, as much for social and organizational reasons as for clinical indications. In recent years, the potential for transmission of vCJD via surgical equipment has prompted the UK Department of Health to implement improvements in decontamination procedures for surgical instruments, as advised by the Spongiform Encephalopathy Advisory Committee (SEAC).78 In addition, adenotonsillectomy was identified as a procedure suitable to pilot the use of single use equipment as patients are generally young, a limited number of instruments are used and patient safety was unlikely to be compromised. Disposable surgical and anaesthetic equipment was introduced in 2001, but increased reports of adverse incidents prompted a withdrawal of the directive in December of that year. Although the return to reusable equipment was applied to anaesthetic and surgical equipment, it was subsequently decided, presumably on the basis of relative risk, that this change did not apply to anaesthetic equipment.79 Currently then, anaesthetic equipment placed in the mouth or respiratory tract during tonsillectomy must be disposable or covered by a disposable protective sheath.



Airway endoscopy procedures Anaesthesia for children undergoing airway endoscopy procedures requires an understanding of the medical condition and likely airway pathology, an appreciation of the surgical requirements for the procedure and the functions, hazards and limitations of bronchoscopic and laser equipment, close cooperation between all members of the theatre team and, not infrequently, a degree of ingenuity and adaptation.



DIAGNOSTIC AIRWAY ENDOSCOPY



Infants and children presenting for diagnostic laryngoscopy and bronchoscopy usually exhibit some degree of airway obstruction, stridor being the most frequent presenting feature. In many cases, a careful history of the onset and features of the obstruction will suggest a likely cause. Investigations, including chest and neck x-rays, barium swallow studies, respiratory function studies (although these are often difficult and unreliable in small infants) may further assist diagnosis. Preoperative information is useful but preparation should be made for unexpected findings and a clear appreciation of the requirements of the surgeon and a cooperative approach are essential to allow accurate diagnosis to be made safely. The anaesthetic technique must afford assessment of both fixed and dynamic elements of the airway whilst



] 517



ensuring airway maintenance and adequate oxygenation and ventilation of the child. An unobstructed view of laryngeal, tracheal and bronchial structures is necessary in a still, spontaneously breathing patient. Depth of anaesthesia must be sufficient to control the intense stimulation associated with laryngoscopy but also allow assessment of vocal cord and cricoarytenoid function without laryngospasm in the almost awake patient. It is essential that a clear understanding exists between anaesthetist, surgeon and operating theatre personnel as to the conduct of the procedure. A range of airway equipment and experienced assistance should be available and all staff should be familiar with the assembly and use of bronchoscopic equipment. Sedative premedication is usually avoided in children with airway obstruction although may be useful in carefully selected patients (for instance frequent attenders) if anxiety is likely to worsen existing airway obstruction. Despite the general reduction in the use of anticholinergic premedication in children, most anaesthetists continue to use these agents in airway endoscopy procedures. The benefits include control of secretions, reducing the incidence of breath holding and laryngospasm, attenuation of vagal responses to airway instrumentation and deep inhalational anaesthesia, as well as more effective topical anaesthesia of the larynx.80 Traditionally, atropine has been used preoperatively for this purpose, either orally or intramuscularly, and certainly is effective. The unpleasant sequelae of this practice for the child (i.m. injection, dry mouth, and diplopia) should be considered and many anaesthetists now employ intravenous atropine or glycopyrrolate at the time of induction of anaesthesia. The technique most commonly employed for diagnostic airway examination involves a volatile anaesthetic agent in combination with topical anaesthesia of the larynx. Maintenance of spontaneous ventilation, certainly until it is established that positive pressure ventilation is possible, is paramount and, in general, an inhalational induction is recommended using 100 percent oxygen and an increasing concentration of volatile agent. Historically, halothane has been the agent of choice but is increasingly being superseded by sevoflurane, which affords a more rapid induction. Debate continues over the place of halothane in paediatric anaesthesia,81 although its availability and cost mean that it is still widely used worldwide. Airway obstruction prolongs the induction of anaesthesia and gentle assistance of ventilation with the application of CPAP may speed the process and improve gas exchange. Intravenous access, if not already secured, is obtained and depth of anaesthesia increased. Some anaesthetists choose to intubate the trachea following induction of anaesthesia, affording airway security for transfer to theatre and setting up of equipment. Intubation, however, may be difficult and muscle relaxants should not be used to facilitate this unless positive pressure ventilation is shown to be possible. Others avoid this to allow a surgical view of the unsullied



518 ] PART 8 PERIOPERATIVE MANAGEMENT larynx. If intubation is performed at this point, the nasotracheal route is convenient, allowing the tube to be withdrawn into the nasopharynx for endoscopy. Topical anaesthesia of the larynx, commonly using lignocaine, is an essential component of the anaesthetic technique to reduce coughing and laryngospasm and to allow assessment of vocal cord and cricoarytenoid function at very light planes of anaesthesia and during awakening. Although there are some concerns that topical anaesthesia may produce subtle changes in glottic function, and one study reports exaggeration of the magnitude of the signs of laryngomalacia in sedated children undergoing flexible bronchoscopy,82 the benefits outweigh these considerations. The glottis, vallecula and trachea are anaesthetized, usually by the application of lignocaine spray. There is considerable risk of laryngospasm during spontaneous ventilation and often a short-acting muscle relaxant is used if intubation has not been performed. It appears, however, that the use of relatively dilute (2 percent lignocaine) solutions without preservative and flavourings found in multi-use metered dose sprays, significantly reduces the risk of laryngospasm (personal observation). Systemic effects from absorption of local anaesthetic agents from the mucosa may occur and it is recommended that dosage is limited to 4–5 mg/kg. During the examination of the airway, spontaneous ventilation is maintained and anaesthesia is continued by insufflation of oxygen and a volatile agent, usually via a tracheal tube placed in the nasopharynx (Figure 41.4). This ‘open’ technique is well established, allows spontaneous ventilation to be maintained and has the merit of simplicity. Of major concern, however, is contamination of the working environment with volatile anaesthetic agents and occupational exposure of theatre personnel to concentrations considerably higher than currently recommended health regulation guidelines permit.83 Alternative techniques for maintenance of anaesthesia, using intravenous agents such as propofol in place of volatile agents, are well described84 and increasingly employed, even in small infants. If intubation is not undertaken during anaesthetic induction this might form part of the examination to ‘size’ the larynx and subglottis allowing documentation of any stenotic lesions based on the external diameter of the endotracheal tube. This information is important when serial examinations are undertaken and also extremely useful anaesthetic information should the child present for other surgical procedures. At the end of the procedure the patient is allowed to waken with the laryngoscope in the vallecula to assess vocal cord function fully. Careful monitoring of cardiovascular and respiratory function is imperative during endoscopy procedures and it is important that the operator, as well as the anaesthetist, are aware of any changes, particularly hypoxia or hypoventilation that may necessitate interruption of the examination or urgent intubation of the



trachea. In this respect pulse oximetry with an audible tone and a camera allowing the anaesthetist to view the surgical field are particularly helpful in promoting patient safety. Following the examination, careful observation and continued monitoring is necessary as instrumentation may exacerbate the pre-existing airway problems. Although infrequently necessary, facilities for immediate tracheal intubation should be at hand. Humidified oxygen is often used during the recovery phase and nebulized adrenaline may be useful if airway oedema is suspected. If topical anaesthesia has been applied to the larynx, oral fluids should be withheld for one to two hours postoperatively. Many anaesthetists administer intravenous steroids such as dexamethasone to children undergoing these procedures in an attempt to reduce the development of oedema. Although evidence suggests that this may reduce post-extubation stridor in children and reintubation rates in neonates in intensive care units,85 the value of steroids during airway endoscopy is questionable. Limiting the number and duration of airway instrumentations and ensuring that appropriate calibre endoscopes are used is probably of greater importance.



BRONCHOSCOPY



Anaesthetic considerations and techniques for rigid bronchoscopy are essentially the same as for diagnostic laryngoscopy. The use of Storz scopes with Hopkins telescopic rods is almost universal (Figure 41.5) affording excellent optical quality and ease of anaesthetic delivery via a side arm that allows connection to the anaesthetic breathing system. Insertion of the telescope effectively closes the system allowing ventilation of the trachea, although there is a considerable resistance to airflow, particularly with small scopes, and assisted ventilation is often required in infants. Occasionally, it is necessary to withdraw the telescope temporarily to allow adequate oxygenation or ventilation. Adequate topical anaesthesia of the lower airways is often difficult to achieve and coughing may be a problem when the scope is advanced to the carina and beyond,



Figure 41.5



‘Storz’ bronchoscopy equipment.



Chapter 41 Paediatric anaesthesia



despite deep anaesthesia. Intermittent neuromuscular blockade may be used, provided adequate ventilation via the bronchoscope is assured, but this may preclude a full dynamic assessment of the lower airways. As in laryngoscopy, constant communication and cooperation between surgeon and anaesthetist is essential. Fibreoptic bronchoscopes may also be used to evaluate the upper and lower airways,86 increasingly so with the availability of ultra-thin scopes suitable for use in infants. Although the procedure may be performed under sedation, anaesthesia may be necessary. The use of an LMA through which the bronchoscope is passed allows safe airway management whilst affording easy passage and manoeuvring of the fibrescope.87



FOREIGN BODY REMOVAL



Removal of an inhaled foreign body is a common indication for bronchoscopy in toddlers. Foreign body inhalation may precipitate acute airway obstruction and require urgent intervention or, more commonly, presentation may be delayed if a small object that passes beyond the main bronchi is inhaled. The conduct of anaesthesia is largely the same as for diagnostic endoscopy although the procedure may be prolonged and require multiple instrumentation of the airway. A spontaneous respiration technique, as described above, is frequently employed with volatile or intravenous anaesthetic agents. Some advocate controlled ventilation via the bronchoscope with neuromuscular blockade although caution should be exercised if a tracheal foreign body is likely to produce distal air trapping.88 Whatever anaesthetic technique is employed, both anaesthetist and endoscopist should be aware of the potential difficulties involved. Airway inflammation and oedema, particularly when an organic foreign body is present, exacerbated by prolonged instrumentation and if removal is difficult, merits careful observation and follow up.



Operative procedures of the airway Advances in microscope and laser technology have allowed the development of operative procedures on the paediatric airway that require particular attention from the anaesthetist. Frequently, patients with varying degrees of airway obstruction undergo repeated procedures and the special risks associated with the use of lasers must be taken into account. Although the techniques described for diagnostic examination of the airway are satisfactory for many surgical procedures, alternative methods for airway management have been employed with varying degrees of success. The simplest of these involves intubation with a small tracheal tube that can be withdrawn into the nasopharynx when surgical access is impeded and



] 519



anaesthesia is continued with insufflation of oxygen and volatile agents as described above. Intermittent apnoeic oxygenation is a modification of this technique using neuromuscular blockade and intravenous maintenance of anaesthesia. Following a period of ventilation, the endotracheal tube is withdrawn to allow surgical access and reinserted repeatedly to allow intermittent ventilation, surgery being interrupted for this. Although successful and safe provided that close monitoring and a cooperative approach are ensured, infants tolerate only short periods of apnoea and repeated intubation may traumatize the airway. Venturi jet ventilation involves entrainment of air produced by the delivery of high pressure gas boluses from a cannula attached to the suspension laryngoscope89 or placed in the trachea. Muscle paralysis is maintained and ventilation monitored by observation of chest movement. Great care is taken to ensure that the cannula is correctly aligned and unobstructed at all times as gastric distension or direct injury to the airway from the high pressure jet may occur. This technique is not suitable when there is significant obstruction at glottic level as serious barotrauma may result. Trans-tracheal high frequency jet ventilation uses a similar principle where ventilation is provided by a high frequency jet ventilator via a cannula inserted percutaneously into the trachea via the cricothyroid membrane.90 The driving pressure is similar or less than that used during conventional ventilation, reducing the risk of barotrauma, and the technique is suitable for use when there is significant obstruction at laryngeal level. Injuries such as extensive surgical emphysema may occur if the catheter is misplaced and a special jet ventilator is required. Use of these techniques will allow adequate anaesthetic and surgical conditions for the majority of procedures. In some children, however, the degree of airway obstruction may necessitate tracheostomy for treatment to be undertaken safely. The availability of tracheostomy equipment and the ability to secure the airway rapidly is mandatory when undertaking laryngeal surgery in children.



Laser endoscopy Use of the carbon dioxide and other laser technology is well established for the treatment of laryngeal papillomatosis and other lesions of the larynx and trachea in children.91 The hazards of laser use (see Chapter 58, Laser principles in otolaryngology, head and neck surgery) must be appreciated by all theatre personnel and strict guidelines followed during laser activation. Apart from general considerations, the major difficulty with laser surgery is safe airway management and the avoidance of airway fires caused by ignition of nonmetallic materials. The choice of anaesthetic technique



520 ] PART 8 PERIOPERATIVE MANAGEMENT



Figure 41.6 Laser endotracheal tubes. ‘Lasertrach’ copper wound red rubber tube (above), ‘Laserflex’ uncuffed and double-cuffed tubes (below).



will depend on the lesion being treated, patient size and, to some extent, the preferences of surgeon and anaesthetist. Often a tubeless technique with insufflation of oxygen via a nasal catheter and maintenance of anaesthesia with volatile or intravenous agents is satisfactory, safe and reduces the risk of mishaps. This technique is widely employed and appears safe and efficient for children undergoing treatment of recurrent laryngeal papillomatosis.92 Although concerns have been raised about the use of high oxygen concentrations in the presence of laser energy, many children require this to avoid hypoxia and, provided all sources of combustion are avoided, the risk appears minimal. If tracheal intubation is necessary, disposable metal or foil protected tubes are available in a range of sizes down to 3 mm internal diameter (Figure 41.6). All of these tubes, however, are either lined with PVC or have a rubber core and so there is a small risk of ignition. Cuffs on the larger tubes are also made from PVC or rubber and should be filled with saline rather than air. The construction of these tubes is such that the external diameter is large when compared to a standard tube of the same internal diameter and care should be taken to avoid trauma to the airway. In children with a tracheostomy, the tube should be removed or replaced with a metal tube before the laser is activated. Similarly, if trans-tracheal jet ventilation is used, the cannulae should be metallic. Although every effort is made to reduce the risk of ignition, prompt action is required in the event of an airway fire to minimize the damage. Any combustible material should immediately be removed and oxygen delivery stopped. The fire should be extinguished with saline or water and ventilation resumed by mask. A thorough and careful examination of the airway is performed to assess damage and formulate a management plan that may include prolonged intubation or urgent tracheostomy.93



ARYEPIGLOTTOPLASTY



Laryngomalacia is a common indication for laryngoscopy in infants and aryepiglottoplasty may be indicated. Following diagnostic endoscopy, anaesthesia can be continued with an insufflation technique or, commonly, a nasotracheal tube is inserted.94 The child can be extubated at the end of surgery and although careful observation is necessary, postoperative problems are rare. Some surgeons prefer to use the carbon dioxide laser rather than microlaryngeal instruments for supraglottoplasty, in which case full laser precautions are necessary. LARYNGEAL CYSTS



Laryngeal cysts are uncommon but may give rise to intermittent airway obstruction. Anaesthetic considerations are essentially the same as for any other cause of upper airway problem, although positive pressure ventilation may be impossible if the cyst produces a ballvalve obstruction to the larynx. Although intubation is usually possible, urgent aspiration of the cyst may be required if the larynx is completely obscured. Following definitive deroofing, extubation and recovery are usually uneventful. LARYNGEAL CLEFTS



Posterior deficiency of interarytenoid tissue may prevent complete glottic closure and result in repeated episodes of aspiration and recurrent pulmonary problems. This may be attributed to other pathology frequently present in these infants. Minor clefts may be closed by direct suture using a microlaryngeal technique under orotracheal anaesthesia. Long clefts are associated with significant morbidity and mortality and usually require tracheostomy and gastrostomy for pulmonary protection.



Chapter 41 Paediatric anaesthesia



Anaesthetic and surgical management of these patients is complex and repeated procedures may be necessary.



Laryngotracheal stenosis and reconstruction ANTERIOR CRICOID SPLIT



Cricoid split procedures are usually undertaken in neonates (often premature) with mild stenosis of the cricoid ring. These infants frequently present as extubation failures following a period of neonatal intubation and ventilation. Endotracheal intubation and positive pressure ventilation are continued during the procedure and, following laryngofissure, a larger sized tube is placed to stent the larynx for five to ten days postoperatively. The tip of the tube must be positioned distal to the split and securely fixed. Postoperatively, meticulous intensive care management is required to ensure that reintubation, which may cause extensive damage, is avoided. Should extubation fail following a cricoid split, tracheostomy is usually necessary.



LARYNGOTRACHEAL RECONSTRUCTION



Reconstructive procedures on the larynx usually involve laryngofissure and interposition of an anterior and/or posterior cartilage graft to widen the lumen. The procedure may be single-stage, following which the larynx is stented with an endotracheal tube for a short period postoperatively or as a staged procedure if a tracheostomy has already been performed to secure the airway. The anaesthetic technique for the single-stage procedure is as for cricoid split and postoperative management in an intensive care environment is essential to ensure tube patency and minimize laryngeal trauma. A similar technique is employed when treating suprastomal collapse in patients with a tracheostomy. To allow surgical access the head is extended and the existing tracheostomy tube is replaced with a reinforced tube that can be sutured or taped to the chest during surgery. If possible, a cuffed tube is used to avoid tracheal soiling with blood and debris and also to prevent anaesthetic gas leakage via the opened airway. Dilation of the tracheostome may be necessary to allow insertion of an appropriate calibre endotracheal tube. In smaller patients the cuff may impinge into the surgical site or a risk of endobronchial intubation may necessitate the fashioning of a shortened preformed tube. In this case, leakage of anaesthetic gases may make adequate ventilation of the patient difficult at some stages of the procedure and close monitoring of gas exchange is vital. However the trachea is managed, constant vigilance is required to maintain airway patency and avert displacement of the tube during surgical manipulations.



] 521



If an anterior costal cartilage graft is taken, the possibility of a pneumothorax must be recognized. Significant pain may also result from the graft site and infiltration with local anaesthetic agents and postoperative opioid analgesia are often required. A chest x-ray is essential at the end of the procedure to exclude a pneumothorax. Postoperatively, airway secretions are increased and aspiration may occur if the stent causes glottic incompetence.



Tracheostomy Many children undergoing tracheostomy are intubated and anaesthesia is continued via the existing tracheal tube. Initially, spontaneous ventilation and topical anaesthesia of the larynx may be advised if it is proposed to extubate the child for a brief examination of the airway before reintubation and tracheostome formation. In the unintubated child, inhalational induction of anaesthesia is usual, followed by intubation with an appropriate sized endotracheal tube. A range of tubes should be available if subglottic stenosis is present, the smallest Portex tube having an internal diameter of 2 mm. Occasionally, a pinhole subglottic airway will preclude intubation even with the smallest tube in which case an LMA or face mask with spontaneous ventilation may prove an adequate technique. Before surgery an appropriate sterile breathing system and connectors should be selected and the ability to connect the tracheostomy tube to the breathing system is checked. Secure intravenous access and full cardiovascular and respiratory monitoring are essential during the procedure and attention should be paid to temperature maintenance, particularly in small babies. Surgical drapes should be placed in such a way that immediate access to the endotracheal tube is assured should airway difficulties occur during surgery. The child is ventilated with 100 percent oxygen prior to tracheal incision and under direct vision the tip of the tracheal tube is withdrawn to allow insertion of the tracheostomy tube. The tube should not be removed from the larynx until correct placement of the tracheostomy is checked, to allow reinsertion should difficulty be experienced in tracheostomy insertion. Anaesthesia is continued via the tracheostomy using a sterile connector and breathing system passed out from the surgical field. Careful inspection of chest wall movement, chest ausculatation, oxygen saturation and end tidal carbon dioxide measurement are imperative to assure correct placement of the tracheostomy tube. At the end of surgery, the tracheostomy tube is secured with ties or tapes and careful suction of the pharynx and tracheostomy tube is performed. If tracheal ‘stay’ sutures have been placed in the tracheal wall, these should be carefully marked and instructions issued for the management of accidental tube dislodgement.



522 ] PART 8 PERIOPERATIVE MANAGEMENT



KEY POINTS  ENT surgery provides the core of many anaesthetists’ paediatric practice.  Local organization of surgical services for children is in a state of flux.  Careful preoperative assessment and preparation is vital for the safe conduct of anaesthesia.  Relatively recent advances in technology and understanding of paediatric physiology have led to significant improvements in anaesthetic practice.  Attention to detail, particularly regarding airway management, is essential.  Close cooperation between anaesthetist and surgeon and a mutual understanding of the problems faced by each affords better care for patients.  Objective evidence to support clinical practice in paediatric anaesthesia is limited.



REFERENCES 1. Department of Health. Welfare of children and young people in hospital. London: HMSO, 1991. 2. McNichol R. Paediatric anaesthesia – who should do it? The view from the specialist hospital. Anaesthesia. 1997; 52: 513–5. 3. Rollin A-M. Paediatric anaesthesia – who should do it? The view from the district general hospital. Anaesthesia. 1997; 52: 515–6. 4. The Report of a Joint Working Group. The transfer of infants and children for surgery. London: British Paediatric Association, 1993. 5. Campling EA, Devlin HB, Lunn JN. The report of the confidential enquiry into perioperative deaths 1989. London: NCEPOD, Royal College of Surgeons of England, 1990. 6. Holzman RS. Morbidity and mortality in pediatric anesthesia. Pediatric Clinics of North America. 1994; 41: 239–56. 7. Tait AR, Malviya S, Voepel-Lewis T, Munro HM, Siewert M, Pandit UA. Risk factors for perioperative adverse events in children with upper respiratory tract infections. Anesthesiology. 2001; 95: 299–306. 8. Parnis SJ, Barker DS, Van Der Walt JH. Clinical predictors of anaesthetic complications in children with respiratory tract infections. Paediatric Anaesthesia. 2001; 11: 29–40. 9. Biavatti MJ, Manning SC, Phillips DL. Predictive factors for respiratory complications after tonsillectomy and adenoidectomy in children. Archives of Otolaryngology and Head and Neck Surgery. 1997; 123: 517–21.



10. Gerber ME, O’Connor DM, Adler E, Myer 3rd CM. Selected risk factors in pediatric adenotonsillectomy. Archives of Otolaryngology and Head and Neck Surgery. 1997; 123: 116–7. 11. Rosen GM, Muckle RP, Mahowald MW, Goding GS, Ullevig C. Postoperative respiratory compromise in children with obstructive sleep apnea syndrome: can it be anticipated? Pediatrics. 1994; 93: 784–8. 12. Wilson K, Lakheeram I, Morielli A, Brouillette R, Brown K. Can assessment for obstructive sleep apnea help predict postadenotonsillectomy respiratory complications? Anesthesiology. 2002; 96: 313–22. 13. McColley SA, April MM, Carroll JL, Naclerio RM, Loughlin GM. Respiratory compromise after adenotonsillectomy in children with obstructive sleep apnea. Archives of Otolaryngology and Head and Neck Surgery. 1992; 118: 940–3. 14. Helfaer MA, McColley SA, Pyzick PL, Tunkel DE, Nichols DG, Baroody FM et al. Polysomnography after adenotonsillectomy in mild pediatric obstructive sleep apnea. Critical Care Medicine. 1996; 24: 1323–7. 15. Phillips S, Daborn AK, Hatch DJ. Preoperative fasting for paediatric anaesthesia. British Journal of Anaesthesia. 1994; 73: 529–36. 16. Splinter WM, Steward DA, Muir JG. The effect of preoperative apple juice on gastric contents, thirst and hunger in children. Canadian Journal of Anaesthesia. 1989; 36: 55–8. 17. Litman RS, Wu CL, Quinlivan JK. Gastric volume and pH in infants fed clear fluids and breast milk prior to surgery. Anesthesia and Analgesia. 1994; 79: 482–5. 18. Meakin G, Dugnall AE, Addison GM. Effects of fasting and oral premedication on the pH and volume of gastric aspirate in children. British Journal of Anaesthesia. 1987; 59: 678–82. 19. Schreiner MS, Triebwasser A, Keon TP. Ingestion of liquids compared with preoperative fasting in pediatric outpatients. Anesthesiology. 1990; 72: 593–7. 20. Gajraj NM, Pennant JH, Watcha MF. Eutectic mixture of local anaesthetics (EMLAR) cream. Anesthesia and Analgesia. 1994; 78: 574–83. 21. Lawson RA, Smart NS, Morton NS. Evaluation of an amethocaine gel preparation for percutaneous analgesia before venous cannulation in children. British Journal of Anaesthesia. 1995; 75: 282–5. 22. Kain ZN, Mayes LC, Bell C, Weisman S, Hofstadter MB, Rimar S. Premedication in the United States: a status report. Anesthesia and Analgesia. 1997; 84: 427–32. 23. Parnis SJ, Van Der Walt JH. A national survey of atropine use by Australian anaesthetists. Anaesthesia and Intensive Care. 1994; 22: 61–5. 24. Rautakorpi P, Manner T, Kanto J. A survey of current usage of anticholinergic drugs in paediatric anaesthesia in Finland. Acta Anaesthesiologica Scandinavia. 1999; 43: 1057–9. 25. Shaw CA, Kelleher AA, Gill CP, Murdoch LJ, Stables RH, Black AE. Comparison of the incidence of complications at



Chapter 41 Paediatric anaesthesia



26.



27.



28.



29. 30. 31. 32. 33. 34.



35.



36.



37. 38.



39.



40.



41.



42.



43.



induction and emergence in infants receiving oral atropine vs. no premedication. British Journal of Anaesthesia. 2000; 84: 174–8. Korpela R, Korvenoja P, Meretoja OA. Morphine-sparing effect of acetaminophen in pediatric day case surgery. Anesthesiology. 1999; 91: 442–7. Anderson BJ, Rees SG, Liley A, Stewart AW, Wardill MJ. Effect of preoperative paracetamol on gastric volumes and pH in children. Paediatric Anaesthesia. 1999; 9: 203–7. Ayre P. Endotracheal anesthesia for babies with special reference to hare-lip and cleft palate operations. Anesthesia and Analgesia. 1937; 16: 330–3. Jackson Rees G. Anaesthesia in the newborn. British Medical Journal. 1950; 2: 1419–22. Conway CM. Anaesthetic breathing systems. British Journal of Anaesthesia. 1985; 57: 649–57. Miller DM. Breathing systems reclassified. Anaesthesia and Intensive Care. 1995; 27: 281–3. Moyle JTB, Davey A, Ward C. Ward’s anaesthetic equipment, 4th edn. London: W B Saunders, 1997. Meakin G, Perkins RJ. Anaesthetic breathing systems for children. Paediatric Anaesthesia. 1996; 6: 346. Brain AIJ. The development of the laryngeal mask – a brief history of the invention, early clinical work from which the laryngeal mask evolved. European Journal of Anaesthesiology. 1991; 4: 5–17. Brimbacombe JA, Brain AIJ, Barry AM (ed.). The laryngeal mask airway instruction manual, 3rd edn. Maidenhead: Intavent Research, 1996. Eckenhoff JE. Some anatomic considerations of the infant larynx influencing endotracheal anaesthesia. Anesthesiology. 1951; 12: 401–10. Penlington GN. Endotracheal tube sizes for children. Anaesthesia. 1974; 29: 494. Khine HH, Corddry DH, Kettrick RG, Martin TM, McCloskey JJ, Rose JB et al. Comparison of cuffed and uncuffed endotracheal tubes in young children during general anesthesia. Anesthesiology. 1997; 86: 627–31. Ho AM-H, Aun CST, Karmakar MK. The margin of safety associated with the use of cuffed paediatric endotracheal tubes. Anaesthesia. 2002; 57: 173–5. Ring WH, Adair JC, Elwyn RA. A new pediatric endotracheal tube. Anesthesia and Analgesia. 1974; 54: 273–4. Murdoch L, Bingham R. Venous cannulation in infants and small children. British Journal of Hospital Medicine. 1990; 44: 405–7. Russell SH, Freeman JW. Prevention of hypothermia during orthotopic liver transplantation: comparison of three intraoperative warming methods. British Journal of Anaesthesia. 1995; 74: 415–8. Kataria B, Epstein R, Bailey A, Schmits M, Backus WW, Schoeck D et al. A comparison of sevoflurane to halothane in paediatric surgical patients: results of a multicentre international study. Paediatric Anaesthesia. 1996; 6: 283–92.



] 523



44. Recommendations for standards of monitoring during anaesthesia and recovery. London: Association of Anaesthetists of Great Britain, Ireland, 1994. 45. American Society of Anesthesiologists. Standards for basic anesthetic monitoring. In: ASA Directory of Members. Park Ridge, IL: American Society of Anesthesiologists, 1994: 735–6. 46. Denman WT, Swanson EL, Rosow D, Ezbicki K, Connors PD, Rosow CE. Pediatric evaluation of the Bispectral Index (BIS) monitor and correlation of BIS with end-tidal sevoflurane concentration in infants and children. Anesthesia and Analgesia. 2000; 90: 872–7. 47. Johnston DF, Wrigley SR, Robb PJ, Jones HE. The laryngeal mask airway in paediatric anaesthesia. Anaesthesia. 1990; 45: 924–7. 48. Watcha MF, Garner FT, White PF, Lusk R. Laryngeal mask airway vs. face mask and Guedel airway during paediatric myringotomy. Archives of Otolaryngology and Head and Neck Surgery. 1994; 120: 877–80. 49. Splinter WM, Roberts DJ, Rhine EJ, MacNeill HB, Komocar L. Nitrous oxide does not increase vomiting in children after myringotomy. Canadian Journal of Anaesthesia. 1995; 42: 263–6. 50. Bennie RE, Boehringer LA, McMahon S, Allen H, Dierdorf SF. Postoperative analgesia with preoperative oral ibuprofen or acetaminophen in children undergoing myringotomy. Paediatric Anaesthesia. 1997; 7: 399–403. 51. Tan CL, Tan S. Diclofenac or paracetamol for analgesia in paediatric myringotomy outpatients. Anaesthesia and Intensive Care. 2002; 30: 55–9. 52. Nargozian C, Ririe DG, Bennun RD, Mulliken JB. Hemifacial microsomia: anatomical prediction of difficult intubation. Paediatric Anaesthesia. 1999; 9: 393–8. 53. Uezono S, Holzman RS, Goto T, Nakata Y, Nagata S, Morita S. Prediction of difficult airway in school-aged patients with microtia. Paediatric Anaesthesia. 2001; 11: 409–13. 54. Jones SE, Dickson U, Moriarty A. Anaesthesia for insertion of bone-anchored hearing aids in children: a 7-year audit. Anaesthesia. 2001; 56: 777–80. 55. Thomsen KA, Terkildsen K, Arnfred I. Middle ear pressure variations during anesthesia. Archives of Otolaryngology. 1965; 82: 609. 56. Green JD, Schuh MJ, Maddern BR, Haymond J, Helffrich RA. Cochlear Implantation in Jervell and Lange-Nielsen syndrome. Annals of Otology Rhinology and Laryngology Supplement. 2000; 185: 27–8. 57. Prasad M, Ward RF, April MM, Bent JP, Froehlich P. Topical mitomycin as an adjunct to choanal atresia repair. Archives of Otolaryngology and Head and Neck Surgery. 2002; 128: 398–400. 58. Asaf T, Reuveni H, Yermiahu T, Leiberman A, Gurman G, Porat A et al. The need for routine pre-operative coagulation screening tests (prothrombin time PT/partial thromboplastin time PTT) for healthy children undergoing elective tonsillectomy and/or adenoidectomy. International Journal of Pediatric Otorhinolaryngology. 2001; 61: 217–22.



524 ] PART 8 PERIOPERATIVE MANAGEMENT 59. Wood P. Difficulty in extubation. Anaesthesia. 1987; 42: 220. 60. Williams PJ, Bailey PM. Comparison of the reinforced laryngeal mask airway and tracheal intubation for adenotonsillectomy. British Journal of Anaesthesia. 1993; 70: 30–3. 61. Webster AC, Morley Forster PK, Dain S, Ganapathy S, Ruby R, Au A et al. Anaesthesia for adenotonsillectomy: a comparison between endotracheal intubation and the armoured laryngeal mask airway. Canadian Journal of Anaesthesia. 1993; 40: 1171–7. 62. Hern JD, Jayaraj SM, Sidhu VS, Almeyda JS, O’Neill G, Tolley NS. The laryngeal mask airway in tonsillectomy: the surgeon’s perspective. Clinical Otolaryngology. 2000; 25: 240. 63. Kretz FJ, Reimann B, Stelzner J, Heumann H, LangeStumpf U. The laryngeal mask in pediatric adenotonsillectomy. A meta-analysis of medical studies. Anaesthetist. 2000; 49: 706–12. 64. Southall DP. (ed.). Prevention and control of pain in children. A Manual for Healthcare Professionals. Report of the working party of the Royal College of Paediatrics and Child Health. London: BMJ Publications; 1997. 65. Royal College of Anaesthetists. Guidelines for the use of non-steroidal anti-inflammatory drugs in the perioperative period. London: Royal College of Anaesthetists, 1998. 66. Harley EH, Dattalo RA. Ibuprofen for tonsillectomy pain in children: efficacy and complications. Archives of Otolaryngology and Head and Neck Surgery. 1998; 119: 492–6. 67. Romsing J, Ostergaard D, Walther-Larsen S, Valentin N. Analgesic efficacy and safety of preoperative versus postoperative ketorolac in paediatric tonsillectomy. Acta Anaesthesiologica Scandinavia. 1998; 42: 770–5. 68. Splinter WM, Rhine EJ, Roberts DW, Reid CW, MacNeill HB. Preoperative ketorolac increases bleeding after tonsillectomy in children. Canadian Journal of Anaesthesia. 1996; 43: 560–3. 69. Cardwell M, Siviter G, Smith A. Non-steroidal antiinflammatory drugs and perioperative bleeding in paediatric tonsillectomy. Cochrane Database of Systematic Reviews. 2005 (2) CD003591. 70. Atallah N, Kumar M, Hilali A, Hickey S. Post-operative pain in tonsillectomy: bipolar electrodissection technique vs. dissection ligation technique. A double-blind randomised prospective trial. Journal of Laryngology and Otology. 2000; 114: 667–70. 71. Nunez DA, Provan J, Crawford M. Postoperative tonsillectomy pain in paediatric patients: electrocautery (hot) vs. cold dissection and snare tonsillectomy – a randomised trial. Archives of Otolaryngology and Head and Neck Surgery. 2000; 126: 837–41. 72. Temple RH, Timms MS. Paediatric coblation tonsillectomy. International Journal of Pediatric Otorhinolaryngology. 2001; 61: 195–8. 73. Giannoni C, White S, Enneking FK, Morey T. Ropivacaine with or without clonidine improves pediatric tonsillectomy



74.



75.



76.



77.



78.



79.



80.



81. 82.



83.



84.



85.



86.



87.



pain. Archives of Otolaryngology and Head and Neck Surgery. 2001; 127: 1265–70. Hung T, Moore-Gillon V, Hern J, Hinton A, Patel N. Topical bupivacaine in paediatric day-case tonsillectomy: a prospective randomised controlled trial. Journal of Laryngology and Otology. 2002; 116: 33–6. Hollis LJ, Burton MJ, Millar JM. Perioperative local anaesthesia for reducing pain following tonsillectomy (Cochrane Review). In: The Cochrane Library, Issue 2 2003. Oxford: Update Software. Morton NS, Camu F, Dorman T, Knudsen KE, Kvalsvik O, Nellgard P et al. Ondansetron reduces nausea and vomiting after paediatric adenotonsillectomy. Paediatric Anaesthesia. 1997; 7: 37–45. Holt R, Rask P, Coulthard KP, Sinclair M, Roberts G, Van Der Walt J et al. Tropisetron plus dexamethasone is more effective than tropisetron alone for the prevention of postoperative nausea and vomiting in children undergoing tonsillectomy. Paediatric Anaesthesia. 2000; 10: 181–8. Department of Health. £200 million for NHS equipment to protect patients against possible variant CJD risk. London: Department of Health, press release January 4, 2001. The Royal College of Anaesthetists. Anaesthesia equipment and tonsillectomy. London: Royal College of Anaesthetists, press release April 2, 2002. Whittet HB, Hayward AW, Battersby E. Plasma lignocaine levels during paediatric endoscopy of the upper respiratory tract. Relationship with mucosal moistness. Anaesthesia. 1988; 43: 439–42. Wark H. Is there still a place for halothane in paediatric anaesthesia? Paediatric Anaesthesia. 1997; 7: 359–61. Nielson DW, Ku PL, Egger M. Topical lidocaine exaggerates laryngomalacia during flexible bronchoscopy. American Journal of Respiratory, Critical Care Medicine. 2000; 161: 147–51. Westphal K, Strouhal U, Kessler P, Schneider J. Workplace contamination from sevoflurane. Concentration measurement during bronchoscopy in children. Anaesthetist. 1997; 46: 677–82. Thaung MK, Balakrishnan A. A modified technique of tubeless anaesthesia for microlaryngoscopy and bronchoscopy in young children with stridor. Paediatric Anaesthesia. 1998; 8: 201–4. Markovitz BP, Randolph AG. Corticosteroids for the prevention and treatment of post-extubation stridor in neonates, children and adults (Cochrane Review). In: The Cochrane Library, 1, 2002. Oxford: Update Software. Wood RE. Spelunking in the pediatric airways: explorations with the flexible fibreoptic bronchoscope. Pediatric Clinics of North America. 1984; 31: 785–99. Bandla HP, Smith DE, Kiernan MP. Laryngeal mask airway facilitated fibreoptic bronchoscopy in infants. Canadian Journal of Anaesthesia. 1997; 44: 1242–7.



Chapter 41 Paediatric anaesthesia 88. Holzman RS, Mancuso TJ. Point/counterpoint. spontaneous vs. controlled ventilation for suspected airway foreign body. In: Society for Pediatric Anesthesia, Summer 2001 newsletter. 89. Grasl MC, Donner A, Schragl E, Aloy A. Tubeless laryngotracheal surgery in infants and children. Laryngoscope. 1997; 107: 277–81. 90. Depierraz B, Ravussin P, Brossard E, Monnier P. Percutaneous transtracheal jet ventilation for paediatric endoscopic laser treatment of laryngeal and subglottic lesions. Canadian Journal of Anaesthesia. 1994; 41: 1200–7. 91. Rimell FL. Pediatric laser bronchoscopy. International Anesthesiology Clinics. 1997; 35: 107–3. 92. Stern Y, McCall JE, Willging JP, Mueller KL, Cotton RT. Spontaneous respiration anesthesia for respiratory papillomatosis. Annals of Otology Rhinology and Laryngology. 2000; 109: 72–6. 93. McCall JE. Anesthetic techniques with laser endoscopy. In: Myer 3rd CM, Cotton RT, Shott SR (eds). The pediatric airway: an interdisciplinary approach. Philadelphia: JB Lippincott, 1995.



] 525



94. Baxter MR. Congenital laryngomalacia. Canadian Journal of Anaesthesia. 1994; 41: 322–9.



FURTHER READING



 Myer CM, Cotton RT, Shott SR (eds). The pediatric airway. An interdisciplinary approach. Philadelphia: JB Lippincott, 1995.



 Sumner E, Hatch DJ (eds). Paediatric anaesthesia, 2nd edn. London: Arnold, 2000.



 Gregory GA (ed.). Pediatric anesthesia, 3rd edn. New York: Churchill Livingstone, 1994.



 Baum VC, O’ Flaherty JE (eds). Anesthesia for genetic, metabolic and dysmorphic syndromes of childhood. Philadelphia: Lippincott Williams, Wilkins, 1999.  Katz J, Steward D (eds.) Anesthesia and uncommon pediatric diseases, 2nd edn. Philadelphia: W. B. Saunders, 1993.  Cote CJ, Todres DI, Ryan JF, Goudsouzian NG (eds). A practice of anesthesia for infants and children, 3rd edn. Philadelphia: W. B. Saunders, 2001.



42 Adult critical care GAVIN G LAVERY



Introduction Definition of ICU, high dependency unit and critical care Staffing, structure and role of ICU and HDU Organ system support Interface between critical care and ENT



526 526 527 527 533



INTRODUCTION The development of the intensive care unit (ICU) and intensive care medicine (ICM) has been driven by advances in medicine/surgery, the need to treat acute illness in an older, sicker population and increased public expectation. The origin of the ICU (which is the official term now used in the UK in preference to ITU – intensive therapy unit) is usually attributed to the large Danish polio epidemic in 1952, when there were insufficient ‘iron lungs’ available to provide respiratory support to the victims. Mortality was decreased significantly by the use of life support techniques previously only applied in the operating theatre. Techniques such as positive pressure ventilation required the constant attendance of medical and nursing staff and therefore these patients were brought together in a specific area of the hospital. Since then, increasingly complex medical problems have been successfully managed due to the increasing ability of the ICU to take over the role of a failing (or failed) organ system until recovery occurs.



DEFINITION OF ICU, HIGH DEPENDENCY UNIT AND CRITICAL CARE As explained above, the ICU is a geographically defined area in the hospital which utilizes specialized personnel and equipment to provide care for the critically ill.



Severe sepsis/SIRS Recent evidenced-based strategies in critical care End of life issues References



534 535 536 537



Indications for admission to ICU are:1  patient requiring (or likely to require) mechanical ventilation;  patients requiring the support of two or more systems;  patients with significant chronic impairment in one or more organ systems who require support for an acute reversible failure of another organ system. Patients with no chance of survival due to irreversible disease should not normally be admitted to ICU. The admission of those whose survival would result in an ‘unacceptable quality of life’ may also be unwise. However, since such issues revolve around subjective judgement and opinions, it is recommended that such cases be considered by very senior and experienced staff before ICU admission is granted (or denied). The high dependency unit (HDU) may be defined as an area for patients requiring more intensive observation, treatment and/or nursing care than can normally be provided on a general ward.2 Such units would not normally accept patients requiring mechanical ventilation for more than a few hours. Costs and staffing of such units are significantly less than that of ICU since the usual nurse:patient ratio is 1:2 or 1:3. It has a role as a ‘stepdown’ unit providing an appropriate level of care for ICU patients who have reached a level of function which makes further ICU management unnecessary, but who are still too ill/unstable for care in a general ward. The HDU may also provide a ‘step up’ facility for patients who are



Chapter 42 Adult critical care



too unstable to be safely treated at general ward level. An HDU should ideally be adjacent to (or in a designated area within) an ICU and have strong clinical and managerial links with the ICU. Some ICUs have a mixed caseload of both ICU- and HDU-type patients and would not be viable if delivering only ICU support. Thus the management of illness in ICU and HDU can be seen as a continuum and the term ‘critical care’ is now used to cover this.3



STAFFING, STRUCTURE AND ROLE OF ICU AND HDU It is no longer appropriate to view the ICU solely as a place where patients receive mechanical ventilation. The ICU should possess the means to support all the major organ systems including circulatory support, haemodialysis and other forms of renal support, nutritional support and the ability to treat coagulation disorders, severe infection and metabolic derangement. ICU should gather in one area equipment, technical expertise and a high concentration of appropriately trained nursing staff which cannot be duplicated elsewhere in the hospital. Although ICU is always one of the most expensive areas for patient treatment (£1200–2500 STG/day), every hospital expected to treat severe acute illness requires this capability. Concentrating resources in one area is the most cost effective way to do this. Patient management in ICU requires a team approach. Medical and nursing staff are the basis of this team but it also includes technical, physiotherapy and dietetic staff in addition to the input of radiology, bacteriology and other laboratory-based personnel. To weld this group together requires a medical director who has specific training and an ongoing interest in critical care and who spends the bulk (if not all) of his/her clinical time there.4 The other senior medical and nursing staff should also be specifically trained and be focused on critical care. Medical responsibility for the patients in ICU varies from unit to unit and between medical systems. At one extreme, treatment is dictated by the referring service at all times. This is usually inappropriate since the people making decisions concerning treatment do not see the patient continuously and may not see ‘all the pieces in the jigsaw’. Conversely, in some units all clinical decisions are taken by the ICU medical staff without any reference to the referring service. This may work and is a common model in Australia. At worst, however, it may lead to professional friction, poor staff relations and poor care. Many well-run units steer a middle path in which the decisions are made by the ICU team after discussion with the other relevant clinicians/support services. An ICU should be large enough to ensure that it is never empty since this would result in the staff being deployed elsewhere and the unit being unable to respond



] 527



quickly to clinical needs. Therefore, a minimum of six bed spaces has been recommended. The average size of ICUs in many countries has been increasing as demand increases and units of 201 beds are now commonplace. ICU requires a high nurse:patient ratio – 1:1 is viewed as appropriate within the UK.4 To achieve this, each bed requires approximately 6.5–7.0 nurses to cover shifts, weekends, holidays, sick and study leave. In addition, each unit needs 24-hour resident medical cover, i.e. a doctor who has no other clinical committments, present physically in the unit at all times. Likewise, there must be 24 hour technical back-up available rapidly. Most units also have their own ‘in house’ laboratory which provides blood gas analysis (at least) and measurement of blood/ plasma electrolytes, glucose, lactate, osmolarity and some drug levels.



ORGAN SYSTEM SUPPORT A substantial part of the benefit of ICU admission is due to attention to detail with immediate intervention and correction of deranged physiology. In ICU or HDU, patients have continuous monitoring of vital signs, frequent blood gas analyses, frequent checks on haematology and biochemistry, hourly calculations of fluid balance and trained staff immediately to hand 24 hours a day. This ability to have real time information about a patient’s condition and to make appropriate changes in management/support is the hallmark of good critical care. Clinical management often involves initiation of or adjustment to organ support. We will now briefly consider some of the commonest forms of organ support in critical care.



Respiratory support At its simplest, this may involve the use of high concentrations of well-humidified (warm) oxygen delivered via standard or continuous positive airway pressure (CPAP) mask in conjunction with aggressive regular physiotherapy. However, in ICU rather than HDU, it is likely to involve tracheal intubation and mechanical (positive pressure) ventilation. In health, the work of breathing (WOB) accounts for only 2–3 percent of total oxygen consumption. With respiratory disease, required minute volume may quadruple, lung compliance is decreased and the efficiency of chest wall movements is reduced. The result is that WOB may require more than 20 percent of total oxygen consumption and precipitate tissue hypoxia and cardiac failure. WOB cannot be easily quantified but is associated with the clinical findings of increased respiratory rate, sweating, tachycardia and the use of accessory muscles of respiration. Following assessment, a patient with



528 ] PART 8 PERIOPERATIVE MANAGEMENT respiratory difficulties may receive one or more of the following.  Increased concentration of inspired oxygen by means of a well applied face mask, with a reservoir bag, high oxygen flow (12–15 L/min) and initiation of pulse oximetry. A patient with oxygen saturations below 90 percent, particularly with appreciably increased work of breathing, may require transfer to ICU or HDU.  CPAP, applied via a tightly fitting face mask or hood and a suitable circuit with oxygen enriched air.  Tracheal intubation, with maintenance of spontaneous respiration, usually via the oral route (with or without CPAP). Indications for tracheal intubation are as follows: – protection of the airway; – long-term correction/prevention of airway obstruction; – to facilitate positive pressure ventilation; – to facilitate broncho-pulmonary toilet.  This approach is useful when airway patency is in question or when removal of bronchial secretions is difficult. However, tracheal intubation alone may be of minimal clinical benefit in many cases since it does not address the problem of increased WOB.  Mechanical ventilation (MV). The general indications for institution of mechanical ventilation are: – depressed respiratory drive; – inefficient respiratory effort/increased work of breathing; – abnormal pulmonary physiology.  More specifically the need for mechanical ventilation is often associated with one or more of the following findings: o6 kPa on 60 percent inspired oxygen; PaO – 2 – PaCO2 >8 kPa with pH o7.2; – evolving respiratory fatigue; – cardiovascular decompensation; – recent major/prolonged surgery; – multiple organ dysfunction. Much of the benefit of MV is that it can completely abolish WOB since the patient may not have to initiate any spontaneous breaths. Unlike spontaneous respiration, MV requires the generation of positive intrathoracic pressure which may produce ventilator-induced lung injury (VILI) due to the effects of pressure (barotrauma) or lung stretch (volutrauma).5 Positive intrathoracic pressure may also reduce venous return to the heart and thus precipitate hypotension – particularly in the hypovolaemic patient. Modern ICU ventilators, with many different modes of ventilation,6 strive to reduce the complications of positive pressure ventilation, but there is no scientific evidence that any specific mode of ventilation is superior.



WEANING FROM MECHANICAL VENTILATION



This is a programme of gradual reduction in respiratory support which is specifically intended to (eventually) result in the patient breathing spontaneously (without any outside support) 24 hours/day. It suggests that the original cause of acute respiratory failure has been controlled or eradicated. Weaning may be achieved over a matter of hours, days or (occasionally) weeks. It is usually achieved by gradually reducing the support supplied by the mode of ventilation being used. There is no convincing evidence for one mode of ventilation being superior to others as a method of weaning, although pressure support weaning has gained converts in recent years. ICU patients who have had short periods of ventilation (24–48 hours) may wean perfectly well by being taken off ventilation abruptly to breathe spontaneously on a CPAP circuit (T-piece weaning).7



Cardiovascular support The primary function of the circulation is to transport oxygen, nutients, carbon dioxide and other waste products to and from cells. This requires adequate amounts of haemoglobin and oxygen and appropriate cardiac output/perfusion. The latter depends on circulating volume (preload), cardiac contractility and the vascular tone in the arterial/arteriolar system (afterload). Normally, arterial blood (100 percent saturated with oxygen) contains approximately 20 mL O2/100 mL. With a normal adult cardiac output of 5000 mL/min, this gives a total oxygen delivery to the tissues of 1000 mL O2/min. Resting healthy adults use approximately 250 mL O2/min, i.e. the extraction ratio is only 25 percent and venous blood is (on average) 75 percent saturated with O2 (15 mL/100 mL blood). There is, therefore, a large safety margin regarding the balance between O2 supply and O2 demand in health. This safety margin is felt to be eroded or absent in critical illness due to increased metabolic rate and decreased efficiency at the cellular level in utilizing available O2. Shock can be defined as inadequate perfusion (or O2 delivery) for the needs of the tissue resulting in cellular hypoxia. Shoemaker et al.8 suggested that, since O2 delivery does not keep up with increased tissue demand in critically ill patients, augmentation of O2 supply by increasing cardiac output would lead to greater, beneficial tissue O2 consumption. Although his specific targets (cardiac index 44.5 L/m2/min, O2 delivery 4600 mL/m2/min and O2 consumption 4170 mL/m2/min) have declined in popularity,9 the principle of maintaining good tissue oxygenation is the foundation of cardiovascular support. For this reason, the end-points of cardiovascular resuscitation are not based on heart rate and blood pressure, but on flow related clinical parameters such as capillary refill, urinary output, serial measurement of arterial lactate or plasma bicarbonate and objective measures of cardiac index (cardiac output/ body surface area).



Chapter 42 Adult critical care



MONITORING OF CARDIAC INDEX AND PERFUSION



Perfusion can be assessed clinically using the clinical parameters above.10 In many situations, such serial assessments are all that is required. In a minority of cases, the severity of illness or the number of organ dysfunctions mean that more objective measures of cardiac function blood flow and perfusion are required. The cardiac index (CI) may be measured noninvasively using ultrasound probes, either placed in the oesophagus behind the left atrium or in the suprasternal notch. A pulsed Doppler shift technique allows the measurement of flow velocity and (since the cross-sectional area of the aorta is also measured) the flow/unit time. Difficulties may arise in maintaining a good probe position. Transthoracic impedance plethysmography assesses CI by measuring the change in voltage across the chest in response to a high frequency current (usually 100 kHz). Changes in transthoracic resistance are assumed to be due to changes in blood volume which reflects cardiac filling and emptying. Unfortunately, the correlation between results obtained using the above techniques and the gold standard (invasive) technique is variable. Invasive measurement of CI requires the placement of a catheter in the pulmonary artery. The principle use to measure CI is the reverse Fick principle. Short bursts of an electrical microcurrent is used to cause a heating effect in the plasma. Quantifying the resultant change in blood temperature allows continuous measurement of CI. It is important to balance the risks of insertion of insertion of a pulmonary artery catheter (PAC) with the benefit obtained. Changes in management prompted by information obtained from PACs has been shown to result in improved outcome in shocked patients who were not responding to standard therapy.11 In contrast, failure to use PACs (and the information they provide) appropriately has been shown to be associated with poorer outcomes12 and the risk of infection increases with time in situ.13 When it is no longer providing information which is being used to make clinical decisions, a PAC should be removed. The use of PAC has decreased in recent years and this trend is unlikely to change. The largest clinical study on the use of the PAC found no evidence of benefit or harm associated with its use.14 Increasingly, clinicians are using less invasive methods of measuring cardiac index at the bedside, e.g. (1) oesophageal Doppler and (2) lithium indicator dilution and pulse power analysis.



FLUID THERAPY – WHAT TYPE OF FLUID SHOULD WE USE?



All forms of shock, except cardiogenic shock, involve either true or relative hypovolaemia. Thus, expanding the blood volume by administration of fluids is the first step in cardiovascular resuscitation. The controversy regarding crystalloid or colloid fluid resuscitation has raged in the literature for decades.15, 16, 17 Fortunately, most clinicians



] 529



are not so extreme in their views and use a mixture of crystalloid and colloid fluids. Some points seem beyond argument:  colloids are more expensive than crystalloids;  colloids may cause anaphylactoid reactions (rarely);  the use of either solution in initial resuscitation allows blood to be used more effectively and safely later in the patient’s care;  colloids are more time-efficient at restoring intravascular volume;  plasma oncotic pressure is better maintained with colloid – this may or may not clinically advantageous. The most important point in resuscitation is to ensure that the patient has received adequate fluid. This will often mean transfusing three to four times the volume of blood lost when using crystalloid replacement or one and a half to two times the volume lost using colloid. When appropriate, filling pressures and/or cardiac index should be measured using a central venous line or PAC. Many studies have suggested that resuscitation to supranormal levels of CI and tissue oxygen delivery may improve survival.8, 18 Even if global perfusion appears satisfactory, regional hypoperfusion may exist, particularly in the splanchnic bed.19, 20 The best clinical guide to good organ perfusion is a urinary output of at least 1 mL/kg/hour without the influence of diuretics or hyperglycaemia. Oliguria is an important clinical sign of hypovolaemia and should frequently lead to a fluid challenge and not administration of a loop diuretic. The sequelae of persistent hypovolaemia are tissue hypoxia, multiple organ dysfunction and death. The sequelae of fluid overload (congestive cardiac failure, pulmonary oedema) may carry a lower associated mortality.



INOTROPES



Inotropes are agents which increase the force of myocardial contraction by a direct effect on the myocardium. Those used clinically are administered intravenously by continuous (accurately controlled) infusion and most are catecholamines or their synthetic analogues. The use of dopamine, for many years the most widely used inotrope, has declined due to doubts about its supposed dose-dependent effects,21, 22 and concerns about endocrine and immunosuppressant side effects.23 The choice of inotrope is often determined not by its inotropic effects but by its other effects – either desirable or undesirable.24 Often a major consideration is whether the drug is being given to improve arterial pressure or improve tissue perfusion. Dobutamine is an inotrope and vasodilator and is the preferred inotrope in many instances. It is a drug which improves perfusion although it has minimal and unpredictable effects on arterial pressure. It is often used in association with noradrenaline.



530 ] PART 8 PERIOPERATIVE MANAGEMENT Adrenaline may improve perfusion and/or pressure and is sometimes used as a single agent in the treatment of septic shock. However, its unpredictable mix of alpha and beta adrenergic receptor stimulation and its potential to produce cardiac arrhythmias is a drawback. Noradrenaline also stimulates both alpha and beta adrenergic receptors but is particularly potent at alpha-receptors, thus it is an inotrope with vasoconstrictor effects. It will raise blood pressure and is often used to treat the hypotension associated with the generalized vasodilatation of severe sepsis/septic shock. It is often used in association with dobutamine in sepsis as dobutamine augments cardiac contractility in the face of sepsis-induced depression and the increased afterload due to noradrenaline. The importance of the splanchnic circulation has raised interest in dopexamine. This dopamine analogue predominantly stimulates dopamine and beta-adrenergic receptors. It is a (weak) inotrope, vasodilator and (possibly) increases splanchnic blood flow. Studies have claimed that dopexamine beneficially improves oxygen flux in the postoperative period25 and that it promotes better colonic mucosal perfusion,26 though this has been questioned by other work.27 A multicentre trial failed to show any outcome benefit from the use of dopexamine.28 The manipulation of the cardiovascular system should follow a logical sequence. In general this sequence is 1. Is the cardiac output adequate? An inadequate cardiac output is suggested by hypotension, oliguria, cold peripheries, an elevated plasma lactate. 2. Is the low cardiac output due to hypovolaernia or poor myocardial function? Central venous pressure (CVP) measurement may be useful. A very low value (0–4 mmHg) suggests hypovolaemia and the CVP trend will guide fluid replacement. A high CVP (415 mmHg) suggests a cardiogenic (pump) problem which may improve with dobutamine (5–20 mg/kg/min). 3. The use of a PAC will give much greater information and should be considered whenever intelligent manipulations based on CVP measurements has failed to produce improvement. 4. Patients with sepsis have rapidly changing haemodynamic profiles. Invasive monitoring with appropriate pharmacological support should be initiated early in patients with septicaemia/septic shock.



Renal support There are many factors associated with the development of acute renal failure (ARF) in the critically ill patient. These often coexist and are summarized in Table 42.1. Common pharmacological causes of ARF are high doses



of some radiological contrast media, toxic plasma levels of some antibiotics, e.g. the aminoglycosides, the use of nonsteroidal antiinflammatory agents, cyclosporin A and antimitotics, e.g. cis-platinum and methotrexate.29 As the understanding of the pathophysiology has advanced, strategies have been developed which might reduce the incidence of ARF. Adequate fluid replacement guided by central venous or pulmonary artery pressure, adequate blood transfusion, attention to oxygen transport and early nutritional support may all be important. The use of dopexamine (0.5–1.0 mg/kg/min) may increase renal blood flow and promote urine flow. Although such a strategy is often used in at-risk patients, there is no objective evidence that it prevents ARF. Previously, dopamine (2–3 mg/kg/min) was thought to have the same effect but the evidence for its potential benefit is now viewed as less convincing. Many clinicians now use n-acetyl cysteine as a pretreatment which may mitigate the renotoxic effects of radiological contrast. Indications for renal replacement therapy (dialysis) can be summarized as the presence of hyperkalaemia, fluid overload, metabolic acidosis and a high plasma level of urea and creatinine. The absolute values of the latter which trigger the decision to dialyse varies between centres. Until recent years, haemodialysis (HD) was usually performed for a three or four hour session on a daily or alternate day basis. Originally performed through surgical or plastic arteriovenous connections, short-term HD is now usually achieved using a double lumen catheter inserted in a large vein, e.g. subclavian or femoral. ICUs are increasingly performing continuous HD (often termed CRRT – continous renal replacement therapy) over a 24 hour period.30, 31 This produces less cardiovascular instability as fluid and electrolyte shifts occur more gradually. The disadvantages are increased costs, need for continuous heparinization and the need for ICU staff training. No satisfactory comparison of these techniques has been performed, although there is evidence that CRRT produces normalization of plasma electrolytes, etc., more quickly than HD.32 Although there is some suggestion of better outcome with CRRT,31 this has been refuted by other studies.33



Nutritional support Simple starvation leads to the exhaustion of glycogen stores within 48 hours and an increase in glucose production from amino acids derived from breakdown of body proteins (gluconeogenesis). The glucose provided is used to fuel the brain and red blood cells which cannot normally utilize other substrate. This produces a rate of body protein breakdown which would be lifethreatening if unchecked. Fortunately, in persistent uncomplicated starvation, the body switches to fat as the predominant energy source and the brain adapts to



Chapter 42 Adult critical care Table 42.1



] 531



Causes of acute renal dysfunction.



Renal dysfunction Pre-renal dysfunction Decrease in blood volume



Decrease in effective blood volume (Renal hypoperfusion)



Cause



Haemorrhage Dehydration Burns GI losses (vomiting/diarrhoea) Renal losses (diuretics, diabetes insipidus, adrenal insufficiency) Cardiogenic shock Liver disease with ascites Hypoalbuminaemia Pancreatitis Peritonitis Hypercalcaemia Catecholamines Cyclosporine Amphotericin B Cyclooxygenase inhibition Angiotensin-converting enzyme inhibitors Renal artery stenosis/occlusion Hyperviscosity syndromes



Intrinsic renal dysfunction Glomerulonephritis Vasculitis Acute tubular necrosis (ischaemia, drugs/toxins, pigments) Interstitial nephritis (antibiotics, NSAIDs, ACE inhibitors) Pyelonephritis Infiltration (sarcoidosis, lymphoma, leukaemia) Radiation nephritis Post renal dysfunction Papillary necrosis Ureteric obstruction (stone, tumour, retroperitoneal fibrosis) Bladder obstruction (tumour, prostatic enlargement, neurogenic dydfunction) Urethral stricture Posterior urethral valves Phimosis



use ketones as a metabolic fuel, thus allowing gluconeogenesis to slow.34 Unfortunately, in critical illness (particularly when complicated by sepsis), gluconeogenesis is not suppressed and this has to be fuelled by rapid protein breakdown – the so-called hypermetabolic or hypercatabolic state in which the body is subject to a process which has been termed ‘autocannibalism’. Such protein breakdown has significant deleterious effects on organ function and outcome. By providing exogamous glucose, amino acids and other nutrients, we hope to modify this process. Enteral nutrition (EN) is nutritionally, immunologically and metabolically superior to feeding using the total parenteral nutrition (TPN) route. The present preference for EN could be justified on the basis that it is less expensive, does not require central venous cannulation with its attendant complications and is less likely to



produce fluid overload, hyperglycaemia or hypophosphataemia. To fully appreciate the superiority of EN, we should consider the effects of critical illness and lack of luminal nutrients on the gastrointestinal tract. The mucosal cells of the gastrointestinal (GI) tract have one of the highest turnover rates of any body tissue. An intact gut mucosa requires a balance between cell renewal and exfoliation. Its prime function is to allow the controlled absorption of nutrients while providing a barrier to the passage of toxins and bacteria from the lumen of the gut into the portal circulation. Barrier function is due to a series of elements: the intact mucosal layer, tight junctions between cells, lymphocytes, macrophages and neutrophils in the submucosa and Peyer’s patches and gut-generated IgA. The absence of nutrients within the gut lumen appear to compromise both gut mucosal health and barrier function.



532 ] PART 8 PERIOPERATIVE MANAGEMENT Moore and colleagues35 published a meta-analysis of eight studies comparing EN and TPN in surgical patients. Infective complications were significantly less in EN groups, with lower incidences of pneumonia, intraabdominal abscesses and (not unexpectedly) catheter sepsis. It should be stressed that, even when catheterrelated sepsis is excluded, TPN is associated with a significantly higher incidence of infective complications than EN. In animals, EN is associated with less intestinal mucosal atrophy, liver dysfunction, bacterial translocation, greater immunocompetence and better survival after a variety of systemic insults.



Table 42.2 Typical daily fluid, electrolyte and nutritional requirements in critical illness. Variable Water Sodium Potassium Calories Nitrogen Calorie:nitrogen ratio



Requirement 35–40 mL/kg 1–1.5 mmol/kg 0.6–0.8 mmol/kg 20–30 kcal/kg 0.2–0.3 g/kg (increased in severe illness) 150:1 (moderate illness) 100:1 (sepsis/MODS/hypercatabolism)



IMMUNONUTRITION TIMING AND METHOD OF DELIVERING EN



Which clinical findings tell us when we might introduce enteral feeding? The presence of bowel sounds, though comforting, is irrelevant. Many patients with no bowel sounds may tolerate and absorb nasogastric feeds. In contrast, gastric stasis, leading to high aspirates and vomiting, may occur even when bowel sounds are present. Often, a low nasogastric aspirate (o250–300 mL in 24 hours) is the most appropriate prompt to initiate feeding. The common problem of gastric stasis may require the increasing use of feeding (surgical) jejunostomy tubes and the selective use of nasoduodenal or nasojejunal tubes. Persuading feeding tubes to negotiate the pyloric canal and remain in position can be a problem. Endoscopic guidance or fluoroscopy are useful – the latter having a 95 percent success rate – but may be logistically difficult in critically ill patients. With such strategies it is possible to feed patients via the enteral route within 24 hours of major surgery. Graham and colleagues,36 in a controlled trial, showed that aggressive early jejunal feeding was associated with fewer infective complications and a shorter ICU and hospital stay than standard nasogastric feeding in head injured patients. EN is usually administered by continuous infusion into the GI tract. This reduces GI side effects but may lead to increased Gram-negative colonization of the upper GI tract. Stopping enteral infusion for four hours to permit a (bactericidal) decrease in intragastric pH has been suggested as a strategy to reduce the incidence of Gramnegative pulmonary infection.37 However, an extensive systematic review, while confirming the superiority of early EN over TPN, could not make recommendations regarding composition of feed, bolus versus infusion feeding and many other issues.38 When patients have suffered a prolonged period of poor nutrition, initiation of feeding may produce severe hypophosphataemia and other electrolytic disturbances which may lead to acute respiratory failure, cardiac dysfunction and neurological problems – the so-called refeeding syndrome.39 The daily maintenance requirements vary between patients and at different stages of their illness. Typical daily values for ICU patients are shown in Table 42.2.



Immunonutrition seeks to use nutrients to modify the immunological response to illness and injury. Potential immunonutrients given by the enteral route include glutamine, arginine, omega-3 fatty acids and nucleotides. Glutamine and omega-3 fatty acids have also been administered parenterally. Their potential role is the preservation of cellular immune function and to beneficially alter the production of inflammatory mediators. Glutamine is required both as a primary fuel and in the synthesis of nucleotide precursors by the immune system. It appears to be conditionally essential in the critically ill. Despite a large reserve, excessive demand for glutamine in such patients may cause glutamine deficiency, resulting in decreased intestinal integrity and cellular immune function.40 Administration of parenteral glutamine has been associated with improved outcome in intensive care patients and patients having major elective abdominal surgery.41, 42 In a small study, the use of enteral glutamine in burns patients was found to reduce the incidence of positive blood cultures and mortality but, puzzlingly, had no effect on length of care or polymorph phagocytosis.43 A welldesigned large study of enteral glutamine in ICU patients failed to show any benefit in terms of incidence of severe sepsis, death or a number of other secondary outcomes.44 The Canadian Critical Care Group systematic review recommended glutamine in critically ill burns and trauma patients, but did not recommend the addition of arginine to EN.38 An enteral feed rich in omega-3 fatty acids has been shown to decrease the requirement for supplemental oxygen, the period of ventilatory support, ICU length of stay and new organ failure in patients with moderate or severe acute respiratory distress syndrome.45 The same group has more recently studied the mechanism of the improvement in pulmonary function and found a decrease in indices of alveolar membrane protein permeability and reduced levels of IL-8, leukotiene B4 and neutrophils in broncho-alveolar lavage fluid in patients fed on an enteral feed containing omega-3 fatty acids and enhanced antioxidants.46 Parenteral omega-3 fatty acids show immune benefits and antiinflammatory



Chapter 42 Adult critical care



effects but no effect on infection rate or mortality in surgical patients.47 Galban et al.48 showed a reduction in mortality in critically ill patients with sepsis receiving immunonutrition, but benefit was confined to those with relatively low illness severity. Another study showed significantly increased mortality in critically ill patients receiving immunonutrition, an effect that was more pronounced in patients with sepsis.49 The most recent meta-analysis found that the use of immunonutrition reduced infections and hospital stay, particularly in surgical (rather than critically ill) patients.50 Despite these apparent benefits of immunonutrition, none (of three) meta-analyses have identified a significant effect on mortality, either globally or across any patient grouping.50, 51, 52



INTERFACE BETWEEN CRITICAL CARE AND ENT Postoperative care after surgery Like other surgical specialties, ENT may refer patients for critical care management following (and occasionally prior to) elective or emergency surgery. This may be due to:  the nature of the acute illness (e.g. major haemorrhage with shock after tonsillectomy or acute mastoiditis/sinusitis with signs of intracranial infection);  the nature of the surgery (e.g. major head and neck surgery after which airway swelling and respiratory compromise might be expected in the postoperative period;  patient factors (e.g. severe cardiac or respiratory disease in a patient requiring prolonged and/or extensive surgery). Often, several of these factors coexist. In such cases the critical care phase may last just one to two days, or may be much longer – particularly in patients with severe cardiorespiratory disease in whom weaning from mechanical ventilation may be difficult. The development of severe infection or systemic inflammatory response syndrome (SIRS) is relatively frequent during critical illness and may lead to the development of multiple organ dysfunction syndrome (MODS) and, potentially, an increased length of stay (LOS) in ICU, morbidity and mortality. Good communication between the ICU staff, ENT staff and the relatives is important, particularly if this phase of care is prolonged and/or complications occur.



] 533



may be required after major vascular or airway injury. Treatment of nasal injury or the drainage of blood-filled sinuses may be required later in the management of such patients. The assessment of actual or possible injury to the tympanic membrane or other parts of the auditory mechanism may also be required.



Management of the difficult airway There are two common clinical scenarios. First, acute management of partial or complete airway obstruction. This is often in cases were it is suspected that tracheal intubation will be difficult or impossible. Airway obstruction may occur at any level:  supralaryngeal – tumour, infection, anaphylaxis;  laryngeal – infections, tumours, surgery, anaphalaxis;  infralaryngeal – tumour, infection, tracheomalacia, subglottic stenosis. The techniques required will vary. In many cases, the role of the ENT surgeon will merely be to observe, providing back-up to the anaesthetic staff in the (rare) event that tracheal intubation is not achieved. It is important, however, that ENT expertise is immediately available, as attempts to intubate the trachea may lead to increase swelling/inflammation and worsen airway obstruction. When ENT intervention is required, this may be to facilitate tracheal intubation via the oral route (e.g. use of an anterior commissure laryngoscope). In other cases, tracheostomy or cricothyroidotomy may be required. In cases where tumour or infection/abscess is the basis for the airway obstruction, resection/debulking of tumour or evacuation of an abscess cavity may be carried out after the airway problem has been resolved. The second common clinical scenario is the late management of patients with continuing need for tracheal intubation. This is usually managed by tracheostomy. It should be remembered that since many tracheostomies are now performed as a bedside, percutaneous procedure, referral to ENT suggests the procedure is expected to be difficult, often due to anatomical factors or coagulation disorders. The indications for late tracheostomy is usually due to the need for prolonged tracheal intubation for the following reasons:  failure (or predicted failure) to wean from mechanical intubation;  lack of airway protection (bulbar palsy, cord dysfunction, coma);  need for tracheobronchial toilet (decreased level of consciousness).



Post-head/neck trauma Assessment of cord function The management of injuries to the head, face or neck may require ENT surgical intervention, either in the acute phase or after many days in ICU/HDU. Early intervention



In patients with problems concerning airway patency/ stridor, inability to cough effectively or suspected



534 ] PART 8 PERIOPERATIVE MANAGEMENT (repeated) pulmonary aspiration, ENT may be asked to rule out vocal cord dysfunction. This may be required after surgery on the thyroid, parathyroid or nearby structures when recurrent laryngeal nerve injury may occur. Some neck, mediastinal and apical lung tumours may also produce dysfunction of the recurrent laryngeal nerve. Assessment usually involves visualizing the cords (without neuromuscular blockade) either using a fibreoptic or rigid bronchoscope. The former may be carried out in ICU as a bedside procedure.



Investigation of source of sepsis Attempts to identify an occult source of severe or recurrent sepsis is a common process in ICU. After exclusion of the usual sites by culturing blood/sputum/urine/cerebrospinal fluid (CSF), imaging the chest and abdomen and removing invasive lines, the question of sinus infection must be ruled out. Since most ICU patients have feeding and/or other tubes passing through the nasal passages, they are predisposed to retention and superinfection of sinus secretions. This may be particularly so after head and facial trauma when there may have been bleeding into the sinus cavities. In such circumstances, ENT services may be asked to perform antral lavage to provide material for culture and remove a nidus of infection.



SEVERE SEPSIS/SIRS The triggering of the inflammatory response is designed to be (and usually is) protective. The usual trigger is invasion by microorganisms (infection) and is termed sepsis. However, tissue damage from trauma, burns or major surgery may also act as the trigger to this inflammatory response, in which case it is termed systemic inflammatory response syndrome (SIRS). Both SIRS and severe sepsis may lead to septic shock whose clinical manifestations include hypotension, tachycardia, pyrexia and warm flushed peripheries. The early cardiovascular changes can be summarized by stating that there is peripheral vasodilatation with an elevated cardiac index, low arterial pressure and low filling pressures. Younger patients with more cardiac reserve may well not become hypotensive due to their ability to substantially increase cardiac index. The criteria for SIRS, sepsis and severe sepsis are well established, as are the definitions of severe sepsis and sepsis with shock.53 Management should include restoration of circulating volume, correction of tissue hypoxia and prevention of avoidable insults to organ function and has been well described elsewhere.54, 55 Good infection control policies and practices have been shown to reduce nosocomial infection rates.56 Early treatment of sepsis with appropriate antimicrobial therapy has been shown to improve survival in critical illness.57, 58



Resuscitation The aggressive infusion of colloid and/or crystalloid under the guidance of central venous or pulmonary artery occlusion pressures will stabilize the condition in many patients. Should it prove impossible to attain an adequate blood pressure, then the problem is usually one of profound vasodilatation. In such cases a controlled infusion of a vasoconstrictor, such as noradrenaline or phenylephrine, may restore pressure and return coronary perfusion to normal. Vasoconstrictors are often used in conjunction with dobutamine to sustain cardiac contractility and general perfusion. When using noradrenaline in this way, it may be desirable to use a means of quantifying cardiac output and global perfusion (e.g. pulmonary artery catheter) to ensure that excessive vasoconstriction does not take place. Even this does not give any guide to the flow through regional capillary networks. The guiding principles of haemodynamic manipulations in SIRS/solidus sepsis are:  restore normovolaemia;  commence dobutamine;  use the smallest dose of noradrenaline which is compatible with acceptable blood pressure;  augment cardiac contractility if required (e.g. dobutamine);  check haemodynamic status by serial measures of arterial lactate and/or measurement of cardiac index. Even with the above measures, a minority of patients remain grossly vasodilated, hypotensive and may die of refractory cardiovascular failure. Others improve haemodynamically with the above measures only to die later following the development of acute lung injury, renal dysfunction and other manifestations of SIRS/sepsisinduced multiple organ dysfunction syndrome (MODS).



Antimicrobials Antibiotics should be administered as soon as possible. At this point the nature of the microorganisms concerned is unknown. Therefore, it is common currently standard practice to use double or even triple antibiotic therapy to cover all possible causative organisms. Fungi should not be forgotten as possible causative agents, especially in patients who have been on long-term broad spectrum antibacterial agents. The next step is to identify and remove potential sources of sepsis. Wounds should be examined and probed if necessary. All intravascular lines and the urinary catheter should be changed.



Investigating the source of infection The source of sepsis is often obvious. In other cases, however, full examination fails to reveal a problem. This



Chapter 42 Adult critical care



group of patients needs persistent investigation. Initial cultures should be taken from blood, sputum, urine, CSF and any surgical drains, if possible before the administration of antibiotics. A chest x-ray may reveal any pulmonary infection and/or the presence of fluid in the pleural cavity. If the latter is present, then a diagnostic tap and/or drainage via a chest drain will be necessary. If cultures, x-ray and clinical examination are unrewarding, the location of an occult source of sepsis may lie in the abdominal cavity (including the pelvis) or the intracranial space or the paranasal sinuses. These possibilities can be narrowed further by ENT endoscopy or double antral punctures and lumbar puncture. The abdominal cavity can be assessed noninvasively in several ways using ultrasound scan, CT scan or labelled white cell scan. After all three of these investigations have been performed and repeated, there should be an 80 percent ‘pick up’ rate for septic foci within the abdomen/pelvis.59 Once located, a septic focus may be obliterated either by radiologically guided percutaneous drainage or by surgical laparotomy. There is still an indication for the investigative laparotomy which will occasionally find a lesion previously missed by other forms of investigation. Laparotomy may need to be repeated after an interval before all sources of sepsis can be excluded.



RECENT EVIDENCED-BASED STRATEGIES IN CRITICAL CARE Rigid glycaemic control Hyperglycaemia is commonly associated with critical illness and has been considered an adaptive response to ensure an energy source for brain cells and erythrocytes. However, high titres of insulin-like growth factor binding protein 1 are associated with decreased insulin production and increased mortality.60 The use of insulin infusion in over 1500 (mainly surgical) ICU patients to maintain blood glucose in the (low) normal range (4.4–6.1 mmol/L) showed a significant benefit over conventional treatment (mortality rates of 4.6 and 8.0 percent, respectively).61 This benefit was most pronounced with prolonged ICU stay (Z5 days). Although this effect might have been due to other effects of insulin, e.g. on plasma free fatty acids, further work has suggested control of blood glucose and not the dose of insulin used, was the factor conferring survival advantage.62



Target haemoglobin levels/blood transfusion Anaemia is common in critical illness for many reasons, including a blunted erythropoietin response to acute anaemia.63, 64 The first issue is whether or not this anaemia is harmful? Studies would indicate that moderate anaemia



] 535



is potentially beneficial65, 66 and that red cell transfusion to maintain a higher than necessary level of haemoglobin may be detrimental.67, 68 What is unclear is whether the harmful factor is a relatively higher level of haemoglobin or the transfusion of donated stored allogeneic blood. While exogenous erythropoietin has been shown to reduce transfusion requirements and result in increased haemoglobin concentration in critically ill patients,69 we should not assume that this will necessarily lead to improved outcome. We need to develop appropriate triggers for transfusion which maximize the potential advantage of anaemia and also reduce the need for allogeneic red cell transfusion with its well-known risks. Other strategies, such as use of blood conservation policies or devices, should be investigated,70 although there is little evidence to date that the latter are effective.71, 72



Early goal-directed therapy The underlying principle of early goal-directed therapy (EDGT) is that the early resuscitation of patients with severe haemodynamic compromise may be incomplete if blood pressure is used as the end-point (see under Cardiovascular support above). This may leave many patients with ‘normal vital signs’ but occult shock with ongoing tissue hypoperfusion/hypoxia and increased risk of subsequent organ failure. The EGDT approach uses fluid/blood transfusion and dobutamine and perfusion related parameters as the end-points of resuscitation. Specifically, central venous oxygen saturation (ScvO2) is monitored via a central venous line. EGDT commenced in the emergency room has been shown to reduce mortality in patients with severe sepsis.73



Low tidal volume ventilation Excessive positive pressure and/or lung stretch has been linked to VILI as discussed under Respiratory support. This may be due to production of inflammatory mediators promoted by use of higher tidal volume ventilation.74, 75 Clinical studies have suggested that limiting the tidal volume and/or the ventilatory pressure improves outcome in patients with acute lung injury.76, 77 Hence, there is an increasing tendency to use low tidal volume (6–8 mL/kg) rather than the more traditional 10–12 mL/kg for MV in critical illness. However, methodological differences78 mean it is difficult to compare studies and some work suggests increased morbidity and poorer outcome with a low tidal volume/low inflation pressure strategy.79



Activated protein C Sepsis, decreased protein C levels and high mortality have been linked,80, 81 while infusion of activated protein C has



536 ] PART 8 PERIOPERATIVE MANAGEMENT been shown to improve survival in animals infused with live coliforms.82 Use of recombinant human activated protein C (APC) in severe sepsis has been shown to achieve an absolute reduction in mortality of 6.1 percent and has an even greater benefit in those patients with associated multiple organ failure (absolute mortality reduction 7.4 percent).83 Although the anticoagulant, profibrinolytic and antiinflammatory (reduction in IL-6) actions of APC are known, its beneficial effects in severe sepsis are still not fully explained.



Steroids Critical illness can unmask and even cause a relative failure in adrenal function. The adrenal gland is vital in the adaptive response to critical illness. It produces glucocortoids (cortisol) and mineralocorticoids from the cortex of the gland and catecholamines from the medulla. It is common in ICU to use catecholamines to support the circulation. However, glucocorticoids are often avoided because of their side effects and so unrecognized adrenal dysfunction may lead to life-threatening cortisol deficiency. Predisposing factors include major surgery, trauma, coagulopathy, but particularly severe sepsis.84 The clinical manifestations of the condition are hypotension unresponsive to treatment and other relatively nonspecific findings such as electrolyte abnormalities (hyperkalaemia, hyponatraemia) and generalized weakness which are common in the ICU population. Although investigation of the use of steroids in severe sepsis has produced well over 100 papers, there have been few methodologically rigorous studies.85 Relative adrenal insufficiency may exist in Z50 percent of patients with septic shock which is refractory to inotropic support.86 Patients with sepsis exhibit reduced plasma cortisol binding and decreased glucocorticoid receptor binding.87, 88 Two studies seemed to confirm that moderate doses of steroid in patients with septic shock improved responsiveness to catecholamines.89, 90 This prompted a multicentre French trial using a seven day course of moderate dose (200–300 mg/day) hydrocortisone and fludrocortisone in patients whose adrenal function was assessed using an ACTH stimulation test.91 Relative adrenal insufficiency was found in 229 of 300 subjects. In this subgroup, 115 received steroids and 114 placebo. The mortality rates were 63 and 53 percent, respectively (po0.023). In ACTH responders, steroids had no survival advantage, nor excess attributable mortality. A subsequent cross-over study showed hydrocortisone restored haemodynamic stability in critically ill patients with septic shock and promoted an antiinflammatory, but not immunosuppressed, state.92 Diagnosis of relative adrenal insufficiency requires a high index of suspicion and blunted response to an ACTH or Synacthen stimulation test.84 Treatment is by steroid replacement during the stressed period – 100 mg of hydrocortisone six hourly is a widely used regime.



END OF LIFE ISSUES Communication with relatives Critical care patients are often unaware of their ICU stay until they are almost ready for discharge. Their relatives often bear a heavy emotional burden, their degree of confusion and helplessness made greater by the complexity and the seriousness of the clinical situation. This may be compounded by the numbers of doctors, nurses and clinical teams involved and the length of time the patient may remain in a life-threatening state. It should be remembered that even a short ‘routine’ admission to ICU is one of the most stressful events ever experienced by most families. Relatives’ stress can be reduced by reducing the number of staff through whom they receive information. Interviews with relatives should occur in a quiet, private and comfortable counselling room adjacent to the ICU. An individual’s interpersonal skills and personality will have a significant influence on their approach and success in counselling relatives. The experience of others, however, can give valuable lessons. These include:  always sit down when talking to relatives (this may mean sitting on the floor when dealing with a large family) otherwise you may appear unfriendly or in a rush;  always introduce yourself and any other accompanying member of staff. If possible, bring the patient’s bedside nurse;  know before you enter the room who will be there and their relationship to the patient;  check the patient’s name (including first name), age and address. Remind yourself of the date of hospital and ICU admission and other details of the case. It is advisable to have the clinical notes with you;  start your delivery with a question such as ‘Before I update you, can I ask what you understand the position to be?’ This achieves several goals. You find out what is already known. You will therefore be able to judge whether the family are inappropriately optimistic or pessimistic. They will focus on the issues which they feel are important or which most worry them. This will allow you time to tailor your information to bring them up to date and (hopefully) correct any misconceptions. You are also able to judge the degree of understanding within the family and the level at which to pitch your discussion. In cases involving multiple clinical teams there is an increased danger of families receiving ‘mixed messages’ and being further distressed and confused. It may be helpful, at an early interview, to point out that each clinician involved will tend to focus on their own particular area/role and that the ICU staff are in the best position to give an overall view of the patient’s condition and short-term prognosis.



Chapter 42 Adult critical care



Limitation of treatment Limitation in this context means to either withhold or withdraw one or more treatments from a patient – presumably because in the view of those clinicians making such a decision, the treatment(s) are not in the patient’s best interest. In ICU, this usually means that the patient has no prospect of survival (irrespective of the treatment offered) or that survival would be unacceptable (in terms of quality of life). The latter needs to be considered very carefully since quality of life is a subjective judgement. Although withholding or withdrawing treatment do not differ ethically, some feel that withdrawal is more difficult as the cessation of treatment appears to be more closely associated with death than withholding, when the progression of disease appears to cause the death of an individual. The opposing view is that having started a treatment and then withdrawn it, we are in a better position to know that it was a futile intervention and one which was not going to alter outcome. Potential difficulties arise because of the clash of ethical and legal principles, medical policies and the emotional responses of both relatives and ICU staff in a situation where the views of the patient are often unknown and unobtainable. Such problems are always more difficult in the ICU environment where many kinds of organ support are made possible by medical technology. Mostly, such interventions are totally appropriate. Occasionally they are obviously inappropriate. However, it is when there is disagreement on appropriateness of treatment that problems may arise and a structured approach to decision making is required.93 Although limitation of treatment is common in ICU,94 doctors, nurses and the public differ in their view of who should have input into such decisions.95 Doctors and nurses in ICU may also disagree on how to conduct endof-life decision-making and who should be involved.96 If limitation of treatment is to be acceptable in ICU, the conduct surrounding such orders (often termed do not resuscitate (DNR) orders) should adhere to the following.  DNR orders should be made in concert by several of the most senior ICU clinicians. The process should include consultation with other members of the ICU team. The views of the next-of-kin should be sought but it should be made clear that they are not being asked to make the decision.  Such decisions should be indicated by a written, unambiguous entry in the patients clinical notes.  The relatives and other clinical services involved should be informed of the decision before it is initiated.  The decision should be reviewed and (if appropriate) renewed every day.



] 537



Effective communication and a caring and open approach will prevent and/or resolve many of the problems which may arise in an ‘end-of-life’ situation.



REFERENCES 1. Department of Health. Report of the Working Group on Guidelines on admission to and discharge from the intensive care and high dependency units. London: Department of Health, 1996. 2. Association of Anaesthetists of Great Britain and Ireland. The high dependency unit. Acute care for the future. London: AAGBI, 1991. 3. Department of Health. Comprehensive critical care. A review of adult critical care services. London: Department of Health, 2000. 4. Intensive Care Society. Standards for intensive care units. London: Intensive Care Society, 1997. 5. Gattinoni L, Caironi P, Carlesso E. How to ventilate patients with acute lung injury and acute respiratory distress syndrome. Current Opinion in Critical Care. 2005; 11: 69–76. 6. Tobin MJ. Current concepts, Mechanical ventilation. New England Journal of Medicine. 1994; 330: 1056–61. 7. Esteban A, Alia I, Gordo F, Fernandez R, Solsona JF, Vallverdu I et al. Extubation outcome after spontaneous breathing trials with T-tube or pressure support ventilation. The Spanish Lung Failure Collaborative Group. American Journal of Respiratory and Critical Care Medicine. 1997; 156: 459–65. 8. Shoemaker WC, Appel PL, Kram HB, Waxman K, Lee T-S. Prospective trial of supranormal values of survivors as therapeutic goals in high-risk surgical patients. Chest. 1988; 94: 1176–86. 9. Russell JA, Phang PT. The oxygen delivery/consumption controversy: approaches to management of the critically ill. American Journal of Respiratory and Critical Care Medicine. 1994; 149: 533–7. 10. Shephard JN, Brecker SJ, Evans TW. Bedside assessment of myocardial performance in the critically ill. Intensive Care Medicine. 1994; 20: 513–21. 11. Mimoz O, Rauss A, Rekik N, Brun-Buisson C, Lemaire F, Brochard L. Pulmonary artery catheterization in critically ill patients: a prospective analysis of outcome changes associated with catheter-prompted changes in therapy. Critical Care Medicine. 1994; 22: 573–9. 12. Connors AF, Speroff T, Dawson NV, Thomas C, Harrell FE, Wagner D et al. The effectiveness of right heart catheterization in the initial care of critically ill patients. Journal of the American Medical Association. 1996; 276: 889–97. 13. Mermel LA, and Maki DG. Infectious complications of Swan-Ganz pulmonary artery catheters. American Journal



538 ] PART 8 PERIOPERATIVE MANAGEMENT



14.



15.



16.



17.



18.



19.



20.



21.



22. 23.



24.



25.



26.



27.



of Respiratory and Critical Care Medicine. 1994; 149: 1020–36. Harvey S, Harison DA, Singer M, Ashcroft J, Jones CM, Elbourne D et al. Assessment of the clinical effectiveness of pulmonary artery catheters in management of patients in intensive care (PAC-man): a randomised controlled trial. Lancet. 2005; 366: 472–7. Ross AD, Angaran DM. Colloids vs crystalloids – a continuing controversy. Drug Intelligence and Clinical Pharmacy. 1984; 18: 202–12. Shoemaker WC, Schlucter M, Hopkins JA, Appel PE, Schwartz S, Chang PC. Comparison of the relative effectiveness of colloids and crystalloids in emergency resuscitation. American Journal of Surgery. 1981; 142: 7381. Hillman K. Colloid versus crystalloid fluid therapy in the critically ill. Intensive and Critical Care Digest. 1986; 5: 7–9. Edwards JD, Brown GCS, Nightingale P, Slater RM, Faragher EB. Use of survivors’ cardiorespiratory values as therapeutic goals in septic shock. Critical Care Medicine. 1989; 17: 1098–103. Gottlieb ME, Sarfeh IJ, Stratton H, Goldman ML, Newell JC, Shah DM. Hepatic perfusion and splanchnic oxygen consumption in patients postinjury. Journal of Trauma. 1983; 2: 836–43. Edouard AR, Degremont A-C, Durantaeu J, Pussard E, Berdeaux A, Samii K. Heterogeneous regional vascular responses to simulated transient hypovolaemia in man. Intensive Care Medicine. 1994; 20: 414–20. Baldwin L, Henderson A, Hickman P. Effect of postoperative low-dose dopamine on renal function after elective major vascular surgery. Annals of Internal Medicine. 1994; 120: 744–7. Thompson BT, Cockrill RA. Renal dose dopamine: a siren’s song? Lancet. 1994; 344: 7–8. Van den Berghe G, de Zegher F, Lauwers P, Veldhuis JD. Growth hormone secretion in critical illness: effect of dopamine. Journal of Clinical Endocrinology and Metabolism. 1994; 79: 41–6. Lavery GG, McMurray TJ. Inotropic drugs. In: McCaughey W, Clarke RSJ, Fee JPH, Wallace WFM (eds). Anaesthetic physiology and pharmacology. Edinburgh: ChurchillLivingstone, 1997: Chapter 34. Boyd O, Grounds RM, Bennett ED. The use of dopexamine hydrochloride to increase oxygen delivery postoperatively. Anesthesia and Analgesia. 1993; 76: 372–6. Baguneid MS, Welch M, Bukhari M, Fulford PE, Howe M, Bigley G et al. Randomized study to evaluate the effect of a perioperative infusion of dopexamine on colonic mucosal ischemia after aortic surgery. Journal of Vascular Surgery. 2001; 33: 758–63. McGinley J, Lynch L, Hubbard K, McCoy D, Cunningham AJ. Dopexamine hydrochloride does not modify hemodynamic response or tissue oxygenation or gut permeability during abdominal aortic surgery. Canadian Journal of Anaesthesia. 2001; 48: 238–44.



28. Takala J, Meier-Hellmann A, Eddleston J, Hulstaert P, Sramek V. Effect of dopexamine on outcome after major abdominal surgery: a prospective, randomized, controlled multicenter study. European Multicenter Study Group on Dopexamine in Major Abdominal Surgery. Critical Care Medicine. 2000; 28: 3417–23. 29. Parsons V. Recent advances in the management of acute renal failure. In: Ledinghan MI (eds). Recent advances in critical care medicine, Vol. 3. Edinburgh: Churchill Livingstone, 1980. 30. Bellomo R, Ronco C. Continuous haemofiltration in the intensive care unit. Critical Care. 2000; 4: 339–45. 31. Kellum JA, Angus DC, Johnson JP, Leblanc M, Griffin M, Ramakrishnan N et al. Continuous versus intermittent renal replacement therapy: a meta-analysis. Intensive Care Medicine. 2002; 28: 29–37. 32. Uchino S, Bellomo R, Ronco C. Intermittent versus continuous renal replacement therapy in the ICU: impact on electrolyte and acid-base balance. Intensive Care Medicine. 2001; 27: 1037–43. 33. Tonelli M, Manns B, Feller-Kopman D. Acute renal failure in the intensive care unit: a systematic review of the impact of dialytic modality on mortality and renal recovery. American Journal of Kidney Diseases. 2002; 40: 875–85. 34. Meguid MM, Collier MD, Howard U. Uncomplicated and stressed starvation. Surgical Clinics of North America. 1981; 61: 529–43. 35. Moore FA, Feliciano DV, Andrassy RJ, McArdle AH, Booth FV et al. Early enteral feeding, compared with parenteral, reduces post-operative septic complications. Annals of Surgery. 1992; 216: 172–83. 36. Graham TW, Zadronzy DB, Harrington T. The benefits of early jejunal hyperalimentation in the head-injured patient. Neurosurgery. 1989; 25: 729–35. 37. Lee B, Chang RWS, Jacobs S. Intermittent nasogastric feeding: a simple and effective method to reduce pneumonia among ventilated ICU patients. Clinical Intensive Care. 1990; 1: 100–2. 38. Heyland DK, Dhaliwal R, Drover JW, Gramlich L, Dodek P. Canadian Critical Care Clinical Practice Guidelines Committee. Canadian clinical practice guidelines for nutrition support in mechanically ventilated, critically ill adult patients. Journal of Parenteral and Enteral Nutrition. 2003; 27: 355–73. 39. Crook MA, Hally V, Panteli JV. The importance of the refeeding syndrome. Nutrition. 2001; 7: 632–7. 40. Andrews FJ, Griffiths RD. Glutamine: essential for immune nutrition in the critically ill. British Journal of Nutrition. 2002; 87: 3–8. 41. Griffiths RD, Allen KD, Andrews FJ, Jones C. Infection, multiple organ failure, and survival in the intensive care unit: influence of glutamine-supplemented parenteral nutrition on acquired infection. Nutrition. 2002; 18: 546–52.



Chapter 42 Adult critical care 42. Jian ZM, Cao JD, Zhu XG, Zhao WX, Yu JC et al. The impact of alanyl-glutamine on clinical safety, nitrogen balance, intestinal permeability, and clinical outcome in postoperative patients: a randomized,double-blind, controlled study of 120 patients. Journal of Parenteral and Enteral Nutrition. 1999; 23: S62–6. 43. Garrel D, Patenaude J, Nedelec B, Samson L, Dorais J, Champoux J et al. Decreased mortality and infectious morbidity in adult burn patients given enteral glutamine supplements: a prospective, controlled, randomized clinical trial. Critical Care Medicine. 2003; 31: 2444–9. 44. Hall JC, Dobb G, Hall J, de Sousa R, Brennan L, McCauley R. A prospective randomized trial of enteral glutamine in critical illness. Intensive Care Medicine. 2003; 29: 1710–6. 45. Gadek JE, DeMichele SJ, Karlstad MD, Pacht ER, Donahoe M, Albertson TE et al. Effect of enteral feeding with eicosapentaenoic acid, g-linolenic acid, and antioxidants in patients with acute respiratory distress syndrome. Critical Care Medicine. 1999; 27: 1409–20. 46. Pacht ER, DeMichele SJ, Nelson JL, Hart J, Wennberg AK, Gadek JE. Enteral nutrition with eicosapentaenoic acid, gamma-linolenic acid, and antioxidants reduces alveolar inflammatory mediators and protein influx in patients with acute respiratory distress syndrome. Critical Care Medicine. 2003; 31: 491–500. 47. Weiss G, Meyer F, Matthies B, Pross M, Koenig W, Lippert H. Immunomodulation by perioperative administration of n-3 fatty acids. British Journal of Nutrition. 2002; 87: 89–94. 48. Galban C, Montejo JC, Mesejo A, Marco P, Celaya S, Sanchez-Segura JM et al. An immune-enhancing diet reduces mortality rate and episodes of bacteremia in septic intensive care unit patients. Critical Care Medicine. 2000; 28: 643–8. 49. Bertolini G, Iapichino G, Radrizzani D, Facchini R, Simini B, Bruzzone P et al. Early enteral immunonutrition in patients with severe sepsis: results of an interim analysis of a randomized multicentre clinical trial. Intensive Care Medicine. 2003; 29: 834–40. 50. Heyland DK, Novak F, Drover JW, Jain M, Su X, Suchner U. Should immunonutrition become routine in critically ill patients? A systematic review of the evidence. Journal of the American Medical Association. 2001; 286: 944–53. 51. Beale RJ, Bryg DJ, Bihari DJ. Immunonutrition in the critically ill: a systematic review of clinical outcome. Critical Care Medicine. 1999; 27: 2799–805. 52. Heys SD, Walker LG, Smith I, Eremin O. Enteral nutritional supplementation with key nutrients in patients with critical illness and cancer – a meta-analysis of randomized controlled clinical trials. Annals of Surgery. 1999; 229: 467–77. 53. American College of Chest Physicians/Society of Critical Care Medicine. Consensus Conference: definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Critical Care Medicine. 1992; 20: 864–74.



] 539



54. Guidelines for the management of severe sepsis and septic shock. The International Sepsis Forum. Intensive Care Medicine. 2001; 27: S1–134. 55. Task Force of the American College of Critical Care Medicine Society of Critical Care Medicine. Practice parameters for hemodynamic support of sepsis in adult patients in sepsis. Critical Care Medicine. 1999; 27: 639–60. 56. Misset B, Timsit JF, Dumay MF, Garrouste M, Chalfine A, Flouriot I et al. A continuous quality-improvement program reduces nosocomial infection rates in the ICU. Intensive Care Medicine. 2003; 30: 395–400. 57. Harbarth S, Garbino J, Pugin J, Romand JA, Lew D, Pittet D. Inappropriate initial antimicrobial therapy and its effect on survival in a clinical trial of immunomodulating therapy for severe sepsis. American Journal of Medicine. 2003; 115: 529–35. 58. Kollef MH. The importance of appropriate initial antibiotic therapy for hospital-acquired infections. American Journal of Medicine. 2003; 115: 582–4. 59. Hinsdale JG, Jaffe BM. Reoperation for intra-abdorninal sepsis. Annals of Surgery. 1984; 199: 31–6. 60. Van den Berghe G, Wouters P, Weekers F, Mohan S, Baxter RC, Veldhuis JD et al. Reactivation of pituitary hormone release and metabolic improvement by infusion of growth hormone-releasing peptide and thyrotropin-releasing hormone in patients with protracted critical illness. Journal of Clinical Endocrinology and Metabolism. 1999; 84: 1311–23. 61. Van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M et al. Intensive insulin therapy in the critically ill patients. New England Journal of Medicine. 2001; 345: 1359–67. 62. Van den Berghe G, Wouters PJ, Bouillon R, Weekers F, Verwaest C, Schetz M et al. Outcome benefit of intensive insulin therapy in the critically ill: Insulin dose versus glycemic control. Critical Care Medicine. 2003; 31: 359–66. 63. Corwin HL, Surgenor SD, Gettinger A. Transfusion practice in the critically ill. Critical Care Medicine. 2003; 31: S668–71. 64. Krafte-Jacobs B, Levetown ML, Bray GL, Ruttimann UE, Pollack MM. Erythropoietin response to critical illness. Critical Care Medicine. 1994; 22: 821–6. 65. Hebert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion Requirements in Critical Care Investigators, Canadian Critical Care Trials Group. New England Journal of Medicine. 1999; 340: 409–17. 66. Hebert PC, Yetisir E, Martin C, Blajchman MA, Wells G, Marshall J et al. Transfusion Requirements in Critical Care Investigators for the Canadian Critical Care Trials Group. Is a low transfusion threshold safe in critically ill patients with cardiovascular diseases? Critical Care Medicine. 2001; 29: 227–34.



540 ] PART 8 PERIOPERATIVE MANAGEMENT 67. Corwin HL, Gettinger A, Pearl RG, Fink MP, Levy MM, Abraham E et al. The CRIT Study: Anemia and blood transfusion in the critically ill – Current clinical practice in the United States. Critical Care Medicine. 2004; 32: 39–52. 68. Vincent JL, Baron JF, Reinhart K, Gattinoni L, Thijs L, Webb A et al. ABC (Anemia and Blood Transfusion in Critical Care) Investigators. Anemia and blood transfusion in critically ill patients. Journal of the American Medical Association. 2002; 288: 1499–507. 69. Corwin HL, Gettinger A, Pearl RG, Fink MP, Levy MM, Shapiro MJ et al. EPO Critical Care Trials Group. Efficacy of recombinant human erythropoietin in critically ill patients: a randomized controlled trial. Journal of the American Medical Association. 2002; 288: 2827–35. 70. Fowler RA, Berenson M. Blood conservation in the intensive care unit. Critical Care Medicine. 2003; 31: S715–20. 71. MacIsaac CM, Presneill JJ, Boyce CA, Byron KL, Cade JF. The influence of a blood conserving device on anaemia in intensive care patients. Anaesthesia and Intensive Care. 2003; 31: 653–7. 72. Thorpe S, Thomas AN. The use of a blood conservation pressure transducer system in critically ill patients. Anaesthesia. 2000; 55: 27–31. 73. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B et al. Early Goal-Directed Therapy Collaborative Group. Early goal-directed therapy in the treatment of severe sepsis and septic shock. New England Journal of Medicine. 2001; 345: 1368–77. 74. Parsons PE, Eisner MD, Thompson BT, Matthay MA, Ancukiewiez M, Bernard GR et al. Lower tidal volume ventilation and plasma cytokine markers of inflammation in patients with acute lung injury. Critical Care Medicine. 33: 1–6. 75. Ranieri VM, Suter PM, Tortorella C, De Tullio R, Dayer JM, Brienza A et al. Effect of mechanical ventilation on inflammatory mediators in patients with acute respiratory distress syndrome: a randomized controlled trial. Journal of the American Medical Association. 1999; 282: 54–61. 76. Amato MB, Barbas CS, Medeiros DM, Magaldi RB, Schettino GP, Lorenzi-Filho G et al. Effect of a protectiveventilation strategy on mortality in the acute respiratory distress syndrome. New England Journal of Medicine. 1998; 338: 347–54. 77. The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. New England Journal of Medicine. 2000; 342: 1301–8. 78. Petrucci N, Iacovelli W. Ventilation with lower tidal volumes versus traditional tidal volumes in adults for acute lung injury and acute respiratory distress syndrome. Cochrane Database of Systematic Reviews. 2004: CD003844. 79. Stewart TE, Meade MO, Cook DJ, Granton JT, Hodder RV, Lapinsky SE et al. Evaluation of a ventilation strategy to



80.



81.



82.



83.



84. 85.



86.



87.



88.



89.



90.



91.



92.



prevent barotrauma in patients at high risk for acute respiratory distress syndrome. Pressure- and VolumeLimited Ventilation Strategy Group. New England Journal of Medicine. 1998; 338: 355–61. Yan SB, Helterbrand JD, Hartman DL, Wright TJ, Bernard GR. Low levels of protein C are associated with poor outcomes in severe sepsis. Chest. 2001; 120: 915–22. Mesters RM, Helterbrand J, Utterback BG, Yan B, Chao YB, Fernandez JA et al. Prognostic value of protein C concentrations in neutropenic patients at high risk of severe septic complications. Critical Care Medicine. 2000; 28: 2209–16. Taylor Jr. FB, Chang A, Esmon CT, D’Angelo A, ViganoD’Angelo S, Blick KE. Protein C prevents the coagulopathic and lethal effects of Escherichia coli infusion in the baboon. Journal of Clinical Investigation. 1987; 79: 918–25. Bernard GR, Vincent JL, Laterre PF, LaRosa SP, Dhainaut JF, Lopez-Rodriguez A et al. Recombinant Human Protein C Worldwide Evaluation in Severe Sepsis (PROWESS) Study Group: Efficacy and safety of recombinant human activated protein C for severe sepsis. New England Journal of Medicine. 2001; 344: 699–709. Oelkers W. Adrenal insufficiency. New England Journal of Medicine. 1996; 335: 1206–12. Cronin L, Cook DJ, Carlet J, Heyland DK, King D, Lansang MA et al. Corticosteroid treatment for sepsis: a critical appraisal and meta-analysis of the literature. Critical Care Medicine. 1995; 23: 1430–9. Annane D, Se´bille V, Troche´ G, Raphael JC, Gajdos P, Bellissant E. A 3-level prognostic classification in septic shock based on cortisol levels and cortisol response to corticotropin. Journal of the American Medical Association. 2000; 283: 1038–45. Beishuizen A, Thijs LG, Vermes I. Patterns of corticosteroid-binding globulin and the free cortisol index during septic shock and multitrauma. Intensive Care Medicine. 2001; 27: 1584–91. Koo DJ, Jackman D, Chaudry IH, Wang P. Adrenal insufficiency during the late stage of polymicrobial sepsis. Critical Care Medicine. 2001; 29: 618–22. Briegel J, Forst H, Haller M, Schelling G, Kilger E, Kuprat G et al. Stress doses of hydrocortisone reverse hyperdynamic septic shock: a prospective, randomized, double-blinded, single-center study. Critical Care Medicine. 1999; 27: 723–32. Bollaert PE, Fieux F, Charpentier C, Levy B. Baseline cortisol levels, cortisol response to corticotropin, and prognosis in late septic shock. Shock. 2003; 19: 13–5. Annane D, Sebille V, Charpentier C, Bollaert PE, Francois B, Korach JM et al. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock. Journal of the American Medical Association. 2002; 288: 862–71. Keh D, Boehnke T, Weber-Cartens S, Schulz C, Ahlers O, Bercker S et al. Immunologic and hemodynamic effects of ‘low-dose’ hydrocortisone in septic shock: a double-blind,



Chapter 42 Adult critical care randomized, placebo-controlled, crossover study. American Journal of Respiratory and Critical Care Medicine. 2003; 167: 512–20. 93. Levack P. Live and let die? A structured approach to decision-making about resuscitation. British Journal of Anaesthesia. 2002; 89: 683–6. 94. Prendergast TJ, Claessens MT, Luce JM. A national survey of end-of-life care for critically ill patients. American Journal of Respiratory and Critical Care Medicine. 1998; 158: 1163–7.



] 541



95. Sjokvist P, Nilstun T, Svantesson M, Berggren L. Withdrawal of life support – who should decide? Differences in attitudes among the general public, nurses and physicians. Intensive Care Medicine. 1999; 25: 949–54. 96. Ferrand E, Lemaire F, Regnier B, Kuteifan K, Badet M, Asfar P et al. Discrepancies between perceptions by physicians and nursing staff of intensive care unit end-of-life decisions. American Journal of Respiratory and Critical Care Medicine. 2003; 167: 1310–5.



43 Paediatric intensive care HELEN ALLEN AND ROB ROSS RUSSELL



Introduction General principles ENT emergencies in PICU Major surgery Summary



542 542 544 546 547



Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



547 547 548 548



SEARCH STRATEGY Data for this chapter are mostly generic management strategies for children in intensive care. The data were gathered through reference to standard texts on paediatric intensive care, notably Textbook of pediatric intensive care1 and Fluid balance and volume resuscitation for beginners.2 Specific information on ENT trauma in children was obtained through PubMed using the key words laryngeal trauma and laryngeal injury, limiting the search to children 0–18. This revealed 65 review articles of which five were found relevant and scrutinized. Information on bronchoscopy was located on PubMed using the key words fibreoptic bronchoscopy and intensive care, limiting the search to children 0–18. This only yielded two papers (one from 1980), but searching on the related links button located a number of relevant articles.



INTRODUCTION Children are liable to present acutely with ENT pathologies, many of which are specific to the paediatric population and may require intensive care management. Those children undergoing major ENT surgery also frequently require admission to a paediatric intensive care unit in the postoperative period. This chapter aims to outline the general management principles of ENT patients within the paediatric intensive care setting. Important differences in the anatomy and physiology of children compared to adults are reviewed. A systematic approach to evaluation of children admitted to the paediatric intensive care unit (PICU) is outlined, together with general principles of fluid and analgesia management in children. Specific issues relating to the care of the major paediatric emergencies, as well as the postoperative management of patients undergoing major ENT surgical procedures, are discussed. Finally, indications and



procedures for performing bronchoscopy within the unit are detailed.



GENERAL PRINCIPLES Anatomical and physiological considerations Care of the paediatric patient requires an understanding of the anatomy and physiology of children, which have some important differences from adulthood. Anatomically, the child has quite different body proportions and shape. The head of the infant and small child is proportionately larger, accounting for nearly 10 percent of body surface area, compared to the 3.5 percent in adults. The neck of a child is shorter, and these two factors tend to increase neck flexion, particularly in the emergency setting. A relatively large tongue and small mandible can make access to the airway difficult.



Chapter 43 Paediatric intensive care Table 43.1



Physiological parameters at different ages.



Age



Newborn 1 year 2–5 years 5–12 years >12 years



Heart rate (bpm) 110–140 90–110 75–90 60–90 50–80



Respiratory rate (bpm) 50–60 30–40 25–30 20–25 15–20



Mean BP (mmHg) 50–60 70–90 80–100 90–110 100–120



In all young children the epiglottis is rounded (‘horseshoe shape’) and the larynx sits high (opposite the third cervical vertebra, compared to the fifth or sixth in an adult) and anterior. The larynx itself is conical, unlike the cylindrical adult larynx, and the narrowest part lies at the cricoid cartilage. Finally, the tracheal cartilage is often less rigid than in adults, and may compress if the neck is overextended.3 Infants (up to approximately six months old) are also obligate nasal breathers. When this is combined with the relatively small airways and risk of mucous obstruction, airway compromise is common. Between three and six years of age, tonsillar and adenoidal hypertrophy further contribute to airway obstruction. Physiologically, there are also substantial and important differences. Respiratory muscle function is affected by the position of the ribs in infants, which lie perpendicular to the spine rather than running caudally as they do in adults. This makes the diaphragm the only effective inspiratory muscle. Infants and children have a faster respiratory rate than adults (see Table 43.1), which is partly related to an increased metabolic rate. In infants, increased chest wall compliance can increase the work of breathing substantially. In the cardiovascular system, a limited stroke volume causes a faster heart rate in children, and they also display a lower mean blood pressure. Their response to cardiovascular stress is limited, and cardiac output can only be increased significantly by increasing the heart rate, hence tachycardia is common.



Evaluation Admission of a child into the PICU is a difficult and important moment in the child’s care. The care of the child is transferred to a new team of doctors and nurses, and the family are moved to a new and stressful environment. It is essential that this handover is careful and complete. Ideally, there should be a face to face handover between the medical and nursing teams involved in the care. The background history, current problems and care plan should be clearly documented. The family of the child need to be included in this handover. Where the admission is planned, families



] 543



should be offered the chance to visit the intensive care unit prior to the surgery so that they have met some of the staff and seen the environment. There should be an area for them to stay in while procedures are being carried out and they need regular updates from medical and nursing staff on progress. The doctors admitting the child also need to evaluate the child carefully on admission. This should follow standard ‘ABC’ rules, assessing the airway, breathing and circulation. Issues of fluid balance (including any feeding plan), analgesia, sedation and blood tests should also be clarified (Table 43.2).



Fluid management Although the principles of fluid balance in children are the same as in adults, there are a number of complicating factors that make life more difficult for the junior doctor. The first and most obvious is an unfortunate tendency for children to grow! Any fluids that need to be given are therefore based on the child’s weight or occasionally the body surface area. Metabolic requirements, including insensible water loss, vary with age and this must be taken into consideration. The distribution of fluids between compartments is different for adults and children, and an understanding of this difference is important in deciding on appropriate volumes and types of fluid. The infant has approximately one-third of their body weight as interstitial fluid – twice as much as in an adult. Consequently, fluids constitute approximately 75 percent of body weight in infancy, compared to 60 percent in the adult. THE NEWBORN INFANT



The full-term infant requires approximately 150 mL/kg per day of fluid (although less in the first few days of life). Of this, approximately 30 mL/kg will be accounted for by insensible losses. Normal fluid intake is entirely as milk, which provides both fluid and nutrition for the first few months of life. Fluid requirements (per kg) gradually



Table 43.2 PICU.



Issues to be considered on admission of a child to Issues to consider



Airway Breathing Circulation Fluids Analgesia/ sedation Tests



Clear? ET tube position? ET tube secured properly? Air entry? Ventilation adequate? Peripheral perfusion? Adequate access? Drug infusions? Adequate sugar? Urine output? Pain control AND sedation adequate? Paralysis needed? Blood gas? Electrolytes? FBC?



544 ] PART 8 PERIOPERATIVE MANAGEMENT Table 43.3 Age



1 day 5 days 1 month 6 months 1 year Over 1 year



Fluid and electrolyte requirements at different ages. Daily fluid requirements



60 mL/kg 150 mL/kg 120 mL/kg 110 mL/kg 100 mL/kg 1000 mL for first 10 kg, 500 ml for next 10 kg then 20–25 mL/kg for rest of body weight



Sodium/potassium requirements 2–3 mmol/kg/day 2 mmol/kg/day 1–2 mmol/kg/day



reduce thereafter, as shown in Table 43.3. It should be emphasized that these requirements are based on normal needs and that they frequently need adjustment in disease. Electrolyte requirements will also vary with age, although not so dramatically as fluid volume. In the infant, immature renal function increases the salt requirements, and 3 mmol/kg of sodium and potassium are usually required daily. In older children this falls to approximately 2 mmol/kg/day of each ion. Calcium, magnesium and other ions are broadly given in the same (per kg) dose as in adults. Glucose requirements are critical, especially in young children. The daily calorie requirement of infants is approximately 100 kcal/kg, more than double that in adults. Maintenance of adequate glucose levels in infants is critically important and hypoglycaemia (even for short periods) can cause permanent brain damage. In infants and small children, enough glucose must be given to maintain blood glucose levels.



THE UNWELL CHILD



Fluid requirements following surgery can be quite variable. In most units, children are started on a regime restricting their input to approximately two-thirds normal requirements. This is especially true if oral feeds are not to be used and fluids are being given intravenously. Intravenous fluids should be chosen with care. Recently, concerns have been raised about the use of hypotonic fluids (especially 0.18 percent saline) as this tends to induce hyponatraemia which can cause substantial problems including seizures and indeed death.4, 5 Solutions containing adequate sodium (such as 0.45 percent saline, 2.5 percent dextrose) are therefore recommended. Whatever regime is used, it needs to be remembered that children are prone to become dehydrated or fluid overloaded quite rapidly and need to be monitored carefully. In infants especially, blood glucose levels need to be measured regularly (up to four hourly). [**/*]



Analgesia Appropriate analgesia needs to balance the sedative and respiratory suppressive effects of opiates against the need for good pain control. Oral paracetamol, or other nonsteroidal antiinflammatory drugs (NSAIDs) may often be adequate, but opiates (e.g. morphine) may be needed. An infusion (e.g. morphine at 10–30 mg/kg/hour) is easily administered. In older children, patient controlled analgesia (PCA) pumps are helpful. Background rates (e.g. morphine at 10 mg/kg/hour) with boluses as needed (e.g. morphine 10 mg/kg up to hourly) is often a good combination but should be discussed with the anaesthetist involved. Sedation may also be needed for patients in the PICU, especially if ventilated or undergoing painful procedures. Units vary between using infusions (e.g. midazolam 50–250 mg/kg/hour) or boluses (e.g. midazolam 0.1 mg/kg, p.r.n. up to four hourly). Current trends favour intermittent bolus dosage as this reduces total sedative load, but patients need close observation to avoid distress. [*]



ENT EMERGENCIES IN PICU Trauma All trauma cases, including those who have sustained craniofacial trauma, should be managed acutely according to advanced life support guidelines,3 with immediate evaluation and management of airway, breathing and circulation. Cases of blunt laryngeal trauma are rare. They may occur if a child is unrestrained in a car involved in a road traffic accident if they hit the dashboard when the head and neck are in a hyperextended position. In children presenting with laryngotracheal injuries in the absence of such a history, the possibility of nonaccidental injury with a direct blow to the anterior neck should be considered as a possible aetiology. A protocol for the management of acute laryngeal trauma was devised and published by Schaefer in 1982.6 This advocates immediate protection of the airway by emergency tracheostomy for all but the most minor of cases, with endoscopy to evaluate the injuries. In patients with moderate injuries, the degree of traumatic damage can be further assessed using CT scanning once the airway has been protected by tracheostomy. In cases where the injuries are severe, open exploration is recommended. [**/*]



Upper airway obstruction The small cross-sectional area and conical shape of the paediatric airway renders children particularly susceptible to upper airway obstruction. A relatively small reduction



Chapter 43 Paediatric intensive care



in the radius of the airway through mucosal oedema, secretions or a foreign body, results in a significant increase in airway resistance (resistance a diameter).6 Stridor on inspiration is indicative of upper airway obstruction. In children where severe upper airway obstruction is suspected as a diagnosis, it is imperative to avoid all invasive procedures and aspects of the examination liable to distress the child as this may precipitate complete upper airway occlusion. Children should be kept warm and calm while experienced anaesthetic and surgical staff are brought together. Examination can then be undertaken in a controlled manner. A full examination of the child is also needed at a convenient point. Birthmarks, and especially haemangiomas (which may indicate a laryngeal lesion), should be recorded. If a child is dysmorphic, a genetic opinion may help identify a syndrome that may have airway implications.



Table 43.4



Features of croup versus epiglottitis. Croup



Age Cause Prodrome



Fever Appearance



Stridor Hypoxia Severity



] 545



Epiglottitis



1–3 years Parainfluenza viruses 1–2 days coryza



2–7 years H. influenzae Hours Sore throat Dysphagia o381C 4381C Lethargic Pale and toxic Drooling and dysphagic Sits with neck extended Barking cough Muffled stridor Loud stridor Stertorous breathing Unusual Frequent o3% hospitalized cases All require intubation require intubation



EPIGLOTTITIS



The advent of the Hib vaccine and its inclusion in childhood immunization schedules to protect them against Haemophilus influenzae type B infections, has led to a marked decline in the incidence of epiglottitis so that it is now a rare condition. However, occasional cases do still occur in unimmunized children or in those in whom the vaccine has failed. Clinical presentation is with a toxic child who is stridulous and drooling. The priority of immediate management is to secure the airway by intubation or, in extreme cases, emergency tracheostomy. They will then remain ventilated in intensive care whilst receiving treatment with intravenous antibiotics (third generation cephalosporin) until the epiglottic swelling has been reduced, usually within 24–36 hours.7 [*] CROUP



Acute viral laryngotracheobronchitis (viral croup) is the most common cause of stridor in children and the vast majority of patients can be managed either at home or on the general paediatric wards. Only the most severe cases require admission to a high dependency unit (HDU) or PICU. In such cases, short-term relief of the upper airway obstruction can be achieved by nebulizing epinephrine (1 mL/kg of 1:1000 up to 5 mL maximum dose) with oxygen. Steroids, either in the form of nebulized budesonide or systemic dexamethasone or prednisolone, can be given to reduce the upper airway inflammation. A very small minority of children with croup require intubation to protect their airway (o5 percent), but the administration of systemic steroids to this group of patients has been shown to reduce both the duration of intubation and the need for reintubation (Table 43.4).8



In all children with stridor, the consideration of other diagnoses, such as a postpharyngeal abscess, needs to be undertaken. [****/***/**/*]



Bacterial tracheitis Bacterial tracheitis, or pseudomembranous croup, is much less common than viral croup but it is potentially life threatening and the majority of patients will require intubation to secure their airway whilst receiving treatment with intravenous antibiotics. The pathogens are usually Staphylococcus aureus or Streptococci, which infect the tracheal mucosa causing necrosis and the production of purulent secretions that may occlude the upper airway. Clinically, a child with bacterial tracheitis is septic with a high fever, croupy cough and progressive signs of upper airway obstruction.



Foreign body inhalation A history of possible foreign body inhalation can usually be elicited, but in the absence of such information it should be suspected as a cause if there is sudden onset of stridor and upper airway obstruction with no history of preceding fever or illness. It occurs most commonly in children between the ages of one and three years, with foodstuffs being the usual cause. In such cases, if the child presents with increasing dyspnoea, apnoea or loss of consciousness which has not responded to simple airway opening manoeuvres, then the advanced paediatric life support guidelines for management of a choking child should be instituted.3 In an older child the Heimlich manoeuvre can safely be attempted or, if the child is unconscious and supine,



546 ] PART 8 PERIOPERATIVE MANAGEMENT abdominal thrusts may be equally effective at dislodging the foreign body. The foreign body may in some cases pass through the larynx and become lodged lower down in the bronchial tree. As discussed earlier, the larynx in the prepubertal child is conical, and objects passing the vocal cords can lodge at the level of the cricoid.3 It is essential that one does not use blind finger sweeps in the mouth for the foreign body as this may push it tighter into the larynx.3 Objects may also pass right into the lungs. On clinical examination there may be unilateral wheezing, reduced air entry on one side or signs suggestive of unilateral lung collapse. A chest x-ray may show hyperinflation of the affected lung with mediastinal shift due to the foreign body causing gas trapping on expiration, or alternatively lung collapse. These patients require a general anaesthetic for rigid bronchoscopic removal of the foreign body and may need a period of ventilation on the PICU afterwards. [**/*]



MAJOR SURGERY Craniofacial surgery and cleft lip and palate Children who undergo craniofacial surgery or a cleft lip or palate repair, may require admission to a PICU or HDU, particularly if there is a risk of the airway becoming compromised by postoperative swelling. If the child returns from theatre ventilated, sedation and analgesia need to be maintained for the whole period of ventilation. A partially sedated child may struggle with an endotracheal tube (ETT) in situ, and this may damage a palatal or laryngeal repair. It may be necessary to paralyze the child once they are adequately sedated. This is often best carried out with short-acting anaesthetic agents, particularly when there is a planned extubation, as this will allow the child to be woken promptly. With longer-acting drugs there may be accumulation of the drug, and a period of suboptimal sedation/wakefulness as it is metabolized. Other important considerations include the use of orogastric rather than nasogastric tubes, best sited under direct vision at the time of the operation, as nasogastric tubes can traumatize a palate repair. Nursing staff should be aware not to perform blind oropharyngeal suction with a rigid Yankaeur sucker, but should use a soft wide bore flexible suction catheter to remove visible secretions only. Infection, particularly with streptococcus, can be devastating to palatal repairs. As well as careful nursing and cleanliness, patients should not be nursed close to other children who may be carrying pathogens. [*]



Surgery of the nose and upper airway Patients who have surgery to the nose rarely require an admission to PICU in the postoperative period. If they do,



then the general principles of evaluation and management should be applied. The exception may be with infants who require surgery for choanal atresia/stenosis. In these children, maintaining a patent nasal stent may require close nursing observation for the first few days. Those undergoing surgery to the upper airway are, however, frequently admitted to PICU for a period after the operation due to the high risk of developing upper airway obstruction as oedema evolves around the site of surgery. This includes operations for laryngeal reconstruction, but also for more minor procedures such as aryepiglottoplasty. It is therefore important to ensure that a PICU/HDU bed is available for the patient before proceeding with surgery. Children who have major upper airway surgery, such as laryngotracheal reconstruction, are electively ventilated for a period of five to seven days after the operation to allow the oedema to resolve before the ETT is removed. In these patients, sedation and analgesia are again critical (see Craniofacial surgery and cleft lip and palate). Paralysis is usually used in this patient group to prevent trauma to the site of surgery and exacerbation of upper airway oedema due to movement of the ETT whilst in situ. Close observation of cardiovascular parameters in paralyzed patients will detect tachycardia and hypertension suggestive of inadequate sedation or pain underneath the paralysis. It is important that a significant leak is allowed to develop around the ETT before extubation is attempted as this may signify that the postoperative oedema is resolving and that extubation is more likely to be successful. [*]



Laryngeal surgery and tracheostomy Managing a tracheostomy in the PICU involves the management of the newly formed stoma, care of an established tracheostomy and training requirements for families looking after long-term tracheostomies. In the first week after a tracheostomy is formed, the airway is dependent on the tracheostomy stent (i.e. the tube) to maintain the airway open. Loss of the stent over this period can allow the airway to close, and the reinsertion of a tube through an immature tracheostomy can be difficult, as the tracheal opening can be obscured (especially in obese patients) and the tube can easily be passed into the wrong plane. This results in obstruction of the airway and ventilation of the tissues of the neck. For the first week, a new tracheostomy should therefore be carefully secured. In our unit, stay sutures are attached to the edges of the trachea at surgery, allowing the trachea to be brought up to the surface of the neck in an emergency and thus facilitating tube reinsertion.9 A tracheal dilator and spare tracheostomy tubes (of the same size and one size smaller) should be kept at the bedside. If a tracheostomy tube does become dislodged in this period and cannot easily be reinserted, then (assuming the larynx permits this) a conventional endotracheal tube



Chapter 43 Paediatric intensive care



should be put in through the mouth or nose to secure the airway. This allows a measured replacement of the tracheostomy tube either in the PICU or in theatre. Following the first tracheostomy tube change (usually at seven to ten days), the tracheostomy is more secure. Care of an established stoma involves the establishment of routine tube changes, tape care and training for the carers (where appropriate). The development of difficulties with tube insertion or of bleeding should raise the possibility of granulation formation. Parents who are to go home with a child who has a tracheostomy will need careful training. As well as confidence in tube and tape changes, training in resuscitation is necessary, as is an ability to perform adequate tracheal suction. Preparation of equipment and provision of support at home may take many weeks to organize. [*]



] 547



 Assisted intubations. In particularly difficult airways, a flexible bronchoscope or laryngoscope can be extremely useful in locating the airway. By threading an ETT over the scope prior to entering the airway, the ETT can then be placed into the correct position.  Collection of bronchoalveolar lavage fluid. In patients who develop severe lung disease, and particularly in patients who are immunosuppressed, the diagnosis of lung pathogens can be very difficult. Bronchoalveolar lavage may be very useful in this situation. It can either be carried out ‘blind’10 by passing a nasogastric tube into the lungs, washing in a small volume of 0.9 percent saline, and then aspirating, or using a fibreoptic scope to direct the lavage to a particular part of the lung. [**/*]



SUMMARY Bronchoscopy in PICU FIBREOPTIC BRONCHOSCOPY



This is a very useful technique, as it is extremely simple to pass a fibreoptic scope through an ETT and directly into the lung. Its advantages include simplicity and access. The scope can visualize all lobes of the lung and identify both static (e.g. mucous plugs) and dynamic (e.g. bronchomalacia) problems. The disadvantages include a narrow suction channel that blocks easily, and the inability to effectively ventilate through the scope. The latter can be circumvented if the scope is passed through a rubber valve in the ETT, allowing continued ventilation around the scope, but is difficult if the diameter of the scope is close to that of the ETT. RIGID BRONCHOSCOPY



This technique is much more effective at removing debris from the airway, and may be easier to use in the very unstable patient as ventilation can be continued through the scope during the procedure. There are several indications for either fibreoptic or rigid bronchoscopy in PICU:  Persistent collapse of lobe/lung. Ventilated children may develop complete collapse of a lung, usually secondary to blockage of the major airways with secretions and mucous plugs. This can lead to major ventilatory problems with shunting. Fibreoptic bronchoscopy down the ETT can often identify the degree of blockage but a rigid bronchoscope may be needed to clear thick secretions.  Assessment of malacia. Clinical tracheo- or bronchomalacia may be difficult to confirm. Flexible bronchoscopy may help locate the site and degree of weakness, but needs to be carried out with the patient self-ventilating as positive pressure ventilation can stent the airway open artificially.



Whilst principles of care for children are similar to those of adults, there must be careful consideration of the different physiological responses seen in children and how they vary with age. Some ENT procedures are also very specific to the younger age group, and in units which are not undertaking paediatric surgery frequently, well maintained guidelines for the care of this group of patients is essential.



KEY POINTS  Foreign body inhalation should be suspected in any child with unexplained stridor or respiratory symptoms.  Paediatric anatomy and physiology is significantly different from adults.  Fluid and electrolyte prescriptions for children must be adapted to the individual and calculated according to weight.  Frequent monitoring and adjustments of fluid and electrolyte prescriptions are required.  All patients with suspected epiglottitis require intubation.  Trauma and infection are the greatest threats to palate repairs.  Good evidence to support clinical practice in the PICU is limited.



Best clinical practice General Principles [ Anatomical and physiological differences between adults and children need to be recognized.



548 ] PART 8 PERIOPERATIVE MANAGEMENT [ Fluid and electrolyte administration must be age-and weight-appropriate. [ Appreciation of family carers and their involvement is essential.



ENT emergencies in PICU [ Laryngeal trauma in children usually requires a tracheostomy.



[ Epiglottitis is now rare in the UK following the introduction of the Hib vaccination.



[ Foreign body inhalation is an important, if uncommon, cause of acute respiratory symptoms in children.



Major surgery [ Paralysis may be needed following palatal surgery to avoid damage from a struggling child.



[ Infection is a major cause of palatal breakdown in children following surgery.



[ Flexible bronchoscopy in children can clear secretions, but the size of the bronchoscope limits its suctioning ability. [ Suspected foreign bodies should be removed through a rigid bronchoscope.



Deficiencies in current knowledge and areas for future research Many of the clinical practices on PICU have evolved through sharing of experience rather than being substantiated by medical evidence or clinical trials and the potential for research projects aimed at providing evidence to support or refute current practices is huge. However, in reality there are significant practical limitations to undertaking such research, for example the ethics of research in children, difficulties in obtaining parental consent for entry into trials, and recruiting sufficient numbers of patients to achieve adequate power. Much of current research focuses on establishing optimal sedation and analgesia regimes for intensive care patients.



REFERENCES 1. Rogers MC, Nichols DG (eds). Textbook of pediatric intensive care, 3rd edn. Baltimore, MD: Lippincott Williams and Wilkins, 1996. 2. Park GR, Roe PG. Fluid balance and volume resuscitation for beginners. London: Greenwich Medical Media Ltd, 2000. 3. Advanced Life Support Group. Advanced paediatric life support. London: BMJ Publishing Group, 2000. 4. Berleur MP, Dahan A, Murat I, Hazebroucq G. Perioperative infusions in paediatric patients: rationale for using Ringerlactate solution with low dextrose concentration. Journal of Clinical Pharmacy and Therapeutics. 2003; 28: 31–40. 5. Bohn D. Problems associated with intravenous fluid administration in children: do we have the right solutions? Current Opinion in Pediatrics. 2000; 12: 217–21. 6. Schaefer SD. Primary management of laryngeal trauma. Annals of Otology, Rhinology, and Laryngology. 1982; 91: 399–402. 7. Damm M, Eckel HE, Jungehulsing M, Roth B. Management of acute inflammatory childhood stridor. Otolaryngology and Head and Neck Surgery. 1999; 121: 633–8. 8. Stannard W, O’Callaghan C. Management of croup. Paediatric Drugs. 2002; 4: 231–40. 9. Burke A. The advantages of stay sutures with tracheostomy. Annals of the Royal College of Surgeons of England. 1981; 63: 426–8. 10. Ashton MR. ‘Blind’ bronchoalveolar lavage. Lancet. 1992; 340: 1104.



PART



9



SAFE AND EFFECTIVE PRACTICE EDITED BY MARTIN J BURTON



44 Training, accreditation and the maintenance of skills Paul O’Flynn



551



45 Communication and the medical consultation Damian Gardner-Thorpe and Richard Canter



559



46 Clinical governance: Improving the quality of patient care Debbie Wall, Patrick J Bradley and Aidan Halligan



568



47 Medical ethics Katherine Wasson



581



48 Medical jurisprudence and otorhinolaryngology Maurice Hawthorne and Desmond Watson



594



This page intentionally left blank



44 Training, accreditation and the maintenance of skills PAUL O’FLYNN



Introduction Ear, nose and throat surgical training in the UK Continuing professional development and the maintenance of skills



551 552 553



INTRODUCTION Over the past 20 years, surgical practice has changed almost beyond recognition. Ear, nose and throat (ENT) surgery, and particularly rhinology and rhinological surgery, are excellent examples of this. The technical aspects of surgery have changed as the materials, investigations and instruments available have evolved. Techniques have radically changed as completely new approaches have developed. A simple illustration of this is the use of rigid endoscopes in nasal and sinus surgery, both for diagnostic and therapeutic purposes. Prior to the mid-1980s, almost all sinus surgery was undertaken using a headlamp. At that time, the illumination was relatively poor and access to parts of the nose and sinuses restricted. In the 1980s, a number of surgeons started to see the benefit of using the illumination and magnification of the operating microscope to deal with sinus disease. However, within a few years rigid endoscopes became increasingly available and in less than ten years entered widespread use in the USA, UK and Europe. This change from relatively crude surgery to accurate dissection, based on ever-improving imaging and now image-guided systems, has entirely changed our knowledge and practice. The concepts that underpin our understanding of sinus disease have changed. A far greater emphasis is placed on medical, allergic and immunological aspects, as well as the relationship between the upper and the lower airway. The transition described above has not been without problems. In both endoscopic sinus surgery and laparoscopic surgery, overenthusiasm for the method with



Guidance for surgeons on appraisal and revalidation: specialty-specific guidance for ear, nose and throat surgeons References



556 558



inadequate training has lead to serious adverse outcomes. The need for structured training was recognized even for experienced surgeons. This change in technology and the rapidly expanding availability of computerized tomography (CT) scanning, giving the surgeon far more information about what they could expect to find anatomically and pathologically in the sinuses, has driven a need for on-going training for established practitioners and new curriculum for trainees. The Hill Surgical Workshop at the Royal College of Surgeons was one response to this need. Surgeons today, both trainees and experienced, have to continue to acquire substantial new knowledge, new diagnostic skills and new technical skills throughout their career. We, as surgeons, have a duty to do this.1 All these technical matters need to be seen within the broader context of change. The public, patients, legislators, healthcare professionals and the Government have all started to voice ‘needs’ within the healthcare system. Public confidence was shaken by the actions of Dr Shipman and the apparent lack of scrutiny of them.2 ‘Disasters’, such as the cases highlighted in Bristol, where well-meaning surgeons fell below an acceptable standard, and the subsequent reports and inquiries have heightened the public taste for scrutiny of the medial professions.3 Patients, in general, want to be fully informed before consenting to a surgical procedure and we as surgeons are bound by the requirements of Good surgical practice to obtain informed consent.4 The issues surrounding consent are in themselves highly complex and outside the scope of this chapter, but highlight yet another change. A good doctor–patient relationship is a fundamental part of the care of a surgical



552 ] PART 9 SAFE AND EFFECTIVE PRACTICE patient. Some surgeons have an innate ‘skill’ of communication, others simply do not. Attention to this aspect of the surgeons overall ability is now well recognized and is assessed and examined in both the MRCS and Intercollegiate FRCS examinations. Good hand–eye coordination is not sufficient to make a modern surgeon. The Royal College of Physicians and Surgeons of Canada published a substantial document in 1996 subtitled ‘Skills for the new millennium’ in which seven ‘roles’ were described as being important in a medical practitioner.5 The precise mixture of the roles and the extent to which they are relevant will vary from specialty to specialty and will be dependent on the stage of development of the practitioner, but to some extent all roles will be relevant to some degree in every practitioner. The roles defined are:       



medical expert; communicator; collaborator; manager; health advocate; scholar; professional.



Methods of evaluating these areas are also suggested. The role of medical education, training and revalidation can be seen as a broad development of various roles largely defined by (though not exclusively) the list above.



EAR, NOSE AND THROAT SURGICAL TRAINING IN THE UK Historic perspective Until the mid-1990s, surgical training in the UK, and largely in those countries associated with the Commonwealth, had been based on an apprenticeship model. This tradition has assumed that over time the apprentice (young surgeon) would:  acquire knowledge relevant to their particular specialization;  develop clinical skills;  gain exposure to conditions and diseases and understand their natural history and standard treatment;  see complications and their resolution. Over time, the novice would gain experience, skill and knowledge to ultimately master his or her specialty. The range of skills and knowledge acquired would be tested by examination and in-training assessment. In the UK, this individual would then progress to positions that are more senior and ultimately become a consultant ‘master practitioner’. The cycle would then continue with the consultant training young surgeons and handing on skills and knowledge.



The mark of a consultant surgeon is the ability to ‘practice independently’. The real meaning of this phrase is not well defined and does not fit well with the likely organization of the surgical professions in the future where ‘team working’ is much more emphasized.6 One view of the organization of surgical teams involves groups of specialists in the generality of an area of surgery ‘where team members work within a closely defined range of skills’, rather like aircrew. Pilots cannot change between levels of competence or types of aircraft without demonstrating competence and usually after specific training. The apprenticeship model, despite considerable criticism, served the UK well. The following are the principle advantages of this system:  it is not time constrained, therefore the apprentice can learn at his or her own pace;  training is close to the ‘coal face’, so that common conditions are seen and treated frequently;  a close working relationship often develops between master and apprentice, helping with ongoing personal and professional development. Critics of this system point to a number of weaknesses, including:  the system fails to meet changing the needs of the profession;  progress to more senior positions depends on availability of such a post and the degree of competition at that time;  criteria against which to measure competency are generally lacking;  poor trainer–trainee relationships could lead to underperformance. Prior to the Calman Report, a typical career pathway in otolaryngology could be summarized as follows: 1. medical school (five to six years) to obtain medical degree. Provisional registration with the General Medical Council (GMC); 2. pre-registration house officer (one year), full registration with GMC; 3. senior house officer (approximately two to three years) including time spent in non-ENT posts such as general surgery, plastics, neurosurgery and accident and emergency. Obtain Primary FRCS (examination in anatomy, pathology and physiology); 4. registrar in otolaryngology (approximately three years). Entry into post by competitive interview. Obtain FRCS in otolaryngology during tenure; 5. senior registrar (approximately three years). Entry into post by competitive interview; 6. Maintain log book. Obtain Certificate of Completion of Higher Surgical Training; 7. consultant post. Competitive interview. Virtually permanent tenure.



Chapter 44 Training, accreditation and the maintenance of skills



Although the tenure of each post was fixed, the time in that grade was not. There was an imbalance in the numbers of posts available and a bottleneck between registrar and senior registrar grades. Over time, the UK system led to excessively long periods of ‘training’. In some specialties, so few consultant posts were available that experienced and able trainees were stuck in the senior registrar grade for many years. This situation failed to satisfy the needs of the National Health Service or the trainees. European Working time legislation7 and the ‘New deal on junior doctors’ hours’8 have forced a rethink of UK training.



The Calman Report In the early 1990s, Sir Kenneth Calman, as Chief Medical Officer, set up a working party that produced a report on the future direction of training.9 As part of the implementation of that report the grades of registrar and senior registrar were merged into the specialist registrar grade between 1995 and 1997. The new grade had protected training time and was not to be seen as part of service provision. Other features of the grade were to be a defined curriculum with clearer educational objectives and assessment. Protected training time has been widely implemented. Specialist registrars, however, remain important service providers. The bottleneck is now at the senior house officer (SHO) level. In ENT, the specialist registrar grade is tenable for six years; after this time, trainees ‘fall off ’ the training scheme whether or not there is a post available or suitable for them. After four years of higher surgical training and satisfactory intraining assessments, trainees sit the Intercollegiate FRCS in Otolaryngology.



Current and future training in otolaryngology/ head and neck surgery At the time of writing, further changes to training are in the pipeline. The Chief Medical Officer, Sir Liam Donaldson, has published Unfinished business10 and Modernising medical careers.11 The Working Time Directive continues to bite, reducing the hours available for training further. In the current proposals, the senior house officer grade is to be abolished and the preregistration house officer time replaced by two post-graduate foundation years. The remaining period of specialty training will be confined to six years: 1. medical qualification; 2. two years’ post-graduate training – foundation years; 3. specialist registrar – six years’ specialty training based on new curriculum. Specialty-specific



] 553



MRCS after two years. FRCS (Otolaryngology) towards the end of the training period; 4. specialist in generality of ENT. At present, there is no consensus as to the precise nature of what a specialist in this context will be or be able to do. ‘Working in teams within a defined range of competencies’ is the phase currently used. This is very different from ‘independent practice’. Terms such as ‘emergency safe’ have been widely adopted. Emergency safe in all aspects of ENT represents a high level of training. To deal with adult and paediatric airway problems, acute middle ear disease and orbital cellulitis and even severe nose bleeds requires a wide range of skills. There is no clarity about provision for superspecialized training in areas such as otology, paediatrics, skull base and head and neck surgery.



Curriculum development project Under the auspices of the Surgical Royal Colleges, each specialty has been working on curriculum development. The new curriculum is much more explicit than any previous attempt and contains standards, criteria and assessment modalities. The ENT curriculum is near completion. The curriculum covers both generic and specialist skills. Using Good surgical practice4 as a guide, the following generic skills are specified and assessed:      



good clinical care; communication skills; maintaining good medical practice; maintaining trust; working with colleagues; teaching and training.



For example, communication skills have a specified objective and the appropriate knowledge, skills, attitudes and means of assessment are explicit. Table 44.1 shows an example of the curriculum document relating to ENT communication skills. In the specialty-specific parts of the curriculum, a similar approach is applied and the knowledge, skills, professional attributes and assessment are specified. In addition, an indication of the depth of the ability or knowledge at each stage in training is given, where (1) is introductory level and (4) is advanced. Table 44.2 shows an example of the paediatric ENT curriculum document.12



CONTINUING PROFESSIONAL DEVELOPMENT AND THE MAINTENANCE OF SKILLS All practitioners are obliged to keep up to date and monitor their performance. These obligations are laid out



554 ] PART 9 SAFE AND EFFECTIVE PRACTICE Table 44.1 ENT communication skills. Objective: To communicate effectively with patients, relatives and colleagues in the different spheres of clinical practice. Circumstance



Knowledge



Skills



Attitudes



Assessment



Within a consultation



Structuring the interview by identifying the concerns of patients and, where applicable, parents of paediatric patients



Demonstrate empathy with patients and family. Be aware of importance of utilizing these skills. Be aware of importance of involving patients and parents in decisions. Offering choices. Respecting other views



Observed practice on ward rounds (teaching and business). Theatre practice and ‘ontake’. Progress and outcomes recorded in trainer’s report. Might include attendance at taught course



Breaking bad news



Structuring the interview and organizing setting. Normal process of bereavement and associated behaviours. Awareness of local policy on organ donation procedures and role of transplant coordinators How and when to communicate effectively with other members of the care team, including the use of handover rounds Awareness of local policy for dealing with complaints



Active listening: Appropriate use of different question types. Use language appropriate for developmental stage of child. Check patients and where appropriate parents and child’s understanding. Avoid jargon. Appropriate use of interpreters As above with emphasis on the avoidance of jargon. Encourage questions. Appropriate manner, avoiding undue optimism or pessimism



Demonstrate empathy sensitivity and honesty



Observed involvement in the clinical setting. Might include attendance at taught course. Information forwarded to trainer’s report



To outline clinical case concisely. Identifying patients’ and parents’ concerns



Be aware of who requires information and expected level of detail. The contribution others can make to clinical decisions Demonstrate honesty and sensitivity in dealing with complaints



Observed practice in clinical setting. Information forwarded to trainer’s report



With colleagues



Complaints



Deal with complaints in oral and written communication. Contribute where required to local complaints procedure



in detail in Good medical practice and Good surgical practice. The key points are:  ‘you must maintain the standard of your performance by keeping your knowledge and skills up-to-date throughout your working life’;1  ‘all surgeons must maintain their knowledge base and performance by fulfilling the continuing professional development (CPD) requirements of the Senate of Surgery and registering this with their college’.4 There are European and UK policy and frameworks for CPD.13, 14 The Senate of Surgery expects that all surgical



Observed practice and review of complaints with trainee as part of appraisal process



staff, other than those in training, will participate in CPD. This includes substantive and locum consultants, associate specialists and staff grades.



Requirements for continuing professional development The minimum requirement for CPD in the UK is 50 credits per annum. One credit represents one hour of activity. CPD is viewed over a five-year cycle. Full- and part-time practitioners require full CPD (not pro rata).



Chapter 44 Training, accreditation and the maintenance of skills Table 44.2



] 555



Management of common paediatric conditions.



Subject/topic



Neck masses in infants and children



Knowledge



Clinical skills



Relevant professional skills



Assessment



The embryology and anatomy of the neck with particular reference to the thyroid gland and branchial arch structures4



Elicit clinical history and signs. Be able to deduce a differential diagnosis and formulate strategy to reach final diagnosis Able to examine the oral cavity, oropharynx and neck4 Be able to perform excision of thyroglossal and branchial cleft cysts competently4



Be able to advise the patient and parents of the treatment options4



Knowledge-based assessment



Understanding of the classification of neck masses4 Awareness of the various hypotheses relating to the aetiology of neck masses4 Understand the principles of the medical and surgical management of neck masses and their complications4



Recognize rarer causes of neck masses, such as first arch abnormalities, tumours, e.g. lymphomas, granulomatous conditions and lymphangiomas3



Clinical assessment in out patients



Discuss risks and potential benefits, potential complications and obtain informed consent4 Be aware of specific issues that relate to the management of children in hospital4



Supervised operating, including pre- and postoperative patient management Evidence of attendance on course dealing with issues relating to the surgical child



To understand the aetiology, presenting signs, symptoms and management of common paediatric ORL conditions. This module gives some indication of the breadth and depth of required knowledge and surgical skills. The list should not be considered to be fully inclusive or exhaustive.



A number of categories of CPD exist, the principle ones being internal and external. Practitioners need to participate in a mix of activities, such as outlined below:  25 external credits: for example, attendance at courses or conferences for which CPD credits have been awarded;  25 internal credits: – 10 hospital-based activities: for example, teaching rounds, pathology and radiology meetings; – 10 independent study: reading journals, writing papers; – 5 audit: contributing to and attendance at meetings. At the time of writing, no CPD return has been checked, validated or audited. There is no policing of the system. The number of returns completed overall has been unsatisfactory, 63 percent in 2000. ENT surgeons have done better with returns of approximately 70 percent. Of the returns received by the Surgical Royal Colleges, little more than a half meet all the requirements.



Table 44.3 indicates the categories and amount of continuing medical education (CME) required by the Surgical Royal Colleges. Table 44.4 summarizes the CPD returns held by the Royal College of Surgeons of England in ENT and surgeons in general.



Accreditation of continuing professional development points The system for conferring CME credits is as follows. Applications are made to either the Surgical Royal College or specialist association. The application form requires basic details of the course or conference content. There is an obligation on the course organizer to keep an attendance register and to collect feedback from the participants. Credits are awarded, on behalf of the British Association of Otolaryngologists – Head and Neck Surgeons, by the Chair of the Education and Training Committee. Feedback from the course is then collected. Despite the current lack of supervision and enforcement,



556 ] PART 9 SAFE AND EFFECTIVE PRACTICE Table 44.3 Annual continual professional development (CPD) requirements. Division of CPD activity Internal CPD, hospital-based A Postgraduate meetings B Research meetings C Journal clubs D Other CPD Internal CPD E Independent study Internal CPD F Clinical audit External CPD, meetings and courses G External meetings H I J K



Points awarded



Annual requirement



Minimum 10 CPD credits



Minimum 10 CPD credits Minimum 5 CPD credits Minimum 25 CPD credits



External courses Distance learning Postgraduate examinations work Visiting another unit



Internal CPD (A–F) = 25 CPD credits, external CPD (G–K) = 25 CPD credits. Fifty credits per annum are in line with European and US norms. However, it is unclear why this number was chosen. In educational terms, the quality and appropriateness of CPD is considered more important than the quantity.



Table 44.4 Continuing professional development participation 2000–2001.



No. surgeons registered for CPD All ENT Percentage of returns received All ENT Percentage of those returned who achieved (all) Minimum CPD requirements for ENT



Table 44.5 CME points awarded by BAO-HNS for applications received between October 2003 and April 2004.



2000



2001



5342.0 579



5888.0 573



63 71 46



60 72 46



52



48



Note. As of February 2003, there are 719 ENT surgeons registered for CPD with the Senate of Surgery of Great Britain and Ireland, 80% of these are consultant ENT surgeons, 20% are non-consultant career-grade (NCCG) doctors.



all surgeons would be well advised to ensure that their CPD affairs are in order. These credits are reviewed at annual appraisals for National Health Service (NHS) consultants. A review of applications for CPD credits to the British Association of Otolaryngologists over a six-month period is shown in Table 44.5.



Number Number Number Average



of events of days of points awarded points per day



32 73 375 5.14



GUIDANCE FOR SURGEONS ON APPRAISAL AND REVALIDATION: SPECIALTY-SPECIFIC GUIDANCE FOR EAR, NOSE AND THROAT SURGEONS The Joint Committee on Continuing Professional Development, chaired by Professor Graham T. Layer (reporting to Senate) is reviewing the system of CPD. New methods of registration and monitoring are likely. As part of this overhaul, specialty-specific guidelines are being drawn up. These specialty-specific guidelines should be read in conjunction with the parent generic document.



Good clinical care DEFINITION



The ENT surgeon possesses a defined body of knowledge and skills within the area of the head and neck. These are used to collect and interpret data, make appropriate clinical decisions and carry out diagnostic and therapeutic procedures within the boundaries of their expertise. Their care is characterized by up-to-date, ethical clinical practice and effective communication with patients, other health workers and the community. INDIVIDUAL STANDARDS



 Demonstrate diagnostic and therapeutic skills for ethical and effective patient care in the area of the head and neck.  Access and apply relevant information to clinical practice.  Demonstrate effective consultation services with respect to patient care and education.  Establish therapeutic relationship with patients/ families. PRESENTING YOUR EVIDENCE



 Demonstrate awareness and application of good practice guidelines such as those produced by the British Association of Otorhinolaryngologists – Head and Neck Surgeons (BAO-HNS) (vestibular schwannoma, head and neck cancer documents



Chapter 44 Training, accreditation and the maintenance of skills



] 557



 Demonstrate an understanding of preferred learning methods in dealing with students and trainees.



available on www.entuk.org) and others, e.g. Good practice in cosmetic surgery.  Proof of involvement in audit, including national audit, if appropriate, e.g. National Sino-nasal Audit, National Prospective Tonsillectomy Audit.



DEMONSTRATING YOUR EVIDENCE



Maintaining good surgical practice



 Summary of formal teaching activities.  Results of feedback from trainees on the effectiveness of teaching/training.



DEFINITION



The ENT surgeon recognizes the need to constantly review activity and ensure that part of their regular CPD focusses on the generality of ENT, as well as on their own area of subspecialization. INDIVIDUAL STANDARDS



 Attendance at annual general meeting of BAO-HNS (or equivalent body) at least every other year (home or abroad).  Attendance at subspecialty professional surgical meeting at least every other year (home or abroad).  Attendance at practical skills revision workshop, perhaps every five years.  Involvement in regular audit meetings on a local/ regional basis with involvement in national audit/ attendance at national audit meeting every five years.  If involved in teaching, attendance at ‘Training the trainers’ or equivalent course.  Attendance at a critical appraisal course or equivalent.  Regular review of a relevant journal, e.g. Clinical Otolaryngology, Journal of Laryngology and Otology or equivalent journal. DEMONSTRATING YOUR EVIDENCE



   



Certificates of attendance at courses. Copies of published articles, book chapters, etc. Guidelines written. Logbooks.



Relationship with patients DEFINITION



Communication skills are essential for the functioning of an ENT surgeon and are necessary for obtaining information from, and conveying information to, patients and their families. INDIVIDUAL STANDARDS



 Ability to establish and maintain rapport with patient characterized by understanding, trust, empathy and confidentiality.  Ability to inform and counsel a patient in a sensitive and respectful manner, while fostering understanding, discussion and the patient’s active participation in decisions about their care.



DEMONSTRATING YOUR EVIDENCE



 Evidence of involvement of patient groups when composing guidelines/information leaflets.  Summary of complaints/letters of thanks.



Working with colleagues DEFINITION



It is essential for an ENT surgeon to be able to collaborate effectively with a multidisciplinary team of expert health professionals for provision of optimal patient care.



Teaching/training DEFINITION



ENT surgeons have a responsibility for the training of students, trainees and other members of the healthcare team (including staff working in audiology, speech and language departments).



INDIVIDUAL STANDARDS



 Help others define learning needs and directions for development.  Provide constructive feedback.



INDIVIDUAL STANDARDS



 A collaborative approach includes the ability to recognize the limits of personal expertise, understand the roles and expertise of other individuals involved and explicitly integrate the opinions of those individuals into the management plan.  Contributing effectively to other interdisciplinary activities implies the ability to recognize team members’ areas of expertise, respect the opinions and roles of individual team members, contribute to healthy team development and conflict resolution, and contribute his/her own expertise to the team’s task.



558 ] PART 9 SAFE AND EFFECTIVE PRACTICE DEMONSTRATING YOUR EVIDENCE



 Evidence of involvement in multidisciplinary team meetings/clinics, e.g. head and neck oncology clinics, voice clinics, etc.  Evidence of liaising with colleagues regarding emergency on-call cover, cover for absence, etc.



Probity, health and managerial See generic continuing professional development document.



Future directions for training and revalidation Surgery and its practice are evolving. The needs of society are changing and the pace of this change is accelerating. The modern surgeon has to respond to this situation. The demands on trainees and established surgeons will continue to grow. Resources will always be restricted. We will have to train the next generation of surgeons in considerably less time than at present. Ahead lies a simple choice between depth and breadth of training. Few will ever be able to acquire the full range of skills and experience of their predecessors. Those embarking on a surgical career will become ‘specialists’ in a shorter time. They are likely to have a proscribed range of skills and will from time to time need to demonstrate their on-going competency. Assessment of competence and demonstration of on-going performance requires considerable effort. One needs to determine precisely what is to be tested, devise ways to do so, ensure consistent assessment and determine appropriate standards.15 Revalidation of specialists and consultants may take many formats: bench top models, observation of practice, formal examinations or peer reviews. Whatever method is used, most surgeons will ‘pass’ the test. If they do not, then there is either something wrong with the test or the service will collapse. The General Medical Council is currently considering what form revalidation will take.16 In the future, a doctor will need to be registered, but will also need a licence to practise. The licence will be subject to revalidation. Revalidation will be possible by two routes, an appraisal route open to those working in a hospital or academic environment and an independent route in which the practitioner will have to demonstrate evidence of adopting the standards of Good medical practice and undertaking continuing medical and



professional development. Revalidation is due to commence in 2005.



REFERENCES 1. General Medical Council. Good medical practice. London: General Medical Council, May, 2001. 2. The Shipman Inquiry. Chair Dame Janet Smith. First report, July, 2002. 3. The Inquiry into the Management of Care of Children Receiving Complex Heart Surgery at the Bristol Royal Infirmary. Chair Professor Ian Kennedy, July, 2002. 4. The Royal College of Surgeons of England. Good surgical practice. London: The Royal College of Surgeons of England, September, 2002. 5. Skills for the New Millennium: Report of the Societal Needs Working Group. The Royal College of Physicians and Surgeons of Canada’s Canadian Medical Education Directions for Specialists 2000 Project. September, 1996. 6. Consultant Surgeons. Teamworking in surgical practice. The Senate of Surgery of Great Britain and Ireland, May, 2000. 7. European Working Time Directive, No. 93/104/EC. Council of the European Union. 8. Department of Health. New deal for junior doctors. London: Department for Health, 1991. 9. Department of Health. A guide to specialist training. London: Department of Health, 1993. 10. Department of Health. Unfinished business: proposals for the reform of the SHO grade. London: Department of Health, 2002. 11. Department of Health. Modernising medical careers. London: Department of Health, 2003. 12. ENT Curriculum Document. Work in progress by SAC in otolaryngology. Chair: Mr Kevin Gibbin, 2004. 13. Academy of Medical Royal Colleges. A framework for continuing professional development. London: Academy of Medical Royal Colleges, February, 2002. 14. European Union of Medical Specialists. Basel declaration and policy on continuing professional development. UEMS (European Union of Medical Specialists), October, 2001. 15. The Royal College of Surgeons of England and Smith and Nephew Foundation. Surgical competence, challenges of assessment in training and practice. London: The Royal College of Surgeons of England, 1999. 16. General Medical Council. A licence to practise and revalidation. London: The General Medical Council, 2003.



45 Communication and the medical consultation DAMIAN GARDNER-THORPE AND RICHARD CANTER



Introduction The biomedical approach to patient care The patient-centred approach Core communication skills



559 559 560 564



INTRODUCTION Every day of our working lives we rely upon communication skills to build up relationships with our patients and colleagues. These core clinical skills are undeniably important in establishing trust and rapport with those who come to us for help and advice. To understand the increasing importance of communication skills, it is helpful to consider the way in which healthcare delivery has altered within western culture. The traditional biomedical approach to healthcare, which defines illness largely in terms of abnormal anatomy and physiology, has become outmoded because of its failure to recognize the full emotional cost of illness and the wider effects of disease upon the individual. The patient-centred approach to healthcare tackles the problem by focussing upon the patient’s perspective. There is, of course, an additional reason why the otolaryngologist takes a particular interest in communication issues. Whilst most of us are able to take our primary senses for granted, a proportion of our patients live with significant hearing loss or speech impairment, and this imparts upon us a special responsibility to remain sensitive to these differences without underestimating the ability of our patients to adapt to their unique environment. Just as our patients experience frustration in some clinical encounters, we too become aware of communication difficulties when our usual channels of conversation are blocked. To meet the challenge, we must acknowledge and accept our role in difficult encounters and devise strategies to meet them.1 In this chapter, we discuss the traditional biomedical approach, its limitations in maintaining a balanced consultation and the concept of patient-centred care. We hope to explore aspects of the consultation, which



Conclusions Key points References



566 566 567



present particular challenges, and to understand the patient’s perspective to generate a better awareness of our own behaviour in difficult clinical encounters.



THE BIOMEDICAL APPROACH TO PATIENT CARE Over 50 years ago, Parsons examined the relationship between doctors and patients in order to describe their roles within society in the context of the prevailing attitudes towards healthcare.2 Most patients accepted their responsibility to overcome illness as soon as possible, to seek professional advice from their doctor and to cooperate willingly with any advice offered. The doctor was expected to apply his or her knowledge and specialist skills to affect a cure. The doctor had a duty to act in the interests of the patient before his or her own and to adopt an emotionally detached and professional manner. In return for this, he or she was granted the right to enquire into intimate areas of personal life, to perform physical examinations and to assume a position of authority in relation to the patient. Whilst these values are still recognizable, the majority of patients now expect a very different style of healthcare and doctors have learnt to modify their traditional practice to meet an increased demand for a mutualistic approach to the consultation. Intrigued by the changing face of healthcare, Byrne and Long3 tape-recorded over 2500 general practice consultations and analysed the behaviour of the doctors and the patients. They identified two extreme consultation styles, which they labelled ‘patient-centred’ and ‘doctor-centred’. The patient-centred approach avoids an



560 ] PART 9 SAFE AND EFFECTIVE PRACTICE authoritarian style and encourages the patient to take a pro-active part in the consultation. Key aspects of this include listening, moments of reflection, facilitation and the use of silence, but were found in no more than one quarter of the taped consultations. One characteristic of a doctor-centred relationship is the frequency with which the doctor will interrupt a patient’s account of their illness. In a patient-centred encounter, the patient is encouraged to participate actively in the consultation whilst the doctor provides ‘expert’ knowledge, contextualizes appropriate management options and facilitates decision-making. Other investigators have identified more sophisticated models of doctor–patient experiences. One such model describes four categories of interaction reflecting the balance of power that exists within a consultation. These categories are paternalism, consumerism, mutuality and default (Figure 45.1).4 Paternalism describes a traditional model of a passive patient and a dominant physician. This relationship is generally outmoded, but can offer significant support to frightened or anxious patients who derive comfort in such a traditional doctor–parent figure. Paternalism is especially important when patients are sick and vulnerable. It is of interest that when they are ill, physicians themselves prefer to hand over responsibility for their care to their colleagues and are most comfortable with a paternalistic model.5 However, they have usually chosen their attending physician with care. A consumerist approach is one where the power relationship is reversed and the patient takes a more proactive role in the consultation. The doctor adopts a passive role and accedes to the patient’s requests, for example for a prescription, procedure or second opinion. Although the model at first sight espouses patient control, it may also reveal situations where the doctor is reliant upon the patient for goodwill or financial settlement. There are some groups in society, including the young, the better educated and those with previous experience of unhappy consultations, who are more likely to confront the doctor’s traditional authority and expect a consumerist approach.



Low physician control



High physician control



Low patient control



Default



Paternalism



High patient control



Consumerism



Mutuality



Figure 45.1 Consultation styles resulting from the level of interaction between the patient and the doctor. Redrawn from Ref. 4, with permission from Sage Publications.



A default situation occurs when neither doctor nor patient takes a lead in the consultation. This situation arises when the patient adopts an overly passive style and the doctor fails to assume control. This inevitably leads to a lack of direction and purpose in the encounter. Mutuality exists when the physician brings their experience and expertise to the consultation, the patient brings his or her expectations, feelings and symptoms, and both parties participate in a joint venture to explore the possible outcomes and treatment options. Mutuality succeeds because of shared decision-making responsibility and because of a willingness of both parties to adhere to any agreed plan. The style has been described as a ‘contractual’ approach to patient care because within this arrangement the patient is free to seek care elsewhere when his or her expectations are not satisfied, and the doctor may withdraw from the relationship when patient requests are considered inappropriate, ethically untenable or impossible to meet. Of course it is possible, indeed likely, that more than one model will be employed during a course of treatment. At the outset, when investigations need to be performed and the diagnosis made, a paternalistic model may be appropriate. This is particularly the case when patients are anxious, for example if there is the possibility of cancer. Later, when treatment options are discussed, a model of consumerism or mutuality would be more appropriate. There is a case for negotiating these different models of consultation style with the patient. The physician may therefore suggest that to begin with he or she ‘takes over the running of the care’, but later when treatment options are to be discussed, he or she will switch to a more mutualistic consultation style.



THE PATIENT-CENTRED APPROACH There are several aspects of the consultation which we might wish to consider in the context of the evolution of the patient-centred approach. These include issues of power, an understanding of basic ethical principles and the factors which determine compliance. A basic understanding of these may help to understand why communication sometimes fails.



Issues of power Doctors, perhaps subconsciously, control the behaviour of their patients. Sometimes the reverse is also true. The balance of ‘power’ depends upon several patient factors including the extent of the patient’s medical knowledge and their capacity to choose between options. Nevertheless, an appreciation of power dynamics helps avoid the risk of applying unintentional or unjustified influence on our patients.



Chapter 45 Communication and the medical consultation



To understand the role of power in the clinical encounter, let us define power as the means by which A gets B to do something, where A might be the doctor and B might be the patient. Power is exerted upon patients in several ways. The most obvious form is expert power, in which a doctor exerts influence by virtue of his or her possession of specialist knowledge. Expert power is rarely used alone, but is instead exercised in conjunction with other forms of power including power of legitimacy (power devolved to a doctor from the Royal Colleges or the General Medical Council), coercive power, charismatic power and of course the offer of rewards, including relief of symptoms.6 Most clinicians exercise these elements of power to suit their consultation style and to achieve a degree of patient compliance. Nevertheless, this is a simple model of power and it is not that helpful in understanding power in the clinical encounter. Lukes7 produced a generalized model of power that does shed light on the clinical situation. He saw power as operating in three dimensions: first dimensional power is when A forces B to do something, the second is that in which A controls the agenda in any interaction with B and the third dimension, in which A controls the world view of B. So how does this power model apply to the clinical encounter? First dimensional power is analogous to coercive power. It might be exercised appropriately in emergency situations, in certain acute psychiatric states or when patients are extremely distressed or anxious and do not wish to debate alternative choices. First dimensional power is frequently exercised in the emergency arena, for example, when a stridulous patient is told that he needs a tracheostomy or a patient with a brisk epistaxis is told that he must undergo immediate nasal packing, even if the treatment itself causes agitation or distress. Second dimensional power is exercised when the doctor deliberately steers conversation towards or away from certain topics in order to influence final outcome. Patient choice will be biased because of a selective presentation of information. This may happen when consultation time is limited or when doctors make it difficult for patients to ask questions in clinic. Second dimensional power may also result from the increasing reliance upon guidelines and protocols which dictate outcome goals without offering any flexibility for alternative options within the process of the consultation. It also includes the power to silence or make it difficult for the patient to ask certain questions. Third dimensional power is harder to appreciate. It manifests itself when a doctor constructs for his or her patient a world view of disease and treatment. The patient believes he or she moves autonomously, but is in fact under the influence of the flow of medical knowledge as presented by the doctor. Consider for a moment the situation faced by a patient with malignant disease of the



] 561



larynx, who has to make choices about the treatment options of palliation, radiotherapy, chemotherapy, surgery or various combinations of each. The encounter is framed within a definition of malignant disease in a conventional biomedical model. A discussion follows, looking at the effectiveness of each of these treatment options, based upon the physician’s understanding of results reported in the medical literature together with his or her experience of previous cases. The patient is apparently free to choose, but these choices have been shaped and polished by the manner in which they have been presented. The biomedical world view can be demonstrated quite easily, because within a conventional biomedical framework model there would be no place for herbal medicine, acupuncture, osteopathy or any other of the many complementary medical alternatives. Indeed, when these issues are raised within a conventional medical consultation, they are frequently problematic.



An ethical framework An ethical framework is required to place the patient’s interests at the centre of the decision-making process. Because of the imbalances that exist between doctors and patients, not only in terms of power but also in terms of knowledge and skills, there is a very real benefit in having a set of principles, sanctioned by society, to guide the doctor in making the best decisions with or on behalf of his or her patients. Such ethical principles have existed since Hippocratic times and have evolved through history in response to novel treatments, new dilemmas and changing attitudes within society. The most widely employed ethical framework is the four principle framework, or Principlism. It comprises a simple, logical and culturally sensitive approach to ethical problems and is useful for most clinical situations where there is a difficult conflict of interest. The four components are autonomy, justice, beneficence and nonmaleficence.8 The first of these, autonomy, is the capacity to think and to have freedom to act upon these free thoughts without let or hindrance. Autonomy empowers patients to make decisions for themselves. Full self-determination is not possible within the constraints of a civilized society as we cannot each expect to live according to our own personal rules. Instead we must act within the rules of society and exercise our right to autonomy within cultural boundaries and the laws of the land. In the context of the medical consultation, the individual’s right to autonomy should be respected as long as it does not adversely affect other members of the community. This means that the patient has the absolute right, in almost all situations, to choose their preferred treatment and be free to reject any intervention without prejudice. There are, of course, situations where the patient delegates his or her autonomy to the doctor – ‘You decide for me,



562 ] PART 9 SAFE AND EFFECTIVE PRACTICE doctor’ – in which case the doctor must defer to the basic Hippocratic principles of beneficence (doing what is good) and nonmaleficence (not doing that which is harmful). Situations exist where the doctor bypasses the principle of autonomy. Usually this results from the patient’s lack of capacity to exercise autonomy resulting from immaturity, disability or disease, but it also happens when the individual’s actions impinge upon the well-being of others and in this instance the doctor must consider the rules of justice. Nevertheless, it follows that we need to be honest and open with our patients, consult them before we do things to them, keep patient information confidentially and respect a patient’s right to choose what is and what is not done to them whenever possible. Beneficence and nonmaleficence are the oldest two principles. Beneficence is the duty to do that which is best for the patient – even if this is at some personal cost to the doctor. Beneficent traits include basic qualities of a doctor including friendliness, good manners and good time-keeping. Nonmaleficence is the duty to do no harm. However, it is all too easy to envisage a situation where an intervention has the potential to improve an outcome, but only with the risk of serious side effects. We are responsible for considering the best option for our patients, ‘all things considered’. We run into difficulties, especially in the modern technological world of medicine, when we cause harm to our patients through an intention to do well. Although the principle of nonmaleficence tempers us in our decisionmaking process, we must try to avoid the tendency towards inappropriate paternalism in deciding the best management option for our patients – to do so takes choice away from patients and risks a breach in the duty of autonomy. Consider this in the context of another, perhaps more traditional approach where arrogance is held as a virtue, preferable indeed to the virtue of humility: ‘if you agree that the physician’s primary function is to make the patient feel better, a certain amount of authoritarianism, paternalism and domination are the essence of the physician’s effectiveness’.9 Justice, or a ‘just’ distribution of health care resource focusses upon making fair decisions when there are competing claims. Justice is particularly important in rationed health care systems where available services are shared out among many individuals. Doctors are sometimes responsible for deciding how to share out these resources, which include their time and skills, as well as expensive investigations and treatments, and this may not necessarily mean being even-handed to every patient.



Compliance Compliance has been defined as the ‘extent to which the patient’s behaviour coincides with medical or other health care advice’ and applies equally to lifestyle advice,



accepting prescription medication and following advice to undergo surgery.10 Amongst doctors, there is perhaps a naive expectation that patients should comply with their advice, though in practice it is hard to determine the reasons why patients do not act upon guidance. Not surprisingly, there is increasing interest in the factors which underpin compliance. The three prime factors are believed to be a good understanding of advice, a high level of satisfaction and an ability to recall information handed over by the clinician. While it is clear that compliance depends upon a good level of understanding between doctor and patient, there is a tendency for health care providers to overestimate the patient’s medical knowledge. When a group of patients were asked to describe the approximate position of vital organs within the body, fewer than one-fifth were able to locate the stomach and fewer than half knew where to find the heart or liver.11, 12 Medical jargon is often confusing and despite our best intentions we frequently lapse into a technical vocabulary which patients cannot comprehend. The word vertigo means very different things to different individuals and the ear, nose and throat (ENT) specialist will also be aware of the imprecise usage of this term throughout the profession. We need to be sure that patients understand our definition of terms and that we appreciate their use of descriptive language too. One straightforward way to check the level of this understanding is to ask patients to repeat back in simple terms their interpretation of the situation. Some expressions run the risk of trivializing the serious nature of a medical problem, for example ‘just pop on the couch’ or ‘we’ll just have a quick look at your ear’. Nevertheless, we need to remain sensitive to the patient’s level of understanding and contextualize our language according to patients’ individual needs. There are circumstances where rapport is strengthened by setting out a clear agenda at the start of a consultation, by explaining, for example that we may interrupt in order to focus on to particular aspects of a history. An example of this might be during a consultation for ‘dizziness’ where a fastidious approach to history taking is required to reach a reliable differential diagnosis. The capacity for information recall varies greatly between individuals. The factors which influence recall include the volume and complexity of the message, as well as the impact that it has upon the individual. Evidence from primary care suggests that surprisingly little factual information is retained by the patient following the end of the consultation. For example, in one study over a third of patients failed to recall the name of their prescription medicine and a quarter failed to appreciate the duration of the course of their medication.13 Given the particular importance of accurate information transfer in surgical specialties, one approach to reinforce handover of knowledge is to dictate the clinic letter in the presence of the patient, in plain and unambiguous English, and to offer the patient a copy for his or her own records.



Chapter 45 Communication and the medical consultation



A number of interesting cognitive factors explain how patients behave in particular health settings. Heuristics are ‘rules of thumb’ which explain patterns of behaviour, for example in relation to perception of risk or threat from disease. Weinstein14 coined the term ‘unrealistic optimism’ to account for the observation that individuals tend to underestimate the risk of a threat to which they are specifically vulnerable. He tested the popular belief that people think themselves invulnerable, by demonstrating in a group of college students a systemic error in estimating personal risk from negative and positive threats. For example, the students overestimated their chances of enjoying their first job or owning their own home, but underestimated their risk of contracting a venereal disease, having a drink problem or getting lung cancer. Their misguided perceptions result from a complex interplay of factors including differing egocentric biases, personal experiences and stereotyped beliefs. Other heuristics influence susceptibility judgements; for example, the tendency for individuals to overestimate their personal risk from less likely causes of death, but to underestimate the risk from more common ones.15 An appreciation of risk perception is especially pertinent to those taking informed consent for surgery. Simple and effective techniques for increasing compliance include provision of written information, keeping explanations simple, repeating important points and the judicious use of follow-up appointments. We should remember too that patients tend to recall the first and last pieces of information that they receive – the primacy and recency effects.12, 16



Consultation models In accepting the validity of the patient-centred approach, we can consider basic consultation theories which help explain the decision-making process from the patient’s perspective. A number of these models overlap the field of health psychology and explore the role of the patient’s motivation, health beliefs and the factors which underpin compliance. Within the traditional paternalistic encounter, the consultation essentially involves the transfer of knowledge from expert to lay person. Whilst there may be many straightforward medical complaints where psychological and emotional components are less important, it is increasingly recognized that patient compliance and The doctor establishes a relationship with the patient



The doctor defines the reason for attendance



The doctor performs a physical or verbal examination



] 563



satisfaction improve when the doctor contextualizes the disease within a more holistic approach. Byrne and Long3 outlined the essence of the doctorcentred relationship, largely analogous to the ‘paternalistic’ approach (Figure 45.2). Pendleton et al.17 devised a patient-centred model which encourages the doctor to explore the patient’s ideas, concerns and expectations and to frame the consultation in a problem-solving approach. One vital tenet of Pendleton’s model is the maintenance of a healthy doctor–patient relationship:  define the reasons for the patient’s attendance;  consider other problems (including risk factors for current complaint);  negotiate the correct course of management of each of the presenting problems;  achieve a shared understanding of the problem – both doctor and patient;  involve the patient in management and encourage patient responsibility for care;  use resources and time efficiently;  establish and maintain a healthy doctor–patient relationship. Neighbour,18 in perhaps the most widely used patientcentred model, espouses an appreciation of subtle clues in the consultation, verbal and nonverbal. The model establishes good doctor–patient rapport and helps the clinician identify and address underlying concerns. This elegant model uniquely recognizes the importance of the role of the doctor and reminds the clinician of his personal risk of stress, anxiety and burn-out:  connecting: where the doctor establishes rapport;  summarizing: clarification of the patient’s reason for attendance;  handing over: joint negotiation of the correct course of management;  safety netting: managing uncertainty and planning for the unexpected;  housekeeping: staying in touch with emotions and awareness of burn-out. While none of these consultation models provides a comprehensive or universally reliable guide to a successful outcome, each has its place in dealing with difficult clinical encounters. Through the evolution of consultation modelling there has been a gradual shift towards a patient-centred emphasis which is likely to be The doctor, with or without the patient, considers the diagnosis



The doctor outlines treatment or investigation



Figure 45.2 The traditional doctor-centred model. Redrawn from Ref. 3, with permission from HMSO.



The consultation is terminated, usually by the doctor



564 ] PART 9 SAFE AND EFFECTIVE PRACTICE the trend, at least for the current generation of healthcare providers.



CORE COMMUNICATION SKILLS (a)



In an encounter between two individuals, it is said that less than 35 percent of the contact is established through verbal means, whilst 65 percent or more is conveyed through nonverbal cues including facial expression and eye contact.19 Patients are particularly sensitive to nonverbal cues from their doctors, but doctors are often unaware of their influence in this respect. It is important to recognize the importance of subtle behavioural cues which are highly influential in determining the direction of the consultation and the subsequent outcome. It is obvious that our patients are likely to prefer a more relaxed consulting environment. Since nonverbal methods of communication include the use of eye contact and body language, we may consider how these are influential in the consulting room. These factors include the layout of furniture, the relative positions of doctor and patient and the distance between them. A relaxed interaction is more likely when chairs are positioned at the same height to ensure good eye contact. A desk represents a potential barrier between the doctor and patient and may be an important factor in limiting the free flow of conversation. In one interesting experiment, the influence of the desk was analysed during consultations between a cardiologist and his patients. On alternate days the cardiologist removed the desk from his consulting room to assess the impact upon nonverbal cues and body language. When the desk was absent, more than half of all patients sat back and relaxed, but when the doctor was positioned behind the desk fewer than one in ten assumed a relaxed body posture and this corresponded with a drop in the level of meaningful interaction between the two.19 Knapp20 looked into the importance of seating configurations in more detail, not only in the context of the medical encounter but also for its significance in business meetings and any other number of social occasions. He identified four independent seating positions for any interaction: the corner position, the cooperative position, the competitive-defensive position and the independent position (Figure 45.3). In the corner position, two individuals are positioned in such a way that the corner of the desk intervenes. It usually results in casual friendly conversation. It allows unlimited eye contact and a full appreciation of gesture. The corner of the desk behaves as a partial barrier but does not intrude as a territorial division. The cooperative position differs from the corner position by the absence of any barrier and a straight face-to-face seating posture. Whilst this favours a mutualistic relationship, a number of individuals feel threatened by the lack of a firm boundary.



(b)



(c)



(d)



Figure 45.3 Examples of seat and desk arrangements. (a) The corner position; (b) The cooperative position; (c) The competitive–defensive position (the territorial line); (d) The independent position. Redrawn from Ref. 20, with permission.



In the competitive-defensive position, the desk constitutes a firm physical barrier. Not only does this prevent intimacy, but it may create competition for control of the consultation. An imaginary territorial line is evident between the doctor and patient. In contradistinction, the independent position is adopted by those who want to avoid interaction, just as people do when they sit in a library or on a train. This style would clearly be alien to a sensitive medical encounter and cannot be recommended! Listening is of course a core communication skill in clinical medicine, though little attention is paid to its full importance. Burnard21 describes three levels of listening. The first is at the linguistic level, comprising an understanding of words, phrases and metaphors. The second level is paralinguistic, where the listener takes cues from timing, volume, pitch and accent including the use of ‘umms’ and ‘ahhs’ in the fluency of speech. The third level is an appreciation of facial expression, gesture, touch, body position, proximity of doctor and patient, body movement and eye contact.



Chapter 45 Communication and the medical consultation



The first level involves the content of speech without any emotional connection. Second level skills allow the doctor to develop a ‘free floating’ attention, observing patient cues other than words alone. At the third level, the doctor maintains a ‘free floating’ attention whilst noticing paralinguistic features and nonverbal cues, as well as noting his own internal thoughts, sensations and feelings. Watkins22 describes this as ‘resonance’ between doctor and patient, whereby information determined from the history is weighted according to nonverbal cues and is characterized in such a way that any underlying anxieties are exposed and addressed. Eye contact serves many functions, and used appropriately, establishes trust, demonstrates integrity, controls the level of intimacy and assists in information transfer. The comfortable level of eye contact depends on the context of the consultation, the physical distance between the participants and their degree of familiarity. An appropriate distance between the doctor and patient is hard to gauge, but it has been established that as distance is reduced, many start to feel uncomfortable, and a distance of less than 30 cm produces tension in most individuals.23



Frustrating encounters Notwithstanding our best attempts to construct the ideal consultation, each and every doctor encounters patients from time to time who provoke an inner sense of frustration or irritation, and this may be for no discernable reason. These patients have been identified within the general practice setting as ‘heart-sink’ patients. While general practitioners are more likely to see their patients repeatedly, the same emotional response to patients is encountered by doctors within hospital practice, especially when patients return with new symptoms on repeated occasions or when they fail to respond to treatment. One interesting aspect of this problem is that a heart-sink patient to one is not necessarily a heart-sink to another. This is explained by understanding that frustration is produced by both doctor- and patient-related factors, and probably represents a generalized failure of the doctor–patient interaction at a fundamental level. Of the many factors involved in a heart-sink encounter, doctor factors not surprisingly include low job satisfaction, the perception of an excessive workload and a lack of training or experience in counselling or communication skills.24 The heart-sink may be avoided by improved communication strategies, including a commitment to the ongoing development of consultation skills and identification of situations which cause stress. A number of special skills exist to combat the heart-sink encounter including a ‘holding strategy’ for listening and understanding the patient’s problems without feeling the need to tackle all issues in one sitting.



] 565



Breaking bad news to patients Sharing bad news with our patients is a difficult task to do well, so it is a little surprising that so often this has been the responsibility of relatively inexperienced members of a surgical team. A number of simple techniques may help to undertake this more sensitively. Some appear obvious, such as planning the consultation in advance, checking the facts and avoiding interruptions to ensure privacy. Breaking bad news can take a significant amount of time and it is important to avoid being rushed. We need to use simple straightforward language and avoid jargon. Many patients value the invitation to bring a relative or close friend for support, though we should be clear in establishing the nature of the relationship in order to understand how the news will impact upon them as well. Nonverbal messages are as important as verbal ones. Patients need to be given sufficient time for silence, tears or anger. Even when the future looks bleak, be careful in using the phrase ‘nothing can be done’ since there are always ways of encouraging hope in any dire situation. Nevertheless, it is equally important to admit when we do not know the answers to difficult questions, such as life expectancy after a diagnosis of terminal disease. Most patients value the offer of ongoing support and follow up is especially valuable.



Dealing with loss Surgical patients inevitably experience concerns about the potential disfigurement and loss of function which result from our actions. We have a clear duty to address these issues in appropriate depth. Within our specialty, patients undergo radical neck surgery, lose the functions of speech, swallowing, balance and the ability to hear properly. Perhaps not surprisingly, those who cope less well with loss of body function are those with a history of premorbid anxiety or depression, and these are factors that are easy to overlook in a busy consultation.25 Consequently, a professional assessment of mental health may be a helpful part of preoperative planning for some of these patients prior to major surgery. The reaction to a loss of function depends upon many complex factors. A sudden unexpected loss may magnify shock and increase difficulty in coping, and our colleagues in primary care will testify to the longer-term consequences of mental health problems after serious disease. Doctors are advised to be honest and frank to their patients in the early stages of disease, especially when there is little hope for recovery. There are circumstances in which patients are disenfranchised from their normal pattern of grief and these include elements described as hidden and concealed losses. Hidden loss takes place when patients are unable to share their experiences with even close friends or relatives, perhaps because of the shame of human



566 ] PART 9 SAFE AND EFFECTIVE PRACTICE immunodeficiency virus (HIV) or because of religious or social stigma. Concealed loss happens when relatives or doctors fail to share information with patients because of a misplaced belief, often with good intentions, that the patient cannot cope with bad news. This rationale for concealment is usually poorly grounded and an opportunity to come to terms with reality may be missed. Many patients have a remarkable capacity to adapt to bad news and may become angry if they later learn that information has been withheld.26 Whether consciously or unconsciously, avoiding grief can lead to difficult outcomes both for doctors and for patients. While doctors are conditioned to remain calm in adversity, they often experience difficulty describing their emotions once a stressful situation has passed. Those who learn to express their emotions report greater satisfaction at work and presumably function better as doctors.26



Loss of hearing The limitations caused by hearing loss depend greatly on the age of onset and on the extent of any residual function. Prelingual deafness is that in which impairment of hearing predates language acquisition and by definition necessitates development of communication skills by alternative means. It is misleading to consider prelingual deafness a loss of function without considering the rich alternative language skills that arise through the use of sign language. The visual–spatial axis of sign language may be developed in the first few years of life, during the critical period of language assimilation, to the same extent as the more common linguistic route. The distinction between truly deaf individuals and those who become deafened later in life is important. Properly nurtured, the congenitally deaf individual is able to develop normal communication skills within the deaf community without any requirement for the spoken word, whereas the individual who encounters hearing loss after normal development suffers an immeasurable handicap without augmentation or alternative aids. THE DEAF COMMUNITY



Absence of hearing does not necessarily correspond to handicap. Deaf communities thrive throughout the world, enjoying their unique cultural existence with a vibrant and expressive language. Different forms of sign have developed across the world independent of spoken language. In 2003, the UK Government formally recognized British sign language (BSL) as a language in its own right, banishing the fallacy that it represents a rudimentary translation of spoken English. The ruling affords BSL the status of a native British language, equal to Welsh or Scots Gaelic as a minority language. As many as 70,000 people in the UK communicate in BSL, a rich and complex



language employing hand shapes, facial expressions and body language within defined spatial boundaries. To understand the importance of this landmark recognition, one must recall the bigotry and persecution experienced by generations of deaf individuals, many of whom were forced to learn to vocalize language and suppress their natural tendency to sign during the critical time of language acquisition. Our role as health care providers is to encourage and respect the autonomy of the deaf community and to assist with technological innovation or general professional services, when required. ACQUIRED HEARING LOSS



In routine practice we frequently encounter patients with acquired hearing loss. There are a number of simple methods for improving communication within the consultation. Slow, deliberate speech of adequate volume in a room with low ambient noise will assist those with a modest hearing loss. A face-to-face position and clear enunciation of words enables lip reading in those with longstanding loss. Other simple but underused communication aids include diagrams, models and picture boards.



CONCLUSIONS There has been a cultural shift in the way in which doctors and patients interact and this has resulted in the evolution of the patient-centred approach to health care. Despite this, there are many elements of the consultation which are difficult to recognize, including issues of empowerment. Since our prime responsibility is to act in our patients’ best interests, it is important for us to understand the ways in which we influence our patients, whether consciously or not. We also need to recognize our own feelings and anxieties when we deal with difficult situations and to appreciate the particular difficulties which arise when we deal with patients with sensory deficit. To learn good communication techniques takes time and experience and inviting feedback from patients should be encouraged rather than feared. It is a useful idea to encourage feedback from patients either during or at the end of the consultation. As the old saying goes ‘asking saves a lot of guesswork’.



KEY POINTS  The traditional approach to the consultation is being superseded by a more patient-centred ethos.  Power dynamics play an important but often unappreciated role in the doctor–patient encounter.



Chapter 45 Communication and the medical consultation



 The four-principle framework is comprised of beneficence, nonmaleficence, autonomy and justice.  Compliance is related to the level of patient understanding, patient satisfaction and recall.  Consultation models offer a technique for understanding the doctor–patient encounter.  Listening to patients is an underused clinical skill.  Most doctors experience heart-sink situations at some point – it is important to recognize the signs.  Remember the grief reaction: denial, anger, despair and acceptance.  Simple techniques, such as models and diagrams, are highly effective methods of communication.



REFERENCES







1. Fitzgerald RG, Parkes M. Coping with loss – blindness and loss of other sensory and cognitive functions. British Medical Journal. 1998; 316: 1160–3. 2. Parsons T. The social system. Glencoe, IL: Free Press, 1951. 3. Byrne PS, Long BEL. Doctors talking to patients. London: HMSO, 1976. 4. Stewart M, Roter D (eds). Communicating with medical patients. London: Sage, 1989. 5. Ende J, Kazis L, Ash A, Moskowitz MA. Measuring patients desire for autonomy: Decision making and information seeking preferences among medical patients. Journal of General Internal Medicine. 1989; 4: 23–30. 6. French JR, Raven B. The bases of social power. In: Cartwright D (eds). Studies in social power. Ann Arbor: University of Michigan Press, 1959: 150–67. 7. Lukes S. Power: A radical view. London: British Sociological Association, Macmillan Press, 1974. 8. Beauchamp TL, Childress JF. Principles of biomedical ethics, 5th edn. Oxford: Oxford University Press, 2001. 9. Ingelfinger FJ. Arrogance. New England Journal of Medicine. 1980; 303: 1507–11. 10. Haynes RB, Sackett DL, Taylor DW. Compliance in health care. Baltimore: Johns Hopkins University Press, 1979. 11. Ley P. Professional non-compliance: a neglected problem. British Journal of Clinical Psychology. 1981; 20: 151–4.







] 567



12. Ley P. Improving patients understanding, recall, satisfaction, and compliance. In: Broome A (ed.). Health psychology. London: Chapman and Hall, 1989. 13. Bain DJ. Patient knowledge and the content of the consultation in general practice. Medical Education. 1977; 5: 347–50. 14. Weinstein ND. Unrealistic optimism about future life events. Journal of Personality and Social Psychology. 1980; 39: 806–20. 15. Slovic P, Fischoff B, Lichtenstein S. Behavioural decision theory. Annual Review of Psychology. 1977; 28: 1–39. 16. Ley P, Morris LA. Psychological aspects of written information for patients. In: Rachman S (eds). Contributions to medical psychology. Oxford: Pergamon Press, 1984: 117–49. 17. Pendleton D, Schofield T, Tate P, Havelock P. The consultation: Developing doctor–patient communication. Oxford: Oxford University Press, 1984. 18. Neighbour R. The inner Consultation: How to develop an effective and intuitive consulting style. Lancaster: MTP Press, 1987. 19. Pietroni P. Language and communication in general practice. In: Tanner B (ed.). Communication in the general practice surgery. London: Hodder and Stoughton, 1976. 20. Knapp M. Non-verbal communication in human interaction, 2nd edn. New York: Holt, Rhinehart and Winston, 1978. 21. Burnard P. Effective communication skills for health professionals: Therapy in practice no. 28. London: Chapman and Hall, 1992. 22. Watkins J. The therapeutic self. New York: Human Sciences Press, 1978. 23. Argyle M, Dean J. Eye-contact, distance and affiliation. Sociometry. 1965; 28: 289–304. 24. Mathers N, Jones N, Hannay D. Heartsink patients: A study of their general practitioners. British Journal of General Practice. 1995; 45: 293–6. 25. Maguire P, Parkes CM. Coping with loss – surgery and loss of body parts. British Medical Journal. 1998; 316: 1086–8. 26. Parkes CM. Facing loss. British Medical Journal. 1998; 316: 1521–4. 27. Butler C, Evans M. The heart sink patient revisited. British Journal of General Practice. 1999; 49: 230–3. 28. Crichton EF, Smith DL, Demanuele F. Patients’ recall of medication information. Drug Intelligence and Clinical Pharmacy. 1978; 12: 591–9.



46 Clinical governance: Improving the quality of patient care DEBBIE WALL, PATRICK J BRADLEY AND AIDAN HALLIGAN



Introduction What is clinical governance? Why clinical governance? Setting standards for delivery on quality Patient and public involvement The NHS Clinical Governance Support Team Multidisciplinary teamworking



568 569 570 571 574 575 576



INTRODUCTION John Shaw is a 45-year-old ear, nose and throat (ENT) consultant and has spent the last eight years working in the ENT unit of a large London hospital. During that time, he has rarely questioned the way that he communicates with his National Health Service (NHS) patients or reflected on his attitude towards the other health care professionals who work with him. John’s behaviour at directorate meetings is proof positive that at least one of the consultants’ unwritten rules has survived the new millennium: ‘Turn up, do the work, don’t get noticed!’ After all, if John maintains his distance, does the job he’s always done and doesn’t stick his head above the parapet, he will avoid getting sucked into management issues. The multidisciplinary team (MDT) that John says he’s part of is ‘multidisciplinary’ and a ‘team’ in name only. In reality, team members are neither accountable to each other for their clinical practice nor do they work together in a multidisciplinary way. They are a group of individuals who share the occasional clinical meeting. The consultants themselves work in silos, in separate clinics, rarely coming together to coordinate patient care – although they often treat the same patients. If you asked John, he would not know whether his practice varied from Sarah’s across the corridor or whether he was making different decisions about treatment of the same patient.



Good practice case studies Conclusion Note Key points References Further reading



576 578 578 578 578 580



Recently, John had a ‘near miss’ when he divided a facial nerve trunk during a parotidectomy. Even though he had resutured the injury at the time, he felt that it was better that he kept this complication to himself. He knew from past experience that if he reported it, he would be blamed. Similarly, he did not want his mistakes exposed to colleagues through his department’s new system of multidisciplinary audit. After all, it may draw attention to deficiencies in his surgical experience and patient selection for surgery, and he would in all probability be held responsible. What’s more, audit threatened to interfere with his clinical work and generally took up too much of his time. Things had been like this in ENT as long as John could remember. The service was under pressure from all sides: long waiting lists, lack of funding, insufficient resources, including incompatible information technology systems and too few consultants to provide the modern service expected. As far as John Shaw is concerned, he does the best that he can for his patients under the circumstances. He does not hear nor want to hear of any complaints, so he assumes that his patients must be satisfied with their treatment and care. Mind you, he has never asked them. John feels that people and the system always want more from him, always want changes to the historic order of things. He has become more sceptical about anything labelled modernization or innovation, because he has seen it all before: too many initiatives come and go in the NHS.



Chapter 46 Clinical governance: Improving the quality of patient care



A few years ago John’s hospital was involved in a clinical governance review. The Commission for Health Improvement (CHI) undertook reviews of clinical governance arrangements in NHS organizations throughout England and Wales until April 2004, when it was replaced by the Commission for Healthcare Audit and Inspection (Healthcare Commission (HC)) (www.healthcarecommission.org.uk). Clinical governance is one initiative that has not come and gone. The clinical director continues to emphasize it within the department. What exactly is clinical governance to people like him working hands on with patients? The definition he has seen says that it is: A framework through which NHS organisations are accountable for continually improving the quality of their services and safeguarding high standards of care by creating an environment in which excellence in clinical care will flourish.1



But, what does that mean for the way that he works in the unit?



WHAT IS CLINICAL GOVERNANCE? ‘People can’t grasp that clinical governance is an integral part of your job – it’s just what you’re doing.’ These words from a clinician show an awareness that clinical governance is a unifying concept; it brings together the aspects of



] 569



care implicit in the work of any health professional or support worker who day-in, day-out strives to ensure that the service they offer patients is safe, and of the highest quality that they are able to provide. Clinical governance is not just about risk management. In an organization where it is fully integrated, there are not only effective processes for identifying and managing risk; there are clear lines of accountability; there is alignment of processes with systems and behaviours (Figure 46.1). In the clinical governance organization:  patients come first;  care is delivered with the highest standards of patient safety;  clinicians work collaboratively, in multidisciplinary teams;  there are clear lines of communication between staff;  staff, regardless of their position, feel valued, included, listened to and empowered;  there is regular clinical audit of services;  clinical decisions are supported by research evidence;  patients’ complaints and clinical errors are reviewed and acted upon;  there are professional development opportunities through education and training;  staff have access to current information technology systems;  data and information are used effectively to review service quality and support decisions and processes;  the organizational culture is open, questioning and blame-free;



Risk avoidance



Clear procedures ce an rm rfo n pe itio or gn co on Po i e t r n rly rve e Ea int elf ive es cis v De cti ion t e Eff ula on reg ck ba ance ed Fe rform pe



Safe environment



ss e ce Ac idenc ev ed to llow e a an s Tim pl d o e t i n g a ng trate s ini gy Tra ent olo m hn lop tec tice ve on rac p de ati s t re or m or Inf upp ctu s tru ras Inf



Open and participative Good leadership Education and research valued



G pra ood sp ctice rea Cli d ev nical i de p Le o nc ss e b licies on Im as sl pro ed e a ve rne me d nt fro pro m fai ce Qu lur ss e ali e s ty i n teg me rat tho ed ds



Go



Co he als ren of ce org ind an ivid iza ua tio l n a , tea lig m a co Exc ne mm ell d nd un ent i c at E pa xtern ions r tn al e for rsh ge ip d



Well-trained staff



Patient partnership Ethos of teamwork Culture



Figure 46.1 Integrating approaches of clinical governance. Redrawn with permission from Ref. 1.



570 ] PART 9 SAFE AND EFFECTIVE PRACTICE  clinical governance is high on the corporate agenda;  there is strong leadership throughout the organization. An organization where clinical governance is mainstream would be able to provide support for clinicians like John Shaw where it is needed. This would enable them to be more accountable for the care they deliver, to re-examine their ways of working and to reflect on whether changes in their attitude and behaviour would not only benefit their relationships with colleagues, but place patient rather than professional concerns at the centre of their thinking. However, developing an organizational culture, where staff feel they can freely report and learn from their mistakes, does not happen overnight. NHS trust chief executives and clinical governance leads have identified cultural difficulties as a major barrier impeding the successful implementation of clinical governance and there is much scope for improvement in communications between trust boards and clinical teams.2 For clinical governance to be fully integrated throughout local NHS organizations, such barriers need to be overcome, but this is not easy if individuals are defensive about their clinical performance and continue to relate to others in ways that maintain hierarchies between professional groups.3 Hackett et al.4 have identified three features of organizational life which need to be tackled by chief executives if a successful clinical governance framework is to be established: culture, leadership of changes to introduce the systems and dealing with established networks of power. They also state that any effort to implement clinical governance must ensure that consultant medical staff, as the key organizational stakeholders, are carried with the change, since the chances of success are limited if they are not ‘signed up’ to the process, its aims and values. (This point also applies to the NHS programme for information technology, Connecting for Table 46.1



Health. It has been widely reported that the programme has had difficulty in winning over the very clinicians who will apply new IT systems in the future, because they have not been fully engaged from the outset of the programme.)5 The values of clinicians and their managers may well differ and without constructive dialogue, this may affect doctors’ attitudes towards clinical governance arrangements in their hospitals (see Table 46.1): The clinician’s traditional valuesyprofessional autonomy, the focus on individual patients, the desire for self-regulation, and the role of evidence-based practiceyThose of managers: the emphasis on populations, the need for public accountability, the preoccupation with systems and the allocation of resourcesywe should not pretend there are no differences between the way that doctors and managers see the world.6



The goal however, whether doctor or manager, should be a common one: continuous improvement in the quality and safety of care for patients and for those who use health services. The required change in organizational culture to make clinical governance mainstream is movement towards a ‘learning organization’.7 This means that NHS organizations learn from their mistakes, including adverse incidents and near misses;8 that the results of clinical audit are widely shared and acted upon; that evidence-based practice is encouraged and that staff themselves are encouraged to learn and develop their skills as healthcare professionals.



WHY CLINICAL GOVERNANCE? Clinical governance sits within a national quality framework which drives, supports and reinforces the



Clinical governance: barriers to implementation in trusts.



Behavioural Reluctance of professions to change ‘yet again’ ‘Why bother’ syndrome – created by lack of suitable evidence-based guidelines combined with lack of time locally Resistance to CG systems as another vehicle for managerialism of professions Absence of good clinical leadership in some specialities locally and resources to invest in a satisfactory output Erosion of powerbases: Threat to power, prestige and status of internal stakeholders in organization



Adapted from Ref. 4.



Cultural Concern that clinical time will be invested in clinical governance (CG) systems at the expense of patient care Belief that systems based on evidence, guidelines and protocols may discourage clinical judgement and freedom for individual cases CG as a management agenda to improve performance by reducing or controlling costs Perceived tension between the mechanisms for self-regulation by doctors and the management culture of accountability, hierarchy and control Reticence to share CG systems outputs with the public as this will undermine professions, clinical services and individual doctors Incident reporting undermined by increase in litigation and the culture of medicine which emphasizes professional autonomy, collegiality and self-regulation



Chapter 46 Clinical governance: Improving the quality of patient care



introduction and implementation of local NHS clinical governance activities. Before explaining what systems have been established to improve standards of treatment and care, and monitor and inspect the quality of services provided, it is necessary to establish why it was felt necessary to introduce clinical governance in the NHS – a public service where for years the notion of ‘quality’ was implicit. The recruitment of well-trained staff, good facilities and equipment were considered synonymous with high standards of patient care (see Table 46.2). The NHS Plan20 and Delivering the NHS Plan21 further emphasized the importance of successful implementation of quality structures and mechanisms to help reinforce the NHS quality strategy. A range of national structures and mechanisms was established to develop national standards, provide effective monitoring and inspection of progress and ensure local delivery of this national policy for high quality health care. Clinical governance is central to the delivery of this strategy and provides the framework by which local organizations work to improve and assure the quality and safety of clinical services for patients.18



SETTING STANDARDS FOR DELIVERY ON QUALITY Standards are the cornerstone of quality assurance. The National Service Frameworks (NSF), the National Institute for Health and Clinical Excellence (NICE) and Table 46.2 Date 1948 1980s Early 1990s End of 1990s



1997 1998 1999 1999



] 571



healthcare standards are instrumental in establishing and maintaining them across the NHS. One of the key aims of the Health Care Standards Unit (HCSU) is to develop standards for better health (SfBH).22, 23 These 24 core standards and the optional developmental standards, describe the level of quality that health care organizations, including foundation trusts, and private and voluntary providers, are expected to meet. All NHS organizations in England are required to take them into account when developing, providing and commissioning healthcare. They are grouped into seven domains including safety, governance, patient focus and accessible and responsive care. The standards are grounded in clinical governance as is demonstrated by the outcome specified for ‘governance’: Managerial and clinical leadership and accountability, as well as the organisation’s culture, systems and working practices ensure that probity, quality assurance, quality improvement and patient safety are central components of all the activities of the health care organisation.24



National Service Frameworks are considered part of the developmental standards of SfBH, and the NHS and local authorities need to demonstrate they are progressing towards achieving the service quality described in them too. NSFs were launched in April 1998, as a rolling programme of national standards designed to address unacceptable variations in health services. They identify key interventions for particular services or care groups. By



Clinical governance: a brief history. Event NHS founded: In its first 40 years, quality improvement initiatives had mixed success; efforts to improve the quality of patient care remained fragmented and disparate. Managers and policy-makers tried to apply industry-based approaches of total quality management (TQM) and continuous quality improvement (CQI).9 Initiatives not widely accepted within health service organizations. Medical and clinical audit introduced, but criticized as being dominated by medical professions. Benefits not readily apparent to the wider health service or to patients.10 Public confidence in the NHS undermined and serious doubts raised about the quality of care available to patients. A series of clinical and organizational failures, included bone tumour diagnosis in Birmingham, UK, and paediatric cardiac surgery at the Bristol Royal Infirmary.11, 12 Small number of clinicians continued to practise, while their poor performance and professional conduct was tolerated by colleagues and went unchallenged by regulatory bodies.13, 14 Urgent action required to systematically address the quality and safety of patient care across all NHS organizations. Systems and processes were failing when credible individuals were able to do incredible things to patients to whom they had a duty of care.15 The NHS policy paper, The new NHS modern, dependable was published.16 The consultation document, A first class service: Quality in the new NHS was published.17 NHS Executive Health Service Circular, Clinical governance: Quality in the new NHS was published.18 Health Act19 made ‘quality’ a legal duty and introduced corporate accountability for clinical and organizational performance across the NHS. The concept of clinical governance was introduced and implemented. Every NHS Trust chief executive became ultimately responsible for assuring the quality of healthcare services provided by their organization



572 ] PART 9 SAFE AND EFFECTIVE PRACTICE 2006, there were NSFs for mental health, coronary heart disease, older people (including stroke), paediatric intensive care, children, diabetes, long-term conditions, cancer and renal services.25 Other standards are produced by the Medical Royal Colleges and the General Medical Council, while guidelines and advice specific to ENT are issued through organizations, such as the British Association of Otorhinolaryngologists, Head and Neck Surgeons (BAOHNS), the British Association of Head and Neck Oncologists (BAHNO) and the Department of Health (www.entuk.org; www.bahno.org.uk/bulletin.htm; www.dh.gov.uk). A nationwide body, the National Institute for Health and Clinical Excellence (NICE), was set up in 1999 for the appraisal of new and existing health treatments (for example, drugs and medical devices).26 NICE provides guidance on clinical procedures and develops clinical guidelines for disease management. For example, in June 2006, guidance was published on the use of the ultrasonic scalpel in tonsillectomies and in July, on tonsillectomy using laser.27, 28 NICE also uses evidence about clinical effectiveness and cost effectiveness to inform decisions in the health sector. Seven national collaborating centres have been established by NICE; one of which helps develop clinical guidelines and service guidance for the NHS in England and Wales on cancer treatment and care, including head and neck cancer. Guideline development groups are composed of doctors, other health professionals and patient/carer representatives.29 The Healthcare Commission sponsors a head and neck cancer audit, part of the National Clinical Audit Support Programme (NCASP). The aim is to deliver risk-adjusted, clinical audit data for head and neck cancer treatment centres in England through the Data for Head and Neck Oncology (DAHNO) audit service.30 Clinical audit is an important component of clinical governance. In 2006, the Chief Medical Officer cited clinical audit as one of a number of quality improvement methods necessary to ensure good clinical governance and a valuable tool for clinical engagement in quality asssurance and improvement. However, he also acknowledged that not only does it fall short of its potential (it needs ‘re-energizing’ locally), but there are concerns about the lack of mechanisms for findings of local clinical audits to be systematized into national service improvements.31 Head and neck cancer services are receiving local feedback reports from the DAHNO with the intention that collecting these data will enable comparative analyses to be made of cancer services across England and the identification of those areas where the quality of patient care could be improved.



Monitoring and inspecting clinical governance As part of the greater emphasis on quality assurance and improvement, and patient safety in healthcare, a



monitoring and inspection system was put in place in November 1999 to independently regulate NHS performance in England and Wales. The Commission for Health Improvement placed the clinical governance arrangements of all NHS trusts under its scrutiny and completed clinical governance reports on each trust. CHI review teams looked at the effectiveness of the policies, systems and processes an organization put in place to deliver clinical governance, gathered data on patient care and gained the views of staff, patients and carers. It also conducted investigations into serious service failures, including the unacceptable performance and behaviour of individual clinicians. NHS employers may initially refer a poorly performing doctor to the National Clinical Assessment Service (NCAS) which assesses concerns about doctors and offers advice (www.ncas.nhs.uk). In 2003, CHI became responsible for publishing NHS performance ratings. All NHS trusts were rated from zero to three stars based on data collated by CHI, including each trust’s clinical governance report. These star ratings were replaced by an annual health check in 2005 after an independent body with statutory powers, the Healthcare Commission superseded CHI. Under the ‘health check’, each NHS Trust Board in England is required to produce an annual report in which it rates the extent to which its services meets the 24 core standards of SfBH.32 This annual rating of performance has two parts: one looks at the quality of services available to patients and the public; the other at trusts’ management of finance and other resources. The system is one of self-assessments, although the focus on the quality of an organization’s leadership and management and other key elements of clinical governance are retained. The Healthcare Commission has made the interests, welfare and experiences of patients paramount among its concerns and adopted a patient-centred approach to improvement in the quality of the NHS and independent healthcare. It has an integrated approach to assessing the standards of services provided for patients and carers and its remit concentrates the functions of what were several organizations, into one. This includes:  coordinating inspection by registering and inspecting private and voluntary healthcare services, including private hospitals and exclusively private doctors;  national studies of the efficiency, effectiveness and economy of healthcare;  production of clinical audit tools to support clinicians in their local clinical governance and audit activities;  independent assessments of patient complaints and arrangements in place for promoting public health;  joint inspections of integrated health and social care services with the Commission for Social Care Inspection (CSCI) (from 2008, the HC and CSCI will merge);  working with the Mental Health Act Commission to assess the provision of care in mental health (the



Chapter 46 Clinical governance: Improving the quality of patient care



Healthcare Commission is expected to take over many of its functions in 2007);  surveys of patients’ and users’ experiences of healthcare.



The National Patient Safety Agency Clinical governance has been recognized as a key driver in improving the safety culture within NHS trusts.33 The National Patient Safety Agency (NPSA)34 was set up in 2001 to build a safer NHS, promoting an open and fair culture in which patient safety incidents and ‘near misses’ from across the NHS would be reported, collected and acted on to improve the quality and safety of care for patients. The NPSA also aims to initiate preventative measures so healthcare organizations nationally can learn from these reports, as well as from the mistakes and problems that affect patients’ satisfaction. Although most information is extracted from existing local risk management systems, an online form allows health professionals across England and Wales to report directly to the system and retain anonymity. The agency’s work also includes safety aspects of hospital design, cleanliness and food, as well as having responsibility for research ethics committees and the National Clinical Assessment Service. By early 2005, all NHS trusts were linked to an NPSA National Reporting and Learning System (NRLS), although by 2006 the lack of timely deliverance of the system was the subject of a number of critical reports.33, 35 There remained a lack of accurate information on serious incidents and deaths; at least 35 trusts had not submitted data and trusts were provided with only limited feedback and solutions to reduce serious incidents. Few trusts formally evaluated their safety culture so incidents were not being seen as opportunities for learning that prevent reoccurrence. These findings were reiterated in the Chief Medical Officer’s proposals for doctors’ regulation and patient safety. By 2006, the systematic attempts to improve adverse event detection were still ‘in their infancy’ and events were still not reported by healthcare workers because of fear of blame.36 Not only was under-reporting of incidents still a problem, but doctors were less likely to report an incident than other staff groups. The Healthcare Commission’s 2006 report on the state of healthcare also expressed concern that there was no reliable information nationally about the avoidable deaths of patients and about serious injuries. It is, therefore, to work with the NPSA and others to improve the collection, analysis and action around reported patient safety incidents.37



Assuring the quality of a doctor’s practice The General Medical Council (GMC) is the regulatory body for doctors who work in the UK. It sets the



] 573



educational requirements for entry on to the medical register and the professional standards expected of doctors once they are on it. These professional standards are made explicit in Good medical practice,38 which describes the principles of good practice and standards of competence, care and conduct expected of doctors in all aspects of their professional work. Serious failure to follow the guidance will put registration at risk. At the time of writing the GMC is undergoing the most extensive reform of medical regulation since it was founded in 1858. Part of these reforms is intended to ensure that clinical governance systems are joined up and effective through the maintenance of high standards of professional practice and patient protection. The GMC oversees the quality assurance of standards of practice. It is the duty of doctors to promote patient safety and to take part in systems of quality assurance and quality improvement,38 that is, to participate in activities related to clinical governance. The GMC, however, has been the subject of mounting criticism over its inability to adequately safeguard high standards of patient safety and care.39 A number of investigations into the practice of individual doctors, culminating in Dame Janet Smith’s inquiry reports into the murders committed by Dr Harold Shipman, showed both a ‘past culture of inaction’ and a ‘conspiracy of silence’, whereby well-known concerns about a doctor were denied or avoided and the interests of the patient were subordinated to other considerations.40 The concerns expressed in the fifth report of the inquiry41 led to the postponement of the launch of doctors’ revalidation (a system, including annual appraisal, to enable doctors to have their licence to practise renewed every five years) – and the commissioning of the Good doctors, safer patients report. The latter report advocated a new approach to promoting and assuring good medical practice and to protecting patients from bad practice. A radical reshaping of the role, structure and functions of the GMC was proposed with increased lay representation and public accountability. The report also recommended that educational and standard-setting bodies for doctors be given a more formal role in medical regulation. Among the needs it addresses are to:  design an effective interface between local healthcare systems for assuring good clinical governance and patient safety and the system of regulating the practice of individual doctors;  introduce an effective system of regular assessment of doctors’ practice which overcomes the weaknesses in earlier revalidation proposals;  create standards of medical practice that are transparent and capable of assessment;  develop methods of assessment that can measure a doctor’s performance against a predetermined standard;  reduce the climate of blame, retribution and disciplinary action that usually attends poor medical



574 ] PART 9 SAFE AND EFFECTIVE PRACTICE performance and introduce stronger elements of prevention and earlier recognition of problems, retraining and rehabilitation;  eliminate situations where poor practice is not recognized and acted upon because of adverse organizational culture and weak local clinical governance.42 Future measures will be implemented within a developing environment which recognizes that, inevitably, even good doctors make errors, which encourages open discussion and honesty about performance of individuals and clinical teams and which recognizes the importance of professional accountability for delivering care that is safe for patients and of high quality. Such an environment, in which weaknesses can be identified and individuals can be helped to learn and improve, rather than be blamed, may offer the best chance for sustained quality. Underpinning all of the above is the maintenance of clinicians’ professional knowledge and skills through professional development, improved access to information and communication technology and evidence-based research to support patient treatment and care. Good medical practice also stipulates that it is the duty of a doctor to put the care of patients first, to listen and respect their views and to respect the rights of patients to be fully involved in their care. Relationships with patients should be based upon openness, trust, good communication, politeness, honesty and consideration.



PATIENT AND PUBLIC INVOLVEMENT The NHS of the past ‘has been characterised by paternalistic providers and compliant patients’.43 Clinical governance, however, aims to change this and put the patient at the centre of delivery of their care. Robinson and Dixon43 acknowledged the changes of culture occurring in the NHS. Over the last decade, there has been a radical change in public knowledge and attitudes towards health. As healthcare consumers, patients have expectations that: information will be readily available about the service they are being offered; their consent will be gained prior to treatment; they will be respected; and they will be genuinely involved in services. In 2003, the Isaacs Report provided evidence that as part of medical research programmes undertaken in Manchester, UK (1985–1997), the organs and tissues of hundreds of people were retained from coroners’ postmortems and community mortuaries without relatives’ knowledge or consent. These practices were carried out in the belief that they were ‘in the public interest’.44 Practices carried out ‘for the public’, but without their consent are no longer acceptable. Seeking consent is an essential feature of good clinical practice. The availability of health information for patients is due, in no small part, to the internet. Many patients and



carers in the UK and elsewhere have access to an ever increasing number of public-friendly health websites. As people become more knowledgeable, they are more able to make informed choices about their care and gain recognition from healthcare professionals that the patient is a partner in making treatment choices and not simply a passive recipient who takes what they are given. In January 2003, a statutory duty was placed on every NHS organization to involve and consult patients and the public in the ongoing planning and provision of NHS services.45 The NHS Community Health Councils, representing patient and public interest in the NHS, were replaced by Patient Public Involvement Forums, designed to give the public a voice in decisions that affect their health and that of their community. These provide a useful contact point for health professionals who seek public involvement and representation, for example when planning and designing new ENT services. In addition, Patient Advice and Liaison Services (PALS) were introduced in trusts for ‘on the spot’ support to patients, carers and their families with concerns relating to their care and treatment, and to provide information on the NHS complaints procedure. The Commission for Patient and Public Involvement in Health (CPPIH) set up in January 2003 to support and manage the patient forums, and the forums themselves, are to be replaced by Local Involvement Networks (LINks), to be established for every local authority with social service responsibilities. A stronger local voice46 sets out these plans for the future of patient and public involvement (PPI), giving people a greater say in health and social care service provision, consultation about changes to services and about design of new services. In addition, the NHS Centre for Involvement was established in 2006 as a ‘one-stop shop’ for information and advice on PPI.47 This aims to support the development of local PPI networks for professionals, patients and the public by providing practical resources and information. In order to promote, lead and support sustained patient and public involvement across the NHS, it will work with PPI advocates from diverse sectors of the community including from PPI forums (whilst they continue), voluntary and community organizations, statutory authorities and local authority Overview and Scrutiny Committees (OSCs). (The latter scrutinize the activities of organizations that provide local health and social care services to see whether services are appropriate to local needs.) The OSCs should be consulted if there are major changes or variations to health services in the community, and, subject to legislation, will be encouraged to focus their attention on the work of commissioners of services.



Patient and public involvement in ENT The care of patients is the primary concern of a good doctor.38 In head and neck surgery, as across all



Chapter 46 Clinical governance: Improving the quality of patient care



] 575



specialties, if patients are to be at the centre of care, then professionals need to be accountable not only to their colleagues, their professional colleges and associations, but directly to the patients themselves. The BAOHNS guidance on clinical governance goes as far as to state, ‘Clinicians should hold themselves accountable to their patients first and their profession second.’48 Patient-centred care in ENT involves:



such as communication, respect, affording patients dignity and involving patients in treatment decisions. Understanding that patients and carers are likely to have a different viewpoint about what constitutes high quality and safe services is key to effective clinical governance in ENT, as in all services.



 working well with other professional groups and disciplines across health and social care, to meet patient needs as a whole;  openly exploring treatment options with patients, including patients’ preferences and personal circumstances;  involving patients and carers in decision-making;  responding to patients’ needs for clear information about their care and understanding of what the treatment process involves:  listening to patients and carers and answering their questions honestly, directly and sensitively;  recording at every stage of the care process;  communicating diagnosis and treatment decisions quickly to the patient’s general practitioner (GP) and to other professionals involved in their care;  involving patients and the public in all aspects of monitoring and developing the service.



THE NHS CLINICAL GOVERNANCE SUPPORT TEAM



An example of patient involvement in service development concerns cancer care standards used by clinicians. Birchall et al.49 involved 40 head and neck cancer patients and 18 carers in a series of focus group meetings. They showed that although participants’ views differed substantially from the professionals, the patients generated information that translated into improvements in existing national, professionally derived, standards of care for head and neck cancer (2000 national consensus standards). The above patients’ views may have differed from those of professionals, but differences may also occur when patients, carers and the public are asked what should be done to improve health services. A team of health professionals on a Clinical Governance Development Programme sought patients’ and carers’ views about the service they provided and how it could be improved. They did this through interviews, questionnaires and workshops. Repeatedly, patients and carers came up with the same concerns about the hospital environment, in particular the state of the waiting room and a lack of decent drink facilities. The team of professionals were perplexed by the responses they received because they expected patients and carers to be concerned about the same things as staff, that is, the clinical aspects of care. Similarly, when the public were surveyed about the assessment of doctors for Good doctors, safer patients,50 they wanted a doctor’s assessment to go beyond their technical skills and whether or not the doctor was up to date. They placed a high value on interpersonal qualities



The team of health professionals in the hospital environment example above was attending a programme run by the NHS Clinical Governance Support Team (CGST).51 The CGST was set up in 1999 to help implement and embed clinical governance in the NHS at a local level and support the delivery of high quality, patient-centred care. It also aims to improve awareness and understanding of clinical governance across and throughout NHS organizations, by acting as a focus of expertise, advice and information on clinical governance. Until 2005, much of the CGST’s work focused on delivering a series of national development programmes and activities designed for multidisciplinary healthcare teams and the board teams of NHS trust organizations.52 The CGST’s earliest activity was to establish a Clinical Governance Development Programme that supported multidisciplinary teams of frontline NHS staff to review their local services and to implement changes in healthcare quality for patients and for staff. In addition, there were programmes for trust boards to ensure that strategic leadership and clinical governance were high on the board’s agenda, and a programme, utilizing learning from the aviation industry, to establish effective teambased working through training ‘team coaches’. A website at www.appraisalsupport.nhs.uk was also established to provide appraisal support and a discussion forum for all doctors, regardless of specialty. This developed out of an earlier general practitioner appraisal programme which trained GPs across England as appraisers. The CGST continues to be a source of expertise on patient and carer involvement for frontline NHS staff and supports more effective means of engaging patients and communities in their local services. In 2006, the work of the Clinical Governance Support Team became the subject of review. Its main development programmes ceased in 2005, although two modular elearning programmes were established for community pharmacists and primary care managers. The CGST also worked to follow up the clinical governance aspects of the Shipman Inquiry41 and to support medical appraisal and revalidation for all doctors.53 Two further workstreams provided customized support to referred NHS organizations operating under challenging circumstances and worked with a small number of trusts to identify ‘known’ leaders and establish a team of people with the skills to drive though clinical governance improvement projects.



576 ] PART 9 SAFE AND EFFECTIVE PRACTICE Developing teams and supporting the way that they function as units has been an important aspect of CGST work. Team development continues to be a difficult task in the NHS, especially where there have been barriers between different professional groups, multiple lines of management, perceived status differentials and a lack of organizational systems and structures for supporting and managing teams of people working together. ENT is a specialty that often brings people together to work in multidisciplinary and multiprofessional ways involving, for example, oncologists, audiometricians, maxillofacial surgeons, speech and language therapists, and so on. A poorly functioning team can not only undermine clinical performance, but also compromise the safety of the patients in its care.



MULTIDISCIPLINARY TEAMWORKING The philosophy of clinical governance has patient safety at its very core. Delivering care that is both high quality and patient-centred must, of necessity, also be safe. Poor team-working can endanger patient safety54, 55 and put patients’ lives at risk,56 whereas effective teamworking has been linked to lower patient mortality.57 As treatment programmes become more complex, patient safety is ever more dependent on the effective, collaborative working of multidisciplinary teams. Good medical practice recognizes that most doctors work in teams with colleagues from other professions, that the skills and contributions of colleagues must be respected, communication must be effective and that team roles and responsibilities should be understood by both colleagues and patients.38 Modern healthcare is increasingly complex and requires coordinated and integrated delivery, none more so than in the treatment of cancer patients, where care can be poorly coordinated and confusing for those patients who experience both a wide range of services and different professionals involved in their care pathway.58 Multidisciplinary team-based working is a way of tackling this potential fragmentation of patient care. There is growing evidence that it makes a critical contribution to the delivery of effective, innovative and safe healthcare in the NHS.54, 57 The 1995 Calman-Hine Report on Cancer Care supported the widespread establishment of multidisciplinary teamworking across all cancer professions to address patients’ needs from diagnosis onwards. The underlying principles for cancer services developed in the report were based on high quality, localized, accessible and patient-centred care. A follow-up review of the implementation of the Calman Report showed that of a sample of 22 trusts/hospitals visited, MDTs who reported as meeting regularly, usually using case notes, were increasingly the norm in cancer care. However, differences between the nine common cancers were highlighted.



Whilst 85 percent of breast cancer teams regularly met to review patients, only 45 percent of head and neck MDTs did so.59 Birchall et al.’s 2003 study58 identified a shift towards multidisciplinary teamworking in head and neck cancer in the UK, although there was some indication of differing work patterns within teams, multiple teams in the same hospital and overlapping teams. This begs the question, what is a team? Many practising head and neck surgeons consider themselves to be working in teams, but what passes as a team, particularly a MDT, may be neither multidisciplinary nor a team. Mohrman et al.60 have defined a team as: y a group of individuals who work together to produce products or deliver services for which they are mutually accountable. The team members share goals and are held mutually accountable for meeting them, they are interdependent in their accomplishment, and they affect the result through their interactions with one another. Because the team is held collectively accountable, the work of integrating with one another is included among the responsibilities of each member.



Case study 1 illustrates the benefits for the patient, their carers and the professionals involved, of making a ‘team’ in a head and neck cancer clinic in a NHS trust in England (a team more like John Shaw’s) into an effective, well-functioning MDT. Working from a multidisciplinary combined oncology clinic or team has many advantages, including:  increased continuity of patient care;  reduced multiple appointments with different professionals;  increased professional knowledge and understanding through pooled expertise;  development of professional peer support and communication;  cost-effective use of time and resources;  individual consultants being accountable and not working in isolation;  decisions being audited and treatment aligned with agreed procedures;  reduced likelihood that mistakes and errors are hidden and will reoccur.



GOOD PRACTICE CASE STUDIES Case study 1: Making the team multidisciplinary Case study 1 illustrates part of a broad portfolio of clinical governance activities being undertaken in local organizations across the NHS to improve patient safety and care.



Chapter 46 Clinical governance: Improving the quality of patient care



HOW THINGS WERE



 A typical head and neck cancer team was made up of single surgeons (e.g. ENT, maxillofacial, plastic) who worked surgically alone, with occasional help from colleagues.  Surgeons referred patients to radiotherapists for radiotherapy after surgery, or if unable to perform surgery, for radiotherapy.  Patients were clinically treated in parallel rooms at best, in different hospitals at worst.  Clinicians ran different clinics and were largely unaware of colleagues’ activities.  Managers tried to encourage more multidisciplinary activity (as recommended by the NHS Cancer Plan, 200061). They initially suggested the team met at lunchtime to minimize the effect on clinical activity.  Minimal equipment for viewing radiology images and histopathology specimens was available at such sessions and there was ‘no regular room availability’ to accommodate such meetings.



] 577



 The team attending the patient now ‘sing to the same hymn sheet’.  Younger members of staff can seek advice and discuss patient care.  Allied health professionals attend weekly meetings in increasing numbers.  A clinic nurse coordinator was appointed to ensure that histopathology and x-ray results are available on the day, that patients have been informed that their case will be discussed and that they can attend the subsequent clinic.  Patients are no longer inconvenienced by a time delay in decision planning. Only new patients, planned postoperative patients and more difficult patient cases are seen in a ‘controlled environment’. It has been agreed that in such a clinic environment, we agree on one principle: that we need a combination of good communication and effective team working to get results. Consultant



WHAT WAS DONE



 Some specialists work across specialties and locations and MDT meetings are often difficult to schedule outside normal clinical working hours.  Most clinicians attended an ‘old style’ weekly combined head and neck clinic for one to several hours. They decided to have the MDT meeting as a part of the normal head and neck clinic and discuss problem and new patients.  Patients who had been discussed or who had previously agreed a combined treatment to include radiotherapy would be seen following the MDT meeting, in a combined clinical MDT environment.



WHAT HAPPENS NOW



 The MDT meeting has a planned structure and is executed ‘like clockwork’.  The MDT usually meets in a room in the radiology suite. A move closer to the clinic is planned, to improve transit time from the MDT meeting to the MDT clinic.  The case for each patient (most are recently diagnosed) is presented by each team leader to an audience of peers and medical students.  Clinical pictures prior to biopsy, radiology and pathology are shown and each patient’s management is openly discussed.  The diverse group assembled may include surgeons, pathologists, radiologists, physiotherapists, dieticians, speech and language therapists, Macmillan nurses, the pain management team and social workers.  A consensus is recommended and agreed prior to the patient’s treatment.



Case study 2: Improved pathways for children with grommets As the second case study shows, by addressing the needs of patients when building services, and considering how their journey through the care pathway may be reduced, it is possible to make changes which not only improve the patient experience, but improve the experience of healthcare professionals too.



HOW THINGS WERE



 A child with grommets inserted as treatment for secretory otitis media (glue ear) was examined by the ENT consultant about six weeks after surgery.  The consultant checked that the grommets were still in position and infection had not set in, before sending the child to the audiologist for a hearing test.  There were follow-up visits to the consultant to check the grommets; each required a test to ensure hearing gain had been maintained.  The audiometrician or audiological scientist was not involved in the patient’s examination or party to the consultation.



WHAT HAPPENS NOW



 Despite some resistance to change, a new process was implemented.  The audiologist runs a clinic in parallel with the consultant, examines the child, takes a history from



578 ] PART 9 SAFE AND EFFECTIVE PRACTICE







 







the parent, performs and interprets the post-grommet insertion hearing test. If results are normal, an audiometrician or an audiological scientist recommends further management of the child following agreed practice guidelines. This allows for direct patient/parent feedback there and then. If hearing is abnormal, the consultant can see and review the patient at the same session – sometimes even in the audiology booth. If the grommets have been rejected, the tympanic membrane is intact and hearing is satisfactory, the child is discharged back into the care of their family doctor. The new process means at least two visits to the doctor are cut out; clinic slots that can be reused, so that new patients can be seen sooner.



NOTE The information in this chapter, particularly with regard to the work of the NHS Clinical Governance Support Team and other national bodies associated with delivering the quality and safety agenda, is subject to change beyond 2007, following a period of review. However, the principles of good clinical governance remain applicable to all involved in continually improving patient safety and care. For the latest clinical governance information, see the CGST website at www.cgsupport.nhs.uk.



KEY POINTS  It is the duty of a doctor to promote patient safety and take part in systems of quality assurance and improvement.  Clinical governance provides the framework by which NHS organizations work to improve and assure the quality and safety of clinical services.  It is supported by systems established to improve standards and to monitor and inspect the quality of services for patients.  Clinical governance puts the patient rather than the professional at the centre of care.  A national Clinical Governance Support Team helps implement and embed clinical governance in the NHS.  A poorly functioning team can undermine clinical performance and compromise patient safety.  Multidisciplinary team-based working makes a critical contribution to the delivery of effective, innovative and safe healthcare.



CONCLUSION Standards for better health reiterated that ‘quality’ is at the forefront of the agenda for the NHS and for private and voluntary providers of NHS care.23 Clinical governance is integral to this agenda, but is in need of further development if it is to be effective. In 2006, the chief medical officer stated that: The culture of clinical governance needs to be spread to more local NHS organizations, there needs to be more consistent adherence to best practice standards and guidelines; more information on quality of care should be available to patients; and more effective and timely learning from adverse events needs to take place.62



The fundamental challenge in implementing clinical governance is to achieve this culture, which for many requires a transformation of ways of working, of attitudes towards others and of systems at the level of local NHS organizations. Clinical governance invites the re-examination of traditional ways of working within health services. It questions existing professional, organizational and cultural boundaries. It questions individuals like John Shaw, since, when appropriately applied, it is a way of working and a way of rethinking professionalism and partnership with patients. The appeal and impact of clinical governance lies in a ‘whole systems’ approach to improving the quality of patient care. It is essential for organizations to have effective systems and processes in place for clinical governance to be delivered, but these alone are not sufficient. Organizations are only as good as the people who work in them. The true mark of successful implementation is when clinical governance is embedded to the extent of it being ‘the way we work around here’.



REFERENCES







1. Scally G, Donaldson LJ. Clinical governance and the drive for quality improvement in the new NHS in England. British Medical Journal. 1998; 317: 61–5. 2. National Audit Office, Report by the Comptroller and Auditor General. Achieving Improvements through Clinical Governance: A Progress Report on Implementation by NHS Trusts. HC105 Session 2002–2003. London: The Stationery Office, September 2003, cited November 29, 2006. Available from: www.nao.gov.uk. 3. Bradley PJ. The culture of consultant hospital doctors and the difficulties of implementing change – The introduction of clinical governance. MBA thesis, University of Nottingham, unpublished, 2002.



Chapter 46 Clinical governance: Improving the quality of patient care







    



4. Hackett M, Lilford R, Jordan J. Clinical governance: Culture, leadership and power – The key to changing attitudes and behaviours in trusts. International Journal of Health Care Quality Assurance. 1999; 123: 98–104. 5. National Audit Office, Department of Health. Report by the Comptroller and Auditor General. The National Programme for IT in the NHS. HC 1173 Session 2005–2006. London: The Stationery Office, 16 June 2006, cited November 29, 2006. Available from: www.nao.gov.uk. 6. Edwards N, Marshall M, McLellan A, Abbasi K. Doctors and managers: A problem without a solution? No, a constructive dialogue is emerging. British Medical Journal. 2003; 326: 609–10. 7. Garside P. The learning organisation: A necessary setting for improving care? Quality in Health Care. 1999; 8: 211–2. 8. Department of Health. An organisation with a memory. London: The Stationery Office, 2000. 9. Deming WE. Out of the crisis. Cambridge: Cambridge University Press, 1986. 10. Thomson RG, Donaldson LJ. Medical audit and the wider quality debate. Journal of Public Health Medicine. 1990; 12: 149–51. 11. Malcolm AJ. Enquiry into the bone tumour service based at the Royal Orthopaedic Hospital. Birmingham: Birmingham Health Authority, 1995. 12. Kennedy I. Learning from Bristol: The report of the public inquiry into children’s heart surgery at the Bristol Royal Infirmary 1984–1995. London: The Stationery Office, July 2001: Command Paper 5207-1, cited November 29, 2006. Available from: www.bristol-inquiry.org.uk. 13. Dyer C. Obstetrician accused of committing a series of surgical blunders. British Medical Journal. 1998; 317: 767. 14. Roach JO. Management blamed over consultant’s malpractice. British Medical Journal. 2000; 320: 1557. 15. Commission for Health Improvement. Investigation into issues arising from the case of Loughborough GP Peter Green. London: The Stationery Office, 2001. 16. Department of Health. The new NHS: Modern dependable. London: The Stationery Office, 1997(Cm 3807). 17. Department of Health. A first class Service – Quality in the new NHS. London: Department of Health, September, 1998. 18. NHS Executive Health Service Circular. Clinical Governance: Quality in the new NHS. Department of Health, 16 March 1999: 65. 19. The Health Act. Chapter c8. London: The Stationery Office, 1999. 20. Department of Health. The NHS Plan: A plan for investment. A plan for reform. London: The Stationery Office, July, 2000, Command paper 4818-1. 21. Department of Health. Delivering the NHS Plan. London: The Stationery Office, 2002, circular. 22. Health Care Standards Unit, cited November 29, 2006. Available from: www.hcsu.org.uk.



 23. 24. 25. 26. 27.



28.



29.



30.



 31.



32.



33.



34. 35.



36.



37. 38.



39. 40.



] 579



Department of Health. Standards for Better Health. Department of Health, April, 2006. Department of Health. Standards for Better Health. Department of Health, April, 2006: 12. National Service Frameworks, cited November 29, 2006. Available from:www.nelh.nhs.uk/nsf/. National Institute for Health and Clinical Excellence, cited November 29, 2006. Available from: www.nice.org.uk. NICE. Tonsillectomy using ultrasonic scalpel. IP Guidance No. IPG178. Published guidance 28 June 2006, cited November 29, 2006. www.nice.org.uk/ page.aspx?o=IP_242. NICE. Tonsillectomy using laser. IP Guidance No. IPG186. Published guidance 26 July 2006, cited November 29, 2006. www.nice.org.uk/page.aspx?o=IP_57. National Collaborating Centre for Cancer (NCC-C), cited November 29, 2006. Available from: www.wales.nhs.uk/ sites3/home.cfm?orgid=432. The National Clinical Audit Support Programme’s Head and Neck Cancer Audit, cited November 29, 2006. Available from: www.icservices.nhs.uk/ncasp/pages/ audit_topics/DAHNO/default.asp?om=m1. Chief Medical Officer. Good doctors, safer patients. Proposals to strengthen the system to assure and improve the performance of doctors and to protect the safety of patients. London: Department of Health, 2006: 15. Healthcare Commission NHS Performance Ratings 2005/ 2006. October 12, 2006, cited November 29, 2006. Available from: http://annualhealthcheckratings. healthcarecommission.org.uk. National Audit Office. A safer place for patients: Learning to improve patient safety. HC 456 Session, 2005–2006. London: The Stationery Office, November 2005: 2, cited November 29, 2006. Available from: www.nao.org.uk. The National Patient Safety Agency, cited November 29, 2006. Available from: www.npsa.nhs.uk. House of Commons Committee of Public Accounts. A safer place for patients: Learning to improve patient safety. Fifty-first report of session 2005–2006, HC 831, London: The Stationery Office. 6 July 2006. Chief Medical Officer. Good doctors, safer patients. Proposals to strengthen the system to assure and improve the performance of doctors and to protect the safety of patients. London: Department of Health, 2006: 30. Commission for Healthcare Audit and Inspection. State of healthcare 2006. October, 2006. General Medical Council. Good medical practice, 2006, cited November 29, 2006. Available from: www. gmc-uk.org/guidance/good_medical_practice/index.asp. Dyer O. GMC regrets failure to act on police warning about gynaecologist. British Medical Journal. 2004; 328: 1035. Chief Medical Officer. Good doctors, safer patients. Proposals to strengthen the system to assure and improve the performance of doctors and to protect the safety of patients. London: Department of Health, 2006: 3.



580 ] PART 9 SAFE AND EFFECTIVE PRACTICE



 41. 42.



43. 44.



45.



46.



47.



48.



49.



50.



 51. 52.



53.



54.



55.



56.



 57.



Smith, Dame J. The Shipman Inquiry Fifth Report: Safeguarding patients: Lessons from the past – Proposals for the future. London: The Stationery Office, 2004. Chief Medical Officer. Good doctors, safer patients. Proposals to strengthen the system to assure and improve the performance of doctors and to protect the safety of patients. London: Department of Health, 2006: xiii/xiv. Robinson R, Dixon A. Completing the Course. Health to 2010. Fabian Ideas 605. London: Fabian Society, 2002: 53. Department of Health. Isaacs Report: The investigation of events that followed the death of Cyril Mark Isaacs. London: The Stationery Office, 2003. Department of Health. Strengthening accountability: Involving patients and the public. Policy Guidance, Section 11 of the Health and Social Care Act 2001. February, 2003. Department of Health. A stronger local voice: A framework for creating a stronger local voice in the development of health and social services. July, 2006. The NHS Centre for Involvement, cited November 29, 2006. Available from: www.nhscentreforinvolvement. nhs.uk. BAOHNS. Clinical governance and the role of BAOHNS. London: British Association of Otorhinolaryngologists – Head and Neck Surgeons, 2000. Birchall M, Richardson A, Lee L. Eliciting views of patients with head and neck cancer and carers on professionally derived standards of care. British Medical Journal. 2002; 324: 1–5. Chief Medical Officer. Good doctors, safer patients. Proposals to strengthen the system to assure and improve the performance of doctors and to protect the safety of patients. London: Department of Health, 2006: 144. NHS Clinical Governance Support Team, cited November 29, 2006. Available from: www.cgsupport.nhs.uk. Fine words into action: Learning from the first five years of the NHS Clinical Governance Support Team. Clinical Governance: An International Journal. 2006; 11. NHS Clinical Governance Support Team Expert Group. Assuring the quality of medical appraisal, cited November 29, 2006. Available from: www.cgsupport.nhs.uk/ About_CG/Resources/CG_Publications/CG_Guides/. NCEPOD. Functioning as a team? The 2002 Report of the National Confidential Enquiry into Perioperative Deaths (Executive Summary), 2002 London: NCEPOD. Available from: www.ncepod.org.uk/pdf/2002/02sum.pdf. West MA, Borrill C, Dawson J, Scully J, Carter M, Anelay S et al. The link between the management of employees and patient mortality in acute hospitals. International Journal of Human Resource Management. 2002; 13: 1299–1310. Mayor S. Poor team work is killing patients. British Medical Journal. 2002; 325: 1129. Borrill CS, Carletta J, Carter AJ, Dawson JF, Garrod S, Rees A et al. The Effectiveness of Health Care Teams in the National Health Service - Report. Universities of Aston, Glasgow, Edinburgh, Leeds and Sheffield, 2000, cited



58.



59.



60.



61.



62.



November 29, 2006. Available from: http:// homepages.inf.ed.ac.uk/jeanc/DOH-final-report.pdf Birchall M, Brown PM, Browne J. The organisation of head and neck oncology services in the UK: the Royal College of Surgeons of England and British Association of Head and Neck Oncologists’ preliminary multidisciplinary head and neck oncology audit. Annals of the Royal College of Surgeons of England 2003; 85:154-157. Commission for Health Improvement and the Audit Commission. NHS Cancer Care in England and Wales. National Service Frameworks Assessments No. 1. London: The Stationery Office, 2001. Mohrman SA, Cohen SG, Mohrman Jr. AM. Designing team-based organisations. San Francisco: Jossey-Bass, 1995. Department of Health. The NHS Cancer Plan: A plan for investment, a plan for reform. London: Department of Health, 2000. Chief Medical Officer. Good doctors, safer patients. Proposals to strengthen the system to assure and improve the performance of doctors and to protect the safety of patients. London: Department of Health, 2006: 31.



FURTHER READING Clinical governance and ENT resources Clinical Audit Support Centre. www.clinicalanditsupport.com. Copeland G A Practical Handbook for Clinical Audit, NHS Clinical Governance Support Team, March 2005. http:// www.cgsupport.nhs.uk/About_CG/Resources/ CG_Publications/CG_Guides/. ENT and Audiology Services site replacing Action on ENT, http:// www.wise.nhs.uk/cmsWISE/Clinical+Themes/ent-audiology/ services.htm. ENT and Audiology Specialist Library at National Library for Health, www.library.nhs.uk/ent. King’s Fund library resources (search term, clinical governance), www.kingsfund.org.uk. Modernisation Agency Action on ENT Good Practice Guide, 2002, www.institute.nhs.uk/Products/ ActiononENTGoodPracticeGuide.htm. Modernisation Agency Action on ENT Top 10 Tips, 10 July 2006, www.institute.nhs.uk/Products/ActiononENTTop10Tips.htm. Modernisation Agency Improvement Leaders’ Guides, www.wise.nhs.uk/cmswise/default.htm. NHS Clinical Governance Support Team, www.cgsupport.nhs.uk. Pre-operative tests guidelines for ENT, www.entuk.org/ publications. Royal Society of Medicine Clinical Governance Bulletin, www.clinical-governance.com. Supporting Doctors Appraisal forum and website, www.appraisalsupport.nhs.uk.



47 Medical ethics KATHERINE WASSON



Introduction: Why do we need medical ethics? The doctor–patient relationship Duties and responsibilities of doctors Bases for medical ethics and moral decision-making Ethical issues in practice Confidentiality Stopping treatment



581 581 582 583 585 586 587



Research ethics Resource allocation Conclusions Key points Best clinical practice References



588 589 591 591 592 592



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words medical ethics, bioethics, duties and responsibilities, consent, refusal, confidentiality, withholding and withdrawing treatment, resource allocation and research ethics.



INTRODUCTION: WHY DO WE NEED MEDICAL ETHICS? We live in a multifaith, multicultural society. In this pluralist context there is a lack of moral consensus. People do not necessarily share a common moral outlook and determining what is right and wrong is based on a variety of different values and value systems. Not only do different individuals hold different moral views, e.g. an atheist and a Muslim, but one person may draw on different and even contradictory moral values in him/ herself, e.g. naturalism and a technological imperative. In medicine, these differences lead to disagreement both in practice and theory between doctors, other professionals, patients, families and institutions. Difficult cases highlight the differences in moral perspectives and pose challenges for determining a ‘best’ or ‘right’ way forward. In response to these difficulties, theoretical ethics provides a framework for approaching complex dilemmas and difficult decisions. Ethics does not provide easy answers to difficult problems, but it offers a clear structure and more consistent means of making moral



decisions. On the practical level, ethics is not something which is outside medicine, but it is an integral part of doctors’ daily practice, whether in relation to gaining a patient’s consent, observing confidentiality or using professional judgement to assess areas such as best interests. It offers critical analysis of difficult issues and guidance in making decisions in practice.



THE DOCTOR–PATIENT RELATIONSHIP The doctor–patient relationship is the context within which many medical ethical dilemmas arise. In the UK and other countries, there has been a general shift from paternalism towards consumerism in medicine. Patients are less willing to accept what the doctor says without question and are more likely to pressurize doctors to do what they wish. Doctors are seen as ‘need-meeters’ who can and should dispense a pill or offer a treatment for every ailment. This poses difficult decisions for doctors in balancing what patients need, want and even demand.



582 ] PART 9 SAFE AND EFFECTIVE PRACTICE Along with this shift towards consumerism has come an increasing emphasis on individual autonomy and ‘rights’. Patients’ rights have been enshrined in the Patients’ Charter (and the Human Rights Act 1998). This is partly a reaction against the paternalism of medicine where ‘doctor knows best’ and takes little note of patients’ choice, desires or preferences regarding treatment. Respecting patient autonomy is now recognized as a fundamental principle in medical ethics.1 Rather than watching the pendulum swing from one extreme to the other, a balance can be struck between paternalism and consumerism. Both the doctor and patient have valuable input to provide in any medical consultation. A partnership model allows for the patient to provide information about his/her own body and symptoms, as well as personal choices and preferences, while the doctor brings his/her knowledge, skills, expertise and experience to the individual case.2 Together they are able to determine the best way forward for this particular patient. This partnership model requires good communication on both sides and recognizes the role of duties and responsibilities in medicine.



DUTIES AND RESPONSIBILITIES OF DOCTORS Duties and responsibilities of doctors go back as far as the Hippocratic Oath and are set out in professional codes of conduct.3, 4, 5 These codes highlight both the positive and negative duties doctors have to their patients, colleagues, institutions and wider society. Duties and responsibilities are fundamental to the practice of medicine and include nonmaleficence, beneficence, respect for autonomy and justice.1 Doctors have a negative duty of nonmaleficence, or doing no harm, to patients. Both professionally and ethically doctors have a responsibility not to harm their patients. A doctor may harm his/her patient by being negligent or not following correct medical procedures. Yet, it is difficult to uphold nonmaleficence in its literal sense as doctors may inflict initial harm on patients in order to achieve good in the end. This is not doctors being negligent or unethical in their practice. Patients who undergo head and neck surgery experience initial harm and pain, but the aim is to bring about cure from disease. This harm is unavoidable and if doctors could cure the patient through other, less harmful, means then they would. The long-term benefits of cure outweigh the short-term detriments of medical intervention. The intention and motivation of the doctor, as well as the results of a treatment, are crucial in assessing the harm done to patients. Some degree of harm may be unavoidable in order to cure a patient and justifiable if the good done is expected to outweigh the harm. The danger is that nonmaleficence may be simply defensive. In contrast, beneficence is proactive and is



about doing good to patients. Not only are doctors to refrain from harming their patients whenever possible, but they also have a duty to do good to them. The most obvious way of fulfilling this duty is to cure a patient of his/her illness, disease or dis-ease and restore him/her to full health. Cure is not possible in all cases. Both doctors and patients should recognize the reality and the limits of medicine. In some cases, limiting the harm done by a disease or illness may be the best doctors are able to do. Beneficence extends beyond cure and includes caring for patients and treating them with respect and dignity. Part of doing good to patients involves respecting their autonomy. Autonomy literally means self-rule and is understood as self-determination and choice. In health care, respecting autonomy generally means allowing patients the freedom to make decisions for themselves, even if doctors disagree with those choices. A patient might refuse curative maxillofacial surgery because of the degree of disfigurement caused even though doctors advocate it. In modern western societies, the autonomy of individuals is frequently emphasized. The ability and freedom to make our own choices is paramount. This carries over into health care too. Considerations of autonomy highlight at least two points within medicine. The first is that patients are not the only people who possess autonomy. Doctors, nurses and other professionals also have autonomy which can and is exercised. For example, a doctor cannot be forced to give a treatment against his/her professional judgement. The second is that in any society or community, individual autonomy is always limited. Autonomy is limited if there is a risk of, or actual, harm to oneself or others. If a patient is suicidal, he/she can be sectioned to avoid harming him/herself. If a patient attempted to attack someone on the hospital ward, then his/her autonomy would be restricted to prevent harming others. For all of us within health care and society more generally, autonomy is limited. Legislation in dealing with the threat of infectious diseases shows that an individual’s freedom and autonomy can be limited if the risk to others is great. Part of doing good to patients and respecting their autonomy involves being just. Doctors have a duty to treat individual patients fairly and equally. This means treating similar cases in similar ways and dissimilar cases in dissimilar ways. Both types of justice ensure there is a minimum standard of consistency in interactions with all patients.6 Fairness and equality do not allow doctors to discriminate against patients based on clinically irrelevant factors, e.g. religion, race, sex. Doctors not only have to weigh up considerations of justice in relation to individual patients, but also between all of their patients. These key responsibilities of nonmaleficence, beneficence, respect for autonomy and justice are four of the fundamental principles in medical ethics and practice. In



Chapter 47 Medical ethics



recent years these have been seen as the key principles that provide the foundation for doctors’ duties in medical ethics. It can be argued that other ethical duties are significant in medicine. That these principles are necessary in medicine is generally agreed, but whether they are sufficient is debated.7 In looking beyond these four principles, doctors also have a duty to care. This duty is fundamental to each and every interaction a doctor has with patients. It is the basis from which doctors begin and carry out their work. For the doctor, this involves looking after patients to the best of his/her ability and applying his/her knowledge, expertise and skills to care for the patient. Part of this duty to care includes protecting the best interests of the patient. Doctors have a responsibility to act in the best interests of their patients. This does not mean that doctors make decisions about best interests without reference to or discussion with the patient. Yet, deciding what is in the best interests of the patient may not be straightforward depending on how and when best interests are defined and who defines them, e.g. doctor, patient, relatives or other professionals. Doctors may not agree whether surgery, radiotherapy or chemotherapy is in the best interests of a child with a neck tumour and the patient and parents may also have differing views. When examining the duties of doctors, there may be a problem over who defines best interest and how it is defined. There is also a difficulty when different duties and responsibilities conflict, e.g. nonmaleficence and beneficence. Analyzing how to proceed and which responsibilities take priority are key questions in medical practice. A hierarchy of duties may be required to clarify priorities and aid decisions in practice. For example, it might be argued that doing no harm is more important than doing good and that the duty of care is fundamental to all interactions with patients.



BASES FOR MEDICAL ETHICS AND MORAL DECISION-MAKING Individuals make and assess moral decisions on different bases. These form the foundations of our view of what is ‘right’ and ‘wrong’. At their root, such decisions can be primarily based on principles, consequences or virtues or some combination of these.



Principle-based morality Deontological morality is based on principles. It deals with what we ought to do; that moral obligation arises from fundamental ethical principles. Whenever people use principles to make decisions, it is important to ask, ‘from where do these principles come?’ There are various



] 583



sources from which doctors and patients draw moral principles.



NATURE/NATURALISM



Naturalism argues that by looking at and reflecting on nature moral truths or principles can be discerned. This includes both the natural world and nature of human beings.8 In the nature of things and the nature of people, there are moral principles which reason can discern. Goodness and badness are natural features of the world and people. By observing the way the world is and people are, we can see morality in the nature of things and people.9 This is closely tied to what we regard as ‘normal’ function. Where there is damage to the head or neck, the aim of surgical intervention is to restore normal function and enable the natural processes to function without impediment. Yet, people have different ideas and conceptions of what is ‘natural’. These different views may conflict. In fact, doctors often try to put right things that have gone wrong in nature. Medicine struggles against the ravages of nature and tries to prevent natural disease from harming people. It is also clear that medicine and surgery are in a crucial sense fighting against natural disease and breakdown. Determining what is natural, and the implications of a naturalistic moral view, may not be straightforward. RELIGIOUS MORALITY



Religion also provides a key source of moral principles. Christianity, Islam, Judaism, Buddhism and Hinduism all teach certain fundamental moral precepts.10 These are usually enshrined in moral and religious codes such as the Koran and the Ten Commandments. There may be debate about whether such principles are independently valid or dependent on being a follower of a particular religion, but the problem for medicine is whether the different religions all support the same principles or whether they are fundamentally different. At first glance, the Jehovah’s Witness refusal of blood transfusions is very different from the ready acceptance of life-saving treatments by other religious groups. Different religions seem to offer different moral principles.



KANT



The philosopher Immanuel Kant also emphasized the role of principles in his highly influential contribution to moral thinking. For Kant, morality is about asking, ‘What ought I do?’10 In response, he tried to identify fundamental principles, or maxims, of action that all people ought to adopt. The key moral principle is the Categorical Imperative, which states ‘Act only on the maxim through which you can at the same time will that it be a universal law.’11 The key test for moral principles is whether they



584 ] PART 9 SAFE AND EFFECTIVE PRACTICE are universalizable, i.e. anyone in the same moral situation would necessarily make the same choice. This is a check against begging the question in your own favour or being biased. Essentially, medical treatment ought to be given to every individual without bias. For Kant, morality involves respect for people. This respect is based on the intrinsic worth and dignity of human-beings. He requires that we treat humanity, whether ourselves or any other, never simply as a means, but also always as an end.12 People are ends in themselves. We cannot do medical experiments on people for the benefit of future generations without their consent. That would be using people for some other end than for their own good. Kant’s ethics involves moral duties, which apply to ourselves and others. He argues that we have positive and negative moral duties. There are both things we should do as well as those we should avoid, e.g. doctors doing good and not harming patients. Any morality based on principles may run into difficulties when different moral principles, like duties, conflict. Kant’s approach is no exception here. Deciding between principles is often problematic and Kant does not provide a clear way forward. Principles are one key aspect of morality and moral decision-making, but morality also can be based on consequences.



Consequence-based morality Morality based on consequences is called teleological morality. A consequentialist approach argues that morality should and does focus on the potential outcomes of our choices and actions, particularly in terms of what will happen if we do something against doing nothing at all.



UTILITARIANISM



Utilitarianism is the most well known of consequentialist theories, particularly in health care. It is a moral theory concerned with social reform and it argues that society should adopt laws which make people happy and avoid pain. At the heart of utilitarianism is the Greatest Happiness Principle (GHP), which states that morality is about achieving the greatest happiness for the greatest number of people. If this is the case, then we must have some way of measuring happiness or pleasure. To this end, Jeremy Bentham developed a pleasure calculus.8, 13 It includes:       



duration: how long it lasts; intensity: how intense it is; propinquity: how near it is; extent: how widely it covers; certainty: how sure we are it will come; purity: how free from pain it is; fecundity: how much it will lead to more pleasure.



In Bentham’s version of utilitarianism, the pleasure or pain of one person, if particularly acute, could outweigh that of many others. To counteract the implicit lack of justice in Bentham’s theory, John Stuart Mill introduced the notion of fairness.14 In assessing pleasure or pain in society, each person should count for one and not more than one. This was an attempt to ensure that the pleasure and pain of each person was weighed fairly against that of others. Even with the inclusion of justice and fairness, utilitarianism poses difficulties when making moral decisions. One is the danger that the minority will have to sacrifice or forego pleasure for the happiness of the majority and that increasing pleasure may be given a greater moral importance than preventing pain. This view would give equal moral weight to giving people a thrill and alleviating pain. Such critique is particularly important for doctors who are confronted with people who are suffering from illness, disease and in pain. Underlying utilitarianism, and all consequentialist theories, is a focus on the ends rather than the means of moral choices. But do the ends always justify the means or are there some means which are wrong or immoral in and of themselves and cannot be justified even if they achieve a good end? This question is important in medicine where doctors are weighing up the potential benefits and burdens of treatment. Either doctors or patients may think that the means of achieving a potentially good end are not acceptable, as in the case of some cancer treatments. One critique of a morality based on consequences is that it is difficult to predict, nonetheless control, the consequences of our actions. Doctors are faced with this challenge on a daily basis when treating patients. What they intend to achieve may significantly differ from the end result. Both morality based on consequences and principles pose problems for doctors. Morality involves an examination of both, but requires more.



Virtue-based morality In recent moral theory, there have been moves away from a morality primarily based on principles or consequences. The emphasis has shifted toward an appeal to virtues,15 harking back to Aristotle and Aquinas. Virtue theory focusses on growing virtuous people and asks, ‘What is a good person?’ It seeks to identify certain qualities and characteristics of a good person and to foster these in individuals. The way that these characteristics are grown and developed in people is in the context of a community. We may ask, ‘What makes a good doctor?’ The list of qualities might include having specific scientific knowledge (of disease and illness, as well as treatments), professional expertise and skills, being caring, compassionate, a good communicator and wise. This would affect the selection criteria of would-be medical students.



Chapter 47 Medical ethics



Then we should ask how these qualities or virtues of a good doctor would be fostered. This would take place within the professional community which trains medical students and newly qualified doctors. Teaching, training and practice are all crucial in the development and fostering of the virtues of a good doctor. The picture painted in virtue theory is very optimistic. It may not be easy to agree on the characteristics of a good person or good doctor, far less to foster them. Virtue theory does not provide clarity about what to do when these characteristics conflict. In the end, it can seem highly optimistic about how virtuous people really are and how easy it would be to encourage and grow a virtuous person, as well as the nature and role of community in that process. In examining three approaches to morality, based on principles, consequences or virtues, we have seen that all three harbour difficulties. Perhaps a more balanced view of morality should incorporate considerations of all three – recognizing that principles offer fundamental grounds for making moral decisions, but these cannot and should not be taken without weighing up the potential consequences of any decision. This should all happen within the context of trying to foster the development of virtuous doctors who will be able to make ‘good’ and ‘right’ decisions in practice, based on who they are rather than primarily on decision-making processes.



ETHICAL ISSUES IN PRACTICE Three approaches to making moral decisions – principles, consequences and virtues – have been examined within the context of the doctor–patient relationship, as well as duties and responsibilities. It is helpful to consider the specific ethical issues in medicine and how we might approach them, rather than simply analyzing the theories of ethics. These include consent and refusal, confidentiality, stopping treatment, ethical issues in research and resource allocation.



Consent and refusal Consent is fundamental to any and every interaction between doctors and patients. By its nature, gaining and giving consent is a process, not a one-off decision, and patients must be free to change their minds at any point. Consent may be implicit or explicit. A patient may give implicit, or implied, consent by indicating nonverbally that he/she is happy to participate in a procedure, for example holding out his/her arm to have blood taken or opening his/her mouth to have the throat checked. Alternatively, consent may be explicitly stated in verbal or written form. Consent forms provide evidence that a particular process has taken place, but serve only as a minimum standard for recording consent, as they do not



] 585



guarantee the quality of communication or level of patient understanding. The normal standard in medicine is fully informed, valid consent given by a patient who is mentally competent. The ‘fully informed’ part of consent involves the patient being given all relevant information regarding the treatment options and potential consequences of having against not having a particular treatment. In the UK, doctors should provide the competent patient with as much information as he/she needs or desires to make an informed decision.16 This will vary from patient to patient and doctors are required to use their professional judgement about the degree of detail regarding benefits, risks and burdens in each case. They should assume patients wish to be well informed and use a ‘prudent patient’ standard.16 This is in sharp contrast to the United States where doctors must make clear any and every known risk, regardless of its likelihood. When attempting to gain fully informed consent, doctors face the challenge of providing sufficient information for a patient to make a decision and telling the truth about the potential risks, over and against allowing patients to keep hope. To be valid, consent must be freely given, ensuring that patients are free from any improper, undue pressure from doctors or other professionals. How, when and by whom information is given to patients about potential treatment or nontreatment options can have a significant impact on the decisions made. If the consultant emphasizes a particular option as the best course, the patient may be more likely to agree to it. Patients need time to digest the information given to them, ask questions, consult others and weigh up different options. There are some instances in which (as well as groups of patients for whom) gaining valid consent is a more ethically and practically difficult process. In emergency situations, where patients cannot give consent or refuse treatment, doctors are obliged to act in the best interests of the patient. They have both a legal and ethical duty to take measures to secure the patient’s life and health. These may be short-term interventions that then enable the patient to recover sufficiently to make longer-term decisions about further treatment for him/herself. For children in the UK, the legal age of consent is 16 years. While the general standard is that parental consent should be sought for medical interventions with children under 16, children should be encouraged to participate in their care and health professionals should act as their advocates.17 Sometimes, children and young people under 16 may have the capacity to give valid consent. A doctor can give treatment without parental consent if he/she judges the young person to be ‘Gillick competent’ and it is in the patient’s interests.17, 18 The former involves assessing whether the child or young person understands the consequences of having against not having a particular treatment or intervention. In reality, the degree of competence required is proportional to the weightiness of the decision. A 13-year-old might be deemed



586 ] PART 9 SAFE AND EFFECTIVE PRACTICE competent to consent to a tonsillectomy, but not brain surgery. Individuals who enter the armed forces agree to give up a degree of their individual autonomy and freedoms in the interests of the unit as a whole. They may be called on to put their life and/or health at risk for the good of the service and country. Prisoners also forfeit their autonomy and are less free to give or refuse consent, in particular with respect to intimate body searches. Both groups are still entitled to a high ethical standard of medical care and doctors have the same professional and ethical duties to these patients as others.19, 20 Some patients may be incompetent or lack capacity to make choices and decisions for themselves.21 Patients who are mentally impaired or disordered or handicapped may lack sufficient understanding of the nature, purpose, risks and benefits of a proposed treatment to give valid consent. In England, Wales and Northern Ireland, noone can legally give consent on behalf of a competent adult.21, 22 In Scotland, people over the age of 16 may appoint a proxy decision-maker to give consent or refuse treatment for them if they become incapacitated.23 Doctors should not assume that these patients are incompetent without seeking to engage with them and discern the level of competence. Such patients should be encouraged to participate in decisions about their care whenever possible. Refusal is at the heart of consent. Patients are free to refuse any treatment or intervention at any time. An adult Jehovah’s Witness may refuse a life-saving blood transfusion. If a doctor attempts to treat this patient in the face of a valid refusal, he/she can be charged with assault or battery. Valid refusal requires the patient understands the significance of his/her choice and its implications. The ethical basis for this is respect for autonomy. Patients are free to make their own choices regarding their health care and can even refuse life-saving treatment, regardless of whether doctors agree with them or not. Because of the potential significance of refusing treatment, in reality a refusal requires a higher degree of competency than consent to treatment. Good practice means doctors strive to ensure the patient fully understands the implications of his/her refusal. It should be carefully documented to safeguard the doctor. Gaining fully informed, valid consent or exploring refusal highlight ethical issues in medical practice. Along with confidentiality, they are fundamental starting points for the doctor–patient relationship and any treatment.



CONFIDENTIALITY Doctors have a duty to protect patient confidentiality and privacy. Confidentiality matters because it is a key basis for trust between the doctor and patient. In practice, understanding what confidentiality means may vary. Patients may think that in telling their doctor something



in confidence it will not be discussed with anyone else. They may not realize that doctors not only talk to each other and other professionals about particular cases, but the number of health professionals who have access to their ‘confidential’ medical notes. Confidentiality involves issues both of truth-telling and disclosure. Doctors have a positive duty to tell the truth to patients and a negative duty not to lie. Without these duties, patients cannot be certain whether to trust what their doctor tells them. Patients expect doctors to give them information and advice which is true and accurate. Not telling patients the truth fundamentally undermines this trust and means that patients’ confidence in what their doctor says is eroded. One argument for not telling patients the truth is that it is sometimes in their best interests. A doctor may decide not to tell a woman with a malignant tumour of the neck that she has cancer. The rationale may be that it would be too upsetting for her. This attitude is paternalistic and assumes that the doctor knows better than the patient how she will cope with the news. If this patient asks for information about her diagnosis, doctors have an obligation to be clear and truthful. A more difficult area is where patients do not directly enquire about their diagnosis. How much information should doctors give and how detailed should it be? Is it acceptable to be ‘economical with the truth’? Information about diagnosis or prognosis should not be forced on patients if they do not want it, but that differs from doctors denying them the opportunity to have this information or patients being given an inaccurate or misleading account of their condition. There are appropriate and inappropriate ways of telling the truth to patients. A woman with motor neurone disease went to see her general practitioner (GP). She had concerns about her death and wanted to know what would happen to her. The doctor told her she would eventually choke to death. The woman then had recurring nightmares where she would wake up choking. This doctor told the truth, but in a way which was insensitive and inappropriate. There is a balance to be struck between answering questions truthfully and being sensitive to the patient’s situation and anxieties. Any disclosure of information about the patient’s condition must be given to him/her first. Doctors should make sure the information and its significance is clear to the patient. Doctors must balance the disclosure of all relevant information to the patient with being sensitive to what the patient does or does not want to know. Patients are free to indicate they do not wish to know certain information, but they should be given the option to decline. Sometimes relatives believe that they have a ‘right’ to know details of the patient’s condition and pressurize doctors to tell them. Doctors have an ethical duty to protect patient confidentiality and privacy and not to disclose information to relatives without the explicit



Chapter 47 Medical ethics



consent of the patient. Even though it may put doctors in a difficult position with relatives, patients are entitled to expect doctors not to disclose information about their medical condition to relatives. There are times when doctors need to consult colleagues about a patient’s condition. It is ethically acceptable to share relevant information, even if confidential, in order to determine what care or treatment options are in the best interests of the patient. Good practice ensures that the patient is aware of these discussions and disclosures and the reasons for them. When working in a team, patient information may need to be shared to ensure good and consistent care is delivered. Disclosure at these times should be on a ‘need to know’ basis and differs from disclosing confidential information which is not relevant to the clinical situation, e.g. the patient’s sexual orientation or family situation. There are some instances where patient confidentiality can be broken without his/her consent. Under the Public Health (Control of Diseases) Act 1984 in England and Wales and the Infectious Diseases (Notification) Act 1889 in Scotland, doctors are under a legal obligation to report notifiable diseases to the appropriate authority.24, 25 Ideally, this should be done with the patient’s full knowledge and consent and every effort made to encourage the patient to comply. If a patient refuses to agree to the notification, the doctor has a legal and ethical duty to report the disease for the protection of the wider public. Disclosure in the case of HIV/AIDS poses particularly difficult ethical issues. On the one hand, doctors have to weigh up the potential harm done to patients from disclosing their HIV status against the potential risk of infection to others, like health care professionals or sexual partners. If patients perceive that confidentiality will not be kept, then they may not come forward for testing and treatment. Every effort should be made to encourage the patient to disclose this information him/herself. If this is unsuccessful, disclosure may take place if a doctor believes a specific person, including a sexual partner, is being placed at ‘serious and identifiable risk’ of becoming infected.24, 25, 26



Data protection and medical records The Data Protection Act 1984 provides protection of personal information held by others, e.g. medical records on computer. Personal information given for one purpose cannot then be used for a different purpose without consent. The use and handling of this information is restricted. The Act also allows people to have access to personal data at reasonable intervals. Patients can request to have access to their medical notes and read what has been written about them. This has altered both the style and content of what doctors write in the medical notes and the use of medical records for research.



] 587



Protecting patient confidentiality and information ensures that trust is maintained between the doctor and patient. When there is a risk of harm to others, doctors may have a duty to break confidentiality. This should be carefully weighed, the British Medical Association or the Medical Defence Union should be consulted and the facts clearly communicated to the patient. The discussions and rationale for breeching confidentiality should be documented.



STOPPING TREATMENT Withholding and withdrawing treatment One of the most difficult areas for doctors is deciding when to stop active treatment for an individual patient. There are tensions between whether and when to begin and stop treatment. Whether to withhold or withdraw a medical treatment or intervention, such as chemotherapy, artificial feeding or hydration, can be a contentious area for doctors, other health professionals, patients and relatives. Weighing up the benefits and burdens of treatment, assessing best interests and quality of life are necessary features of such judgements.27 Whether to begin or withhold a treatment is the first phase in this debate. Judgements about withholding treatment should be based on the best available evidence and the individual patient’s condition. If the likely burdens and risks of treatment outweigh the potential benefits, the doctor may judge the treatment to be futile. There is much debate about what constitutes futile treatment and who decides, as doctors’ and patients’ interpretations of futility may differ. In the literature, the generally accepted definition of a futile treatment is one that has less than a 1 percent chance of success.28, 29 If a treatment has been started, there may come a point where doctors believe it should be stopped. The British Medical Association (BMA) document Withholding and withdrawing life-prolonging medical treatment offers guidance for doctors in making these decisions. It states that although prolonging a patient’s life usually provides a health benefit, doctors are not obliged to prolong life at all costs with no reference to burdens of treatment or quality of life. If a treatment fails or ceases to provide a net benefit for a patient, the justification for providing it is removed and the treatment may, ethically and legally, be withheld or withdrawn (pp. 1–2).30 In practice, doctors may find it more difficult to withdraw than to withhold treatment. To prevent situations where patients are denied treatments which might be beneficial because doctors think it will be too difficult to withdraw at a later stage, the BMA has argued, ‘Although emotionally it may be easier to withhold treatment than to withdraw that which has been started, there are no legal, or necessary morally relevant,



588 ] PART 9 SAFE AND EFFECTIVE PRACTICE differences between the two actions.’30 This view is debated.31 Philosophically, it can be argued that withdrawing treatment is more like an action (doing something) than an omission (refraining from doing something) and so has a different moral weight and significance.32 Regardless of where one sits within this philosophical debate, in practice there will be cases where it is appropriate to stop or withdraw treatment. Doctors are justified in withdrawing treatment and this does not constitute euthanasia. If treatment is withdrawn and ‘nature takes its course’, it is the disease which brings about the patient’s death, not the doctor. In the UK, the Tony Bland case has become crucial to the debate about withholding and withdrawing treatment. He was a young man who was seriously injured in the Hillsborough football disaster and was in a persistent vegetative state (PVS). After four years, the doctor and family wanted to remove the nasal-gastric tube through which Tony was being fed. They appealed to the coroner, under whose protection all of the Hillsborough victims were placed, who asserted that he would charge the doctor with murder if he stopped feeding Tony. The case eventually went to the House of Lords and two key rulings were given.33 First, the Law Lords ruled that artificial feeding was now a medical treatment and, therefore, could be withdrawn. Second, because Tony was in PVS it was not in his best interests to be kept in PVS. Both judgements have been debated and raised issues of withholding and withdrawing treatment among the wider public. In reaching a decision about whether to withhold or withdraw treatment, good medical practice elicits and incorporates the patient’s views and that of the relatives, provided the patient has given permission for that discussion to take place. If doctors are recommending that treatment be withheld or withdrawn, clear rationale and justifications should be communicated to the patient and recorded in the notes. Patients are free to request treatments and some demand particular interventions, but doctors are not obliged to give treatments against their clinical judgement. Clinical judgements to withdraw treatment should consider futility, best interests and quality of life issues. Doctors have a duty to act in and protect the best interests of their patients. When making treatment or nontreatment decisions, good practice will involve competent patients in that process. For the incompetent patient, the views of the relatives should be sought to ascertain what the patient would have wanted, but doctors are under no legal obligation to do as the relatives wish. Ultimately, the doctor must act in a way which he/she thinks will be in the best interests of the patient. If the patient is a child below the age of consent, then the parents’ views are crucial in deciding what constitutes best interests. Even when the child may be deemed competent to consent or refuse treatment, good practice would include the parents in that decision whenever possible.



Similar issues are raised when attempting to assess quality of life. Doctors will have some view of what constitutes a reasonable quality of life, but this may be very different from the patient’s view. Competent patients should be given the opportunity to express their view of quality of life issues. For incompetent patients, the relatives should be consulted about the patient’s views, but ultimately doctors must make a decision which they believe to be in the best interest of the patient.



RESEARCH ETHICS Good medical practice relies on a solid knowledge base. This requires good research. The government’s emphasis on ‘evidence-based medicine’ has led to the creation of the National Institute for Clinical Excellence (NICE), which monitors the quality of clinical practice. The rationale for treatments and techniques should be supported by research results and not simply past practice. Research is needed to continue to improve practice and develop new techniques and treatments. However, medical research raises ethical questions for doctors, researchers, patients and the wider community. The nature, content and process of research merit ethical, clinical and scientific scrutiny. Research can cover a broad range of activities. It can involve tracking a specific genetic disorder in the population or developing a new surgical intervention for hearing impairment. A key aspect of scientific research is that it tests a thesis or hypothesis and that the results are testable and repeatable.34 If the results cannot be tested or duplicated, there is reason to be sceptical about the research. Peer review is an integral and necessary part of all scientific research. This provides a level of scrutiny for the claims and results of research. For a new technique, intervention or therapy to become a standard treatment, researchers must convince their peers of its merits. A key question in setting up any research study is who will benefit from it? Therapeutic research is intended to benefit the subject, or patient, directly. Nontherapeutic research does not benefit the individual subject, but aims to contribute to medical knowledge and may benefit others in the future.35 Whether the development of innovative treatment should be categorized as research or an extension of usual practice is debated. On the one hand, if the digression is small and focussed on the individual patient’s situation then it may be justifiable and the innovation may be seen as a useful byproduct of effectively treating an individual patient. On the other hand, such innovation may be seen as ‘trial and error’ which is ethically and scientifically unacceptable and puts patients at risk because the procedures are not monitored or scrutinized sufficiently or at all.35, 36 The process of setting up a medical research study involves obtaining ethical approval from the local research



Chapter 47 Medical ethics



ethics committee (LREC). They were introduced to help weigh up the potential benefits and risks of research for individuals and society. LRECs examine the scientific quality of the protocol, whether the investigators are competent and facilities adequate, possible hazards, measures for providing information and seeking consent, adequate compensation arrangements regarding harm from the trial, methods of recruitment, payment to investigators, and storage and use of subject-identifiable information.35 For multicentre trials, multicentre regional ethics committee (MRECs) have been instituted both to speed the process of ethical assessment and to prevent duplication. According to the BMA,17 although there is no legal requirement for researchers to obtain ethical approval for a study, it is almost impossible to get funding and becoming increasing difficult to publish results without it. These committees examine the nature, purpose and process of proposed studies with consent processes being the subject of particular scrutiny. Consent to participate in research must be valid and freely given. According to the Declaration of Helsinki (1964), subjects must be ‘adequately informed of the aims, methods, anticipated benefits and potential hazards of the study and the discomfort it may entail’. Participants need information about the positives and negatives of any study and the alternatives before being able to give valid consent. For minors and children, the same guidelines apply to consenting to take part in research as for treatment. Parental consent should be sought if the child is under 18 years, but it should not override refusal by a competent child. In UK law, parents cannot give consent for a child for procedures which are of no particular benefit and may carry a risk of harm, or refuse treatment which would be in the child’s best interests, e.g. a life-saving blood transfusion.37, 38 Great care is taken in research on vulnerable groups, such as psychiatric patients. Voluntariness is a key element in consenting to research. Subjects must not be pressurized in any way to take part and they must be free to refuse before or at any time during a study without this impacting on their basic care and treatment. This can be difficult, as doctors are in a position of power and may pressurize subjects unintentionally. Subjects may feel guilty if they refuse to participate or fear that a refusal will negatively affect their care. To avoid this situation, recruitment of subjects should be undertaken by someone other than the person carrying out the research, e.g. a research nurse who explains the options.36 It should be made clear to subjects that a refusal will not be counted against them by their doctor. Written information for subjects and their GPs is particularly important. Randomized control trials (RCT) are viewed by many as the gold standard for medical research, but are criticized by others.36 This method randomly allocates participants into different groups who then receive



] 589



different treatments. It enables researchers to compare different interventions and determine which one produces the most beneficial results. If they are doubleblinded, then neither the participant nor researcher knows which subjects are given the different interventions. This approach attempts to remove bias from the research. Yet, it also removes choice regarding treatment options from the participant and raises ethical concerns. The use of placebos in research has come under scrutiny as it denies some subjects the potential benefits of the intervention. It is no longer considered ethical to use placebos if there are known treatments which may offer some benefit to the subject.39 Whether or not placebos should be used to test the effectiveness of an existing treatment, where no other exists, is still debated.40 Research also raises ethical concerns about confidentiality and the handling of medical records and information. To undertake a study, participants medical records may become available to a number of people other than the doctor, research team or practice. The means of protecting confidentiality both during the course of the study and future use of the findings must be addressed. The BMA17 states that whenever patient-identifiable information is used, a medically qualified person should be identified to ensure the information is handled correctly and confidentiality is protected. A vital part of good research is that it offers some benefit to the individual or society, that consent is freely given on the basis of detailed information, that confidentiality of patient-identifiable information is protected and that the research constitutes good science which is ethically sound. To improve treatments, develop new techniques and interventions, medicine must rigorously assess current practice. This should not be at the expense of subjects’ health, life or conditions or without careful consideration of the risks and benefits and ethical issues involved. These should be clearly explained to participants and they should be given the option to consent or refuse to participate. Whether and how much money should be given towards research raises ethical questions of resource allocation.



RESOURCE ALLOCATION Regardless of the specialty or setting (National Health Service (NHS) or private sector), the issue of limited resources is a reality for doctors. They face multiple pressures from patients, relatives, colleagues, management and Departments of Health, and doctors must make choices about how they spend their time, energy and money.



590 ] PART 9 SAFE AND EFFECTIVE PRACTICE



Levels of allocation Resource allocation happens on different levels. In macroallocation, central government divides its resources between areas such as health care, education and defence. Monies given to health care are then allocated between different hospitals trusts, primary care trusts and regional health authorities that distribute among primary, secondary and tertiary care. Micro-allocation includes decisions about treatment options, staffing levels and how much each service or department spends on research and administration.



NEEDS



For doctors on the front line, there is a tension between meeting the needs of individual patients and assessing the cost to the wider community. Assessing patient need is a key part of determining treatment options. Who defines needs and how are they defined? Abraham Maslow argued that there is a hierarchy of humans needs, ranging from physiological and safety needs to self-actualization (the ability to fulfil oneself and realize one’s full potential). He also claimed that people seek to have lower level needs met first before focussing on higher level needs.41 Basic needs are those things universally required by all people to survive, such as food, water and shelter. Nonbasic needs are not required for survival, but contribute to our psychological, social and spiritual well-being.42 Within the context of limited resources, doctors may have to focus on the most urgent and pressing basic needs before addressing other needs.



PREFERENCES, DESIRES AND WANTS



Patients have a wide variety of health needs, but they also have desires and preferences about their treatment options. With the move away from paternalism, patients are less willing to do what the doctor tells them without question. Patients can and do communicate what they want and may also place demands on doctors for particular treatments or drugs; for example, parents who insist their child be given grommets for their ears. Of course, sometimes patients want things they do not need, such as cosmetic surgery, and need things they do not want, such as chemotherapy. There is a danger that doctors give in to pressure from patients and the most demanding patients get what they want, while others do not.



RIGHTS CLAIMS



Sometimes patient demands are based on ‘rights’. Claims to rights have become a common part of modern Western culture. In medicine, the Patient’s Charter and the



Human Rights Act (1998) have raised the notion of rights and correlative expectations in the public consciousness. When confronted with rights claims, doctors need to decipher what kind of rights are involved and whether or not they have a duty to fulfil them. Legal rights provide a minimum standard of protection from harm for all people in a society. If anyone breaches another person’s legal rights, then he/she can be taken to court and prosecuted. In medicine, if a doctor is negligent and harms a patient then he/she may be struck off or even sent to prison. Doctors and patients have clear duties to uphold and respect legal rights. In contrast, the duties and responsibilities doctors have in relation to human rights are less clear. The Human Rights Act (1998) includes both the ‘right to life’ (Article 2) and a right to be free from ‘inhuman or degrading treatment’ (Article 3). These may conflict in medicine, particularly when considering issues such as ‘do not attempt resuscitation’ orders or whether or not to switch off the ventilator for an unconscious patient in the intensive therapy unit (ITU). How doctors are to make decisions in light of human rights, particularly when they conflict, is not clear.



ENTITLEMENTS



Arguably, a more useful means of approaching rights, expectations and duties is to focus on entitlements. To what are patients entitled in health care? In the UK, people are entitled to access the NHS and receive good quality care and appropriate treatment. Patients are not necessarily entitled to any and every treatment they want or demand. Doctors have a responsibility to provide care, but should only supply treatment which is likely to be successful, unlikely to cause harm and there are sufficient resources. Doctors should be clear about the extent of their duties and recognize that these may vary depending on the type of rights involved. Patients also have responsibilities which include giving their doctor all information relevant to their condition and wherever possible not needlessly endangering their own health. Patients do pressurize doctors based on needs, desires and rights and doctors must make decisions about how best to allocate their resources. These factors should be weighted against considerations of justice.



CONSIDERATIONS OF JUSTICE



One key moral basis for making and assessing allocation decisions is that of justice. Justice as fairness, equality and equity all have an impact on medical and health care. Fairness requires that a universal and uniform standard of treatment is provided for all.43 It requires consistency in the way people are treated. To be fair, two patients with the same condition should have access to the same types



Chapter 47 Medical ethics



of services and care and be given the same treatment options. Equality, like fairness, provides a minimum standard of consistency in treating people. It requires that similar cases are treated in similar ways and dissimilar cases are treated in dissimilar ways. Equality seeks to eliminate discrimination based on inappropriate grounds such as race or religious belief. Inequality exists where similar cases are treated in different ways. When allocating resources, doctors should be consistent in their dealings with patients and not allow discrimination based on clinically irrelevant factors to determine those decisions. Although equality and fairness highlight important principles of justice and offer protection for individuals because of their focus on uniform treatment, they do not necessarily allow for differences in individual cases to be weighed sufficiently. In contrast, equity allows for differences in treatment between individuals, but these are permitted only for morally justifiable reasons.44 It recognizes that giving people equal consideration is not the same as identical treatment.45 A diversion from the normal standard of fairness and equality must be both justified and justifiable. Inequity exists where differences in treatment are not morally justified. If two patients have the same congenital abnormality of the ear and one is offered surgery and the other is not based on their economic class this would not be justifiable. Doctors must be able to provide a clear rationale for their decisions to the patient, relatives, other colleagues, management, government and wider public. The process and basis for these decisions should be communicated to the patient at the time and recorded in the notes. When weighing up how best to use resources, doctors have to balance the needs, wants, rights and entitlements of the individual patient with their own duties and responsibilities to that individual and the wider community. This involves considerations of justice as fairness, equality and equity.



CONCLUSIONS Doctors experience increasing pressure on their time, skills and resources. Daily, they are faced with difficult decisions, whether in relation to consent, confidentiality or withholding and withdrawing treatment. The moral basis on which they weigh up and make such decisions is crucial to their own practice and the delivery of patient care. In our pluralist society, it may not be clear how and on what basis such decisions are made. Disagreements between doctors and patients, relatives and other colleagues can and do arise when confronted with moral choices and decisions. Reflecting on the specific ethical grounds for such decisions provides a more consistent framework for approaching such decisions. Whether they



] 591



involve principles, consequences, virtue theory or a combination of these, ethical frameworks must be applied in practice. How they are applied may be seen in light of the doctor’s duty to care for his/her patients and protecting patients’ best interests, which involves nonmaleficence, beneficence, being just and respecting autonomy. All of these must be weighed and prioritized in the context of the doctor–patient relationship where both parties bring particular knowledge and skills to the decision. When this relationship breaks down and some consensus about a moral dilemma cannot be reached, there is an increasing tendency to retreat to law to sort it out. This leads to polarization of the parties involved and defensive medicine as doctors seek to protect themselves rather than put the best interests of the patient first. The law can be a blunt instrument, serves as a minimum standard and addresses legal, not ethical, questions. As one way to avoid this retreat to law and in order to resist practising defensively, doctors should be increasing their moral awareness. Addressing ethical dilemmas within medical practice and discussing them with patients and relatives is one means of avoiding ending up in the courts. Doctors should be able to justify their decisions ethically and be clear about the moral basis on which they were made. Such justifications may have to be given to patients, families, colleagues, management, the government and wider society. With increasing pressure from the media, it is important that doctors offer a moral, as well as clinical, justification for their decisions. As medical science and technology continue to advance, ethical challenges will continue to be faced within medicine. Encouraging healthy debate and dialogue within the medical profession and good communication with patients, relatives and colleagues can help avoid defensive practice and legal battles. Doctors should be encouraged to reflect on the moral basis for their decisions, while being aware of the different moral perspectives and values of patient and relatives, and look for some common ground and shared values in the midst of disagreement rather than focussing solely on the differences.



KEY POINTS  Medical ethics is necessary as part of good clinical practice in dealing with patients, families, colleagues and institutions.  The doctor–patient relationship is characterized by partnership and good communication.  The good doctor fulfils the duty of care, acting in the best interests of patients,



592 ] PART 9 SAFE AND EFFECTIVE PRACTICE



 



 



 



respecting their autonomy and acting justly. Good medical moral decision-making draws on principles, consequences and medical virtues. Consent and refusal of treatment are based on competence, the giving of full information, understanding of its significance and freedom of choice. Confidentiality is both legally and morally required, except in extreme circumstances. Withdrawing and withholding treatment is both clinically and morally permissible in situations of futility and in the best interests of the patient. Careful allocation of limited resources means clear setting of priorities and acting fairly, equitably and equally. The conduct of research is subject to strict ethical guidelines to safeguard the patient/ participant and the researcher.



REFERENCES















Best clinical practice [ In all dealings with patients, families and colleagues, [ [ [ [



[ [ [ [ [ [



doctors should recognize the need for high moral standards. Doctors and patients should both act in partnership. The patient’s autonomy should be respected, while recognizing autonomy has limits. Good communication with colleagues and patients is vital. Doctors should not only do no harm, do good, act justly and maximize autonomy, but also have a duty of care and to act in the best interests of patients. Doctors should be able to recognize the need for and make moral decisions on the basis of principles, consequences and medical virtues. Consent and refusal are part of a process based on competence, full information, understanding of its significance and free choice. Confidentiality must be respected, unless legally required to be broken. Withholding and withdrawing treatment may be appropriate when the situation is futile and it is in the best interests of the patient. Resources should be allocated fairly and equitably. Research should be conducted ethically with good clinical and statistical grounds, consent, protection of confidentiality and communication.











1. Beauchamp TL, Childress JF. Principles in biomedical ethics, 5th edn. Oxford: Oxford University Press, 2001. This text provides a good grounding in the theory of medical ethics and the four key principles. 2. Cook ED. Choice and consent. In: Patients’ choice: A consumer’s guide to medical practice. London: Spire, 1993: 18–21. 3. British Medical Association. Rights and responsibilities of doctors, 2nd edn. London: British Medical Association, 1992. In this text, the BMA outlines carefully multiple duties of doctors and offers guidance for ethical practice in relation to issues such as consent and refusal, confidentiality, stopping treatment, resource allocation and research. 4. General Medical Council. Duties of a doctor. London: General Medical Council, 1995. 5. British Medical Association. Medical ethics today: The BMA’s handbook of ethics and law, 2nd edn. London: BMJ Publishing, 2004. This offers useful guidance on a range of clinical and ethical issues for doctors including consent, confidentiality, stopping treatment, resource allocation and research. 6. Wasson K. Ethical arguments for providing palliative care to non-cancer patients. International Journal of Palliative Nursing. 2000; 6: 66–70. 7. Takala T. What is wrong with global bioethics? On the limitations of the four principles approach. Cambridge Quarterly of Healthcare Ethics. 2001; 10: 72–7. 8. Cook ED. The values that surround us. In: The moral maze. London: SPCK, 1983: 21–4. David Cook provides an accessible and clear explanation of different approaches to morality as well as one framework for approaching such decisions. 9. Wasson K, Cook ED. Morality in medicine: Five approaches to ethics. CME Bulletin of Palliative Medicine. 2001; 2: 6–10. This provides a concise overview of five approaches to ethics in medicine-principles, consequences, relativism, reductionism and virtue theory. 10. O’Neill O. Kantian ethics. In: Singer P (ed.). A companion to ethics. Oxford: Blackwell, 1993: 175–85. 11. Paton HJ. The moral law: Kant’s groundwork of the metaphysics of morals. London: Hutchinson, 1961. 12. Kant I. The foundations of the metaphysics of morals. Translated by Beck LW. London and New York: MacMillan, 1985. 13. Bentham J. A fragment on government and introduction to the principles of morals and legislation. Harrison W (ed.). Oxford: Basil Blackwell, 1948. 14. Mill JS. Utilitarianism, liberty, representative government. London: Dent, 1954. 15. MacIntyre A. After virtue: A study in moral theory, 2nd edn. Notre Dame: Notre Dame University Press, 1984. 16. British Medical Association. Consent and refusal. In: Medical ethics today: Its practice and philosophy. London: British Medical Association, 1993: 10.



Chapter 47 Medical ethics 17. British Medical Association. Consent and refusal: Children and young people. Medical ethics today: The BMA’s handbook of ethics and law, 2nd edn. London: BMJ Publishing, 2004: 131–64. 18. British Medical Association. Children and young people. In: Medical ethics today: Its practice and philosophy. London: British Medical Association, 1993: 78. 19. British Medical Association. Chapter 16, Doctors with dual obligations and Chapter 17, Doctors working in custodial settings. In: Medical ethics today: The BMA’s handbook of ethics and law, 2nd edn. London: BMJ Publishing Group, 2004: 565–601 and 602–47. 20. British Medical Association. Chapter 1, Consent and refusal and Chapter 9, Doctors with dual obligations. In: Medical ethics today: Its practice and philosophy. London: BMA, 1993: 21–22, 245–54. 21. British Medical Association. Treatment without consent: Incapacitated adults and compulsory treatment. In: Medical ethics today: Its practice and philosophy, 2nd edn. London: British Medical Association, 2004: 99–126. 22. British Medical Association. Mental health. In: Rights and responsibilities of doctors, 2nd edn. London: British Medical Association, 1992: 99. 23. Adults with Incapacity (Scotland) Act 2000. 24. British Medical Association. Confidentiality. In: Rights and responsibilities of doctors, 2nd edn. London: British Medical Association, 1992: 41–3. 25. British Medical Association. Confidentiality. In: Medical ethics today: The BMA’s handbook of ethics and law, 2nd edn. London: British Medical Association, 2004: 173. 26. General Medical Council. HIV infection and AIDS. London: General Medical Council, 1988. 27. British Medical Association. Caring for patients at the end of life. In: Medical ethics today: The BMA’s handbook of ethics and law, 2nd edn. London: British Medical Association, 2004: 351–87. 28. Schneiderman LJ, Jecker NS, Jonsen AR. Medical futility: Its meaning and ethical implications. Annals of Internal Medicine. 1990; 112: 949–54. 29. Schneiderman LJ, Jecker NS, Jonsen AR. Medical futility: Responses to critiques. Annals of Internal Medicine. 1996; 125: 669–74. 30. British Medical Association. Withholding and withdrawing life-prolonging medical treatment. London: British Medical Association, 1999.



] 593



31. Rhymes JA, McCullough LB, Luchi RJ, Teasdale TA, Wilson N. Withdrawing very low-burden interventions in chronically ill patients. Journal of the American Medical Association. 2000; 283: 1061–3. 32. Cook ED. Stopping treatment. In: Patients’ choice: A consumer’s guide to medical practice. London: Spire, 1993: 125–6. 33. Airedale NHS v Bland (1993) AC 789. 34. Cook ED. Research. In: Patients’ choice: A consumer’s guide to medical practice. London: Spire, 1993: 191. 35. British Medical Association. Research and innovative treatment. In: Medical ethics today: The BMA’s handbook of ethics and law, 2nd edn. London: British Medical Association, 2004: 489–90. 36. British Medical Association. Research. In: Medical ethics today: Its practice and philosophy. London: British Medical Association, 1993: 198–9. 37. Medical Research Council. The ethical conduct of research on children. London: Medical Research Council, 1991. 38. British Medical Association. Accessed on June 30, 2005, www.bma.org.uk/ap.nsf/Content/ Parental#Consentfrompeoplewithparental 39. Michels KB, Rothman KJ. Update on unethical use of placebos in randomised trials. Bioethics. 2003; 17: 188–204. 40. Rothman KJ, Michels KB, Baum M. For and against: Declaration of Helsinki should be strengthened. British Medical Journal. 2000; 321: 442–5. 41. Maslow A. Motivation and personality, 3rd edn. Revised by Frager R, Fadiman J, Reynolds C, Cox R. New York: Harper and Row, 1987. 42. Wasson K. Resource allocation. Christian Medical Fellowship File 17. London: Christian Medical Fellowship, 2002. 43. Rawls J. A theory of justice. Cambridge: Belknap and Harvard University Press, 1971. 44. Downie RS, Telfer E. Caring and curing: A philosophy of medicine and social work. New York and London: Methuen, 1980. 45. Outka G. Social justice and equal access to health care. In: Lammers SE, Verhey A (ed.). On moral medicine: Theological perspectives in medical ethics. Grand Rapids: Eerdmans, 1987: 632–43.



48 Medical jurisprudence and otorhinolaryngology MAURICE HAWTHORNE AND DESMOND WATSON



Introduction Tort Negligence Consent Confidentiality The coroner The role of the expert



594 595 595 600 602 602 603



INTRODUCTION When these authors were first appointed as consultant otolaryngologists, the closest contact with the legal process that they envisaged was the writing of expert witness reports in personal injury cases. Along with this was a realization that there was a small chance that these reports might lead to a court appearance as an expert witness. The regional health authorities and later hospital trusts had by then been forced to accept vicarious liability for the actions of the consultants they employed so the only other contact that apparently was possible was the result of a claim alleging negligence in private practice. Although this event was feared by most consultants, the chances of a successful claim were and still remain low. The last decade and a half has seen some changes in how claims of clinical negligence are decided but the developments in other fields such as consent, confidentiality and clinical performance have been much more significant. The medical profession in general has twice been put in the spotlight by very major events which were subject to intense media interest: the Bristol Royal Infirmary Inquiry into unacceptable complication rates in the paediatric cardiac surgery unit and the trial of Harold Shipman who was convicted of murdering an unknown number of his patients have changed the way that the public views the medical profession. The response from politicians has been to try to tighten the legal framework in which doctors practise so that the public can be reassured that it can never happen again.



Performance and complaints The Human Rights Act The criminal law The future Key points References



605 606 607 609 610 610



The law in England differs slightly from that in Scotland and Northern Ireland but most of the principles are the same. The first major subdivision is into criminal law and civil law. In criminal law, the state prosecutes an individual or an organization (the defendant) alleging breach of a law or statute. If the case is proven, the individual or organization is subject to a sentence, often a fine or a custodial term of some kind. In civil law, one individual or organization (now known as the claimant, formerly the plaintiff) brings a suit for damages against a defendant and seeks damages from that defendant. Ideally, the damages would exactly redress the loss suffered by the claimant. For instance, a claimant proved to have been swindled out of his car by a defendant might be happy merely to have his car back. In medical negligence cases, it is manifestly impossible to restore negligently caused losses so the damages are monetary. There are three important forms of monetary damages in clinical negligence: general, special and punitive or exemplary. General damages are those awarded to compensate the successful claimant for pain and suffering. Special damages are those awarded to take account of loss of income, special nursing needs, etc., arising from the negligence. Punitive damages are exactly what they say they are: damages awarded to the claimant to punish the negligent defendant. These are similar in a way to a fine in criminal cases but are not awarded in clinical negligence cases in the UK where most defendants have insurance to cover any damages awarded against them.



Chapter 48 Medical jurisprudence and otorhinolaryngology



Civil and criminal jurisdictions differ in two other important ways: in the standard of proof required and in the influence of statutes. In criminal cases, the standard of proof is ‘beyond reasonable doubt’ which is often taken to imply a probability of guilt of at least 95 percent. In civil cases the standard of proof is ‘on the balance of probabilities’ which means greater than 50 percent probability. This figure leads to some apparently illogical conclusions in the field of causation where it has been argued that, as far as the law is concerned, a 5 percent fiveyear survival is identical to a 49 percent five-year survival (so no compensation should be paid even if negligence is proven) but completely different from 51 percent five-year survival. The view of the House of Lords, the highest legal authority in England and Wales, on this subject is awaited. The source of the criminal law is founded in the common law and on statutes passed by Parliament and, more recently, European law. Statutes, precedent and European legislation state what actions constitute criminal behaviour. In the course of a criminal trial, counsel may well cite a precedent of what has been said and what decisions were made by courts in the past in similar cases. In a civil trial, there may be no statute that has any relevance to the questions to be decided so the decisions are based almost entirely on what happened in previous cases. This is referred to as the common law. The common law is in a state of continuous development as civil cases are tried by high courts, such as the Court of Appeal and the House of Lords, which can introduce subtle changes in decisions that then become the common law. Much argument may then occur as to what exactly their Lordships meant. Was what they said in a previous case relevant to the one currently being tried? Even more debatable is whether or not the important remark was part of a judgement or some sort of subjunctive aside to the judgement in the form of ‘if this case had been similar to case A then I would have decided as follows’. The next section of the chapter (see under Causation below) illustrates exactly how one word in a judgement can be important when the case referred to as Bolitho1 is considered. Besides major sections devoted to clinical negligence and the duties of an expert medical witness, this chapter examines the two most common questions asked of a medical indemnity organization’s helpline: consent and the doctor’s duty of confidentiality. The criminal law and the role of the coroner as they are likely to apply to surgeons in their professional activities are discussed. The duties of a consultant surgeon as a doctor and the mechanisms that are or shortly will be in place to ensure that these are complied with are described and the additional duties of that consultant as a trainer is also mentioned.



] 595



largely based. This is the law of tort or wrong done by one party to another. The two forms of tort that are relevant to surgeons are negligence and assault and battery. There are four essential elements of the legal tort of negligence: 1. 2. 3. 4.



there must be a duty of care; there must be a breach of that duty; harm must have occurred; the breach must have caused the harm.



Negligence occurs when party A is considered to owe a duty of care to party B, to have breached that duty and, as a result, harm has come to B. It must be shown that A could reasonably have foreseen that the breach might cause the harm. For instance, a driver travelling quickly on a country road should have it in mind that there might be a cyclist just around the corner. A second important aspect of negligence on which many clinical negligence cases fail is causation. This works on the so-called ‘but for’ test. It must be proved that, ‘but for’ the breach of duty by the defendant, the damage suffered by the claimant would not have occurred. It is perhaps a surprise to many people that assault and battery are sources of both civil and criminal litigation. Assault is the threat of physical violence and battery is any unlawful touching or physical contact of one person by another. The law makes an exception for ordinary day-to-day contact such as occurs in a bus queue or when handing over change for a newspaper but this does not apply to medical examination. Only with consent (which will be discussed further under Consent below) does a medical examination become lawful. Without consent, approaching a patient with a shining Lack’s tongue depressor at the ready is an assault and using it to depress the tongue is a battery. Although the act of opening the mouth probably implies consent, the patient has the option of lodging a civil claim alleging battery or even of reporting a possible criminal battery to the police. It is important to note that civil actions for battery, unlike those for negligence, do not require that there has been any harm. Patients seeking damages for lack of adequate consent can sue either in negligence (where the ‘but for’ test applies but consent must be truly informed) or in battery (where causation is not an issue but a much less rigorous consent is enough to refute the allegations).



NEGLIGENCE Standard of skill and care



TORT



In deciding negligence the claimant must establish four things.



Before discussing clinical negligence and consent issues, it is relevant to review the legal principle on which these are



1. A duty of care was owed to the claimant. 2. There has been a breach of that duty.



596 ] PART 9 SAFE AND EFFECTIVE PRACTICE 3. Damage has occurred to the claimant. 4. The breach must have caused the damage. To establish a duty it is necessary for a doctor/patient relationship to be in existence. There is no Good Samaritan law in the UK, and doctors only have a legal responsibility to people they have agreed to treat. In general practice the agreement arises by virtue of accepting a patient onto one’s list. In a hospital setting, the agreement either arises out of the doctor’s Terms and Conditions of Employment, or with private patients, as a result of contract. Patients tend to be seen by several different doctors and, therefore, the question: ‘who owes the duty of care?’ may be a pertinent one. In the National Health Service (NHS), the employing authority will usually be named as first defendant as it would be liable to pay damages arising from any successful action by virtue of the principle of vicarious liability. Individual doctors may then be named as subsequent defendants. An unresolved issue is who should be sued if a patient is placed on a waiting list, but suffers additional damage in the time it takes to receive a first appointment? Is the defendant the referring GP, or the Authority, or NHS Trust at whose clinic the patient is awaiting an appointment? The standard of care is that of the reasonably skilled and experienced doctor. The appropriate test, known as the Bolam (Bolam v. Friern Hospital Management Committee) test,2 states that it is: the standard of the ordinary skilled man exercising and professing to have that special skill. A man need not possess the highest expert skill; it is well-established law that it is sufficient if he exercises the ordinary skill of an ordinary competent man exercising that particular art.



The standard of care relates to the specialty in which the doctor practices. A GP will not be required to possess the skills of a specialist. An inexperienced doctor cannot rely on his lack of experience as a defence to alleged negligence (Wilsher v. Essex Area Health Authority).3 However, a junior doctor may discharge his duty by seeking the help of a superior. While a doctor is under a duty to keep himself appraised of developments in his area, this is subject to the bounds of reasonableness. Failure to read one article, which might have prevented the negligent act, could be excusable, while failure to be aware of new techniques that have become widespread may be inexcusable (Crawford v. Board of Governors of the Charing Cross Hospital).4 A doctor will not be negligent simply because he acted in a way that another doctor would not have done. The Bolam test establishes that a man is not negligent if he is acting in accordance with a practice merely because there is a body of medical opinion which would take a contrary view, provided there is a reasonable body of opinion



which supports the practice. A body of opinion must consist of at least two people. This statement has recently been qualified by the case referred to as Bolitho (Bolitho v. City and Hackney Health Authority).1



Causation In Bolitho, a registrar in paediatrics was accepted to be in breach of her duty of care by failing to visit an infant with breathing difficulties. Her defence was that, given the clinical circumstances at the time she should have visited, she would not have intubated the infant. The Court of Appeal had found that the failure to attend the infant constituted negligence. The House of Lords heard opposite expert opinions, one asserting that there was a body of opinion that would undoubtedly have intubated in those circumstances and the other that there was also a responsible body of opinion that would not have intubated. Their Lordships then reversed the judgement saying that the defendant’s assertion that she would not have intubated, a practice supported by a reasonable body of opinion, meant that on the balance of probabilities the brain damage would have occurred even if the registrar had visited. The causal link from the failure to discharge her duty of care by visiting the patient to the harm suffered was therefore broken so no negligence could be proved. Bolitho is important because it clearly brings Bolam into the field of causation and also because of the dissenting judgement of Lord Browne-Wilkinson. Referring to the reasonable body of medical opinion that would allow a Bolam-based defence, he said ‘The use of these adjectives – responsible, reasonable and respectable – all show that the court has to be satisfied that the exponents of the body of opinion relied on can demonstrate that such opinion has a logical basis’ (our emphasis). Later in judgement he qualified this by saying, ‘it will very seldom be right for a judge to reach the conclusion that views genuinely held by a competent medical expert are unreasonable’. It is not yet clear how these comments will be used as precedents for future cases but they certainly will be. An example might be the practice of stripping all the mucosa from a vocal cord in Reinke’s oedema. There is certainly a body of expert opinion that would condone the practice but, in view of our knowledge of the physiology of the superficial lamina propria, is it logical to operate in this way and can that body of expert opinion be considered responsible?



Proof of medical negligence The onus lies upon the claimant to prove that the surgeon’s negligent treatment caused the injury. It is not sufficient simply to prove that the surgeon’s actions were reprehensible or even reckless if the claimant cannot go



Chapter 48 Medical jurisprudence and otorhinolaryngology



on to show that their injury is directly attributable to the surgeon’s poor performance. Hence, while it was clear that a casualty officer was in breach of his duty of care when he failed to examine a patient attending his department with the obvious signs of poisoning, there was a complete defence (Barnett v. Chelsea & Kensington Hospital)5 when it was shown that there was no antidote to the poison taken by the patient. This is similar to the reasoning in Bolitho above, although that case was much more complicated than Barnett. In Barnett, it was a straightforward and uncontested statement that the workmen who had accidentally ingested arsenic in their tea were doomed whether or not they received medical attention. In Bolitho the courts had to decide whether or not, on balance of probabilities, the hypoxic brain damage would have occurred even if the registrar had visited the child. The standard to which the claimant must prove his case is on the balance of probabilities. This means that the claimant must show that his version of events and expert analysis are more likely to be true than those put forward by the defence. If the case for each side is evenly balanced then the claimant will fail. Hence, where a patient suffers a nerve palsy that could equally have been due to negligence, or could have occurred as an inherent risk of the operation even when performed with proper skill and care, the claimant’s case will fail in the absence of some item of evidence to tip the scales in favour of negligence (Ashcroft v. Mersey Regional Health Authority).6 An apparent reversal of the burden of proof can occur where the likelihood of negligence is so obvious that it ‘speaks for itself ’ (res ipse loquitur). In other words, a defendant may find that he is obliged at least to provide an explanation of the patient’s injury that is consistent with reasonable care having been taken, even where the claimant has no positive evidence of negligence, where:  there is no evidence as to how or why the accident occurred;  the accident is such that it would not occur without negligence;  the defendant is proved to have been in control of the situation in which the accident occurred.7 The doctrine first evolved in simple personal injury cases concerning falling objects (Scott v. London & St Katherine’s Docks,8 Byrne v. Boadle9 and Pope v. St Helen’s Theatre10), but has obvious attractions to a claimant in a medical case, where there may well be very real uncertainty as to how an injury came about, and where the often unconscious patient is entirely under the control of the medical team. Despite the fact that it is often pleaded, in England the doctrine has more often been conceded than litigated in medical cases, and so its ambit is unclear.11 In practice the courts are reluctant to apply a doctrine derived from the relatively simple ‘bumps and



] 597



thumps’ of stevedoring to the complex issues of causation found in medical litigation. It has been said: The human body is not a container filled with a material whose performance can be predictably charted.... because of this medical science has not yet reached the stage where the law ought to presume that a patient must come out of an operation as well or better than when he went into it.12 (Girard v. Royal Columbian Hospital)



Thus, as a matter of law, the onus of proving negligence will almost always fall on the claimant, although in cases where there is a strong and obvious inference of negligence from the very facts themselves (e.g. a retained swab (Mahon v. Osborne)13 or where the claimant woke from his anaesthetic with the septic finger still attached and an adjacent one in the bucket14) the defence may well find that onus discharged unless they are able to provide some alternative theory, not involving negligence, to answer the claimant’s case.



Damages A claimant having established negligence, and having proved injury as a consequence, the question turns to the assessment of the compensation to be paid. The court must award a sum of money that will, as nearly as possible, put the injured person in the same position as he would have been had he not been injured (Livingstone v. Rawyards Coal Company).15 In some jurisdictions (notably the USA) an element of punitive damages may be awarded. Such an approach has been rejected in England and Wales (Kralj v. McGrath)16 and the rest of the UK for mere inadvertent negligence. The argument is based on mens rea being present before the claimant can be given a monetary award by way of punishing the defendant. Damages are broadly classified into two categories.



GENERAL DAMAGES



These include damages for pain, suffering and loss of amenity. This is the aspect of the claimant’s loss which is not ascertainable by a mathematical calculation of economic loss and, in the past, was determined by a jury (Ward v. James).17 Juries were effectively abolished in personal injury trials in the mid-1960s. While the level of general damages will depend on the claimant’s circumstances, and the suffering which the particular injury has caused the individual, a general bracket for a particular injury will be determined with reference to awards (adjusted for inflation) from previously decided cases. Obviously, it is impossible to ‘compensate’ someone for loss of a limb or for a life of continual pain so the awards are essentially conventional figures. As a guide, the



598 ] PART 9 SAFE AND EFFECTIVE PRACTICE current conventional figure for injuries of the utmost severity is between £120,000 and £150,000, and so injuries of lesser significance will be a proportion of that ‘maximum’. In a recent case in Teeside (Levitt v. Hartlepool Area Health Authority),18 a minor born deaf was awarded £50,000 general damages (adjusting for inflation in 2003 this would probably be £60,000), which included her total hearing loss.



SPECIAL DAMAGES



These are the specific ‘out-of-pocket’ monetary expenses and losses which the claimant has incurred up until the date of the trial. These will include loss of earnings, medical expenses, travel costs and the cost of any special equipment consequent upon the injury. Often there will be a claim for losses expected in the future. (Future loss is technically an item of general damages as its assessment is uncertain and it was formerly a matter for the jury to determine. As a matter of practice, the calculation of future loss is inseparable from that of past loss, many of the same arguments rehearsed in both, and one flows naturally to the other.) The amount of money at issue will depend almost exclusively on the particular circumstances of the claimant (Lim v. Camden Health Authority),19 guided by the principle that a claimant should only recover what he has lost as a result of the injury. He must prove that his expenditure was (or will be) reasonably necessary, and that his needs cannot be met more cheaply by other reasonable means. As a rule, English law seeks to compensate a claimant by way of a single lump sum award. This creates obvious problems where a claimant is likely to remain unemployed or requires continuing care for many years in the future. The only certain result is that the claimant will be either under- or overcompensated, depending on how circumstances unfold. This problem has not been solved, but ameliorated in certain cases. Where a claimant suffers a condition that may deteriorate significantly in the future, rather than receive a small sum to represent the risk, they can apply for an order for ‘Provisional Damages’, which leaves it open to the claimant to come back to court at a later date.20 In very large claims where the claimant’s life expectancy will determine the level of the award, the risk can be borne by an insurance company by the purchase of an annuity. This is known as a ‘Structured Settlement’.



Dealing with medical negligence claims What happens when a surgeon is sued?21 The Civil Procedure Rules 199822 are considered below. They require that, in cases of clinical disputes, a pre-action protocol be followed. This sets procedures and time limits for various stages of the process of being sued. There are



also recommendations for Trusts about how they should respond to a claim and what processes should be in place to ensure that they are able to respond in an appropriate and timely way. In effect, the pre-action protocol will penalize a Trust which has such a poor clinical governance structure that it is unable to respond to a claim in a relatively short time. What the individual consultant surgeon will experience if his patient sues him alleging negligence will differ in some ways depending whether the patient was a private patient or an NHS patient. The first stage of the proceedings is the same: the patient, having found some means of financial support for his or her claim against the health care provider (a matter outside the remit of this chapter but considerably more difficult than formerly) will request, via a solicitor, release of notes. In private practice, the consultant will need to release these notes himself and is wise to check with his anaesthetist, the private hospital and other doctors involved in the secondary care that they will do the same. The consultant would be most unwise at this stage if he did not inform his medical indemnity organization and seek their help and advice. If the case is an NHS one, the matter must be referred to the Trust’s legal department immediately. If the clinical governance system in the Trust is working well, there is a good chance that the events have already been reported to the legal section via an incident form and that contemporaneous statements have already been taken and filed in preparation for a potential claim. Alternatively, the claimant may have already used the NHS complaints’ procedure so all the statements from that will be available. A Trust is vicariously liable for the actions of its employees and so the legal department should take over all further responses to solicitors and communication with the defendant legal team. Both Trusts and independent practitioners have 40 days to provide the copy medical records and failure to meet this deadline will prejudice any defence. The potential claimant’s solicitors will now obtain an expert report on the medical records with examination of their client if appropriate and can decide whether or not the claim has merit. If it has, they will serve a letter of claim on the health care provider setting out a summary of facts, the main allegations of negligence, description of injuries and an outline of losses. This may include an offer to settle. The defendant organization or individual must provide a reasoned answer within three months of receipt. In addition, the defendant’s solicitors will be concerned to obtain guidance from experts swiftly and, once received, be able to form an initial view about whether or not the claim is defensible. The assessment of a claim at this stage will be facilitated by recent rule changes which have been made concerning the conduct of personal injury actions, including medical negligence litigation. These changes require that at the time of service of a writ, a claimant must also submit a medical report setting out



Chapter 48 Medical jurisprudence and otorhinolaryngology



the claimant’s condition and prognosis, together with a Schedule of Special Damages said to result from any alleged negligence. Traditionally, such information was not available until much later in the litigation process. The change has resulted in a defendant’s solicitors being able to assess the potential value of a claimant’s claim at an early stage. Accordingly, following that initial expert opinion, a settlement can be proposed early in the proceedings on behalf of the otolaryngologist, if appropriate. If settlement is desired but cannot be negotiated between the solicitors, then the solicitors acting for the otolaryngologist may try to settle the claim by what is known as a ‘payment into court’. In making a payment into court, the defendant can make an assessment of the damages which might be awarded in due course at trial, if negligence is established. If the trial judge awards a sum higher than the level of the payment in, the claimant will receive his or her costs in the usual way. If, however, the sum awarded is the same or a lesser sum then, as a general rule, all costs of the action, including the defendant’s costs after the payment into court, must be borne by the claimant. The claimant has 21 days to decide whether or not to accept the payment in. Thereafter the costs conservancies will operate. In cases where the claimant is legally aided, this device is of more limited assistance to a defendant. In other circumstances, there may well be a strong incentive to a claimant to accept a payment into court through fear of being prejudiced on costs in this way. Following the service of a defence, in response to allegations contained in the statement of claim, both parties may raise requests for further information about their respective pleadings, and the claimant may choose to file a reply to the defence. Directions concerning the conduct of a claim will be given by the court, usually after the pleadings are completed. In the High Court, a formal hearing will take place to consider these directions, but in personal injury actions brought in the County Court, a series of so-called automatic directions is given by the court. These are rarely, if ever, appropriate in medical negligence cases and it is therefore usual for the parties to apply to the court for specific directions to be given. These directions will provide for, among other things, the disclosure of witness statements as to fact and expert reports. In the past, in medical negligence actions, neither expert reports nor witness statements were disclosed by one party to another. Only at trial would the nature of the respective cases become clear. However, as a result of recent cases and changes in the procedural rules, expert reports and witness statements must now be disclosed, and this will be ordered by the court at a directions hearing. It is usual for witness statements to be disclosed, followed shortly thereafter by expert reports, in order that experts can consider this information available from the other side in preparing the report. The exchange of



] 599



both statements and reports gives a further opportunity for review of the case, both by claimant and medical defendant. At this stage the claimant may realize that the claim is weak and the case may be discontinued. Equally, the defence may feel that the case is not defensible and settlement negotiations or a payment into court may follow. TRIAL AND PREPARATION FOR TRIAL



If the exchange of witness statements and expert reports does not promote the settlement of a claim, then the case proceeds towards trial. In medical negligence actions, a fixed date for trial will usually be given by the court because of the significant number of clinicians who may have to make themselves available, either as experts or as witnesses of fact. By way of preparation for trial, both parties should arrange conferences with counsel in order to review the case in detail, and ensure all preparations are complete. If the defence considers that the case should be settled, or there are certain aspects of the case where a defendant may be vulnerable, a payment into court may be made. The payment in may be limited to those aspects of the claim where the defendant could be found liable. The claimant has 21 days within which to accept the payment in before the penalty of costs starts to run. As a significant proportion of the costs of an action result from the trial itself, the defendants are usually anxious to make any payment into court before 21 days in advance of the trial. Thereafter bundles of documents must be prepared for use in the court by all parties, which will include the pleadings, medical records, witness statements and expert reports. A proportion of cases are settled literally at the doors of the court, the last opportunity for compromise before further significant costs of trial are incurred. If no compromise can be reached, the trial will commence with an explanation of the case to the judge by counsel for the claimant, setting out the relevant events and the nature of the allegations. The claimant’s counsel will then call witnesses of fact, usually followed then by expert witnesses. Each witness will be cross-examined in turn by the defence, and then re-examined by claimant’s counsel if necessary. When the claimant’s evidence has been called, the defence case is then put in the same way, and usually in the same order. The trial judge may allow variation in this order of evidence, particularly in complex cases, so that witnesses for both sides are called first, to be followed then by experts. This will allow experts to hear all the evidence of fact before giving a final opinion. However, the present arrangements for disclosure of witness statements make such variation rare. At the conclusion of the case, the judge will usually hear submissions from counsel on the law to be applied and the appropriate level of damages to be awarded if the



600 ] PART 9 SAFE AND EFFECTIVE PRACTICE claimant is successful. It is usual then for the judge to reserve judgement, to be delivered at a later day, as most negligence cases are complex and will require some consideration. Once judgement is given, and if the claimant is successful, the judge will indicate the level of damages to be awarded. Only at that stage is the judge informed about any payment into court which may have been made, and the costs to be awarded can then be considered.



FUNDING OF NEGLIGENCE CLAIMS



Otolaryngologists may well have been concerned at the possibility that medical negligence actions in England and Wales will follow the pattern of the USA, and that the perceived increase in the number of actions here may continue. However, with the reduction in the eligibility for legal aid, many actions which would have been funded by legal aid in the past will now have to be funded privately. The introduction of contingency or conditional fees is unlikely to assist all those who are no longer eligible for legal aid. As an incentive for the claimant’s solicitor, the conditional fee system23 will allow the solicitor to claim twice the rate of fees they might have otherwise obtained in a successful action. There is a risk that the solicitor might receive no payment at all if the case is unsuccessful. As suggested above, solicitors acting for patients are therefore likely to take on fewer claims,24 concentrating on those where liability is obvious and can be quickly established. The result is likely to be that there will be no significant increase in the number of medical negligence actions in the short term. Indeed, it is possible that there will be a reduction at the expense of patients who might otherwise have had successful claims for compensation.



CONSENT Much medical litigation is caused by the practitioner’s failure to disclose adequate information about the risks inherent in a given procedure. Common law recognizes the principle that every person has the right to have his bodily integrity respected. There is a presumption that a person should not be exposed to risk unless he has voluntarily accepted that risk, based on adequate information and adequate comprehension. Obtaining consent to carry out an operation is necessary to avoid three sorts of legal jeopardy. The first and second arise because any touching of another person is a potential battery. (Assault is the threat of causing physical injury. An unconscious person can be battered but cannot be assaulted.) This means that, in very exceptional circumstances, a surgeon performing an operation without consent may be prosecuted in the criminal courts for alleged criminal battery. A more likely



outcome of such surgery is a civil case seeking damages for the tort (or wrong) of battery. There is no need to show that anything went wrong with the surgery, merely that it was carried out without valid consent. The third, and much the most likely legal outcome regarding consent, is that the claimant tries to show that he was not warned of the risks of a proposed therapy that resulted in harm and that, had he been warned of those risks, he would have refused the treatment. An example might be a patient with a dead ear after a competently performed stapedotomy who was not warned of that risk before surgery. He may well be able to claim damages from the surgeon because he can say that ‘but for’ the surgeon’s failure to warn of the risk, he would not have undergone the surgery. It is immediately obvious that any careful surgeon must record what potential adverse outcomes of a proposed operation have been discussed with a patient. If it is not recorded in the notes, courts have become sceptical of a statement such as ‘it is my usual practice to warn of these risks’. The question of how much information must be given to the patient will vary from situation to situation, but is generally set by the professional standard according to the Bolam test, with doctors being arbiters of how much information should be given. This means that, in the example above, if the defendant surgeon could assemble a respectable body of opinion whose practice would not be to warn of the risk of dead ear, his conduct did not breach a duty of care. Notably, in England and Wales, there is no notion of informed consent, as in the USA, whereby the amount of information to be disclosed is dictated by what a patient would want to know. This is modified to a degree by the case of Sidaway (Sidaway v. Board of Governors of the Bethlem Royal Hospital)25 in which the House of Lords held that where the proposed treatment involved a substantial risk of grave or adverse consequences such that, notwithstanding any practice to the contrary, a patient’s right to decide whether to consent to the treatment was so obvious that no prudent medical man could fail to warn of the risk (save in emergency or some other sound clinical reason for nondisclosure), then it would be negligent not to warn. To continue the example above, the claimant could argue that total deafness in one ear is such a disability that any reasonable patient would wish to be told about it preoperatively. Accordingly, the right of the doctor (acting in accordance with a reasonable body of medical opinion) to decide what the individual patient should be told remains enshrined in English case law. Such medical paternalism, which may be in the best interest of some of our patients, is currently and repeatedly being questioned. The doctor should realise that in Sidaway when the question ‘is informed consent a part of English law?’ was put to the five law lords, the answer was not unanimous. Scarman said ‘Yes’, Diplock said ‘No’ and Bridge, Keith and Templeman said ‘Yes with reservations’.



Chapter 48 Medical jurisprudence and otorhinolaryngology



In assessing which material risks should be mentioned, doctors should consider the degree of probability of the risk materializing and seriousness of possible injury if it does. A risk, even if it is a mere possibility, should be disclosed if its occurrence would cause serious circumstances (Hopp v. Lepp).26 Medical evidence will be necessary for the court to assess the degree of probability and the seriousness. A further medical factor upon which expert evidence will also be required is to assess the character of the risk, i.e. is this risk common to all surgery or is it specific to the particular operation? Special risks inherent in a recommended operation are more likely to be material. The legal standard of disclosure required in response to direct questions is also set by the professional standard (Blyth v. Bloomsbury Health Authority).27 Although the amount of information given must depend upon the circumstances, as a general proposition it is governed by the Bolam test (supra). In Australia, the doctrine of informed consent has gone one stage further from Sidaway where the reasonable patient was considered the arbiter of what risks should or should not be disclosed preoperatively. In Rogers v. Whittaker,28 it is clear that the risks that must be disclosed are those that affect that one patient and not the generality of reasonable patients. To take the stapedotomy analogy one last stage further, few surgeons would warn of a dead ear risk of 1 in 2000 operations and probably the reasonable patient would be unlikely to want to know that this was a complication as it is so unlikely. The question arises of whether the patient having that operation on an only hearing ear might view the risk differently and want to be aware of it in making a decision.



Who can give consent? There are three categories of patient described in the Department of Health Reference Guide to Consent for Examination or Treatment: the adult with capacity to consent; the adult without capacity to consent; and children and young people.



] 601



parents, or others acting in loco parentis, may give their consent or refusal to medical treatment. Recent case law suggests that it may still be possible to treat a seemingly competent child who is refusing to give consent, providing someone else with the capacity to consent provides consent on the child’s behalf (In re R, and in re W).30, 31



MENTALLY INCOMPETENT PATIENTS



Where a patient is unable to provide consent on his or her own behalf by reason of mental incapacity, no one else, including a court, may give consent on that person’s behalf. The doctrine of necessity, however, permits a doctor to lawfully operate on or give other treatment to adult incompetent patients, provided that the treatment is in their best interest, either to save their lives or to ensure improvement in their physical or mental health (Re F).32



JEHOVAH’S WITNESSES



Certain groups of patients, including Jehovah’s Witnesses, may refuse to receive blood transfusions or other lifesaving therapies. The Court of Appeal has affirmed patients’ rights to refuse medical treatment, even if this will result in their death; nevertheless, for such a refusal to be effective, the court must be satisfied that at the time of refusal the patient’s capacity is not diminished by illness or medication or given on the basis of false assumptions or misinformation. In the case of Re T,33 Lord Donaldson said: An adult patient who suffers from no mental incapacity has an absolute right to choose whether to consent to medical treatment, to refuse it or to choose one rather than another of the treatments being offered.



A decision to refuse medical treatment does not have to be sensible, obviously rational or well considered, and in the case of a competent patient, the doctor cannot override the patient’s wishes because he believes it to be in the patient’s best interests.



CHILDREN



Section 8(1) of the Family Law Reform Act 1969 provides that a person over 16 may give a valid consent to medical treatment as though he was an adult. As regards children under 16, the general principle is that laid down in the Gillick (Gillick v. West Norfolk and Wisbech Area Health Authority) case,29 that the parental right to determine whether or not a child under 16 should have medical treatment terminates when the child achieves a significant understanding and intelligence to enable him or her to understand fully what is proposed. Until such time, the



OTHER SPECIAL GROUPS



Pregnant women, patients who are human immunodeficiency virus positive or suffering from acquired immunodeficiency syndrome and the elderly do not represent special categories for the purposes of consent. Although the amount of information which has to be given to a patient varies from case to case, a decision to withhold information solely on paternalistic grounds that it may deter the patient from accepting the therapy may not be justified in law.



602 ] PART 9 SAFE AND EFFECTIVE PRACTICE



CONFIDENTIALITY Doctors are well aware of their duty, originally stated in the Hippocratic oath, of confidentiality. What is often not appreciated is the common law and statutory basis of this duty and the circumstances in which confidential information may legally be revealed. There is a common law duty placed upon professionals to guard confidential information. A patient may sue a doctor through the civil courts if there is alleged to have been damage as a result of breach of confidentiality. The Data Protection Act (1997) places a statutory duty on anyone who holds patient identifiable data (in any form) to guard and release that information under strict controls. Any surgeon who keeps any sort of patient identifiable information at his residence (where his employer’s blanket registration under the act will probably not apply) and who is not registered with the data commissioner breaches the provisions of the act. It is particularly important to note in this context that, if disclosure is properly requested, it is an offence to fail to disclose ALL of the records, both computerized and manual. The second statute that obliges surgeons to care for information appropriately is the Human Rights Act (1998). This enshrines in UK law the sixth clause of the European Convention on Human Rights granting citizens an absolute right to privacy and to family life. There are a number of instances in which confidential information can be released. The first is with the consent of the patient. This is sometimes implied consent such as applies when a GP releases confidential information in a referral to a consultant (and so to both secretaries). On other occasions, there needs to be expressed consent, for instance in release of information to insurers when a holiday must be cancelled. Sometimes a doctor is obliged to release information. The statutory examples concern the Terrorism Act 2000, the Abortion Act 1967 and the infectious diseases regulations. A court has power to require disclosure of records and there are several cases where disclosure is appropriate in the public interest. These are to protect the rights of others, such as reporting to prevent or detect serious crime or to prevent suspected child abuse and in reports to the Driver and Vehicle Licensing Authority. In all cases of this nature where the disclosure is not governed by the Data Protection Act, the information released must be no more than that necessary to comply with the public duty. A surgeon with any doubt as to how to resolve the conflict between duties of confidentiality and the public interest is well advised to consult his or her medical indemnity organization and to ensure that the final decision is one with which a responsible body of medical opinion would concur (Bolam). The situation after the death of the patient is different. Doctors have a duty to disclose on death certificates and to national confidential enquiries as well as to the police and courts, especially the coroner. Data protection no longer applies and the relevant statute is the Access to Health



Records Act 1990. This clarifies that executors or, failing them, next of kin can consent to release of medical records. It appears that anyone with a legitimate claim to view the record can see that part of the record that is relevant.



THE CORONER By the time this chapter is published, it is likely that the changes in the role of the coroner will be clearer. For some time, there has been unease about exactly what coroners are expected to do but the moves to reform have been hastened by the Shipman case and the subsequent inquiry. Harold Shipman, a general practitioner, was found guilty of murdering a large number of his patients. For a number of successive years, the local coroner received considerably greater numbers of reported deaths of patients of Dr Shipman than of any other local GP. Despite this, no alarm bells appear to have rung and Dr Shipman’s murderous activities continued unabated. The Shipman Inquiries which followed the criminal trial have recently been completed and they seem likely to make wide-ranging recommendations about the role of the coroner and the General Medical Council (GMC). There is also an independent inquiry into the office of coroner which may possibly arrive at a different conclusion. As things stand at present, the coroner for a particular area is appointed by the Crown and is usually a lawyer. There are a few medical coroners but these are less prevalent than previously. The coroner has a number of duties of which the investigation of certain deaths is the one relevant to this chapter. In the course of these investigations, the coroner has wide powers to call or subpoena witnesses and to make whatever investigation he sees fit to establish the facts relevant to the death. The equivalent officer in Scotland, the Procurator Fiscal, has similar powers, which may be even wider. When the coroner is mentioned, most doctors ask three questions: what deaths must be reported to the coroner; what verdicts can be brought and will these criticize the doctor; what must I do if required to write a report for or appear before the coroner? The following deaths must be reported to the coroner:  violent or unnatural death;  death in custody (even if the prisoner was licensed on leave to attend the hospital);  death where no doctor is able to issue a death certificate. Most jurisdictions also have local agreements about the reporting of deaths within a certain period after hospital admission and after surgical procedures. Coroners usually wish to know about deaths from notifiable diseases, industrial diseases and those in receipt of war pensions. The range of verdicts that the coroner can return is surprisingly limited. The coroner’s court is (in theory at



Chapter 48 Medical jurisprudence and otorhinolaryngology



least) not adversarial and no finger of blame should be pointed by the court. The coroner’s remit is to establish the identity of the deceased, the time and place of death and the mode of death. Although death due to gross neglect is an acceptable verdict, death due to negligence is not. In practice, of course, the family of the deceased and their lawyers may use the proceedings of the coroner’s court to assess whether or not an action in negligence is likely to succeed. What about the surgeon called upon to provide a report to the coroner? The first piece of advice would be for the surgeon to contact his or her defence organization for support. The report itself should be factual and full and directed to a nonmedical audience. The report should not contain opinion but just facts. In particular, a report to a coroner should never contain the words ‘negligent’ or ‘negligence’. Not only are these a matter of opinion and not fact, but they also have special well-defined meanings for lawyers that may not be apparent to doctors. In most cases, the report will be all that the coroner needs and the doctor will not be required to appear. If the coroner feels that the evidence that the doctor can give may be crucial to his investigations, then he will summon the doctor to the inquest. Sometimes the family of the deceased will see this as their opportunity to confront the doctor who they see as the source of their grief. To be fair, sometimes the process works well, the family see that the doctor is not some kind of ogre and the resentment fades. Even if this does not occur, the coroner should ensure that the family do not persecute the doctor and that the proceedings remain directed towards establishing the who, when, where and how but not why the deceased perished. If the doctor anticipates a difficult inquest, he may be accompanied by legal assistance but the solicitor may not answer for him nor is there any cross-examination of the other side. Defence organizations can advise the doctor but there is always a risk that the presence of a legal advisor will make the family (and the coroner) suspect that some sort of cover-up is being attempted. Doctors summoned to the coroner’s court need to follow the usual rules of court appearances – dress smartly but conservatively and do show appropriate respect to the coroner. Bring the notes or fair copies and copies of the reports already supplied and read them through before the case opens. Remember that the doctor is a witness and not an advocate or defendant and so there is no problem in asking for a question to be repeated or for a pause while the notes are consulted before answering.



THE ROLE OF THE EXPERT An expert medical witness is instructed by either the defence or the claimant to give his or her opinion on the medical matters pertinent to the case to the court. This evidence is different from that of a material witness who is



] 603



asked for the facts. The expert’s job is to interpret the facts for the court. In simple terms, it is ‘to give impartial advice and opinion’. Many otolaryngologists will be called upon to give expert opinion on personal injury cases where neither the claimant nor the defendant is a clinician. The many reports on noise-induced hearing loss that have been written by otolaryngologists are a form of written expert testimony in just such personal injury cases. A few otolaryngologists will be asked to prepare reports that assist courts in deciding clinical negligence claims and the special circumstances that pertain to these are discussed below. Until the Civil Procedure Rules 1998 came into force, experts were retained by either the claimant or defendant’s side and to some extent this remains the case. In theory, an expert’s report would be the same whether he was instructed by the solicitors of the claimant or those of the defendant. The 1998 Rules have made it clear that the duty of an expert, whether he is expecting to be paid by the defendant or by the claimant, is to produce for the court an unbiased opinion. The 1998 Rules encourage a court, where practical, to appoint a single joint expert to produce such an opinion. They also give direction about exchange of reports, communication between and with experts and the expert’s right to ask the court for directions. In practice, the expert’s evidence is usually contained in a report and it is rare for an expert to have to appear in court. The form and content of an expert report were previously based upon a well-known civil case. In the Ikarian Reefer (National Justice Compania Naviesa SA v. Prudential Assurance Company Ltd, The Ikarian Reefer),34 Mr Justice Cresswell stated that he considered that a misunderstanding on the part of some of the expert witnesses had taken place concerning their duties and responsibilities which had contributed to the length of the trial. Although this was a shipping case, the seven duties and responsibilities laid down have equal validity for medical experts. 1. Expert evidence presented to the court should be, and should be seen to be, the independent product of the expert uninfluenced as to form or content by the exigencies of litigation. 2. Independent assistance should be provided to the court by way of objective unbiased opinion regarding matters within the expertise of the expert witness. (An expert witness should never assume the role of advocate.) 3. Facts or assumptions upon which the opinion was based should be stated together with material facts which could detract from the concluded opinion. 4. An expert witness should make it clear when a question or issue fell outside his expertise. 5. If the opinion was not properly researched because it was considered that insufficient data was available then that had to be stated with an indication that the opinion was provisional. If the



604 ] PART 9 SAFE AND EFFECTIVE PRACTICE witness could not assert that the report contained the truth, the whole truth and nothing but the truth, then that qualification should be stated on the report. 6. If after exchange of reports an expert witness changed his mind on a material matter then the change of view should be communicated to the other side through legal representatives without delay and when appropriate to the court. 7. Photographs, plans, survey reports and other documents referred to in the expert evidence had to be provided to the other side at the same time as exchange of reports. More recently, the Civil Procedure Rules have modified these directions. The expert should obtain the Civil Procedure Rules Chapter 35 and read sections 1.1 through to 1.6, which is too extensive to reproduce here. Before writing a report the expert should be aware of what the solicitor requires. Reports usually refer to one or more of the following six areas which are: 1. An initial statement on the possible merits of an allegation for a claimant before notes and other evidence are obtained. 2. Liability. 3. Causation. 4. Current condition. 5. Prognosis. 6. Expert opinion on an area of medicine. A quote should be given on the cost of the report in advance. Lawyers may have no concept of the cost in time and research to answer what to them may be the most simple of questions. The solicitor will not be pleased to receive a report of 100 pages with detailed bibliography costing £2,000 when the damages sought are only £500 for a relatively minor event. The report should be double spaced and typed on A4 paper. Each sheet should have the name of the claimant or defendant typed in the top right-hand corner and be separately numbered. It is also helpful to number the paragraphs. The names of the parties should be stated as should the requesting solicitor or insurance company. It is inadvisable to use the word ‘negligent’ in the report. Negligence may be implied by using phrases such as ‘falling below an acceptable standard’ or ‘followed a course of action that could not be supported by any body of medical opinion’. Phrases such as ‘reckless action’ or ‘flagrant disregard’ may have a special meaning and lead to criminal charges rather than a civil case. There is always the possibility that the credentials of the expert will be tested in court. It is extremely unwise for the expert to step outside his field. Should he be forced to comment outside his field of expertise he should add the rider that he is speaking only as an average medical practitioner. The expert will need to be au fait with up-to-date research in his own field.



The medical expert in clinical negligence The medical expert retained in a case of medical negligence is usually asked to elucidate the areas of medical contention within a case. He needs to be aware of the various views on current practice even if these views are held by a minority of doctors, provided always that the minority is a reasonable one and that the practices supported are logical (Bolitho). Although the temptation is to be an arbitrator of medical colleagues, in court, this is the province of the judge. Nevertheless, the expert will be asked to comment as to whether the claimant’s complaint has merit. When acting for the defence the expert will inevitably come across cases where defence is impossible. Here it is the expert’s duty to advise that a speedy settlement be made to the aggrieved patient. This will have the secondary benefit of avoiding a colleague’s professional shortcomings being exposed to public criticism in court. Rarely, some doctors expect their colleagues to defend them whatever the circumstances. Hence, being an expert can lead to criticism or alienation by colleagues. Medical negligence cases can take an inordinate amount of time. Not only may the expert have to inspect all the records, examine the claimant and prepare reports, he will have to do research, attend meetings with solicitors and counsel and attend court. A single case may, in unusual circumstances, take up to 100 hours of time or more. The expert should never take on a case if he cannot afford the time. Subsection 1.6 of the practice direction requires that experts comply with the relevant approved expert’s protocol. As might be expected, there is a detailed preaction protocol for the resolution of clinical disputes. Although this is aimed primarily at the legal profession, it is required reading for any doctors involved in providing expert testimony in clinical negligence cases. In particular, the times allowed by the court for the completion of various stages of the claim and the defence are clearly stated.



The doctor as an expert witness in court The purpose of expert opinion given in court is to persuade the judge that one side of a case has greater merit than the other. It is the judge who will decide between the two sides of the argument. The expert should not be tempted to usurp the role of the judge for this may do untold damage to his own side or at the very least earn a rebuke which may undermine his confidence. In court, the expert should wear conservative clothes. Evidence should be given in a straightforward, unequivocal manner. A personal view may be represented especially in response to a direct question but should always be tempered with information about acceptable



Chapter 48 Medical jurisprudence and otorhinolaryngology



alternative practice and opinion. He should always be prepared to concede points if it is appropriate to do so, and not to adhere rigidly to one view when that cannot be sustained. The expert often has difficult concepts to convey to the judge. The expert should not hesitate to use pictures, models or even video to illustrate a point but should avoid being seen as a flamboyant ‘show off ’ lest he discredits himself by not giving due respect to the court. The usual course of examining a witness in court is that the barrister for the side calling the witness will examine that witness and try to anticipate and pre-empt difficult questions from the other side. Counsel for the opposing side will then cross-examine the witness and may attempt to undermine the evidence given by the expert or the standing of the expert in his profession. Above all, the expert must not see this as a personal insult lest he should lose his temper and hence his dignity. After this potentially hostile cross-examination, the first barrister is permitted to re-examine the witness and try, if necessary, to restore the faith of the judge in that witness.



PERFORMANCE AND COMPLAINTS Besides using the legal system to sue a surgeon in tort, patients have recourse to a number of mechanisms to resolve concerns about how they have been treated or managed. Some of these mechanisms have the potential to involve doctors in disciplinary or performance investigations so the two elements are considered together. A patient or a family unhappy with the outcome of a surgical process can of course sue in the tort of negligence but this is expensive and will succeed in only a few cases. Another option is a complaint to the Trust employing the surgeon. Trusts are obliged to have a complaints procedure and the patient is entitled to a prompt reply from the chief executive. In practice, the time limits are so tight that especially when the complaint is a complex clinical one (rather than the more usual gripes about hospital food or parking for visitors) the first reply is a form letter. Following this the complaint should be answered point by point with an apology if appropriate. If this is rejected by the patient then the complaint is passed to a convenor (usually a nonexecutive director of the Trust) whose job is to decide whether or not to convene an interim review panel to try to resolve the matter. This should be done only if, in the convener’s view, there is no other way to resolve the complaint and there is a good chance that the panel will achieve such a resolution. The convener also sets the terms of reference of this investigation. A glance at the advice to conveners on this suggests that this structure was neither well thought out nor transparently nonpartisan. Conveners who refuse to sanction further investigation are seen as agents of the Trust that they work for. Aggrieved patients tend to see the narrow terms of reference that may be



] 605



required to allow the review panel to function as further evidence of attempts to quash legitimate complaints. Some patients turn at this stage to clinical negligence proceedings; as soon as this happens, the complaints process ceases. The independent review panel will in time produce a report to the complainant, the complained against and the Trust chief executive. The chief executive must then report to the complainant what action has been taken. Among the possible actions are Trust disciplinary procedures or report of a doctor to the GMC. Before considering what threats the GMC poses to surgeons, the Health Service Commissioner must be mentioned. This officer, also referred to as the ombudsman, can review the decisions of the convenor not to set up an independent review. There may also be grounds for appeal to the ombudsman if there are problems with the terms of reference or the conduct of the independent review. Following general dissatisfaction with the process and especially the role of the convenor, there has been a process of consultation followed by a report on how the system might be improved. If the legislation goes through to put this report into practice, then the convenor will disappear and the follow-up to failed local resolution will be the Healthcare Commission. If the investigation of a complaint leads to further investigation of the conduct or performance of a doctor, there are a number of forms that this can take. One is an internal investigation by the doctor’s Trust. Although there are regulations that govern the conduct of such investigations, there is such a variety of approaches that this chapter cannot be dogmatic. It is clear that a Trust must not unjustifiably suspend a doctor pending investigation and then drag its heels in the investigation. There may well be a case against the trust under the Human Rights Act (1998) and the doctor is probably best advised to consult his or her defence organization. Two new organizations may put a doctor’s reputation or right to practise in jeopardy. These are National Patient Safety Agency (NPSA) including the National Clinical Assessment Authority (NCAA) and the Healthcare Commission. The Healthcare Commission arose out of Commission for Healthcare Audit and Inspection which arose from the Commission for Healthcare Improvement (CHI). The NCAA will deal with concerns about the performance of individual doctors by investigating problems and arranging performance assessments where necessary. Memoranda of understanding have been agreed between the NCAA and both CHI and the GMC so it is clear that a suggestion of deficient performance or conduct uncovered by one of the three organizations may well be reported to the other two. The organization that these newer groups feed to and which has disciplinary control over doctors in the UK is the GMC. This organization has a statutory duty to maintain standards and ensure that doctors registered with the GMC do not endanger the public. The GMC is bedevilled by lots of committees with similar acronyms



606 ] PART 9 SAFE AND EFFECTIVE PRACTICE but there are basically three routes through the system. The first stage is usually the reporting to the GMC by a Trust, by a member of the public, by another doctor or the NCAA. Increasingly, the GMC is looking at reports in the media which suggest poor performance or conduct and initiating proceedings on the basis of these. The complaint is screened and the decision made to reject it at this stage or to refer it on for investigation. The screener decides which of the three routes: health, conduct or performance is the most appropriate. Health procedures are aimed to rehabilitate the ill doctor whose illness is affecting his or her work. The four stages of the process are initial assessment, medical examination, medical supervision and health committee action. The possible outcomes are suspension, conditional continued registration to practise or a decision that the doctor is fit to practise without conditions. The performance procedures act where there is ‘a departure from good professional practice . . . sufficiently serious to call into question a doctor’s registration’. Referrals are examined and the doctor is likely to be asked to have a performance assessment. If he or she refuses, the Assessment Referral Committee may order an assessment. The case coordinator uses the outcome of the assessment process to decide that: no further action is required; counselling or training is required; or the case is so serious or the doctor has refused assessment or to agree to retraining such that referral to the Committee on Professional Performance is required. This committee can leave the surgeon’s registration intact, impose conditions or suspend or erase. The third channel through the processes of the GMC is the conduct procedures. Here the first relevant filter is the Preliminary Proceedings Committee (PPC). The PPC meets in private and can clear the doctor; refer into the health process; advise or warn the doctor; or refer to the Professional Conduct Committee (PCC). A PCC hearing is, to all intents and purposes, a court of law with a panel, advised by a legal assessor, sitting in judgement. The GMC and the doctor can be represented by barristers if necessary and, unlike the health process, the tone is adversarial. The PCC can postpone a decision and can admonish, suspend or erase the doctor. Sometimes the case is dismissed or the doctor’s continued registration is made conditional. Formerly, outcomes of all three processes could be appealed to the judicial committee of the Privy Council but now appeals are heard by the High Court. Recently, the government has established the Commission for the Regulation of Healthcare Professionals. If it is felt that a decision of the GMC has been too lenient then it is envisaged that the decision can be overturned by this commission in the interests of protecting the public. There is one more committee of the GMC which is important when the question of a doctor’s continuing practice is in doubt. This is the Interim Orders Committee (IOC) to which a doctor may be referred at



the start of any of the three processes. The IOC has powers to impose conditions and to suspend a doctor while the other committees are deliberating. The rationale of the committee is to protect the public from a doubtful doctor while the full process is completed and a fully argued decision of one of the full committees is reached. All that you have just read about above concerning the GMC is very soon likely to be history. After the Shipman Inquiries a White Paper, Trust Assurance and Safety – The Regulation of Health Professionals in the 21st Century was published on February 21, 2007. The Privy Council has extended the term of the 19 elected medical members of the GMC due to stand down on June 30, 2007 until December 31, 2008. This White Paper has several radical proposals which are listed below.  Doing away with the current standard of proof in fitness to practise cases and replacing it with the civil standard of proof.  Establishment of an independent body to adjudicate on fitness to practise cases.  Revalidation will have two parts – relicensing and recertification (which will apply to those on the specialist register and GPs. The establishment of GMC affiliates to assist employers and the NHS deal with complaints at a local level.



THE HUMAN RIGHTS ACT Perhaps the two most potentially influential recent statutes affecting surgeons are the Data Protection Act (DPA) (1998) and the Human Rights Act (HRA) (1998). The first of these is discussed under the heading of confidentiality where its effects, although far-reaching, are relatively easy to predict. The effects of the HRA are potentially much more profound but are still being discovered and untangled by current legal cases. In a way this is curious since the HRA formally enshrines in UK legislation, the European Convention on Human Rights.



The European Convention on Human Rights Many people understand this convention to be a product of the European Community and compliance with it to be a condition of membership. In reality it is a convention to which the UK government signed up in the 1950s. The requirement until the HRA became law was that national legislation should be interpreted in accordance with the convention. The HRA enshrines the rights defined in the convention in UK law, allowing UK citizens to claim in UK courts that their rights have been breached and seek damages or perhaps a stay of criminal proceedings. A number of rights are described in the HRA and these are divided into qualified and absolute rights. The ones of



Chapter 48 Medical jurisprudence and otorhinolaryngology



most relevance to otolaryngologists are numbers 2, 3, 6 and 8. Article 2 states the right to life. There are a number of ways in which this is relevant. Perhaps this is another route by which the family of a dead patient may seek redress if there is the suggestion that the death occurred due to corporate or individual error. The verdicts that are open to a coroner are discussed under The criminal law below, but, as described, this important field is changing. The coroner is a public body under the terms of the HRA and so must ensure that verdicts are in line with that act. Will we soon see a verdict of death in contravention of Article 2 of the European Convention on Human Rights returned as a halfway house between death by misadventure and unlawful killing? What would be the consequences for a subsequent negligence claim? If the strong lobby seeking to legalize assisted suicide in certain circumstances is successful, how will this sit with the right to life? It is not clear in the legislation how a patient can personally renounce his or her rights under Article 2. Article 3 is the right not to be subject to torture or to inhumane or degrading treatment or punishment. This appears to influence the facilities that must be made available by hospitals for the care of patients as well as powerful implications for the whole consent process. Surely it will not be long before a Trust is sued by a patient alleging that the wait on a trolley for 24 hours in casualty constituted inhumane or degrading treatment. There is also a suggestion that the coroner will have to pay regard to Article 3 as well if, in his view, there was inhumane and degrading treatment leading to a death. There are two other articles with relevance to otolaryngologists: Articles 6 and 8. These two interact with each other when the question of GMC and other legal and paralegal proceedings are considered. Article 6 affirms the right of all citizens to a fair and timely trial. A magistrate in the UK has already stayed a prosecution because he judged that the prosecution had taken an unreasonably long time to bring the case to trial. As the defendant was a doctor and now leaves the criminal court without even having been tried, let alone convicted, it is possible that this judgement will profoundly affect his treatment by the GMC. Since GMC committees are trials within the meaning of the HRA, they are subject to exactly the same legal liabilities as courts and coroners. It is here that Article 6 may begin to conflict with Article 8, which grants the right to respect for private and family life. Consider the position of a surgeon seeking to refute at the GMC committee on professional performance an allegation that there is a consistent pattern of poor professional competence. In order to defend himself and to liaise with his professional advisers (solicitors and counsel), he would surely need access to the notes of the patients that he cared for and to whom he may have given substandard care. Also, the GMC might wish to call the patients as material witnesses to the substandard care. At present,



] 607



there is no obligation on a patient, unwittingly a witness at a GMC committee, to grant consent to release of his or her confidential information to either the doctor or the GMC. The patient has rights under the DPA to know what is being done with his or her data and, except in certain circumstances, to withhold consent for release of that information. The exceptions do not seem to apply to GMC proceedings. Surely a patient who found out that a Trust had divulged notes to a surgeon and his legal advisors or to the GMC, in pursuit of a fair trial, could bring action under both the HRA and the DPA. Perhaps also, a patient or even a doctor other than the one being investigated, obliged by subpoena to attend a GMC hearing as a witness and subsequently hounded by the press, can argue that no respect has been shown to his or her privacy. The whole question of how the HRA (1998) will affect surgeons is under active consideration at present. The only guarantee in this field is that lawyers will continue to use the act to push the common law in new and unusual directions.



THE CRIMINAL LAW Otolaryngologists rarely face criminal proceedings in connection with medical practice, but mention should be made concerning two types of criminal offence which may flow from medical treatment.



Involuntary manslaughter In recent years, an increasing number of doctors have faced the prospect of prosecution following the deaths of their patients, allegedly as the result of medical malpractice. The English law relating to involuntary manslaughter has been in a confused state, and clarification has only very recently been provided by the Court of Appeal. Traditionally, the test for involuntary manslaughter has been one of gross negligence. In the words of the then Lord Chief Justice, Lord Hewart, in the case of R v. Bateman,35 in order to establish criminal liability the facts must be such that ‘the negligence of the accused went beyond a mere matter of compensation between subjects and showed such a disregard for life and safety of others as to amount to a crime against the state and conduct deserving of punishment’. Adjectives such as ‘gross’, ‘wicked’ and ‘criminal’ were used to describe the degree of negligence required. This is effectively something of a ‘gut reaction’ test for the jury. In a crime of this nature, it is arguably best left open to the jury to make the determination of criminality. An attempt was made to further define gross negligence by the House of Lords case of Andrews v. DPP.36 In that instance, ‘recklessness’ was considered to be the adjective which most closely described the concept. In



608 ] PART 9 SAFE AND EFFECTIVE PRACTICE this way, the House of Lords was endeavouring to ascribe as guilty states of mind the appreciation of risk, coupled with a determination to run it, or the wilful indifference to the question of whether such a risk existed. However, ‘mere inadvertence’ was not considered to merit criminal responsibility. The test was thrown into confusion in 1981 as a result of decisions in two cases (R v. Caldwell and R v. Lawrence)37, 38 by the House of Lords which redefined the concept of ‘recklessness’. The cases concerned the offences of criminal damage and causing death by reckless driving. It was determined that the defendant would be reckless if his action created an obvious and serious risk (to a victim in the case of manslaughter) and the defendant either appreciated the risk but went on to run it or failed to appreciate the risk. This test has been criticized as being unduly harsh. It allows no realistic explanation of the defendant’s state of mind to be put forward. Thus, a doctor under pressure, perhaps through overwork and being asked to perform tasks which he or she should not otherwise have been required to do, would have no opportunity to explain those circumstances as part of defence. The mere fact that an obvious and serious risk had been created to the patient, as a result of which the patient had died, would be sufficient to secure a conviction. The test was applied in a case of motor manslaughter (R v. Seymour)39 by the House of Lords in 1983. Thereafter, however, there was significant confusion as to whether the case should extend to all cases as an involuntary manslaughter. As a result, in some cases involving doctors charged with involuntary manslaughter the traditional test of gross negligence was applied, in others the new harsh formulation for recklessness was considered appropriate. In yet other cases, attempts were made to combine the two tests. This unsatisfactory situation has been resolved to a degree by the case of R v. Prentice and Sulliman.40 The case involved two junior doctors, Dr Prentice being a preregistration House Officer who, as part of the chemotherapy treatment, injected vincristine intrathecally in error. As the Court of Appeal observed, the mitigating circumstances in relation to both doctors were many, but a version of the recklessness test was put to the jury, resulting in their convictions. These were quashed by the Court of Appeal (R v. Prentice and Sulliman)41 on the basis that the test of recklessness was not appropriate for cases of involuntary manslaughter where a breach of duty is concerned, particularly for cases involving doctors, where the court observed: ‘Often there is a high degree of danger to the deceased’s health, not created by the defendant, and preexisting risks to the patient’s health is what causes the defendant to assume the duty of care with consent. His intervention will often be in situations of emergency.’ The traditional test for gross negligence was preferred. The court set out the matters which need to be established for a prosecution of manslaughter, namely the



existence of a duty of care, the breach of the duty causing death and gross negligence which the jury considers justifies criminal conviction. The Court of Appeal went on to set out various states of mind which it considered could probably lead the jury to making a finding of gross negligence.  Indifference to an obvious risk of injury to health.  Actual foresight of a risk coupled with the determination nevertheless to run it.  An appreciation of the risk, coupled with an intention to avoid it but also coupled with such a high degree of negligence in the attempted avoidance as the jury considers justifies conviction.  Inattention to or failure to avert a serious risk which goes beyond mere inadvertence in respect of an obvious and important matter which the defendant’s duty demanded he should address. The first three states of mind are not unduly onerous. The fourth, however, may present particular problems to medical practitioners. Medical treatment generally may be considered to amount to an obvious and important matter. Further, many serious risks are inherent in treating patients and, accordingly, medical practitioners remain in danger of falling foul of the criminal law relating to manslaughter. At the conclusion of the judgement of the Court of Appeal in the case of Drs Prentice and Sulliman, the court expressed the view that the Law Commissioners should look at the law in relation to involuntary manslaughter as a matter of urgency, and thus modifications may be made to this area of the law in the near future. In Canada,42 manslaughter may be committed where a doctor causes the death of a patient by, among other things, criminal negligence which may be committed where someone under a duty shows a wanton or reckless disregard for the lives or safety of others. The precise meaning of ‘wanton or reckless disregard’ is in doubt. In a recent case (R v. Tutton and Tutton)43 the Supreme Court of Canada was divided between those who considered that a defendant should have an intention to run a prohibited risk, or a wilful blindness to the risk, and those who considered that a marked and substantial departure from a standard of behaviour expected of a reasonably prudent individual would suffice. This dilemma is similar to that recently addressed by the Court of Appeal in this jurisdiction in the case of Drs Prentice and Sulliman. In New Zealand (R v. Yogasakaran),44 the test for manslaughter is that of mere negligence, and a breach of the civil standard of care resulting in the death of a patient is all that is required. Similarly in Greece, the offence under the Greek Criminal Code45 is of causing death by negligence, mere civil negligence being sufficient. The charge is often seen as a precursor to civil proceedings, where a victim’s family may make complaint to prosecuting authorities, effectively as part of the bargaining process for compensation.



Chapter 48 Medical jurisprudence and otorhinolaryngology



Assault, battery and bodily harm Charges of assault and battery are considered very infrequently by the prosecuting authorities in this jurisdiction in relation to medical practitioners. Courts are reluctant to consider actions in tort for battery arising out of a failure to obtain consent, let alone criminal charges. Very few doctors will intend to inflict harm on a patient, the overall aim being to provide some form of therapeutic benefit. However, where the clinician performs a procedure which goes substantially beyond that to which the patient has consented and that is known to the clinician, or the clinician is reckless as to whether or not the patient’s consent will authorize the procedure actually performed then such an offence may be made out, even if of therapeutic benefit. In circumstances where surgical intervention takes place, an offence of greater seriousness than mere assault may result, for example, assault occasioning grievous bodily harm.



THE FUTURE It is without doubt that the shift in emphasis is away from the doctor and towards patients’ rights. This can clearly be seen in a very recent landmark case (Chester v. Afshar)46 heard by the House of Lords. Carole Chester agreed to undergo lumbar surgery on the advice of Mr Afshar, a neurosurgeon. She suffered partial paralysis. The judge found that Mr Afshar had not performed the surgery negligently but had failed to warn her of the partial paralysis inherent in the operation. The judge found that she would not have undergone the operation had she been warned but at the time would have sought further advice before deciding on what to do. ‘Lord Hope of Craighead said that the claimant could not succeed on conventional causation principles since she did not assert that, if informed, she would never have undergone the surgery and since the defendant’s failure neither affected the risk which eventuated nor was the effective cause of her injury. However the causation issue was to be addressed by reference to the scope of the defendant’s duty, namely to advise his patient of the disadvantages and dangers of the treatment he proposed; that duty was closely connected with the need for the patient’s consent and was central to her right to exercise an informed choice as to whether and, if so, when and from whom to receive the treatment. The function of the law was to enable rights to be vindicated and to provide remedies when duties had been breached. Unless that was done the duty was a hollow one, stripped of practical force and devoid of content. The injury the claimant sustained was within the scope of that duty and might, on policy grounds, be regarded as having been caused, in the legal sense, by the defendant’s breach. Accordingly justice



] 609



required that the claimant be afforded the remedy she sought.’ Lord Steyn and Lord Walker of Gestingthorpe delivered concurring opinions; Lord Bingham and Lord Hoffmann disagreed so the defendant lost his appeal to the House of Lords. Clearly this has a major impact for no longer does the patient have to state that they would not have had the surgery if they had been aware of all the risks but it is now sufficient for the patient to argue that they would have sought further advice. This has increased pressure on ensuring that specific risks are warned and the warning is recorded. It is now no longer good enough just to record that risks have been explained for a case to be defended where consent is a core issue. Punitive (exemplary) damages are not normally awarded in jurisdictions of the UK and Australasia but in 2002 the Privy Council heard the case of A v. Botrill47 from New Zealand. This case involved the wholesale misreading of cervical smears by a private pathologist in Gisborne, New Zealand. Between 1990 and 1994 he examined four smears from Mrs A. After Mrs A had a radical hysterectomy for invasive cancer the smears were reviewed and all four slides had been misread or misreported. Mrs A made a successful claim for accident compensation. Disciplinary proceedings against Dr Botrill found him guilty of conduct unbecoming of a medical practitioner. Mrs A then brought court proceedings claiming exemplary damages. Justice Young dismissed the action applying the principle ‘exemplary damages may be awarded, but only if the level of negligence is so high that it amounts to outrageous and flagrant disregard for the plaintiff ’s safety, meriting condemnation and punishment’. Two events then occurred: Mrs A found another ten women whose slides had been misread and there was public concern about smear reporting in the Gisborne area. This led to a review of Dr Botrill’s slides in Sydney which showed that Dr Botrill’s false reporting rate was 50 percent or higher. In the light of the new evidence Mrs A applied for a retrial which was granted. Dr Botrill then appealed this decision and his appeal was successful. Mrs A sought leave to appeal to the Privy Council which was granted. The argument for Mrs A was that cases of reprehensible wrong doing which are totally unacceptable to the community but arise from inadvertent negligence should not be put beyond the reach of exemplary damages. Dr Botrill’s argument was that there was no conscious, outrageous and flagrant disregard for the plaintiff ’s safety. The Privy Council agreed with Mrs A and allowed her appeal on a majority decision. Although this is a New Zealand case it now brings the possibility of punitive damages for reprehensible wrongdoing even if there was no flagrant disregard of risk. It seems that now inadvertent negligence may lead to exemplary damages in selected cases of the worst kind of professional practice.



610 ] PART 9 SAFE AND EFFECTIVE PRACTICE



KEY POINTS  Clinical negligence is a matter for civil law.  Damages awarded are general and special.  The standard or proof is ‘on the balance of probabilities’.  There are four essential elements of the legal tort of negligence: – there must be a duty of care; – there must be a breach of that duty; – harm must have occurred; – the breach must have caused the harm.  The standard of care is that of the reasonably skilled and experienced doctor.  ‘ It will very seldom be right for a judge to reach the conclusion that views genuinely held by a competent medical expert are unreasonable.’  If the case for each side is evenly balanced then the claimant will fail.  The court must award a sum of money that will, as nearly as possible, put the injured person in the same position as he would have been had he not been injured.  The result of the introduction of contingency fees is likely to be that there will be no significant increase in the number of medical negligence actions in the short term.  Where a proposed treatment involves a substantial risk of grave or adverse consequences such that, notwithstanding any practice to the contrary, a patient’s right to decide whether to consent to the treatment was so obvious that no prudent medical man could fail to warn of the risk, then it would be negligent not to warn.  The parental right to determine whether or not a child under 16 should have medical treatment terminates when the child achieves a significant understanding and intelligence to enable him or her to understand fully what is proposed.  ‘An adult patient who suffers from no mental incapacity has an absolute right to choose whether to consent to medical treatment, to refuse it or to choose one rather than another of the treatments being offered.’  Any surgeon who keeps any sort of patient identifiable information at his residence and who is not registered with the data commissioner breaches the provisions of the DPA.  The coroner’s remit is to establish the identity of the deceased, the time and place of death and the mode of death.



 Articles 2, 3, 6 and 8 of the HRA are those most relevant to surgeons.  Modifications to the law on involuntary manslaughter are awaited; currently the traditional description of gross negligence resulting in death usually applies.



REFERENCES 1. Bolitho v. City and Hackney Health Authority [1997] Weekly Law Report 151. 2. Bolam v. Friern Hospital Management Committee [1957] 1 Weekly Law Reports 582. 3. Wilsher v. Essex Area Health Authority [1988] 1 All England Reports 871, House of Lords. 4. Crawford v. Board of Governors of Charing Cross Hospital [1953] Times Law Reports, 8 December 1953, Court of Appeal. 5. Barnett v. Chelsea & Kensington Hospital [1968] 1 All England Reports 1068. 6. Ashcroft v. Mersey Regional Health Authority [1985] 2 All England Reports 96. 7. Picard E. Legal liability of doctors and hospitals in Canada. In: Kennedy I, Grubb A (eds). Medical law: text and materials. London: Butterworths, 1989: 423. 8. Scott v. London & St. Katherine’s Docks [1865] 3 Hurlstone & Coltman’s Exchequer Reports 596 ExCh (sugar bags). 9. Byrne v. Boadle [1863] 2 Hurlstone & Coltman’s Exchequer Reports 722 (a barrel of flour). 10. Pope v. St. Helen’s Theatre [1947] Kings Bench 30 (the ceiling of a theatre). 11. M.A.M.S. Leigh. Res ipse Loquitur: What does it mean? Medical Defence Union Journal. 1993; 9: 66. 12. Girard v. Royal Columbian Hospital [1976] 6 Dominion Law Reports (3d) 676. 13. Mahon v. Osborne [1939] 2 Kings Bench 1450. 14. Personal correspondence. 15. Livingston v. Rawyards Coal Company [1880] 5 Appeal Cases 25 @ 39, House of Lords per Lord Blackburn. 16. Kralj v. McGrath [1986] All England Law Reports 54. 17. Ward v. James [1965] 2 All England Reports 563, Court of Appeal. 18. Levitt v. Hartlepool Area Health Authority (1993) (Unreported). 19. Lim v. Camden Health Authority [1979] 2 All England Reports 910, House of Lords, per Lord Scarman. 20. Rules of the Supreme Court: Order 37 Rules 7–10. 21. See generally the Rules of the Supreme Court and the County Court Practice for Procedure. 22. Civil Procedure Rules 1998. 23. S.58 Courts and Legal Services Act 1990. 24. Preliminary results of the Law Society Survey of Personal Injury Specialists. LS Gaz, 29 September 1993: 3.



Chapter 48 Medical jurisprudence and otorhinolaryngology 25. Sidaway v. Board of Governors of the Bethlem Royal Hospital [1985] Appeal Cases 871, House of Lords. 26. Hopp v. Lepp [1979] 112 Dominion Law Reports 3d 67. 27. Blyth v. Bloomsbury Health Authority [1993] 4 Medical Law Reports, Court of Appeal. 28. Rogers v. Whitaker [1993] 4 Medical Law Reports 79, High Court of Australia. 29. Gillick v. West Norfolk and Wisbech Area Health Authority [1986] Appeal Cases 112, [1985] 3 All England Reports 402, House of Lords. 30. In re R [1991] 4 All England Reports 177, Court of Appeal. 31. In re W [1992] 3 Weekly Law Reports 758. 32. Re F [1990] 2 Appeal Cases 1, House of Lords. 33. In re T (Adult: Refusal of treatment) [1992] 3 Weekly Law Reports 783. 34. National Justice Compania Naviera SA v. Prudential Assurance Company Ltd. (Ikarian Reefer) [1993] Times Law Reports, 3rd March 1993. 35. R v. Bateman [1925] 19 Criminal Appeal Reports 8, Court of Appeal.



36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47.



] 611



Andrews v. DPP [1937] Appeal Cases 576, PC. R v. Caldwell [1982] Appeal Cases 341, House of Lords. R v. Lawrence [1982] Appeal Cases 510, House of Lords. R v. Seymour [1983] 2 Appeal Cases 493, House of Lords. R v. Prentice and Sulliman (unreported November 1, 1991, Owen J). R v. Prentice and Sulliman [1993] 4 All England Reports 935, Court of Appeal. The Annotated Tremear’s Criminal Code 1992, Sections 219, 220, 222. R v. Tutton and Tutton [1989] 48 Canadian Criminal Cases (3d) 129, Canadian Supreme Court. R v. Yogasakaran [1990] 1 New Zealand Law Reports 399, New Zealand Court of Appeal. Article 302 Greek Penal Code. Chester v. Afshar [2004] Appeal Cases 41, House of Lords A v. Botrill [2002] UKPC 44.



This page intentionally left blank



PART



10



INTERPRETATION AND MANAGEMENT OF DATA EDITED BY MARTIN J BURTON



49 Epidemiology Jan HP van der Meulen and David A Lowe



615



50 Outcomes research Iain RC Swan



633



51 Evidence-based medicine Martin J Burton



645



52 Critical appraisal skills Martin Dawes



649



This page intentionally left blank



49 Epidemiology JAN HP VAN DER MEULEN AND DAVID A LOWE



Introduction What is epidemiology? ‘Streams’ of epidemiology The ‘anatomy’ and ‘physiology’ of epidemiological research



615 616 616



Choosing the study design Key points Acknowledgements References



624 630 631 631



618



INTRODUCTION This chapter presents epidemiology as a methodological discipline that provides important principles for clinical and health services research. It introduces the ‘determinant–occurrence relationship’ as a key epidemiological concept. On the basis of this concept, we will demonstrate how epidemiological methods and techniques can be used to address a wide range of questions. This chapter is intended to inform those who want to read the medical literature and evaluate research evidence. In other words, it has been written especially with the needs of the ‘consumers’ of research in mind. We will concentrate on the choice of a study design for different types of research questions with the ultimate aim of developing a study that produces results that are relatively precise (free of random error) and accurate (free of systematic error or bias). An additional consideration is that the study needs to be efficient (affordable in terms of time and money). Strengths and weaknesses of study designs will be discussed as much as possible on the basis of examples related to diseases of the ear, nose and throat. This chapter is not based on a specific literature search strategy. It amalgamates information and points of view as can be found in major textbooks and reviews of epidemiology, methods of health services research and evidence-based clinical practice.



Example: the indication for tonsillectomy Simple, but well-designed, epidemiological research can be of great importance for clinical practice. A recent



systematic review indicated that the effectiveness of tonsillectomy is uncertain,1 and in turn, that the indications for tonsillectomy are controversial. Nevertheless, tonsillectomy is one of the commonest surgical procedures carried out in children as well as in adults. In the 1930s, an estimated 200,000 were performed annually in England and Wales, a huge number compared to an annual number of approximately 40,000 in the year 2000. In the pre-war period, tonsillectomy had become popular to the point of being fashionable, and there was marked variation in its frequency according to geographical location, social class and sex.2 It was estimated that at least 85 children lost their lives each year as a direct result of tonsillectomy. Another study carried out in the 1940s demonstrated that there was great uncertainty about effectiveness and indications,3 as cited in a book by Sackett and coworkers.4 Among 389 eleven-year-old schoolchildren with intact tonsils examined by a group of clinicians, tonsillectomy was recommended in 174 (45 percent). A second opinion was requested in the 215 schoolchildren for whom tonsillectomy was not recommended, and tonsillectomy was recommended for 99 (46 percent). The remaining 116 children, in whom on the two previous occasions tonsillectomy was not recommended, were then examined for the third time, and tonsillectomy was recommended in 51 (44 percent) of them. The most remarkable finding is that tonsillectomy was recommended in each of the three cycles for approximately 45 percent of the children. These historical examples illustrate that uncertainty about diagnostic and management decisions can lead to overtreatment, which may have serious consequences for patients. Given the fact that there is still no high-quality



616 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA evidence on the effectiveness of tonsillectomy more than 50 years later, the same problem may still exist, albeit to a lesser extent. The only published study on the effectiveness of tonsillectomy that could be included in one systematic review studied 91 children and was affected by important baseline differences in the characteristics of the surgical and the control group.1 It is obvious that this kind of clinical uncertainty can only be solved by epidemiological evidence of high quality.



WHAT IS EPIDEMIOLOGY? Epidemiology is a relative young discipline, but during its short life many definitions of epidemiology have emerged.5 Many of these describe epidemiology in terms of its subject matter. The perhaps most frequently cited definition in this context states that epidemiology is the study of the distribution and determinants of disease frequency in human populations.6 These three closely related components – distribution, determinants and frequency – encompass many epidemiological principles and methods.7 For example, an epidemiological study that would fit perfectly within this definition is that of the frequency of head and neck cancer in a certain geographical area. This study could also consider the distribution of the disease among different subgroups. The determinants of disease occurrence would then derive from these two. Although this definition has its merits in that it covers many epidemiological studies, some of which attracted large media attention (for example, smoking and lung cancer, cholesterol and heart disease, effect of diethylstilbestrol on offspring, unprotected sex and the acquired immune deficiency syndrome), a growing number of epidemiologists feel that the above-cited definition does not cover their work. The reason is that the subject matter of what can be considered epidemiological studies has become rather heterogeneous. Attempts to produce a definition of epidemiology based upon its subject matter therefore produce confusion rather than clarity.



The concept of the determinant–occurrence relationship In the last two decades, an alternative definition has arisen that defines epidemiology as a discipline that studies the functional relationship between the occurrence of disease (or related health outcomes) and its determinants.8 This may seem a trivial step, but this focus on the concept of the ‘determinant–occurrence relationship’ rather than the subject matter has considerably broadened the scope of epidemiology in medicine. One of the major reasons why epidemiology has gained such far-reaching importance lies in the nature of medical knowledge. Medical knowledge is, to a large extent,



derived from the experience obtained in groups of similar patients. Medicine is therefore an empirical science, and many of the empirical relationships in medicine can be considered as determinant–occurrence relationships. It is especially this realization on the basis of which epidemiology has become a basic science in medicine.



Epidemiology as a methodological discipline An alternative approach to clarify the definition of epidemiology is to consider it as a methodological discipline. From this perspective, the subject of epidemiological inquiry is not so much the determinants of the occurrence of disease and other health outcomes, but the principles and methods for the study of determinant– occurrence relationships. As a methodological discipline, epidemiology allows a large number of different questions to be answered. As explained earlier, it may play a role in describing the distribution of health and disease in a population. Who is affected? Where and when does this health problem occur? Why does it occur in a particular population? However, other questions can be considered as well. Once a health problem occurs, epidemiology also provides methods for monitoring the course and outcome of a health problem. What is the outcome of the health problem? It may answer questions regarding the outcomes of interventions. Is intervention A more effective than intervention B? How well does a diagnostic test distinguish between people with and without the target disorder? Epidemiological research is also essential for the assessment of the burden of disease and the need for health services and evaluation of the access to services. How much suffering does this health problem cause in a population? How many people need a certain intervention? Who uses the intervention? What factors explain differences in health care use?



‘STREAMS’ OF EPIDEMIOLOGY Given the extension of the boundaries of epidemiology – based on the introduction of the determinant–occurrence relationship as the key epidemiological concept and the realization that epidemiological principles provide the methodological underpinning for the study of a wide range of questions – one can recognize several ‘streams’ of epidemiological research. These streams differ largely according to the types of questions they address, but they share most of the methodological concepts and principles.



Classical epidemiology The first stream of research is ‘classical epidemiology’, sometimes referred to as ‘aetiological epidemiology’ or



Chapter 49 Epidemiology



‘risk factor epidemiology’, of which the ultimate goal is ‘the elaboration of causes that can explain patterns of disease occurrence’.5 The geographical distribution of a disease, the variations in its frequency over time, and the special characteristics of people affected by it, are typical objects of study. There is often a natural progression in this type of epidemiological research.7 First, there is a concern about the possible influence of a particular factor on the occurrence of disease. This suspicion can have many origins – clinical practice, laboratory research, theoretical speculation – but it often arises from examination of disease distributions, and leads to the formulation of a specific hypothesis about its causes. It can then be further explored in studies of individuals that include an appropriate comparison group. A systematic collecion and analysis of data may reveal that a statistical association exists. It is then essential to assess whether random errors or systematic errors might be responsible for the findings. Finally, a judgement needs to be made about whether an observed association represents a cause–effect relationship. An otolaryngological example of this type of epidemiological research is the study of the distribution of head and neck cancers and the association of the occurrence of this disease with traditional risk factors such as diet, smoking and drinking habits or socioeconomic status. Another more specific example is the testing of the hypothesis that there is a link between human papillomavirus and the occurrence of a subset of these cancers. The result of aetiological research may guide the first steps towards the development of primary prevention that keeps disease from occurring at all. However, the reductionist nature of classical epidemiology has been criticized. It has been argued that epidemiological studies often ignore ‘the interdependence of multiple agents and how human populations become exposed and susceptible to them’.9 Epidemiological research that focuses on the effect of risk factors measurable at the level of the individual neglects the population context and the social and cultural determinants that act at population level.



Clinical epidemiology The second stream of research is ‘clinical epidemiology’. The adjective ‘clinical’ is added because clinical epidemiology ‘seeks to answer clinical questions and to guide clinical decision making with the best available evidence’.10 Clinicians are mainly concerned with problems of individual patients. It may seem paradoxical that results of epidemiological studies that are derived from groups of patients should be applicable to the problems of individual patients. It is obvious, however, that this is because the best evidence to solve a clinical problem is derived from the experience of a large number of similar patients.



] 617



Advocates of evidence-based medicine have addressed this paradox directly. They state that clinicians need to carry out ‘the particularization, to the individual patient, of our prior experiences (both as individual clinicians and collectively) with groups of similar patients’.4 A good clinician should therefore use the best available external evidence together with his own unsystematic clinical experience and intuition – based on a blend of knowledge derived from anatomy, physiology and other basic sciences.11 Evidence alone is never sufficient to make a clinical decision. Decision-makers always trade the benefits and risks, inconvenience and costs associated with alternative management strategies, and in doing so should especially consider the patients’ values and preferences. The significance of the definition of epidemiology as a methodological discipline that addresses determinant– occurrence relationships is based on the reach of the concept of determinant–occurrence relationships in itself. If one considers the presence of disease as the outcome and the diagnostic information as the determinant, then this represents a diagnostic problem. The object of study is, in this case, the functional relationship between diagnostic information and the presence of disease. If one considers the occurrence of a disease or a healthrelated event in the future as the outcome and the presence of certain patient characteristics as the determinants, then this represents a prognostic problem. One of these patient characteristics can be the use of a specific therapy, and in that case, this represents a therapeutic problem. In the latter case however, to achieve accurate results the ‘ceteris paribus principle’ – the condition that all other determinants are equal – is a fundamental notion that we will explicitly address under Randomized controlled trials. A study of the diagnostic accuracy of fine-needle aspiration as a test for malignant disease in patients with nodular thyroid disease is an otolaryngological example of diagnostic research. The influence of the age of patients with an oropharyngeal carcinoma on long-term survival constitutes a prognostic research question. The effect that early surgery compared with watchful waiting for glue-ear has on language development in preschool children represents a therapeutic research question.



Epidemiology and health services and public health research Epidemiology is also one of the core disciplines of health services research and public health. Whereas classical epidemiology and clinical epidemiology focus largely on determinants of health and disease and related conditions in individuals, health services research is more directed towards questions addressing the quality and organization of health care systems and public health research towards the health and health care problems in communities. One



618 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA could consider this as a third stream of epidemiological research. Important questions for health services research are those that address the variations in processes and outcomes of health care services, as well as the determinants of these variations. Research in public health may be concerned with the influence of environmental factors, socioeconomic conditions and health services on health in the community. It will be clear by now that many of these questions take the form of determinant–occurrence relationships, which again confirms the crucial role of an epidemiological approach. A study comparing the outcome of thyroid surgery performed by experienced surgeons and surgeons in training is an example of an epidemiological study directly addressing a determinant of the quality of otolaryngological care.12 Another example is a study of the impact that the publication of a guideline on the treatment of persistent glue-ear in children had on the rate of surgery.13 A systematic review of studies addressing the effectiveness of screening young children to undergo early treatment for glue-ear is a third example of public health research.14



a question that can be translated into a feasible and valid study can be surprisingly hard. What often happens is that instead of a ‘research question’, a ‘topic’ is formulated. ‘Framing’ the research question is the first step for every new project, of which the importance cannot be overestimated. Conversely, many studies fail not so much because the study design is flawed or the execution of study is poor, but because the question was not formulated adequately in the first place. An otolaryngological example of such a ‘topic’ that needs further research is the concern about the rising complication rate after tonsillectomy in the UK. It is clear that we need to provide more detail and structure before we are able to formulate a relevant and answerable ‘research question’. The concept of the determinant– occurrence relationship may provide some guidance in this context. In other words, we need to be specific about what outcomes we want to study as well as what determinants we want to consider. Finally, we need to specify the target population, the kind of people for whom the study should provide answers. In epidemiological research, a good research question therefore has three components. It should define:



THE ‘ANATOMY’ AND ‘PHYSIOLOGY’ OF EPIDEMIOLOGICAL RESEARCH



1. the target population; 2. the comparison or control group, at least if there is any; 3. the outcome(s) of interest.



We cannot discuss the different study designs before we have examined what epidemiological research is made up of and how it works. The easiest way to do this is to describe the components of a study protocol and the way results are used to draw inferences from the study results about the truth in the universe. The following sections are based largely on the introductory chapters of a recent book by Hulley and colleagues about designing clinical research.15



Essential components of a protocol for an epidemiological study A research protocol is a document that provides all essential details of a study. A protocol is necessary for guiding all the decisions that need to be made in the course of the study. The process of writing a protocol itself helps the investigator to enhance the scientific rigour and efficiency of the project. Most research involves teamwork and a written document ensures that all members know how the study should be implemented and what they are expected to contribute. A good protocol provides answers to a number of essential questions. WHAT QUESTION WILL THE STUDY ADDRESS?



The research question defines what you want to achieve. Many experienced researchers will agree that formulating



An example of a ‘good’ research question for a study addressing the complication rates after tonsillectomy could be: ‘In patients undergoing tonsillectomy because of recurrent tonsillitis, does the use of the bipolar diathermy forceps increase the occurrence of tonsil bleeds severe enough to require return to theatre in the first 28 days after surgery, compared to other tonsillectomy dissection types?’ Admittedly, this is a rather convoluted sentence, and is only given here to illustrate how the three components of a research question can be covered in one question. In practice, the question will be shortened to ‘Does bipolar diathermy increase the haemorrhage rate after tonsillectomy’, but in that case the details that define the target population, the comparison and the outcome of interest need to be provided separately. WHY IS THE STUDY QUESTION IMPORTANT?



Good epidemiological research should also pass the ‘so what’ test. In the Introduction or Background section of a protocol, it should be argued that answering the research question will provide a significant contribution to our state of knowledge or, in other words, add to what is already known about the problem. It is therefore important to be on top of the published literature before developing a study. It should also be clear that the results of the proposed study will help to resolve current uncertainties, which may lead to new scientific understanding and influence clinical and public health policies.



Chapter 49 Epidemiology



It is often ‘scholarship’ that will identify the gaps in the current knowledge and how these can be addressed in an optimal way by learning from the work of others. For example, a recent systematic review of the literature on the effect of tonsillectomy in patients with chronic or recurrent tonsillitis concluded a lack of evidence to guide decision making for this surgical intervention in adults or children.1 More importantly in this context, the authors of the systematic review also concluded on the basis of the results of their review that future trials should address the effectiveness of tonsillectomy in subgroups according to age, severity and disease frequency, and that patients should be followed up for at least one year to assess outcomes such as general well-being, behaviour, growth, sleep and eating patterns in addition to severity and frequency of infections. These conclusions based upon a systematic review provide powerful arguments for the direction of future research.



] 619



the next step is to decide whether the study needs a comparison or control group. If so, the study is often called ‘analytical’. If not, it is a ‘descriptive’ study. Descriptive studies are often used as a first step into a new area of study – the scientific ‘toe in the water’,15 and followed by analytical studies to answer questions that can address determinant–occurrence relationships. A more detailed presentation of study designs follows under Choosing the study design.



WHO ARE THE STUDY SUBJECTS AND HOW WILL THEY BE SELECTED AND RECRUITED?



A good choice of the study subjects ensures that the results of the study will accurately represent what is going on in the population of interest, the target population, the set of people best suited to the research question (for example, patients with early laryngeal squamous cell carcinoma for a study comparing radiotherapy and surgery for laryngeal cancer). The protocol must also specify the study sample, which is the subset of the target population available for study (for example, all consecutive patients with this disease condition referred to a regional head and neck cancer centre in a defined period of time). The study sample should be a subset of the target population that can be studied at an acceptable cost and is large enough to control random error and representative enough to control systematic error. In many cases, controlling both random error and systematic error sets conflicting demands, which is sometimes referred to as the ‘precision–bias’ dilemma. For example, a study that aims to evaluate the usefulness of the endoscope compared to the headlamp for sinonasal surgery should carefully consider which patients to include. It is highly likely that the results of the study



HOW IS THE STUDY STRUCTURED?



Choosing the study design is a complex issue. The actual choice depends strongly on the research question. The study should be designed in such a way that it produces results that are relatively precise (free of random error) and accurate (free of systematic error or bias). An additional consideration is that the study needs to be efficient (affordable in terms of time and money). A simplistic ‘taxonomy’ of study designs, presented in Figure 49.1, shows that two fundamental decisions have to be made. First, the investigators have to decide whether they want to assign the determinants themselves in an ‘experimental study’, or whether they want to examine events as nature takes its course in an ‘observational study’. Second, if an observational study design is chosen,



Did investigator assign determinants? Yes



No



Experimental study



Observational study



Random allocation?



Comparison group?



Yes



Yes



No



Randomized controlled trial



Exposure



Cohort study



Nonrandomized controlled trial



outcome



No Descriptive study



Analytical study Direction?



Exposure and outcome at the same time



Exposure Casecontrol study



Outcome Crosssectional study



Figure 49.1 Algorithm for classification of types of clinical research. Reproduced from Ref. 16, with permission from Elsevier.



620 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA will differ according to the extent of the surgery. The comparison might be most relevant for patients who undergo surgery for procedures around the middle meatus and the anterior ethmoid. Including patients who undergo only simple polypectomy, who comprise approximately 50 percent of the total number of patients who undergo a form of sinonasal surgery, may seem an attractive option because it will double the size of the study. However, including these patients will diminish the extent to which the study sample represents the population for which the research question is of interest. In an ideal world without practical and financial limitations, we would study the entire target population. Often, if not always, the target population is too large and the study will be carried out in a sample. Earlier, we gave as an example the study comparing radiotherapy and surgery for early laryngeal carcinoma that was carried out in consecutive patients visiting a regional centre in a defined period of time. This type of sampling, including patients that are easily accessible to the investigator, is called ‘convenience sampling’. It is very frequently used in clinical research. It has obvious advantages in terms of costs and logistics, but its drawback is that it might not sufficiently represent the target population. Consecutive sampling, including without interruption all accessible people, is especially useful in this context given that it reduces the possibility that selection either by the investigators or self-selection by the subjects influences the results. Probability sampling is the gold standard for ensuring that the study sample is representative of the target population, except for the effect of chance variation. There are several probability-sampling methods. With simple random sampling, every individual in a population has the same chance of being included. However, in some cases, a form of stratified sampling (random sampling with a known sample size within subgroups) is desirable. For a study of patient satisfaction of patients visiting an otolaryngology outpatient clinic, for example, the investigators may wish to divide patients in ‘strata’ according to their diagnosis or the procedure that was being carried out, and then sample specified numbers from each stratum. The choices made about the selection of the study subjects (the intended sample) are important as they have an impact on the extent to which the study findings can be generalized. The sampling procedures can affect the generalizability of the results in a number of ways. First, the actual sample might be different from the intended sample. For example, people who were eligible for the study might have refused, and people who participated might be different (more or less healthy, dependent on the study context) from those who did not. Second, the study sample should be sufficiently similar to the target population. For example, the ‘spectrum’ of disease in patients with laryngeal squamous cell cancer might differ from one centre to the next, and in turn the effects of radiotherapy and cancer might differ as well. Third, the



investigator must form an opinion about whether the results can be generalized to people outside the target population. In terms of our example, do radiotherapy and surgery have similar effects in other countries, in patients with more advanced disease, or in patients who are on average older or younger? This will always be a subjective judgement that depends on findings in other studies, more general scientific knowledge, and what is sometimes called ‘a feeling for the organism’. It is not a yes-or-no decision, and may trigger debate among experts. This debate may be informed by the study itself if the study population is diverse enough to explore the constancy of effects in different subgroups within the study. WHAT MEASUREMENTS WILL BE MADE?



The quality of a study depends on how well the variables measured in the study represent the phenomena of interest. Another concern is how the variables of interest can be measured as precisely (free of random error) and accurately (free of systematic error) as possible without making the study unreasonably expensive. It is important to have some understanding about the types of variables that are encountered in epidemiological research, and how they can be measured. Any variable can be considered to be of one of two basic types: continuous variables and categorical variables. This distinction determines the way in which these variables should be measured as well as analysed. As a general rule, with categorical variables, the analysis will involve a description or comparison of the proportion of subjects falling into the various categories. With continuous variables, the descriptions or comparisons are most often presented in terms of average values or medians if the sample is small and the average value is considered to be inappropriate to describe the ‘midpoint’ of the variable’s distribution. Theoretically, continuous variables can take all possible values on a continuum along a specific range. Many clinical parameters are continuous. Variables, such as blood pressure, body weight, body temperature, are not restricted to particular values and are only limited by the accuracy and precision of the measuring procedures. The units in which a continuous variable is expressed specify a uniform difference along the entire length of the scale. A difference in blood pressure of 10 mmHg has the same interpretation irrespective of whether it occurs at the lower or the upper end of the blood pressure scale. Discrete variables can only take certain numerical values – in most cases only integers (whole numbers). If discrete variables have a considerable number of possible values, they can often be treated as if they were continuous. Examples of these ‘quasi-continuous’ variables are counts (such as the number of cigarettes smoked, number of episodes of acute tonsillitis in last year, or number of days spent in hospital) or scores on ‘clinimetric scales’ (such as the 20-item Sino-Nasal Outcome Test) that measures symptom severity of sinonasal conditions on a scale with



Chapter 49 Epidemiology



discrete values ranging from 0 (no symptoms) to 100 (very severe symptoms)17 or the Epworth Sleepiness Scales, used in patients with obstructive sleep apnoea, that can take discrete values from 0 (no daytime sleepiness) to 24 (very severe daytime sleepiness).18 Strictly speaking, considering scores on a clinimetric scale as continuous is somewhat problematic, as the interpretation of a difference of one unit may vary along the length of the scale. For example, does an increase from five to ten on the Epworth Sleepiness Score correspond to an increase from 15 to 20? Phenomena that cannot be measured quantitatively can often be measured by classifying them into categories, and then counting the number of subjects that fall within the defined categories. In its simplest form, there are only two categories (such as man/woman, or dead/alive), and these variables are called dichotomous or binary variables. When there are more than two possible categories, the variables are termed poly- or multichotomous variables. Polychotomous variables can be further classified into those that are ordered, called ordinal variables, and those that are not, called nominal variables. Socioeconomic status, ASA grades describing fitness to undergo surgery,19 and degree of pain are examples of the former, and race, marital status, and blood type are examples of the latter. As a general rule, continuous variables are more ‘informative’ than categorical variables. It is therefore advisable to use continuous variables as much as possible. For example, ‘body temperature’ as a continuous variable should be preferred to a categorical variable dividing patients into those with a temperature below and above a certain threshold. Good measurements are precise and accurate. There are largely three sources of random and systematic error: (1) variability due to the observer; (2) variability due to the subject; and (3) variability due to the instrument used. A number of strategies can be followed to increase simultaneously both precision and accuracy of the measurements. First, the measurement procedures should be standardized and clearly described in the study protocol. Second, those who are involved in taking the measurements should be trained and their performance monitored in the course of the study. Third, the measurements should be carefully chosen in terms of what is known about their performance. Fourth, calibration procedures should be carried out against a ‘gold standard’. This is especially essential for mechanical devices such as weighing scales, thermometers and blood pressure measuring devices. Furthermore, a simple and rather effective approach to increase measurement precision is to repeat the measurements and to use the mean of two or more observations. The latter approach is commonly used in studies measuring blood pressure. An important approach to reduce systematic error is blinding, which conceals information about determinants, or in some cases outcomes, to the observers and/or



] 621



subjects. This reduces the possibility that the observers or the study subjects distort the overall accuracy of the measurements, consciously or unconsciously. Blinding, however, does not ensure overall accuracy of the measurements, but it may eliminate ‘differential bias’ that affects one study group more than another. Blinding is especially relevant for measurements that incorporate some subjective judgement. For example, a study comparing early surgery with watchful waiting for glue-ear in pre-school children used a tester of language development who was unaware of what treatment the children had received.20 HOW WILL THE RESULTS BE ANALYSED?



The main statistical methods that are going to be used to analyse the results should be defined in the study protocol. Choosing the statistical methods after the results have become available will increase the likelihood of finding associations between determinants and outcome on the basis of chance alone. An important element of the analysis plan is the description of how the outcome variable is going to be analysed. For example, a trial on early surgery for glue-ear that measured language development with the Reynell development language scales (a test of expressive language and verbal comprehension abilities in children aged six months to six years) could analyse the Reynell data as a continuous variable.20 Conversely, it could define groups with normal language development and with delayed language development by dichotomizing the Reynell data, and use this categorical variable as the main outcome measure. If a categorical variable is used as the outcome measure, the ‘denominator problem’ should be considered. Figure 49.2 summarizes the questions that have to be answered in this respect. The first question asks whether the numerator is going to be included in the denominator. If the answer is yes, the next question is whether time is included in the denominator or not. Let us first consider the case when the numerator is included in the denominator, but time is not. A first example of this would be a measure of the occurrence of sensorineural hearing loss in neonates. One could simply report the proportion of neonates with hearing loss: the numerator would be the number of neonates with hearing loss and the denominator would be the total number of neonates included in the study. This proportion is often referred to as prevalence. A second example would be a measure of the occurrence of head and neck cancer in people, initially free of the disease, who developed the disease within a specified time period. The numerator would be the number of subjects who developed the disease during follow up and the denominator would be the total number of subjects at the beginning of the study. This proportion is often referred to as cumulative incidence. Suppose that 1000 smokers are followed up



622 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA Table 49.1 ‘Outcome odds ratios’ and ‘determinant odds ratios’ – fictitious example of complication rates after tonsillectomy.



Is numerator included in denominator?



Dissection instruments used No



Yes



Bipolar diathermy Other Total



Is time included in denominator?



Measure:



Yes



No



Rate



Proportion



Odds



Figure 49.2 Algorithm for distinguishing rates, proportions and odds. Adapted from Ref. 16, with permission from Elsevier.



during five years and that 15 develop head and neck cancer, the cumulative incidence would be 1.5 percent. The calculation of a proportion as an estimate of disease occurrence, as described above, assumes that the entire study sample has been followed up for the specified time period. This is often not the case. Some subjects might enter the study later, and others might be lost to follow up. An alternative way of measuring outcomes that takes these varying follow-up time periods into account is to include time in the denominator. The numerator would be exactly the same as for proportions, but the sum of the time that each individual was followed up and at risk of becoming a case would be used as the denominator. In this case, an incidence rate (also called incidence density, force of morbidity/mortality, hazard) could be calculated that is equal to the number of people who develop cancer divided by the total person-time at risk during follow up. For example, on the basis of the previous example of the occurrence of head and neck cancer in smokers, one could have expressed the incidence rate of cancer as approximately three cases per 1000 people per year (or 15 divided by approximately 5  1000 years of person-time at risk). Because the incidence is very low, a simple relationship holds: cumulative incidence Eincidence  average follow-up time. Many epidemiological studies report, in some form or other, outcomes expressed as ‘odds’. It is important to distinguish odds from probability. A probability can be estimated as the proportion of people in whom a particular characteristic, such as the presence of disease, is present. The larger the group of people under study, the better the observed proportion of people with the disease reflects the ‘true’ probability. Odds, on the other hand, represents the ratio of two complementary probabilities.



Postoperative complication Yes



No



Total



36 a 24 c 60 a 1 c



264 b 676 d 940 b 1 d



300 a 1 b 700 c 1 d 1000



For example, if the probability of disease is 0.20, then the odds are 0.20%/0.80 = 1:4 = 0.25. Another way of calculating the odds is by dividing the number of subjects with the characteristic by the number without it. In other words, the numerator is not included in the denominator (see Figure 49.2). Most frequently, it is not the odds itself that is reported in epidemiological studies, but the odds ratio – the ratio of odds in different groups. This can be the ‘outcome odds ratio’, or the ratio of the odds of the outcome of interest in the group with the determinant to the odds of the outcome in the group without the determinant. Conversely, this can also be the ‘determinant odds ratio’, or the odds of the determinant in the group with the outcome compared to the odds of the determinant in the group without the outcome of interest. One reason why epidemiologists find odds ratios so attractive is that ‘outcome odds ratios’ are equivalent to ‘determinant odds ratios’, which is a fact used in the analysis of case–control studies (see Case–control studies). For example, Table 49.1 presents fictitious findings of a study on risk factors for postoperative complications after tonsillectomy. The outcome odds ratio is the same as the determinant odds ratio, which can be easily understood as both odds ratios simplify to the ‘cross product’ of the cell frequencies of a 2  2 table. The ‘outcome odds ratio’ can be calculated as a b aþb = aþb c d cþd = cþd



¼



a=b a  d 36  676 ¼ ¼ ¼ 3:8: c=d b  c 24  264



Similarly, the ‘determinant odds ratio’ can be calculated as a c aþc = aþc b d bþd = bþd



¼



a=c a  d ¼ ¼ 3:8: b=d b  c



When a dichotomous outcome variable is used (for example, alive/dead, or disease present/absent) ‘risk differences’ and ‘risk ratios’ are commonly used as ‘measures of effect’.10 The ‘risk’ in this circumstance can represent ‘prevalence’, ‘cumulative incidence’, ‘incidence’ or even ‘odds’, as described above. One might ask, ‘What is the additional risk of head and neck cancer in people who smoke, over and above that in people who do not?’ The answer to this question is obviously a risk difference. In a previous fictitious example, we supposed that we



Chapter 49 Epidemiology



followed up 1000 smokers for a period of five years and that we found that 15 of them developed head and neck cancer. The cumulative incidence of cancer in this group is therefore 1.5 percent. We can also suppose that we did the same in 1000 nonsmokers and that we found cancer in seven of them. The cumulative incidence in nonsmokers is 0.7 percent. The risk difference attributable to smoking is therefore 0.8 percent (= 1.5–0.7 percent). On the other hand, one might ask, ‘How many times are smokers more likely to get head and neck cancer compared to nonsmokers?’ Following our example, we could calculate the risk ratio as 2.1 (= 1.5 percent/ 0.7 percent). In other words, smokers are approximately twice as likely to develop head and neck cancer over a five-year period as nonsmokers. This risk ratio does not tell us anything about the magnitude of the risk difference. Even for large risk ratios, the risk difference might be quite small if the outcome is uncommon. As a rule of thumb, risk differences are more meaningful as a measure of effect in clinical situations, because it represents the actual additional probability in those exposed. On the other hand, relative risks are more meaningful as effect measures of a causal relationship in aetiological studies. HOW LARGE IS THE STUDY GOING TO BE?



A major problem of many studies is that the study size is too small. Small studies lead to imprecise results. Small studies often increase, rather than reduce, the scientific and clinical uncertainty in a specific area. Although one might argue that small studies have a value of their own because they might be included in a systematic review and meta-analysis, which would mimic the results of a larger study, it should be the goal of every study to include a number of subjects that would in itself produce ‘meaningful’ results. A considerably less frequent problem is that of a study that is larger than necessary to be meaningful. Larger studies are more difficult to carry out and may be more costly than necessary. The most important issue in this context is of course to determine what can be considered meaningful. How can the appropriate size of a study be determined? This question is most commonly answered on the basis of a statistical ‘sample size calculation’. Although the answer of a sample size calculation is exact (for example, the number of subjects to be included in a study), in many cases it depends on a number of rather subjective choices. The way that the sample size of study is to be calculated depends on whether it is an analytical study (study with a comparison or control group) or a descriptive study. For analytical studies, a research hypothesis should be formulated that further refines the study question. This hypothesis sets out the basis for statistical significance testing. Because this hypothesis guides the statistical analysis, it should be simple (addressing one determinant



] 623



or comparison and the occurrence of one outcome) and specific (defining unambiguously the target population, the control and comparison group, and the outcome of interest). For the purpose of statistical significance testing, this hypothesis should be given in the form of a ‘null hypothesis’ that states that there is no association between the determinant and the outcome. A statistical test helps to estimate the probability that an association observed in a study is due to chance (the ‘p-value’). The ‘alternative hypothesis’ states that there is such an association. Sometimes, a distinction is made between one-sided and two-sided hypotheses. Two-sided hypotheses state only that an association exists between determinant and outcome, whereas one-side hypotheses specify the direction of the association. For example, the null-hypothesis of a study comparing early surgery with watchful waiting for glue-ear in pre-school children could be that there is no difference in the language development in the two treatment groups. The one-sided null hypothesis would state that language development with early surgery is poorer than with watchful waiting. One-sided p-values are appropriate in selected circumstances, when only one direction of the association is important. One-sided p-values may seem attractive as, generally speaking, with a one-sided significance test the p-value would be half the size of that with a two-sided test, but, on the other hand, if the results of the study would suggest that the language development is indeed poorer with surgery, then this would produce a large p-value as it is likely to observe a result like that given that the null-hypothesis (language development with surgery is poorer) is true. For this reason, it is convention to use two-sided hypotheses when planning the size of a study as well as two-sided p-values when analysing the results, unless there are well-argued reasons for the contrary. When developing a study protocol for an analytical study, the investigator does not know the size of the effect that the determinant of interest has on the outcome, as one of the objectives of the study is to estimate it. For example, the investigators of the study of the effect that early surgery has on language development in children with glue-ear had to decide the smallest effect of surgery that in their view would be of clinical importance. This quantity is often called effect size or target difference. Determining this effect size is the most arbitrary step in a sample size calculation. The smaller the effect size, the larger the study needs to be, other things being equal. Previous studies could help to make a guess about what can be expected, but they do not provide any guidance about what is an important difference and what is not. Before a sample size can be calculated, two further arbitrary choices need to be made. First, the ‘alpha level’ or ‘significance level’ has to be chosen that defines the cut-off point for the p-value – conventionally 0.05 – to classify a result either as ‘significant’ (if p-valuercut-off), in which case the null hypothesis is rejected, or as ‘not significant’ (if p-value4cut-off), in which case the null



624 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA hypothesis is not rejected. Second, the ‘power’ of the significance test has to be chosen – commonly used values are 0.80 or 0.90 – that defines the probability to obtain a significant result if we assume that the defined effect size or target difference would be the ‘true effect’ or ‘true difference’. Finally, and only if a continuous outcome variable is going to be analysed, a measure of the variability (spread) of that outcome – standard deviation or variance – is needed. The greater the variability in the outcome variable among the subjects, the more difficult it will be to demonstrate a difference between two groups, and the larger the study needs to be. The required sample size can be obtained on the basis of the above-described ingredients either by using the formulae presented in statistical textbooks, from published tables or by using commonly available statistical software packages. The ingredients and the formulae for study size are purely technical. They do not take into account the value of the information obtained in the study. The most important decision is to determine the ‘right balance’ between the value of greater precision in study results against the greater costs. Essentially, this decision boils down to a cost–benefit analysis in which the benefits are more difficult to quantify than the costs. For descriptive studies, concepts such as null hypothesis, effect size, alpha level and power do not apply. Instead, descriptive statistics, such as means and proportions, are presented. The sample size of descriptive studies depends on how precise the investigator wants these descriptive statistics to be. Confidence intervals, such as 95 or 99 percent confidence intervals, are commonly used to represent the precision of the estimates. For example, a study of the five-year survival after surgical treatment in 228 patients with hypopharynx carcinoma found that the survival was 27.2 percent with a 95 percent confidence interval ranging from 21.5 to 33.5 percent.21 This interval indicates that we can be 95 percent confident (which can be considered more or less the same as saying that there is 95 percent chance) that the ‘true’ five-year survival probability lies within this interval. When estimating the sample size for descriptive studies, the investigator specifies the desired width of the confidence interval, and the sample size derives from that and can be read from tables or obtained from statistical software packages. This approach can be followed for any type of descriptive variable for which confidence intervals can be calculated.



CHOOSING THE STUDY DESIGN The taxonomy of study designs presented in Figure 49.1 provides a simple decision aid for choosing the most appropriate design. A number of questions have to be answered with the ultimate goal being to create an affordable study that provides results that are relatively



precise (free of random error) and accurate (free of systematic error or bias). A first distinction that has to be made is that between studies in which the investigator actively assigns the determinants (experimental studies) or those in which this is not the case (observational studies). A second distinction that has to be made is that between observational studies that do or do not have a comparison or control group. If so, the study is called analytical, if not, descriptive. The following sections are largely based on a series of short essays on clinical research published in the Lancet,16 and chapters of recent books about designing clinical research by Hulley and colleagues12 and clinical epidemiology.10 We start this section by describing observational studies and will end with the description of experimental studies (randomized controlled trials).



Minimizing bias and confounding We have indicated earlier that the precision of study results depends mainly on two factors: the size of the study design and the methods used to obtain the measurements. Furthermore, a study should be designed in such a way that it avoids systematic error or bias. A great number of different types of bias have been documented but it is not a major simplification to consider all these types of biases in three categories: selection bias, information bias and confounding. To examine whether selection bias would introduce systematic error in the study results, two questions need to be answered: ‘Do the sample of study subjects sufficiently represent the population of interest?’ and ‘Are the groups that are going to be compared similar in all important respects apart from the determinant(s) of interest?’ The first question is especially relevant when designing a randomized controlled trial, because trials usually enrol patients who tend to be different (often healthier) from the target population and the results tend to overestimate the effects compared to what they would be in routine practice. This contributes to the differences between the efficacy of a therapy observed in the highly controlled circumstances in selected clinical settings and the effectiveness of a treatment in actual practice. The second question refers directly to the comparability of the groups. In other words, do the groups differ importantly aside from the comparison that is being studied? Information bias results from incorrect information about the determinant or the outcome or both. The important question that has to be answered is: ‘Has information been gathered in the same way?’ In cohort studies (see Cohort studies), information about the outcome should be obtained in the same way for those with and without the determinants under consideration. Also, those who collect the information about the outcome should be unaware of (‘blind to’) the determinant status of the subjects as much as possible. In



Chapter 49 Epidemiology



case–control studies (see below), information about the determinant status should be collected in the same way for cases and controls. Confounding occurs when two determinants or risk factors are associated with each other. In that case, the effect of one is confounded with, or distorted by, the effect of the other. Confounding can occur because of selection bias, by chance, or because the two determinants are really associated in nature. Selection bias and confounding are not mutually exclusive, but they are often presented differently as they represent problems at different stages in a study. Selection bias is an issue when the sampling procedures for patients are determined during the design phase of a study, whereas confounding is an issue to consider during the analysis phase. Often in the same study more than one bias operates, as in the following example. Imagine a study that compared the effect that surgical and nonsurgical management had on speech and language development in children with persistent otitis media with effusion. In this example, selection bias could have been present if the children who received surgical treatment came from a different background and had parents who were on average of higher socioeconomic status, or if they had a family history of chronic ear infection. Information bias could have occurred because the assessment of language development might have been influenced by the fact that those who were applying and interpreting the speech and language tests knew the treatment the children had received. Finally, the conclusion that surgical treatment improved speech and language development might have been the result of confounding if the children who received surgical treatment had had more serious or more frequent ear infections. The latter bias is often referred to as ‘confounding by clinical indication’. When designing studies, investigators should try as much as possible to rule out bias. The concerns about bias need to be thought through very carefully. This relates directly to the three components of a good research question, and these components should be represented adequately in the study design. First, the study sample should adequately represent the target population. Second, the measurement of the determinants should adequately represent the comparison of interest. Third, the measurements of the outcomes should adequately represent the outcome(s) of interest. When confounding is anticipated, the investigator should consider design and analysis strategies to control its influence on the outcomes. The simplest approach in the design phase is restriction or specification. For example, investigators could decide to include only nonsmokers in a study on the effect of human papillomavirus and oropharyngeal cancer. The disadvantage of this strategy is that it reduces recruitment and prohibits generalization of the results to smokers. Another potential approach in the design phase is matching which means that subjects with matching



] 625



values for the confounding variables are included. This can be done individually (pair wise matching) or in groups (frequency matching). However, the use of matching has many disadvantages. The recruitment process can become rather cumbersome. Special analytical techniques are required that represent the fact that the subjects are not sampled independently. There is also the danger of overmatching – matching on the basis of a variable that is not a confounder, which would reduce the study’s power. Given these disadvantages, matching should be avoided if other strategies could provide sufficient control for confounding. In the analysis phase, the investigator can use either stratification or adjustment. Stratification ensures that the analysis compares like with like, first by comparing subjects within a group (or stratum) of subjects with similar values of one or more confounding variables and then, in a second step, by calculating an overall estimate by pooling the stratum-specific results. The most important disadvantage of this approach is the limited number of variables that can be controlled for. Adjustment on the basis of a statistical multivariate model does not suffer from this limitation, but its effects depend on the adequacy of the model fit. Both approaches depend on the completeness, precision and accuracy of the measurements of the confounders.



Observational studies DESCRIPTIVE STUDIES



Descriptive studies can describe the experience in only one individual, a case report, or a report on a series of individual cases, the case-series report. Case series are the most frequently published studies in surgery. An important function of descriptive studies is that they can describe the distribution of diseases or disease-related conditions and events in a specified population. Good descriptive research should address the five ‘W’ questions – who, what, why, when and where, although as soon as these questions are addressed on the basis of the data, the distinction between descriptive studies (no comparison) and analytical studies (comparison between groups essential) starts to disappear, but a distinctive characteristic of descriptive studies is that they cannot directly quantify the risk that can be attributed to a specific determinant. For example, a series of three cases was published recently of perichondritis of the pinna as a result of ‘high’ ear piercing, and was used to focus attention on the risks of body piercing.22 The study also presented the increasing trend in the number of hospital admissions for perichondritis of the pinna in the UK and Wales from approximately 600 in 1990–1991 to approximately 1200 in 1997–1998. The latter is an example of a surveillance study, which is another important type of descriptive study. A surveillance study is an ongoing and systematic collection, analysis and interpretation of health



626 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA data essential to the planning, implementation and evaluation of health services and public health practice. The fact that this particular descriptive study cannot quantify the attributable risk might not be a weakness as most cases of perichondritis of the pinna may be due to piercing practices in the first place. The strength of descriptive studies is that they often use existing data and thus are inexpensive and efficient to carry out. Furthermore, few ethical limitations exist. The disadvantages are that there is a danger of overinterpretation. For example, in early 2001 the UK government recommended that single-use instruments should be used for adenotonsillectomy to minimize the risk of transmission of variant Creutzfeldt-Jacob disease (vCJD). In the course of the same year, suspicion was raised about the postoperative haemorrhage rate. It was thought that the increase in the haemorrhage rate was due to the poor quality of the single-use instruments and, in late 2001, the UK government recommended the reintroduction of reusable instruments. The need to consider alternative explanations arose later as more detailed analyses of hospital admission data revealed that there also had been a gradual increase in the posttonsillectomy complication rate from 3.9 percent in 1995–1996 to 7.3 percent in 2000–2001 (unpublished data, Van der Meulen JHP, 2003).



Analytical studies An important distinction that is often made with respect to observational studies that are analytical is whether the determinants and outcomes relate to phenomena that occur at the same time. If so, it is called a cross-sectional study. If not, it is either a cohort study (looking forward in time from determinant to outcome) or a case–control study (looking backward from outcome to determinant).



CROSS-SECTIONAL STUDIES



The single most important characteristic of crosssectional studies is that the determinant(s) as well as the outcome(s) measured occur at the same time, and that there is no follow-up period. Cross-sectional designs are very well suited to describe the prevalence of health and disease-related conditions and their distribution patterns. For example, a study was carried out in 864 school children in the UK to investigate the effect of parental smoking on middle ear disease.23 Data on parental smoking were collected with a questionnaire and glueear was considered to be present in children with a flat tympanogram. The study identified 82 cases in total, 45 in the 407 households with at least one smoker and 37 in the 457 households with no smokers, and a risk ratio of 1.4 (= 11.0 percent/8.1 percent, with a 95 percent confidence interval ranging from 0.9 to 2.1) could be calculated.



The major strengths of cross-sectional studies is that they examine the presence or absence of an outcome and determinant at the same time. Cross-sectional studies are therefore fast, relatively cheap, and there is no ‘loss to follow-up’. A major disadvantage of cross-sectional studies is that they provide only a snap shot of complicated temporal relationships between determinant and outcome. Parental smoking, for example, is likely to be persistent, but chronic middle ear infection in children is an intermittent disease. As a consequence, causal relationships are difficult to establish with cross-sectional studies. For the same reasons, they are not suitable for producing information on prognosis or the natural history of a condition. Studies of the accuracy of diagnostic tests are a special case of cross-sectional studies. A key feature of such studies is that the diagnostic test results and ‘gold standard information’ or reference test results about the presence or absence of the conditions of interest are collected at the same time. The subjects should have both undergone the diagnostic test in question (for example, ultrasound-guided fine-needle aspiration to diagnose malignant disease in patients with nodular thyroid disease), as well as the reference test (histology after surgical resection). The results of one test should not be available to those who are carrying out the other. In terms of the example, the pathologist examining the surgical specimen should be ‘blind’ to the results of the fine needle aspiration. Another important criterion for a valid evaluation of the diagnostic accuracy is that the diagnostic test is evaluated in ‘an appropriate spectrum of patients’ (such as those in whom we would use it in practice). The accuracy of a diagnostic test, a measure of how well the test results help to distinguish between individuals with and individuals without the disease or target disorder, is often expressed in terms of sensitivity and specificity. However, it is important to realize that this is, in most cases, a simplified representation of the diagnostic accuracy. Sensitivity and specificity can only be calculated for a test with dichotomous results, often called ‘positive’ if the result suggests that the disease is present and ‘negative’ if it suggests that disease is absent. The sensitivity is estimated as the proportion of individuals with the disease who have a positive test result. The specificity corresponds to the proportion of individuals without the disease who have a negative test result. The sensitivity and specificity of a ‘perfect’ test are 100 percent. The sum of the sensitivity and specificity is sometimes used as a single measure of the diagnostic accuracy. The sum of the sensitivity and specificity for a ‘worthless’ test, a test that would perform no better than tossing a coin, is expected to be 100 percent and for a perfect test 200 percent. In many cases, however, relying on sensitivity and specificity as measures of diagnostic accuracy is too simplistic. For example, many tests, if not all, have more



Chapter 49 Epidemiology



than two possible outcomes, and dichotomizing the results in a positive and a negative test result reduces the amount of diagnostic information that a test can provide. Furthermore, if one would like to dichotomize the test results, it is not immediately obvious what the cut-off value between positive and negative results should be. Lastly, the actual aim of performing a diagnostic test is not to distinguish between individuals with and without the disease, but to identify those individuals for whom the expected benefit of treatment outweighs the expected harm. This indicates that the classic diagnostic ‘steppingstone approach’, jumping from signs and symptoms to diagnosis and then from diagnosis to treatment, is rather artificial in many cases. This should be reflected in studies assessing diagnostic tests. They should focus on clinical effectiveness (effect on patient outcomes) rather than on diagnostic accuracy alone (for a comprehensive review of methods for the evaluation of diagnostic technologies, see Knottnerus.24) COHORT STUDIES



Cohort studies follow groups of individuals over time. The word ‘cohort’ originates from the Latin word for a group of 300 to 600 soldiers in a Roman legion. This is appropriate as the same rule applies for epidemiological cohorts as once for Roman cohorts: if a person joins a cohort, he will be a member of that cohort forever, which emphasizes the need for the completeness of follow up. Most follow-up studies are analytical (comparing the occurrence of outcomes according to presence or absence of certain determinants), but they can be descriptive as well. The design of many cohort studies is relatively straightforward. A group of people is assembled, none of whom have experienced the outcome of interest, but all of whom could experience it. On study entry, people may be classified according to the determinants of interest. These people are then observed over time to see in which of them the outcome of interest occurs. Despite this simple design, the terminology used to describe these studies can be confusing. For example, terms such as ‘longitudinal study’, ‘prospective study’ or ‘incidence study’ are often used interchangeably. The only real problematic issues in this respect is the use of the term ‘prospective study’, as cohort studies can consider data that are collected prospectively as well as retrospectively. The terms ‘retrospective’ and ‘prospective’ refer to the way that the data have been collected rather than to the study design. The essential characteristic of prospective data collection is that data are being collected on determinants and outcomes that manifest themselves after the establishment of a study protocol. In all other situations, the data collection should be considered to be retrospective. Retrospective data collection can be based on data recorded in the past for other purposes or on the memory of the study subjects, investigators or other



] 627



parties. It is obvious that the precision and accuracy of prospectively collected data can be expected to be better than retrospectively collected data. Cohort studies can be based entirely on retrospective data (using retrospective data on both the determinants and the outcome), based entirely on prospective data (using prospective data on both the determinants and the outcome), or based on a mix of retrospective data on determinants and prospective data on the outcome. Cohort studies have many attractive features. They are the best way of ascertaining the incidence of a disease or health-related event or condition, as well as the natural history of a disorder. They also provide insight into the temporal order of determinants and outcome, which strengthens the inferences that can be made about whether an observed risk factor is a cause of the outcome. Furthermore, multiple outcomes can be considered which is especially relevant if cohort studies are compared with case–control studies (see Case–control studies). However, the dangers of considering multiple outcomes are also obvious since testing many hypotheses may lead to misleading results. Lastly, prospective cohort studies are especially valuable for the study of fatal diseases, or more general studies of the occurrence of disease-related conditions with a short duration. When these kind of conditions are studied retrospectively, the observed occurrences may be an under-representation of all occurrences. The major drawback of cohort studies is that they are relatively expensive and therefore an inefficient way to study rare outcomes. Cohort studies become more efficient as the outcomes become more common. Another disadvantage for cohort studies that collect outcome prospectively is that the results may not be available for a long time. A prospective study of the effect of passive smoking on the head and neck cancer incidence may take more than ten years to come up with relevant results. However, not all prospective cohort studies need to take a very long time to complete. A cohort study could be used, for example, to study the risk factors for complications after tonsillectomy. Such a study could prospectively collect patient characteristics as well as data on surgical technique, instruments used and experience of the surgeons. Every patient could be followed up for a certain period, say one month, and the occurrence of complications could be registered. Such a study would provide insight into the incidence of complications – it might even distinguish between primary haemorrhage, secondary haemorrhage as well as infection, and also in the way this incidence depends on relevant risk factors. CASE–CONTROL STUDIES



Case–control studies can be considered as ‘research in reverse’.16 Many epidemiologists consider them as one of the most important tools in their armamentarium. The design of case–control studies can be appreciated by again



628 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA considering the example of the study on risk factors for complications after tonsillectomy. For this study, we could also use a case–control design. All the patients who experience a complication in the first month after surgery could be considered as ‘cases’, and patients without postoperative complications as ‘controls’. The investigators then have to look back in time to find data on the patient’s risk factors that were present at the time of surgery. The essential feature of a case–control study is that not all controls need to be included but that a selection would suffice. Many case–control studies include only one control for each case, whereas others included more but hardly ever more than four or five. The reason that the number of included controls per case is seldom higher is that including more controls will have only a small effect on the power of the study and on the precision of the estimates (‘law of diminishing returns’). The fact that only information has to be collected for a selection of the controls makes that, in many circumstances, case–control studies are the most efficient design in terms of time, effort and therefore money. Hence, case–control studies are especially relevant if the occurrence of the outcome of interest is low. However, if the frequency of the determinant of interest is low, case–control studies might become inefficient. For example, if the frequency of a certain risk factor of post-tonsillectomy complications is low, investigators have to examine many cases and controls to find some who have been exposed. Some have advocated a ‘rule of thumb’ stating that cohorts are more efficient than case–control studies if the occurrence of the outcome is more frequent than that of the determinant and vice versa. Unlike cohort studies, case–control studies cannot produce an estimate of the occurrence of the outcome of interest since we lack information about the denominator. A case–control study of post-tonsillectomy complications will neither provide information on how many patients underwent the operation nor on what their risk factors were. The relevant effect measure that a case–control study can provide is the odds ratio for a determinant, derived from the proportions in cases and in controls (more precisely, the odds) in whom the determinant is present. Earlier, we referred to this odds ratio as the ‘determinant odds ratio’. This determinant odds ratio is equivalent to the ‘outcome odds ratio’ and the odds ratio obtained from a case–control study can therefore be used as a measure of relative risk. The advantage of case–control studies (efficiency in time, effort and money) comes at a price however, because two methodological issues may introduce major systematic errors: selecting the control group and obtaining information about the determinants. The selection of cases is relatively straightforward provided that the definition of the outcome being studied (the ‘case definition’) is clear. The selection of the control group is more problematic. Two criteria need to be met. First,



controls should be representative of the population at risk of becoming cases. In other words, the controls should have been selected as cases had they developed the disease or outcome of interest themselves. Second, selection of the cases should be independent of the determinant(s) being investigated. Therefore, a case–control study of risk factors for post-tonsillectomy haemorrhage should include as controls only patients who underwent a tonsillectomy themselves. Furthermore, it seems inappropriate to select as controls all patients who underwent a tonsillectomy immediately after patients who developed complications. If that were the case, it would be very likely that the same surgeon using the same technique and the same type of instruments treated these patients. It is left as an exercise for the reader to decide what an appropriate strategy to select controls would be. RANDOMIZED CONTROLLED TRIALS



Randomized controlled trials are cohort studies with prospective data collection. Their distinctive feature is that they use a random allocation scheme to assign the determinant. With randomization, you can expect that the prognostic characteristics of the randomized groups or ‘arms’ of the trial are similar except for differences due to chance variation. Randomization eliminates the influence of both known and unknown confounders that are present at the time of randomization. Without randomized treatment allocation, it cannot be excluded and is in practice very likely that imbalances in prognostic factors between the groups occur that are the result of selection bias – a type of bias often called ‘confounding by clinical indication’. The ethical argument in favour of a random process deciding what intervention patients receive is that there is equipoise, or a state of genuine uncertainty on the part of the clinical investigator regarding the comparative merits of each intervention. If there is genuine uncertainty about which treatment is best, it is not possible to recommend one over the other which justifies that a random process decides. The problem is, however, that equipoise depends on subjective judgements and that therefore experts may disagree. Furthermore, preferences of patients or those who are candidates to receive the intervention have to be taken into account. The process of informed consent should address all these issues to ensure that patients can evaluate the potential risks and benefits of the study from their own perspective before they agree to participate. It is important that the randomization process is ‘concealed’ from the investigators who include the participants into the trial. Proper allocation concealment requires that the investigators do not know the arm to which a participant will be allocated until the participant has definitively been recruited and included in the study. Concealment of the randomization is the only way to prevent the investigators influencing the balance of the



Chapter 49 Epidemiology



prognostic characteristics between the groups that are being compared. For example, prior knowledge of the next allocation may allow investigators to exclude certain candidate participants from the trial because they are perceived to be allocated to an inappropriate group. More directly, the investigators may try to influence the order of inclusion. Concealment of treatment allocation is so important that with inadequate concealment a randomized controlled trial should be considered nonrandomized. Empirical studies have shown that trials with inadequate concealment overestimate the treatment effect by as much as 40 percent on average. Another important feature of many randomized controlled trials is the use of a form of blinding, which prevents the participants and the investigators who are in contact with the participants from being aware of which treatment has been offered. Blinding can help to prevent bias in a number of ways. First, if participants do not know what treatment they are receiving, it is less likely that their perceptions and expectations of the treatment that they receive can influence their compliance and the physical and psychological response to the intervention. For example, most patients expect that a new treatment is better than an existing one. Second, blinding investigators is important, as this would prevent them consciously or unconsciously managing the participants in the trials differently. Also, their attitude towards the treatment can influence how patients respond to the treatment. Third, blinding participants and investigators will prevent outcomes from being assessed differently. The terms ‘singleblind’ and ‘double-blind’ are often used to indicate trials in which only the participants or both the participants and the investigators are blinded. Double blinding is impossible in almost all surgical trials. To avoid biased management, the investigators should then try to standardize other potential treatments as much as possible. Approaches to minimize biased assessment of outcomes could include the use of a third party who is unaware of the treatment originally given. When blinded assessment is not possible, one should try as much as possible to use ‘hard outcomes’ (based on measurements resistant to bias). Another alternative includes the use of standardized outcome measurement scales that can be completed by the participant. This approach is likely to produce less biased outcomes than the judgement of an investigator. Many people also consider the use of a placebo treatment as an essential feature of a randomized controlled trial. Placebo treatment is a form of treatment indistinguishable from the ‘active treatment’ under study, but it does not have a specific known mechanism to influence a patient’s health. Blinding often requires the use of placebo treatment. Apart from that, the choice whether to use placebo treatment depends on the question that the trial tries to answer. First, if the question is whether intervention A is better than intervention B, then it is obvious that one should compare the effects of



] 629



A and B with each other and not against placebo treatment. Second, if the question is whether intervention A is better than no treatment at all, the answers may be different for researchers and clinicians. Researchers are likely to be more interested in the specific effects – effects that can be attributed to the ‘active component’ of an intervention. Clinicians are likely to be more interested in the combined effects of the active and placebo components. On the other hand, it is always useful to know what part of the total effect is due to the active component and what part to the placebo when balancing the potential benefits against potential harms and costs, which are likely to differ between the active and placebo components. It is obvious that placebo treatment is rarely an option for randomized controlled trials in surgery. For example, in a trial comparing tonsillectomy with nonsurgical management, it would require a form of ‘sham surgery’ to provide a placebo treatment. It is left to the reader to decide what the advantages and disadvantages of using such a form of placebo treatment in this context would be. If a substantial number of participants do not receive the study interventions, do not comply with the study protocol or are lost to follow up, the trial is likely to be underpowered and its results biased. Strategies to maximize compliance and follow up should therefore be an integral part of every trial protocol. An obvious strategy is to make participation in a trial as convenient, painless and enjoyable as possible. Some trials have a ‘run-in period’ that can be used to ‘screen out’ patients who may not adhere to the study protocol and the followup procedures. It is essential to ascertain that compliance and completeness of follow up do not differ between the trial arms as this could lead to biased estimates of the effects of the interventions. The results of trials can be analysed in two ways. First, the comparison of the intervention can be carried out according to the intervention to which the patients were randomized (‘intention-to-treat’ analysis) or according to the treatment they actually received (‘per protocol’ analysis). The advantage of an intention-to-treat analysis is that the question that is being addressed corresponds exactly with the one clinicians and patients try to answer in clinical practice. The disadvantage is that if many patients do not receive the treatment they were randomized to, this would obscure the difference between the trial arms. Per protocol analysis, on the other hand, addresses which intervention is better more directly. With this form of analysis, the treatments are being compared according to the treatments that the patients actually received. The problem with this approach is that if many patients do not receive the treatment to which they were randomized, the study no longer represents an experimental study. The analysis should focus on a single outcome – often referred to as the ‘primary endpoint’ – to avoid the problems of interpreting the outcome of multiple



630 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA hypothesis tests. This primary endpoint should also be used for the power calculation in the design phase of the trial. It is often desirable to consider a number of secondary endpoints – outcomes that represent different aspects of the outcome of interest – to provide a more detailed picture of the effects of the interventions under study. For example, a randomized controlled trial of tonsillectomy and nonsurgical management could consider the reduction in the number of episodes of tonsillitis or sore throat as primary endpoint, but could also collect information on postoperative complications, reduction in time off work and reduction in the use of certain drugs, such as analgesics and antibiotics. Subgroup analyses are comparisons between randomized groups in a subset of the patients with specific characteristics. The most important question that subgroup analyses try to answer is whether the effect measure is different in different subgroups by carrying out a statistical test for interaction (or ‘effect modification’). For example, one could investigate – when analysing the results of the trial on tonsillectomy – whether the effects of tonsillectomy on recurrence of tonsillitis are different in children younger than five years compared to children of five years and older. Subgroup analyses can be dangerous and misleading, but they can also provide an important extra insight into the generalizability of the results. One danger arises from the fact that multiple comparisons are carried out, which increases the risk of producing false–positive results. To avoid this risk, only a limited number of subgroup analyses should be carried out and they should be specified in the study protocol, in other words, before the results of the trial are known and on the basis of patients characteristics that are measured before randomization. Furthermore, the actual number of subgroup analyses carried out should be reported. Another approach to minimize the risk of false-positive results is to reduce the significance level of the statistical tests for interaction (for example, from the conventional 0.05 to 0.05 divided by the number of tests carried out as specified by the Bonferroni method). A further danger is the limited power of the subgroup analyses, because the group size is, by definition, smaller than that of the original trial population. An important problem of randomized controlled trials is that of generalizability. Most trials are carried out in highly controlled conditions in a selected group of patients. Their results therefore provide evidence about the efficacy of an intervention (does the treatment work under ideal circumstances?) and not about the effectiveness (does the treatment work in actual practice?). It depends on the question that needs to be answered as to what extent the design of a trial should focus on efficacy or on effectiveness. An explanatory trial (addressing the efficacy question) is needed in the early stage of a new intervention. However, a pragmatic trial (addressing the effectiveness question that aims to create study conditions



that reflect as much as possible actual practice when effectiveness is more important) is more appropriate to answer whether an intervention should be included in the ‘repertoire’ of available services. It is obvious that not all questions about the effectiveness of surgical procedures can be addressed with a randomized controlled trial.25 A number of obstacles have been identified that make randomized controlled trials inappropriate (outcome of interest relatively rare, relevant outcomes far in the future, or randomization affecting the effectiveness of a procedure), impossible (refusal of clinicians or patients to participate, ethical obstacles), or inadequate (low generalizability of experimental studies).26 If the obstacles are insurmountable, carefully designed observational studies should be considered.



Concluding remarks This chapter introduces the ‘determinant–occurrence relationship’ as a key concept for medical research, and epidemiology as a methodological discipline with immediate relevance for otolaryngological research. A basic understanding of epidemiological principles is therefore essential for all people who are involved in clinical research and desirable for all clinicians who want to read the medical literature critically. We have only presented a broad picture of epidemiological concepts and principles, but it will be clear to the reader that studies vary according to the likelihood that their results are accurate (free of systematic error) and precise (random error). For this reason, a ‘hierarchy of evidence’ has been suggested.11 For therapeutic issues, this hierarchy ranges from well-conducted randomized controlled trials at the top – preferably summarized in a systematic review and meta-analysis – to case-reports and expert opinion at the bottom (used throughout this book). For diagnostic and prognostic issues, very different hierarchies are necessary. These hierarchies of evidence are not absolute. If the therapeutic effects are very large in comparison to the potential effects of bias – for example, effects of insulin in ketoacidosis – a description of a series of cases may already provide compelling evidence. Nevertheless, it should be a leading principle that users of research try to look for the best available evidence from this hierarchy.



KEY POINTS  Epidemiology is a methodological discipline. Epidemiological concepts and principles will help to design empirical studies that are relatively precise (free of random error), accurate (free of systematic error or bias) and



Chapter 49 Epidemiology



efficient (affordable in terms of time and money).  The ‘determinant–occurrence relationship’ is a key epidemiological concept, and many clinical research questions can be ‘framed’ as determinant–occurrence relationships.  Epidemiological studies can be distinguished into experimental studies (investigators assign the determinants themselves, e.g. randomized controlled trial) and observational studies (investigators examine events as nature takes its course). Observational studies without a comparison or control group are called descriptive studies.  A research protocol is a document that provides all essential details of a study. It should contain information on: – research question; – study design; – selection and recruitment of the subjects; – measurements; – statistical analysis; – sample size calculation. A ‘hierarchy of evidence’ has been suggested on the basis of the likelihood that results are accurate and precise. For therapeutic issues, this hierarchy ranges from well-conducted randomized controlled trials at the top – preferably summarized in a systematic review and meta-analysis – to case-reports at the bottom.











ACKNOWLEDGEMENTS Jan van der Meulen is supported by a National Public Health Career Scientific Award, Department of Health, NHS R&D, UK. David Lowe was supported by a project grant from the Department of Health, UK.



REFERENCES 1. Burton MJ, Towler B, Glasziou P. Tonsillectomy versus nonsurgical treatment for chronic/recurrent acute tonsillitis (Cochrane Review). In: The Cochrane Library, Issue 1, Oxford: Update Software, 2003. 2. Glover JA. The incidence of tonsillectomy in school children. Proceedings of the Royal Society of Medicine. 1938; 21: 1219–36. 3. Bakwin H. Pseudodoxia pediatrica. New England Journal of Medicine. 1945; 232: 691–7. 4. Sackett DL, Haynes RB, Guyatt GH, Tugwell P. Clinical epidemiology: a basic science for clinical medicine. Boston: Little Brown, 1985.







] 631



5. Rothman KJ, Greenland S. Modern epidemiology. Philadelphia: Lippincott Williams & Wilkins, 1998. Comprehensive and up-to-date textbook that can be used as a reference source for epidemiological principles and concepts. 6. MacMahon B, Pugh TF. Epidemiology: principles and methods. Boston: Little Brown, 1970. 7. Hennekens CH, Buring JE. Epidemiology in medicine. Philadelphia: Lippincott Williams and Wilkins, 1987. 8. Miettinen OS. Theoretical epidemiology: principles of occurrence research in medicine. John Wiley and Sons, 1985. 9. Loomis D, Wing S. Is molecular epidemiology a germ theory for the end of the twentieth century. International Journal of Epidemiology. 1990; 19: 1–3. 10. Fletcher RH, Fletcher S, Wagner EH. Clinical epidemiology: the essentials. Philadelphia: Lippincott Williams & Wilkins, 1996. Accessible and well-written book focussing on methods and techniques for clinical epidemiology. 11. Guyatt G, Haynes B, Jaeschke R, Cook D, Greenhalgh T, Meade M et al. Introduction: the philosophy of evidencebased medicine. In: Guyatt G, Rennie D (eds). Users’ guides to the medical literature: essentials of evidence-based clinical practice. Chicago: AMA Press, 2002. 12. Manolidis S, Takashima M, Kirby M, Scarlett M. Thyroid surgery: a comparison of outcomes between experts and surgeons in training. Otolaryngology and Head and Neck Surgery. 2001; 125: 30–3. 13. Mason J, Freemantle N, Browning G. Impact of effective health care bulletin on treatment of persistent glue ear in children: time series analysis. British Medical Journal. 2001; 323: 1096–7. 14. Butler CC, Van der Linden MK, MacMillan H, Van der Wouden JC. Screening children in the first four years of life to undergo early treatment for otitis media with effusion (Cochrane Review). In: The Cochrane Library, Issue 3, Oxford: Update Software 2003. 15. Hulley SB, Cummings SR, Browner WS, Grady D, Hearst N, Newman TB. Designing clinical research. Philadelphia: Lippincott Williams & Wilkins, 2001. A readable book designed to help beginning investigators to get started in the world of clinical research. The authors have tried to keep it simple and have used numerous examples to explain the options for the design and implementation of clinical epidemiological studies. 16. Grimes DA, Schulz KF. An overview of clinical research: the lay of the land. Lancet. 2002; 359: 57–61. 17. Piccirillo JF, Merritt Jr. MG, Richards ML. Psychometric and clinimetric validity of the 20-Item Sino-Nasal Outcome Test (SNOT-20). Otolaryngology and Head and Neck Surgery. 2002; 126: 41–7. 18. Johns MW. A new method for measuring daytime sleepineess: the Epworth Sleepiness Scale. Sleep. 1991; 14: 540–5. 19. http://www.asahq.org/clinical/physicalstatus.htm, accessed September 2003.



632 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA 20. Maw R, Wilks J, Harvey I, Peters TJ, Golding J. Early surgery compared with watchful waiting for glue ear and effect on language development in preschool children: a randomised trial. Lancet. 1999; 353: 960–3. 21. Eckel HE, Staar S, Volling P, Sittel C, Damm M, Jungehuelsing M. Surgical treatment for hypopharynx carcinoma: feasibility, mortality, and results. Otolaryngology and Head and Neck Surgery. 2001; 124: 561–9. 22. Hanif J, Frosh A, Marnane C, Ghufoor K, Rivron R, Sandhu G. Lesson of the week: ‘High’ ear piercing and the rising incidence of perichondritis of the pinna. British Medical Journal. 2001; 322: 906–7.



23. Strachan DP. Impedance tympanometry and the home environment in seven-year-old children. Journal of Laryngology and Otology. 1990; 104: 4–8. 24. Knottnerus JA (ed.). The evidence base of clinical diagnosis. London: BMJ Books, 2002. 25. McCulloch P, Taylor I, Sasako M, Lovett B, Griffin D. Randomised trials in surgery: problems and possible solutions. British Medical Journal. 2002; 324: 1448–51. 26. Black N. Why we need observational studies to evaluate the effectiveness of health care. British Medical Journal. 1996; 312: 1215–8.



50 Outcomes research IAIN RC SWAN



Introduction What are patient-based outcome measures? Why use patient-based outcome measures in research? Assessment of patient-based outcome measures Types of patient-based outcome measure Generic instruments Specific instruments



633 634 634 634 637 637 640



How to choose patient-based outcome measure Which generic instrument? Key points Deficiencies in current knowledge and areas for future research References



641 642 642 642 642



SEARCH STRATEGY The aim of this chapter is to explain the background to patient-based outcomes research. There is a vast literature on this subject, so there was no formal literature search. The emphasis is on instruments relevant to otolaryngology. The instruments described are simply examples for the reader. The author does not suggest that these are the best instruments in their subject area.



INTRODUCTION Clinical outcomes research examines the outcomes of treatment or of disease. Traditionally, outcomes of medical care are based on clinical observations or laboratory measurements. While these measures provide useful information for the clinician, they are often of limited interest to patients. There is often poor correlation between clinical outcomes and functional capacity and well-being which are the areas of most interest to the patient. There has been increasing recognition that traditional measures need to be complemented by some measure reflecting the impact of the intervention on the patient in terms of health status and health-related quality of life. These terms refer to experiences of illness such as pain, fatigue, disability and broader aspects of the individual’s physical, emotional and social well-being. Medicine, in particular surgery, formerly had the principal objective of reducing mortality and morbidity. These objectives are usually straightforward to assess.



Nowadays, a large proportion of clinical practice is either cancer or chronic disease. There has been little improvement for some time in survival in cancer while treatments often have associated side effects and functional impairment which significantly affect the patient’s quality of life. There is an increasing prevalence of chronic diseases with an ageing society and here the aims of treatment are to arrest or reverse decline in function.1 These factors have led to an increased interest in patient-based outcome measures. At the same time, increased attention is being given to patients’ opinions and wishes in relation to their health. Patients should be involved in decisions about their treatment. To contribute usefully, they need information about the outcomes of treatment, not just in terms of surgical results but in terms of the possible effects on their quality of life. Financial resources limit health care around the world, and increasingly the distribution of these resources is influenced by the benefits perceived by patients, their carers and society as a whole.



634 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA



WHAT ARE PATIENT-BASED OUTCOME MEASURES? Patient-based outcome measures are, in general, questionnaires that ask patients about their perception of their health. Usually, these instruments are made up of a number of items or questions. These items are linked in a number of domains or dimensions. A domain refers to an area of behaviour or experience, such as mobility, self-care, depression, pain, social functioning and general well-being. Many questionnaires focus on physical function, such as ability to walk, climb stairs, wash and dress themselves. Others ask about the impact of health on various areas of an individual’s life, such as ability to socialize with members of their family and friends. These are aspects of health-related quality of life (HRQoL). Overall quality of life is influenced by many factors other than health, such as social, financial and physical factors. Patient-based outcome measures assess only one aspect of quality of life and are not measures of overall quality of life. Patient-based outcome measures assess some aspect of the patient’s subjective experience of health and the consequences of illness – and of treatment. As these experiences are those of an individual patient with an individual personality and lifestyle, they cannot be objectively verified. This point is sometimes raised as a criticism of patient-based outcome measures, but it should be borne in mind that many clinician-based outcomes are the subjective opinion of the clinician.



reliable, objective clinical measures to assess the outcome of treatment. To justify our treatment we need patientbased outcome measures to demonstrate the efficacy of treatment – improvement in HRQoL. We also need patient-based outcome measures to demonstrate the effects of these non-life-threatening conditions on HRQoL. Even when we have objective measures, e.g. closure of a tympanic membrane in myringoplasty, we often know little about the effects on HRQoL. There are also occasions where there is disagreement between clinical measures of success and HRQoL outcomes, e.g. septal surgery for deviated nasal septums. It is likely that the results from one of the measures are unreliable or that they are measuring different things – the validity of both measures should be questioned.



ASSESSMENT OF PATIENT-BASED OUTCOME MEASURES There are eight criteria that should be applied to patientbased outcome measures: appropriateness, validity, reliability, responsiveness, precision, interpretability, acceptability and feasibility3 (Table 50.1). There are few patient-based outcome measures for which there is sufficient evidence to allow judgement on all of these criteria.



Appropriateness WHY USE PATIENT-BASED OUTCOME MEASURES IN RESEARCH? In the early days of research, few if any studies included an assessment of health-related quality of life. It is increasingly argued now that clinical trials should include patient-based outcome measures except where it is clear that these are not relevant outcomes. The UK Medical Research Council (MRC), the European Organisation for Research and Treatment of Cancer (EORTC) and the National Cancer Institute of Canada (NCIC) all have policies stating that the likely impact on quality of life should be assessed, or justification provided for not doing so.2 Patient-based outcome measures have been used as the primary outcome measure in randomized controlled trials in many areas including cancer and heart disease. They are also useful in providing evidence of the overall value of a treatment in a way that allows comparison with other treatments in the same area or in other areas. Patient-based outcome measures are particularly relevant in otolaryngology. Head and neck cancer forms a small proportion of our patients, but a much larger proportion of our clinical workload. The majority of our patients do not have a life-threatening condition and the morbidity is small. Most of our patients simply want us to make them feel better. In many cases we do not have



The first and most fundamental question when selecting a patient-based outcome measure is how to identify one Table 50.1



Assessment of patient-based outcome measures.



Criterion Appropriateness Validity Reliability Responsiveness



Precision Interpretability Acceptability



Feasibility



Meaning Does the content of the instrument match the intended purpose of the trial? Does the instrument measure what it claims to measure? Does the instrument produce the same results when repeated in the same population? Does the measure detect clinically meaningful changes in the patient condition? Can the instrument detect small differences between patient groups? Can results from the measure be interpreted clinically and are they relevant? Is the format of the instrument and the questions acceptable to the planned subjects? Is it feasible to use this instrument in this setting with these subjects?



Chapter 50 Outcomes research



that is appropriate to the aims of the particular trial. The aims of the trial, the patient group being studied, the type of treatment and the relevant quality of life questions should be carefully considered. The instrument should measure aspects of patients’ lives that patients consider important, and should not omit aspects of HRQoL that are important to the patients in the trial.4 Clinicians often think that they know what aspects of HRQoL are important to patients. Many studies have demonstrated, however, that patients’ views often differ from clinicians’. The most effective way of establishing the importance to patients is asking patients their views. A list of aspects of HRQoL can be drawn up by clinicians and patients. A group of patients can then be asked which of these items are problems for them and how important these items are. This is the method commonly used in creating the patient-based outcome measures which are widely used, such as the Medical Outcome Study Short-Form 36-Item Health Survey (SF-36). The purpose of the trial must be specified precisely in order to select an instrument that fits that purpose. In many studies, the rationale for selection of outcome measures is not clear. Careful consideration of content and relevance of a questionnaire to the purpose of the trial should be given. The instrument selected must be as relevant to the health problem and the proposed intervention as possible. It is often recommended that one generic and one disease-specific instrument be used in a trial to increase the likelihood of appropriate assessment of outcomes.



Validity The validity of a measure is an assessment of the extent to which it measures what it claims to measure. Validity is not a fixed property of a measure, but is assessed in relation to a specific purpose and setting.5 It is, therefore, meaningless to refer to a validated outcome measure, as many reports do. Evidence of validity of an outcome measure in one situation does not mean that there will be adequate validity in another research setting. This apparently simple property depends on a range of different types of evidence, including how the content was chosen and its relationship to other variables. There are several different ways of assessing validity of a patientbased outcome measure. No single set of observations is likely to determine validity, so assessment of validity in relation to a specific trial is not straightforward.



CRITERION VALIDITY



Criterion objective measures criterion



validity is the correlation of a measure with an or ‘gold standard’ measure. As gold standard rarely exist in assessment of quality of life, validity is rarely relevant in patient-based



] 635



outcome measures, and validity is judged by a more indirect assessment of content and construct validity. It can be assessed when a shorter version of an instrument is used to predict the results of a full-length version.



FACE AND CONTENT VALIDITY



Face and content validity are among the most relevant issues for the use of patient-based outcome measures in clinical trials. They address whether items clearly address the intended subject matter and whether the range of aspects is adequately covered. They are explained by Guyatt et al.:6 ‘Face validity examines whether an instrument appears to be measuring what it is intended to measure, and content validity examines the extent to which the domain of interest is comprehensively sampled by the items or questions in the instrument.’ In other words, does the questionnaire look right and does it cover the right things? Face and content validity are mainly based on careful examination of the content of the instrument and qualitative judgement rather than statistical criteria. Evidence of how the questionnaire was initially developed is useful. Questionnaires with good validity are constructed in phases with involvement of patients with experience of the particular health problem. The content of poor questionnaires is chosen by ‘experts’. CONSTRUCT VALIDITY



Construct validity is also very relevant but is a quantitative assessment of the relationship of a construct to other variables. A construct is a theoretical idea about the domain to be measured. For example, patients with hearing disability should have poorer audiometric thresholds. Many patient-based outcome measures are multidimensional: they assess, for example, physical, psychological and social aspects of an illness. Those questions related to psychological aspects should correlate with each other much more than with questions assessing physical function. The internal structure of such instruments is established by construct validation, most commonly factor analysis. For a detailed discussion of assessment of construct validity, the reader is referred to Fitzpatrick et al.3



Reliability Reliability is the reproducibility and internal consistency of an instrument. It assesses the extent to which the instrument is free from random error. It is essential to establish that any changes observed in a trial are due to the treatment and not to problems in the measuring instrument. As the random error increases, the size of the sample required to produce an accurate result increases.



636 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA REPRODUCIBILITY



Reproducibility is the degree to which an instrument gives the same results on repeated applications with the same subjects, and is also known as the test–retest reliability. An instrument should produce the same, or very similar, results on two or more administrations. This should be relatively straightforward to assess, but care must be taken with the time interval between tests. Repeat measurements should be far enough apart in time for the subject to forget their earlier answers, but not so far apart that their health status might have changed. This is commonly reported as a correlation coefficient. However, a correlation coefficient measures the strength of association between two measures and not the extent of agreement.7 Bland and Altman7 recommend plotting the scores from the two tests graphically which is certainly a simpler method than a statistical comparison of repeated scores.



INTERNAL CONSISTENCY



More than one question is usually used to measure one domain in a questionnaire because several related observations will produce a more reliable estimate than one. Individual questions in a domain should correlate highly with each other and with the total score for questions in that domain. This is the internal consistency of a patient-based outcome measure. The correlation is often measured using Cronbach’s alpha.8, 9 If all the questions in a domain are the same, Cronbach’s alpha will be 1, while if there is no relationship alpha will be 0. If the correlation is too high, it is likely that the questions are addressing a very narrow aspect of an attribute and some items may be redundant which then reduces the content validity. It is therefore suggested that Cronbach’s alpha should be between 0.7 and 0.9.10



Responsiveness to change Responsiveness is the ability of an instrument to detect clinically important change, even if that change is small.11 This is sometimes called sensitivity to change but the term sensitivity has other, more general uses. It is particularly important in clinical trials when changes might correspond to therapeutic effects of treatment. An instrument can be both reliable and valid but not responsive to change. There are several statistical methods of assessing responsiveness. The simplest method is to compare change scores for an instrument over time with changes in another variable or variables. The other variable should preferably be an objective indicator, such as a physiological measurement or a clinician-based outcome measure. An alternative method of assessing responsiveness is calculation of the effect size in a given clinical situation. This is the size of change in a measure between



assessments, for example before and after treatment, compared with the variability of scores for that measure on one assessment. The effect size is defined as the mean change in a variable divided by the standard deviation of that variable.12 [**/*] The effect size can then be expressed in standardized units that allow comparisons with other outcome measures. It has been suggested that effect sizes can be used to assess the size of change in a study arm: an effect size of 0.2 is small, 0.5 is medium and 0.8 or greater is a large change.12 [**/*] Other more complex statistical measures of responsiveness are described by Fitzpatrick et al.3 One of the main limitations on the responsiveness of an instrument is when the wording of questions does not allow reporting of very good or very poor health states: ceiling and floor effects. Subjects with initial high scores may not show any improvement following effective treatment and those with initial poor scores may not show any deterioration when their clinical situation deteriorates. As with validity, evidence of responsiveness in one situation does not mean that there will be adequate responsiveness in another research setting.



Precision or sensitivity The preferred term is precision as sensitivity has a number of other uses in research. Precision is the ability of the instrument to reflect true differences in health states. As clinical trials often aim to detect small differences between patient groups, precision is a desirable capability. One of the main influences on the precision of an instrument is the format of the answers. The simplest answers are ‘yes’ or ‘no’, but they do not allow any assessment of difficulty or severity. The most commonly used graded response is a Likert scale, such as: 1. 2. 3. 4. 5.



very satisfied; satisfied; neither satisfied or dissatisfied; dissatisfied; very dissatisfied.



There is some evidence that using seven response categories rather than five increases precision, though this is rarely used. The main alternative to a Likert scale is a visual analogue scale, where patients can mark any point on a line to represent their answer. Though this would appear to offer more precision, comparison studies of Likert scales and visual analogue scales have found no advantage of visual analogue scales. In addition, it appears that visual analogue scales are less acceptable to many patients who find it difficult to translate their feelings into numbers. Patients’ responses in patient-based outcome measures are generally converted into numerical values for



Chapter 50 Outcomes research



statistical analysis. Most instruments use simple ordinal values, for example one to five (or seven) for a Likert scale, which are capable of less precision. However, the majority of published reports of health status measures use parametric statistical analysis that is appropriate for interval scales,13 though the interval between one and two (very satisfied and satisfied) may not be the same as between four and five (dissatisfied and very dissatisfied). Some patient-based outcome measures use an explicitly derived weighting system for responses. The weights can be assigned by a panel of patients and health professionals, for example the Nottingham Health Profile,14 or be based on preference measurements obtained from a random sample of the general population, such as the Health Utilities Index (HUI).15, 16 The fact that weighted scoring systems appear to be much more exact with their scoring suggests that they might be more precise, but this may well be deceptive. Several studies have compared weighted and ordinal scoring systems and have not shown any significant difference in precision between these two methods of scoring. Ceiling and floor effects may influence scores of instruments. Some patient-based outcome measures do not include questions that would identify very poor levels of health, so all patients with poor health have similar scores and further deterioration will not be identified. Sometimes patients with minor health problems will be scored as having excellent health so treatment of their problem will not result in any improvement in score. One other important factor in precision is bias. This can be reduced by general aspects of study design, such as making assessments blind to intervention. In many cases it is not possible to keep the patient blind to which treatment arm of a trial he is in, so his judgement of outcomes may be influenced. Patient-based outcome measures may therefore vary in how precisely their scores relate to underlying distributions of patients’ health status. Researchers need to carefully consider factors that might influence precision. The degree of precision required of a patient-based outcome measure will depend on other aspects of trial design such as sample size and the expected differences between arms of the trial.



Interpretability Can results from the measure be interpreted clinically and are they relevant? The interpretability of an instrument is concerned with how meaningful are the baseline scores or a change in scores. Clinically important changes in scores can be estimated by comparison with clinical or laboratory tests in the same patient group. Representative data are available from the general population for some widely used instruments such as the HUI (see under Utility measures) and the SF-36. The scores from the trial patients can be compared with the means and standard



] 637



deviations for the general population. A significant change in the score in the trial could be set at one and a half standard deviations from the population mean.10



Acceptability Clearly it is essential that the format of the instrument and the questions are acceptable to patients. An instrument should not cause distress to patients or be difficult to understand. In general, shorter instruments are more acceptable to patients. The instructions to patients should make it very clear that their answers will not influence their treatment. Acceptability is also important in order to obtain high response rates to questionnaires to make results of trials easier to interpret, more generalizeable and less prone to bias due to nonresponse. When choosing an instrument, it is useful to know if the instrument has been used in similar settings before. Ideally, acceptability of an instrument should be directly tested at the design stage by seeking the views of patients. Subsequently, evidence of acceptability can be found in patient response rates.



Feasibility The feasibility of an instrument is dependent on the time available for its completion and the staff available to help with its completion, either by interviewing patients or explaining the instrument to patients. The data collected have to be entered onto a computer and the time required and ease of entering has to be considered. The measure must be short enough to be completed or administered in the intended setting and with the types of patients and families involved in the trial. This is one of the criteria of patient-based outcome measures that can usually be easily judged by investigators in the research setting. If in doubt, feasibility can be assessed by piloting the study methods.



TYPES OF PATIENT-BASED OUTCOME MEASURE There are two basic types of patient-based outcome measures: generic and specific. Generic instruments access multidimensional health profiles, overall medical condition and personal function. Specific instruments focus on the problems found in individual diseases, disabilities and patient groups.11 Within each of these categories there are different types of instruments (Table 50.2).



GENERIC INSTRUMENTS Generic instruments are designed to access a broad range of aspects of health status and the consequences of illness



638 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA Table 50.2 Category Generic



Types of patient-based outcome measures. Type



Example



Health profile



SF-36, Sickness Impact Profile, Nottingham Health Profile, Glasgow Benefit Inventory Health Utilities Index (HUI), EQ5D Dizziness Handicap Inventory (DHI) Sino-Nasal Outcome Test (SNOT-20) McGill Pain Questionnaire



Utility Specific



Disease specific Site specific Dimension specific



and therefore to be relevant to a wide range of patient groups and conditions. The advantage of generic instruments is that they can be used for a broad range of health problems and this enables comparisons across different groups of patients with diverse conditions. Because of their broad range of content and general applicability, such instruments have been used to assess the health of samples from the general population. Such data have been used to generate normative values across populations with which other groups of patients with specific health problems can be compared. Since generic instruments are used more often than specific instruments, there are usually more data available regarding their reliability and validity. As generic questionnaires cover a broad range of aspects of health status, many items may be irrelevant to a particular condition. These items result in a wide range of scores which are not relevant to the condition being studied. As there are few questions relevant to a specific condition, the instrument may be insensitive to changes that might occur as a result of treatment for that condition.



Health profile SF-36



The most commonly used health profile instrument is the SF-36.17, 18 [**] The SF-36 is a 36-item, self-completed questionnaire which measures health status in eight dimensions: physical functioning, role limitations due to physical problems, role limitations due to emotional problems, social functioning, mental health, energy and vitality, pain and general perceptions of health. It can be completed by the patient in less than ten minutes. Responses are summed to give a score for each dimension: physical component summary (PCS) and mental component summary (MCS).19 It has been used in a wide variety of patient groups and conditions, including many otorhinolaryngological conditions (Table 50.3). Data from trial subjects can be compared with normative data for the population.28 Garratt et al.29 identified 408 papers



Table 50.3



Examples of the use of the SF-36 in otolaryngology.



ORL condition Chronic rhinosinusitis Chronic otitis media Dysphonia Laryngeal cancer Meniere’s disease Rhinoplasty and otoplasty



Reference van Agthoven et al.,20 Gliklich and Metson,21 Piccirillo et al.22 [***/**] Nadol et al.23 [**] MacKenzie et al.24 [****] Stewart et al.25 [**] Smith and Pyle26 [**] Klassen et al.27 [**]



which included aspects of development and evaluation of the SF-36 over a ten-year period (1990–1999). It can also be used in a reduced 12-item version which is a subset of the original SF-36.30 [***] However, the larger instrument gives more reliable estimates of individual levels of health and is therefore the better choice of instrument in small studies. Recently, Brazier et al.31 derived from the SF-36 a preference-based single index measure, the SF-6D, to make the instrument more useful in evaluations of costeffectiveness (see Utility measures below and in Table 50.3). This weighted scoring model could potentially be applied to any SF-36 data set. However, initial validation studies demonstrated some inconsistencies and further assessment of the scoring model is required.32 [**]



SICKNESS IMPACT PROFILE



The Sickness Impact Profile (SIP) is a general healthstatus questionnaire comprising 136 questions answered as either ‘yes’ or ‘no’.33 [***] They are grouped into twelve categories: walking, body care and movement, mobility, social interaction, alertness behaviour, emotional behaviour, communication, sleep and rest, eating, home management, recreational activities and work. Each item is weighted, and the scores of all answered questions are combined. There are twelve category scores, two summary scores (physical and psychosocial) and a total score. The scores are standardized and range from 0 to 100 points, with 100 indicating the poorest function. The test–retest reliability and internal consistency are high and it has good content and construct validity.33 [***] It is, however, a long questionnaire and is usually administered by interview. Its acceptability and feasibility in a trial must be carefully considered.



NOTTINGHAM HEALTH PROFILE



The Nottingham Health Profile is a generic, selfadministered questionnaire designed to measure perceived physical, social and emotional health problems.14, 34 An initial pool of statements was collected from patients at interview and from this pool 38 items were chosen



Chapter 50 Outcomes research



relating to six dimensions: physical mobility, pain, social isolation, emotional reaction, energy and sleep. The scores can be compared with the average scores in a population matched for gender and age.



GLASGOW BENEFIT INVENTORY



As other generic instruments are often insensitive to the nonacute disorders generally seen in an ORL clinic, the Glasgow Benefit Inventory (GBI) was developed for use in patients with otolaryngological conditions.35 [**] The GBI is a post-intervention questionnaire that assesses the effects of interventions on the health status of patient, rather than the actual health status. The GBI has 18 items in three domains: psychological, social and physical wellbeing. Rather than attempt to assess the difference between before and after treatment measures, it asks directly about the change in health status resulting from treatment. It is likely, therefore, to be more sensitive to such change than two separate instruments before and after intervention. In addition, compliance will be significantly higher because patients are only required to complete one questionnaire. The response to each question is based on a five-point Likert scale, for example: ‘Since your operation/intervention, have you found it easier or harder to deal with company?’ The words in italics (operation/intervention) in each question are replaced by words appropriate to the intervention of interest. Responses are scored using a weighted scale from a population sample to give a score between –100 and 1100. The GBI has been used to assess the benefit experienced by patients following various treatments (Table 50.4). Details of the questionnaire and its use can be found at www.ihr.gla.ac.uk.



Utility measures Multi-attribute utility measures access a broad range of aspects of health status, like other generic instruments, but have a particular form of numerical weighting or valuation of health states. Utility measures have been developed from economics and decision theory in order to provide an estimate of individual patients’ overall Table 50.4



Examples of the use of the GBI.



ORL intervention Bone-anchored hearing aids Vestibular schwannoma surgery Speech therapy for dysphonia Rhinoplasty Tonsillectomy Surgery for snoring



] 639



preferences for different health states. They are scored as a single number between 1 (full health) and 0 (death). The weighted scoring method is based on preference measurements obtained from samples of a general population. They are asked to value different aspects of health as defined by the instrument using one of a number of valuation techniques, most commonly standard gamble, time trade off and visual analogue scales (see Brazier et al.,42 for explanation). Utility scores reflect the health status and value of that health status to the patient. As utility measures are scored as a single value, they do not define individual dimensions of health that contribute to the individual’s overall sense of well-being. They are relatively insensitive to small but relevant changes in health and are therefore more suited to studies of large populations. There are several health utility measures that have been widely used. Brazier et al.42 carried out an extensive review of these and recommended the HUI and the EuroQol (EQ-5D) as the instruments of choice. HUI



The HUI-I evolved from studies in neonatal intensive care.43 [**] A second version (HUI-II) was developed to assess outcomes in long-term survivors of childhood cancer, but is suitable for use in a wider range of children.15 Though it is claimed to be a generic instrument, its content is explicitly aimed at children with questions specifically aimed at developmental age. The HUI-II was revised to make it more relevant to an adult population. The HUI-III assesses nine aspects of health status: vision, hearing, speech, ambulation, dexterity, emotion, cognition, self-care and pain.44 It has 15 questions with four to six available responses for each and is easy for patients to complete. The HUI-III has been used in a wide variety of clinical studies and has been shown to be reliable and responsive.45 It is one of the few generic patient-based outcome measures which specifically assesses hearing and speech and is therefore potentially useful in many areas of otolaryngology. Normative data have been collected from large populations and, from these data, weighted scoring scales have been devised.16 It has been used in a cost–utility study of cochlear implants46 and in the UK cochlear implant study.47 [***] EQ-5D



Reference 36



Dutt et al. Fahy et al. 200237 [**] Wilson et al.38 [**] McKiernan et al.39 [**] Bhattacharyya et al.40 [**] Banerjee and Dempster41 [***/**]



The EQ-5D is a self-completed questionnaire with five dimensions: mobility, self-care, usual activities, pain/discomfort and anxiety/depression.48 It is easy to complete and is very acceptable to patients. Responses can be scored using a weighted scale from a large general population sample.49 It is very brief with only five questions with three levels of response for each. Brazier et al.42 compared the EQ-5D and the SF-36 in a survey of 1980 adults and found the EQ-5D to be less sensitive.



640 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA However, it was originally intended to complement other forms of health-related quality of life measures. It has been widely used and its test–retest reliability is good.50



SPECIFIC INSTRUMENTS



they experienced with vertigo.52 [**] It has 25 questions about restriction of physical and social activities and emotional distress, each scored on a five-point Likert scale. It has good reliability and validity but has not been widely used. HEARING



Disease- or condition-specific The aim of these instruments is to provide the patient’s perception of the problems related to a specific disease or condition. It should be remembered that disease may have a broad impact on the patient’s life. To make these instruments comprehensive, a detailed survey of patients suffering from the condition should be conducted when developing the instrument. They may also not detect problems associated with a disease and its treatment that have not been anticipated by the developers if patients have not been involved in the planning. As all of the questions are developed specifically to assess a particular health problem, the content should be very relevant for use in studies of that condition: high validity. They are more likely to detect changes that occur in that particular condition as there should be few if any irrelevant questions: high responsiveness. They are also more acceptable to patients as the relevance of the questions to their condition is obvious so completion rates should be high. The major disadvantage of disease-specific instruments is that they do not allow comparison to be made with other patient groups with other diseases or conditions. Such comparisons require generic instruments designed for use with any health problem, and for this reason it is often useful to combine a specific instrument with a generic one. VERTIGO



The Dizziness Handicap Inventory (DHI) is a 25-item questionnaire designed to measure the self-perceived disability and handicap caused by dizziness or imbalance.51 [**] They reported good internal consistency and test–retest repeatability. The questions were grouped in three domains of functional, emotional and physical aspects of dizziness with three possible answers to each question: yes, sometimes and no. The questions were selected from an initial bank of 37 questions chosen from the case histories of patients. They take a clinical perspective of disability and handicap in daily activities and therefore content validity should be carefully inspected before choosing this instrument for use in a trial. It is preferable to have more input from patients in the selection of items to include in a questionnaire (see Face and content validity above). This instrument has been widely used in studies of imbalance in general. The Vertigo Handicap Questionnaire (VHQ) was developed from patients’ accounts of the problems that



There are many instruments that have been developed for use in audiology. Assessing hearing disability has one major difference from assessing most other disabilities in otolaryngology – there is a gold standard – pure-tone audiometry. One of the most widely used measures is the Hearing Handicap Inventory for the Elderly (HHIE).53 [**] This instrument has been shown to have good face and content validity.54 [***] While there are reliable tools for measuring hearing, assessment of the efficacy of hearing aid provision and comparison of benefits from different prescription strategies are heavily reliant on patient-based outcome measures. The HHIE has been used for this purpose,55 but it has been shown to have poor precision when changes in the HHIE were compared with changes in the Speech Intelligibility Index when using a hearing aid.56 [**] This is probably because the HHIE predominantly assesses the emotional and psychological response to hearing impairment. The Glasgow Hearing Aid Benefit Profile (GHABP) was developed as a measure of hearing disability and the benefit obtained from the use of a hearing aid.57 It assesses unaided disability, handicap, benefit from a hearing aid, residual disability and patient satisfaction in eight listening situations – four specified and four chosen by the patient. It has good validity in assessing the outcome of hearing aid fitting, high test–retest repeatability and good patient acceptability. It is now used as one of the primary outcome measures in the National Health Service (NHS) programme of modernizing hearing aid services in England. Details of the questionnaire and its use can be found at www.ihr.gla.ac.uk. OTITIS MEDIA



The OM6 was developed for use in children with recurrent acute otitis media and otitis media with effusion.58 [**] Six domains (physical suffering, hearing loss, speech impairment, emotional distress, activity limitation, caregiver concerns) are each addressed by a single question. Test–retest repeatability was high, and responsiveness was demonstrated by a significant change in scores after surgery for ventilation tube insertion.59 [**] The criterion validity has been questioned, however, in that it does not correlate well with other markers of disease severity, such as audiometry and severity of symptoms of recurrent acute otitis media.60 Kubba et al.60 suggested that the instrument lacks precision as it does not differentiate between children with otitis media and others with sore throats.



Chapter 50 Outcomes research



CANCER



The European Organization for Research and Treatment of Cancer Quality of Life Study Group developed the EORTC QLQ-C30 as an instrument of 30 items for use in international trials in cancer.61 [***] Further modules have been developed which can be added to the core instrument to provide assessment of specific cancers, for example head and neck cancer.62 [**] This allows collection of data for comparison across cancer groups and additional data that are particularly relevant to specific cancers.



] 641



handicap. The questions use a five-point Likert scale. It has been used to assess the impact of a number of voice disorders, including vocal cord polyps, cord palsy, spasmodic dysphonia and functional disorders. Its construct validity has been confirmed by comparing its subscales with appropriate subscales of the SF-36 in assessment of the health-related quality of life in patients after treatment for laryngeal cancer.25 [**] Significant improvements in VHI scores occurred after treatment of spasmodic dysphonia with botulinum toxin.65 [**]



Dimension-specific Site-specific These instruments contain items that are particularly relevant to patients having treatment for a specific region of the body and should be sensitive to changes experienced by patients following treatment in that region. They are particularly useful in otolaryngology and have the advantage that patient groups are not limited to a specific disease classification. They thus allow comparison of patients with similar symptoms but different pathology. Because they have a narrow focus, they are unlikely to detect any change in broader aspects of health or quality of life following intervention. SINONASAL DISEASE



Many patient-based outcome measures have been designed for use in sinonasal disease. The 20-item Sino-Nasal Outcome Test (SNOT-20)63 is a modification of the previously used 31-item Rhinosinusitis Outcome Measure (RSOM-31).22 [**] The 20 questions refer to specific sinonasal symptoms and some general health questions. It has good internal consistency (Cronbach’s alpha 0.9), and showed good responsiveness to change. Face and content validity seemed good and the construct validity was high when compared with clinical assessment of disease. Test–retest scores were highly correlated (r = 0.9) though this measures the association between tests rather than repeatability (see Reproducibility above). Piccirillo et al.63 report the SNOT-20 to be a valid outcome measure in their particular research setting. The questionnaire is easily completed by the patient and acceptability is high. [**] VOICE



There is no generally accepted objective test to serve as a ‘gold standard’ for the assessment of voice disorders. Various questionnaires have been developed for evaluation of the consequences of dysphonia. The Voice Handicap Index (VHI) was designed to assess the selfperceived effect on quality of life of voice disorders.64 It is a 30-item questionnaire divided into three subscales: functional handicap, emotional handicap and physical



These instruments assess one specific aspect of health status. The most commonly assessed dimension is psychological well-being. Another common dimension, more relevant to otolaryngology, is pain. The McGill Pain Questionnaire is one of the most widely used measures of pain severity for both clinical and research purposes.66 It comprises 20 subclasses of pain descriptors that provide pain severity scores across sensory–discriminative, motivational–affective and cognitive–evaluative dimensions. It has been found to have good short-term repeatability in chronic pain conditions, and it discriminates well between different types of pain syndromes. A short form of the McGill Pain Questionnaire has been developed.67 [***/**] It is highly correlated with the original longer version and has comparable sensitivity. It may be a useful instrument in situations in which the standard questionnaire takes too long to administer. The advantage of dimension-specific instruments is that they provide a more detailed assessment in the area of concern than is possible in more general instruments. They are perhaps most useful in a study where they are used in combination with another instrument.20 [***]



HOW TO CHOOSE PATIENT-BASED OUTCOME MEASURE The choice of patient-based outcome measure depends on the purpose of the study. To assess the effects on healthrelated quality of life of a disease or condition, a generic instrument is required. Specific instruments are of little value as they do not allow comparison with other conditions. A specific instrument may provide additional information by assessing the severity of the disease if no objective measure is available, for example in tinnitus. Comparison of the scores from the specific instrument with the generic instrument may be valuable. If the aim of the study is to assess the efficacy of treatment, it is usually recommended that two instruments be combined, a generic and a specific instrument. The two different measures are likely to produce complementary evidence. A disease-specific measure will be more responsive to the main effects of intervention,



642 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA and therefore produce the evidence most relevant to the clinician. A generic measure may allow comparisons across interventions and disease groups, but is likely to be relatively insensitive to the effects of the intervention. It is possible, however, that the additional burden on the patient may reduce overall compliance, especially if there is overlap between questions in the two instruments.



Deficiencies in current knowledge and areas for future research Development of a valid and acceptable patient-based outcome measure takes a great deal of time and effort. Large numbers of subjects are needed as development requires several stages. Patient involvement in the selection of items is essential. There have been many patient-based outcome measures that have been reported once and never used again. On the other hand, the currently available and widely used patient-based outcome measures are not ideal. We need better ones.



WHICH GENERIC INSTRUMENT? Acceptability to the patient and feasibility of completing the questionnaires are obviously important. The time required to complete the commonly used instruments varies. The HUI Mark II takes three minutes, the SF-36 ten minutes and the SIP 20 minutes.68 [**] Appropriateness depends on the study. Utility measures are, in general, only appropriate for studies with large numbers of subjects. Small studies would probably be better to choose a generic health status questionnaire. Some generic instruments have been shown to be better than others in assessing patient with particular problems. Edelman et al.68 reported that the HUI was better used for evaluating relatively healthy populations because of some floor effects, while the SIP was better for more severely ill populations as their study patients were grouped at the healthy end of the scale. Other authors have similarly reported that the SIP has ceiling effects.69, 70, 71 [***/**] A careful evaluation of reports of the use of these instruments in similar patient groups and study settings should be carried out before making a final choice. In the absence of this, the SF-36 appears to be the safest choice. The HUI Mark III is suitable for large studies and has the advantage for otolaryngology research of including items about hearing and speech.



REFERENCES











KEY POINTS  Patient-based outcome measures assess the patient’s subjective experience of illness and of treatment.  Generic instruments access a broad range of aspects of health status and enable comparisons across different groups of patients with diverse conditions.  Specific instruments are more sensitive to the effects of specific diseases or conditions, but do not allow comparison to be made with other patient groups with other diseases or conditions.  It is often useful to combine a specific instrument with a generic instrument.







1. van den Bos GAM, Limburg LCM. Public health and chronic diseases. European Journal of Public Health. 1995; 5: 1–2. 2. Fayers PM, Hopwood P, Harvey A, Girling DJ, Machin D, Stephens R. Quality of life assessment in clinical trials–guidelines and a checklist for protocol writers: the U.K. Medical Research Council experience. MRC Cancer Trials Office. European Journal of Cancer. 1997; 33: 20–8. Good advice from authors with experience of reviewing grant applications submitted to the MRC. 3. Fitzpatrick R, Davey C, Buxton MJ, Jones DR. Evaluating patient-based outcome measures for use in clinical trials. Health Technology Assessment (Winchester, England). 1998; 2: i–iv, 1-74. Extensive review of the literature to describe the range of patient-based outcome measures available and the criteria for selecting an instrument for use in a trial. 4. Guyatt GH, Cook DJ. Health status, quality of life, and the individual. Journal of the American Medical Association. 1994; 272: 630–1. 5. Jenkinson C. Evaluating the efficacy of medical treatment: possibilities and limitations. Social Science and Medicine. 1995; 41: 1395–401. 6. Guyatt GH, Feeny DH, Patrick DL. Measuring healthrelated quality of life. Annals of Internal Medicine. 1993; 118: 622–9. 7. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986; 1: 307–10. 8. Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika. 1951; 16: 297–334. 9. Bland JM, Altman DG. Cronbach’s alpha. British Medical Journal. 1997; 314: 572. 10. Streiner DL, Norman GR. Health Measurement Scales: a practical guide to their development and use. Oxford: Oxford University Press, 1995. An invaluable practical guide for those who wish to develop a new patient-based outcome measure. Very good and readable explanation of theory behind these instruments.



Chapter 50 Outcomes research 11. Guyatt GH, Veldhuyzen van Zanten SJO, Feeny DH, Patrick DL. Measuring quality of life in clinical trials: a taxonomy and review. Canadian Medical Association Journal. 1989; 140: 1441–8. 12. Kazis LE, Anderson JJ, Meenan RF. Effect sizes for interpreting changes in health status. Medical Care. 1989; 27: S178–89. 13. Coste J, Fermanian J, Venot A. Methodological and statistical problems in the construction of composite measurement scales: a survey of six medical and epidemiological journals. Statistics in Medicine. 1995; 14: 331–45. 14. Hunt SM, McEwen J, McKenna SP. Measuring health status: a new tool for clinicians and epidemiologists. Journal of the Royal College of General Practitioners. 1985; 35: 185–8. 15. Feeny D, Furlong W, Boyle M, Torrance GW. Multiattribute health status classification systems. Health Utilities Index. PharmacoEconomics. 1995; 7: 490–502. 16. Feeny D, Furlong W, Torrance GW, Goldsmith CH, Zhu Z, DePauw S et al. Multiattribute and single-attribute utility functions for the health utilities index mark 3 system. Medical Care. 2002; 40: 113–28. 17. Ware Jr. JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Medical Care. 1992; 30: 473–83. 18. McHorney CA, Ware Jr. JE, Raczek AE. The MOS 36-Item Short-Form Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Medical Care. 1993; 31: 247–63. 19. Ware JE. SF-36 Physical and mental health summary scales: a user’s manual. Boston: Health Assessment Lab. New England Medical Center, 1994. 20. van Agthoven M, Fokkens WJ, van de Merwe JP, Marijke van Bolhuis E, Uyl-de Groot CA, Busschbach JJ. Quality of life of patients with refractory chronic rhinosinusitis: effects of filgrastim treatment. American Journal of Rhinology. 2001; 15: 231–7. 21. Gliklich RE, Metson R. Techniques for outcomes research in chronic sinusitis. Laryngoscope. 1995; 105: 387–90. 22. Piccirillo JF, Edwards D, Haiduk A, Yonan C, Thawley SE. Psychometric and clinimetric validity of the 31-item Rhinosinusitis Outcome Measure (RSOM-31). American Journal of Rhinology. 1995; 9: 297–306. 23. Nadol Jr. JB, Staecker H, Gliklich RE. Outcomes assessment for chronic otitis media: the Chronic Ear Survey. Laryngoscope. 2000; 110: 32–5. 24. MacKenzie K, Millar A, Wilson JA, Sellars C, Deary IJ. Is voice therapy an effective treatment for dysphonia? A randomised controlled trial. British Medical Journal. 2001; 323: 658–61. 25. Stewart MG, Chen AY, Stach CB. Outcomes analysis of voice and quality of life in patients with laryngeal cancer. Archives of Otolaryngology – Head and Neck Surgery. 1998; 124: 143–8. 26. Smith DR, Pyle GM. Outcome-based assessment of endolymphatic sac surgery for Meniere’s disease. Laryngoscope. 1997; 107: 1210–6.



] 643



27. Klassen A, Jenkinson C, Fitzpatrick R, Goodacre T. Patients’ health related quality of life before and after aesthetic surgery. British Journal of Plastic Surgery. 1996; 49: 433–8. 28. Jenkinson C, Coulter A, Wright L. Short form 36 (SF36) health survey questionnaire: normative data for adults of working age. British Medical Journal. 1993; 306: 1437–40. 29. Garratt A, Schmidt L, Mackintosh A, Fitzpatrick R. Quality of life measurement: bibliographic study of patient assessed health outcome measures. British Medical Journal. 2002; 324: 1417. 30. Ware Jr. J, Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Medical Care. 1996; 34: 220–33. 31. Brazier J, Usherwood T, Harper R, Thomas K. Deriving a preference-based single index from the UK SF-36 Health Survey. Journal of Clinical Epidemiology. 1998; 51: 1115–28. 32. Brazier J, Roberts J, Deverill M. The estimation of a preference-based measure of health from the SF-36. Journal of Health Economics. 2002; 21: 271–92. 33. Bergner M, Bobbitt RA, Carter WB, Gilson BS. The Sickness Impact Profile: development and final revision of a health status measure. Medical Care. 1981; 19: 787–805. 34. Hunt SM, McKenna SP, McEwen J, Backett EM, Williams J, Papp E. A quantitative approach to perceived health status: a validation study. Journal of Epidemiology and Community Health. 1980; 34: 281–6. 35. Robinson K, Gatehouse S, Browning GG. Measuring patient benefit from otorhinolaryngological surgery and therapy. Annals of Otology, Rhinology and Laryngology. 1996; 105: 415–22. 36. Dutt SN, McDermott AL, Jelbert A, Reid AP, Proops DW. The Glasgow benefit inventory in the evaluation of patient satisfaction with the bone-anchored hearing aid: quality of life issues. Journal of Laryngology and Otology – Supplement. 2002; 28: 7–14. 37. Fahy C, Nikolopoulos TP, O’Donoghue GM. Acoustic neuroma surgery and tinnitus. European Archives of Oto-Rhino-Laryngology. 2002; 259: 299–301. 38. Wilson JA, Deary IJ, Millar A, Mackenzie K. The quality of life impact of dysphonia. Clinical Otolaryngology and Allied Sciences. 2002; 27: 179–82. 39. McKiernan DC, Banfield G, Kumar R, Hinton AE. Patient benefit from functional and cosmetic rhinoplasty. Clinical Otolaryngology and Allied Sciences. 2001; 26: 50–2. 40. Bhattacharyya N, Kepnes LJ, Shapiro J. Efficacy and quality-of-life impact of adult tonsillectomy. Archives of Otolaryngology – Head and Neck Surgery. 2001; 127: 1347–50. 41. Banerjee A, Dempster JH. Laser palatoplasty: evaluation of patient benefit using the Glasgow benefit inventory. Journal of Laryngology and Otology. 2000; 114: 601–4. 42. Brazier J, Deverill M, Green C, Harper R, Booth A. A review of the use of health status measures in economic evaluation. Health Technology Assessment (Winchester, England). 1999; 3: i–iv, 1-164.



644 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA 43. Torrance GW, Boyle MH, Horwood SP. Application of multi-attribute utility theory to measure social preferences for health states. Operations Research. 1982; 30: 1043–69. 44. Torrance GW, Furlong W, Feeny D, Boyle M. Multiattribute preference functions. Health Utilities Index. Pharmacoeconomics. 1995; 7: 503–20. 45. Furlong WJ, Feeny DH, Torrance GW, Barr RD. The Health Utilities Index (HUI) system for assessing health-related quality of life in clinical studies. Annals of Medicine. 2001; 33: 375–84. 46. Palmer CS, Niparko JK, Wyatt JR, Rothman M, de Lissovoy G. A prospective study of the cost-utility of the multichannel cochlear implant. Archives of Otolaryngology – Head and Neck Surgery. 1999; 125: 1221–8. 47. Barton GR, Summerfield AQ, Marshall DH, Bloor KE. On behalf of the POCIA Collaboration. Choice of instrument for measuring the gain in utility from cochlear implantation. Oxford: Health Economics Study Group, 2001. 48. EuroQol Group. EuroQol – a new facility for the measurement of health-related quality of life. The EuroQol Group. Health Policy. 1990; 16: 199–208. 49. MVH Group. The measurement and valuation of health: final report on the modelling of valuation tariffs. York: Centre for Health Economics, University of York, 1995. 50. Brooks R. EuroQol: the current state of play. Health Policy. 1996; 37: 53–72. 51. Jacobson GP, Newman CW. The development of the Dizziness Handicap Inventory. Archives of Otolaryngology – Head and Neck Surgery. 1990; 116: 424–7. 52. Yardley L, Putman J. Quantitative analysis of factors contributing to handicap and distress in vertiginous patients: a questionnaire study. Clinical Otolaryngology and Allied Sciences. 1992; 17: 231–6. 53. Ventry IM, Weinstein BE. The hearing handicap inventory for the elderly: a new tool. Ear and Hearing. 1982; 3: 128–34. 54. Weinstein BE, Spitzer JB, Ventry IM. Test-retest reliability of the Hearing Handicap Inventory for the Elderly. Ear and Hearing. 1986; 7: 295–9. 55. Newman CW, Weinstein BE. The Hearing Handicap Inventory for the Elderly as a measure of hearing aid benefit. Ear and Hearing. 1988; 9: 81–5. 56. Gatehouse S. Outcome measures for the evaluation of adult hearing aid fittings and services. Scientific and Technical Report to the Department of Health. MRC Institute of Hearing Research, Glasgow, 1997. 57. Gatehouse S. Glasgow Hearing Aid Benefit Profile: derivation and validation of a client-centred outcome measure for hearing aid services. Journal of American Academy of Audiology. 1999; 10: 80–103. 58. Rosenfeld RM, Goldsmith AJ, Tetlus L, Balzano A. Quality of life for children with otitis media. Archives of Otolaryngology – Head and Neck Surgery. 1997; 123: 1049–54. 59. Rosenfeld RM, Bhaya MH, Bower CM, Brookhouser PE, Casselbrant ML, Chan KH et al. Impact of tympanostomy



60.



61.



62.



63.



64.



65.



66.



67. 68.



69.



 70.



 71.



tubes on child quality of life. Archives of Otolaryngology – Head and Neck Surgery. 2000; 126: 585–92. Kubba H, Swan IRC, Gatehouse S. How appropriate is the OM6 as a discriminative instrument in children with otitis media? Archives of Otolaryngology – Head and Neck Surgery. 2004; 130: 705–9. Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, Duez NJ et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. Journal of the National Cancer Institute. 1993; 85: 365–76. Bjordal K, Ahlner-Elmqvist M, Tollesson E, Jensen AB, Razavi D, Maher EJ et al. Development of a European Organization for Research and Treatment of Cancer (EORTC) questionnaire module to be used in quality of life assessments in head and neck cancer patients. EORTC Quality of Life Study Group. Acta Oncologica. 1994; 33: 879–85. Piccirillo JF, Merritt Jr. MG, Richards ML. Psychometric and clinimetric validity of the 20-Item Sino-Nasal Outcome Test (SNOT-20). Otolaryngology – Head and Neck Surgery. 2002; 126: 41–7. Jacobson BH, Johnson A, Grywalski C, Silbergleit A, Jacobson G, Benninger MS. The Voice Handicap Index (VHI): development and validation. American Journal of Speech and Language Pathology. 1997; 6: 66–70. Rosen CA, Murry T, Zinn A, Zullo T, Sonbolian M. Voice handicap index change following treatment of voice disorders. Journal of Voice. 2000; 14: 619–23. Melzack R. The McGill Pain Questionnaire: major properties and scoring methods. Pain. 1975; 1: 277–99. Melzack R. The short-form McGill Pain Questionnaire. Pain. 1987; 30: 191–7. Edelman D, Williams GR, Rothman M, Samsa GP. A comparison of three health status measures in primary care outpatients. Journal of General Internal Medicine. 1999; 14: 759–62. Andresen EM, Rothenberg BM, Panzer R, Katz P, McDermott MP. Selecting a generic measure of healthrelated quality of life for use among older adults. A comparison of candidate instruments. Evaluation and the Health Professions. 1998; 21: 244–64. Coons SJ, Rao S, Keininger DL, Hays RD. A comparative review of generic quality-of-life instruments. Pharmacoeconomics. 2000; 17: 13–35. Good reviews of the three most widely used generic instruments: the SF36, the Nottingham Health Profile and the Sickness Impact Profile. De Korte J, Mombers FM, Sprangers MA, Bos JD. The suitability of quality-of-life questionnaires for psoriasis research: a systematic literature review. Archives of Dermatology. 2002; 138: 1221–7. Good reviews of the three most widely used generic instruments: the SF-36, the Nottingham Health Profile and the Sickness Impact Profile.



51 Evidence-based medicine MARTIN J BURTON



Introduction Definition of EBM The application of EBM



645 645 647



INTRODUCTION Doctors want to give their patients the best treatment they can, i.e. the treatment that is most likely to help the individual patient in front of them. The idea behind evidence-based medicine (EBM) is that healthcare practitioners should only use (prescribe, recommend, implement) therapies that have been proven to be effective, and should not use unproven therapies that may be ineffective at best or, at worst, actually harmful. By ‘harmful’ we do not necessarily mean ‘fatal’ or even imply that the harm is serious. Any outcome which disadvantages the patient – their health, lifestyle, even their wallet or purse – might be considered as a harm, or, if not strictly a harm, a factor to be balanced against any benefit which is accrued as a result of the treatment. The reader may ask – ‘so what’s new?’ Surely the approach described has been at the heart of medical treatment for centuries. It has been suggested that rather than thinking of the pre-EBM era as being unscientific, we should refer to the present as a time in which we make ‘better use of evidence in medicine’. The paternalistic and authoritarian approach that characterized much of medical practice in the past should be replaced by one which acknowledges uncertainty more explicitly, and which operates within a defined framework and with a set of rules for the weighing of scientific evidence. In 1995, Davidoff et al.1 listed five linked ideas in which EBM is rooted. 1. Clinical decisions should be based on the best available scientific evidence; 2. The clinical problem – rather than habits or protocols – should determine the type of evidence to be sought;



Conclusion Key points References



648 648 648



3. Identifying the best evidence means using epidemiological and biostatistical ways of thinking; 4. Conclusions derived from identifying and critically appraising evidence are useful only if put into action in managing patients or making healthcare decisions; 5. Performance should be constantly evaluated.



DEFINITION OF EBM The most familiar definition of EBM was coined by Sackett and his colleagues in 1996: ‘Evidence-based medicine is the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients’.2 This concise statement contains many, but not all, the important features of EBM. Before unpacking this description on EBM, it is worth recalling that ‘EBMy can be seen as a tool, and tools can be used for good and for ill’.3 It is important that the user knows how to use the tool correctly and does not use it unwisely or inappropriately.



Conscientious The conscientious otolaryngologist wants to ‘do the right thing’ for his patient – to do more good than harm. He should also want to do this in the right way. In other words, to ‘do things right’ – do them better, more efficiently, etc. Muir Gray refers to this practice as ‘doing the right things right’.4 The term ‘conscientious’ also brings to mind a certain carefulness and attention to



646 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA detail. These characteristics are clearly a requisite for high quality surgery and many practitioners hone and practice their surgical skills in order to improve them. The same diligence with which they do this, the same desire to master their craft, should be directed to acquiring the skills of critical appraisal and data synthesis necessary to practise EBM successfully.



Explicit Openness and transparency are features of the modern health service in the UK. They are also key elements of rigorous science – prerequisites for allowing experimental work to be evaluated and replicated. One of the principles underlying evidence-based practice is that the methods used are clearly stated and, whenever possible, are stated a priori. This explicitness allows others to evaluate and criticize the methods used and, when appropriate, to make the necessary changes to allow the quality of the work to be improved. Evidence-based practice explicitly evaluates the strength of the evidence underlying a particular intervention or recommendation. By critically appraising a study, specific questions can be answered about the validity of the evidence it contains. How good is it? Can we really believe the results? How certain (or, more usually, uncertain) are we about the results? This notion of explicitness extends to consideration of issues of uncertainty. In the past, otolaryngologists, like many other physicians, may have found being open and explicit about the uncertainty which surrounds much medical practice, difficult. They have found sharing this with their patients even more so. As new concepts of doctor–patient partnerships (in distinction to doctor– patient relationships) have developed, this approach has become untenable. In the modern health service ‘doctors have the difficult task of explaining the nature of uncertainty and risk in the practice of today’s scientific medicine’.5



Judicious The use of the term ‘judicious’ reminds us that clinical practice must be rooted in considered opinions and sensible decisions. In other words, the evidence has been evaluated and ‘weighed’, benefits being considered and balanced against harms. Common sense has not been ignored. It is a truism to say ‘evidence is necessary but not sufficient’. It needs to be integrated with patient’s preferences (of which more below), economic constraints, organizational factors relating to the delivering healthcare organization and ethical obligations. All this is also to be considered in the context of social responsibility and resource allocation prioritization.



Current best evidence It may once have been possible for the conscientious otolaryngologist to keep their knowledge comprehensive and up-to-date by reading a few journals every month and going to occasional conferences. In the twenty-first century it is impossible. There are now in excess of two million articles per year published in the biomedical press. The internet has resulted in an explosion of knowledge sources which must be acknowledged, even if the quality of their contents is at times uncertain. Whereas once this knowledge was available only to the select few, now it is available to everyone – doctor, patient, carer and healthcare provider. Whilst this ‘sea change in the availability of knowledge offers a tremendous opportunity for the health professions to engage the public in taking an interest in and responsibility for their own health y for some medical practitioners, and indeed for some members of other health professions, this democracy in information access can pose troubling questions. How does one negotiate differences in the interpretation of medical evidence between doctor and patients? How does one signal which evidence is reliable and which is not? How does one offer advice in the face of lack of evidence?’5 The practitioner of EBM needs to be able to locate the evidence relating to the clinical question or questions posed by the patient in front of him, and then assess its quality to determine if it is the ‘best’ – and most appropriate – available. In Chapter 52, Critical appraisal skills, Martin Dawes discusses the concept of levels of evidence; suffice it to say at this stage that practitioners need to be aware that just because something is written in a distinguished journal, it does not mean that this represents absolute scientific truth. In the same way that EBM challenges ‘eminence based medicine’ (the concept that age, authority and seniority per se bring with them an inevitable certainty of valid clinical truth), it also challenges a presumption that published studies are always well designed, well conducted and well reported. Concepts of publishing have changed in recent years, but paper-based journals and textbooks continue to be produced, and the best ones can form a starting point for the clinician’s enquiries. However, it is likely that greater use will be made of electronic resources and in particular electronic databases such as MEDLINE and EMBASE. Assistance in identifying and searching appropriate databases is often available from hospital librarians and information specialists. There are a number of electronic resources which house material that has already been assembled, appraised and included because of its methodological quality and can provide a ‘short cut’ to ‘best evidence’. These include the Cochrane Central Register of Controlled Trials (CENTRAL), part of The Cochrane Library, the ENT and Audiology Library of the National Library for Health and Clinical Evidence, published by the BMJ publishing group (Table 51.1).



Chapter 51 Evidence-based medicine Table 51.1



Evidence-based resources.



Resource General Centre for Evidence-Based Medicine, Oxford The Cochrane Collaboration PubMed (MEDLINE provided by the US National Library of Medicine) Clinical Evidence (BMJ Group publication) Evidence-Based On-Call Otolaryngology ENT and Audiology Library of National Library for Health Cochrane ENT Disorders Group



Web address



www.cebm.net/ www.cochrane.org www.ncbi.nlm.nih.gov/ entrez www.clinicalevidence.com www.eboncall.org/ www.library.nhs.uk/ent www.cochrane-ent.org



Making decisions about the care of individual patients The original definition of EBM focuses on individual patients, yet EBM is often criticized because the evidence that its practitioners use is derived from groups or populations. The opponents of EBM constantly emphasize their interest in the individual in front of them. The questions they should be asking are these: is my patient so different from those patients who have been evaluated in well-conducted scientific studies that the valid results demonstrating that the treatment is likely to be effective do not apply to them? Or, when studies do show a treatment to be effective, can I be certain that my patient is so different that I would reasonably expect them to respond positively, and without any harms of that treatment outweighing the benefits? Sackett’s definition might be criticized for the use of the phrase ‘making decisions’. This suggests that the doctor is making the decision on behalf of the patient. It might be considered more appropriate for the patient to make the decision with the help and assistance of the doctor. Equally, the definition says nothing of the integration of the management options with the patient’s values and beliefs. These must always play an important part in the decision-making process. The following modified definition of EBM is proposed: Evidence-based medicine is the conscientious, explicit and judicious use of current best evidence in helping individual patients to make healthcare decisions that respect their personal values and beliefs.



THE APPLICATION OF EBM Since its original description and definition, the role of EBM has been extended from the individual patient to



] 647



groups and populations of patients. It has become an integral part of modern healthcare practice. However, it is not without its critics and a number of particular issues merit consideration. EBM has its limitations and these must be recognized. Unfortunately, over-zealous and enthusiastic promotion and practice, which ignores the underlying methodologies or loses focus on individual patients, can lead it into disrepute. It has been said that EBM provides ‘a natural environment’ for the randomized controlled trial (RCT). The attentive reader will note that this is the first mention of RCTs. Whilst it is true that these help provide the highest ‘level’ of evidence, they are certainly not a sine qua non for the practice of EBM. Indeed, it has been argued that the focus on RCTs may provide a ‘selective imbalance’ in the nature of the evidence available. For example, it is far more likely that there will be RCT evidence available about pharmacological interventions, compared with complex behavioural interventions, physical therapy, surgery, etc. This reflects the fact that a large amount of ‘effort’ (including commercial funding and support) is likely to have been put into generating this evidence and the outcomes are often more easily quantifiable.6 As a result, pharmacological treatments for a condition may appear to be ‘evidence-based’ because high quality evidence does exist, whereas nonpharmacological treatments (which may in fact be as, or more, effective) are seen as being nonevidence based, simply because they have not been evaluated by RCTs. The greater availability of evidence for certain types of intervention reflects biases in the research agenda. This is produced by several factors but includes the lack of a mechanism whereby the agenda can be set by the health needs of communities. EBM has been criticized for focusing too much on ‘mainstream’ medicine. It is undoubtedly true that there is more, and generally better quality, evidence available about traditional medical interventions. However, one advantage of the arrival of EBM has been that those who promote ‘complementary’ medicine understand the nature of the rules by which the effectiveness of these therapies will be judged. As a result, there are increasing numbers of systematic reviews and RCTs of interventions in the field of complementary medicine. In addition to embracing the involvement of complementary practitioners, EBM-oriented organizations, such as the Cochrane Collaboration, have tried to widen and encourage participation by consumers. This involvement, in parallel with the explicit involvement of patients in making their own healthcare choices, has helped engage and empower patients in a way that should increase both their confidence in, and compliance with,5 treatment. Conversely, this has been seen as a threat to the professional autonomy of doctors and a challenge to their therapeutic freedom. Unsurprisingly, the challenge to the ‘paternalistic and authoritarian nature of much medical practice’ has been resisted in some quarters. Interestingly,



648 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA a recent Royal College of Physicians Working Party which examined issues of medical professionalism,5 suggested the abandonment of the word autonomy from the definition of a professional, not least because the word can be open to misinterpretation. Too often the word suggests ‘an appeal to personal authority – that is, the right to pursue a practice that is entirely self-generated’. Whereas ‘the doctor should tailor his or her care to the expressed needs of the patient in the light of reliable scientific evidence’.



CONCLUSION The EBM era is characterized by the better use of evidence in medical practice. Finding that evidence, appraising it and integrating it into the advice given to individual patients is a key part of modern medical practice. Otolaryngologists should apply themselves to the study of EBM methodologies with the same perseverance and diligence as they give to the acquisition of their surgical skills.



KEY POINTS  Evidence-based medicine is a key component of modern medical practice.  Evidence is necessary but not sufficient.  All otolaryngologists should acquire the fundamental skills of searching, appraising and synthesizing required to practice EBM and should do so with the same diligence they apply to learning surgical skills.



REFERENCES 1. Davidoff F, Haybes B, Sackett D, Smith R. Evidence based medicine. British Medical Journal. 1995; 310: 1085–6. 2. Sackett DL, Rosenberg WMC, Gray JAM, Haynes RB, Richardson WS. Evidence-based medicine: what it is and what it isn’t. British Medical Journal. 1996; 312: 71–2. 3. Hope T. Evidence based medicine and ethics. Journal of Medical Ethics. 1995; 21: 259–60. 4. Gray JAM. Evidence-based healthcare. London: Churchill Livingstone, 1997. 5. Royal College of Physicians. Doctors in Society. Medical professionalism in a changing world. London: Royal College of Physicians. Retrieved 18 March 2007, from www.rcplondon.ac.uk/pubs/books/docinsoc/docinsoc.pdf 2005. 6. ter Meulen R, Biller-Andorno N, Lenk C, Lie RK. Evidencebased practice in medicine and healthcare. Berlin: Springer, 2005.



52 Critical appraisal skills MARTIN DAWES



Why do we need critical appraisal? Tips for practising evidence-based health care Basics of critical appraisal Forming answerable questions Appraising diagnosis articles



649 650 651 652 654



WHY DO WE NEED CRITICAL APPRAISAL? Evidence-based practice (EBP) begins and ends with patients. We can harm patients by giving inaccurate information about prognosis, make assumptions about diagnostic test results that are false, give therapy that is ineffective or harmful, or fail to give effective therapy. The only way to prevent this happening is to ensure that we have the most up to date knowledge available so that we can share that information with patients. There were 27 randomized control trials published last year just about otitis media. There is no possibility that by reading a couple of journals per week and attending postgraduate seminars that doctors will be able to keep up to date with everything that is important that may affect their patients. EBP starts with the clinician accepting that there is uncertainty about what we do and that sometimes (when that uncertainty occurs frequently or when it is important) we should search for the latest knowledge. Pragmatically, what we do at present is ask colleagues. The flaw with this is the inability to assess the quality of the evidence that is presented to you by your colleague. Does this matter? If the question is important or frequent then yes, it does matter. Research evidence can be approached from two angles. The first is the clinician who wants to undertake research and the second is the clinician who wants an answer to a clinical question. The latter wants to know what to believe. Is it the randomized trial, or the case series or case–control trial? What papers should they look for to answer their question? Levels of evidence have been proposed to help clinicians identify trials most likely to yield the ‘truth’



Appraising therapy articles Appraising systematic reviews Key points Appendix References



658 661 665 665 668



and associated with these are grades of recommendation (see the tables in the Appendix to this chapter). For questions about the efficacy of a therapy, these state that a systematic review of homogenous trials is likely to be the most believable evidence. The process of EBP therefore requires that one searches for the best evidence. For therapeutic decisions, this is usually going to be a systematic review where the authors have searched for and combined, where appropriate, the data from the randomized control trials on that topic. It has the advantage of doing the hard work for you as well as this combination of data in the form of a metaanalysis. Systematic reviews are important as they include the data from all the trials up to that date. It is this that is important and is the hard work. A normal search for papers to be included in the review will often produce a list of more than a thousand papers, and all the titles and abstracts must be read to select relevant articles. The full text of these relevant articles must then be obtained (in itself a time-consuming and expensive task) and read to ensure they meet the selection criteria. The data then have to be independently abstracted from the selected articles. Undertaking a valid, well carried out systematic review is no easy task, but is essential for clinicians to determine the latest evidence. Once a trial or trials have been found that seems to answer your question, one has to check three things. Is the study valid? What are the results? Can these be applied to my patients? The need to check for validity seems, on the surface, to be bizarre. Surely this is what journals should be doing. That assumption is correct but in reality the job of a journal is primarily to make a profit. Therefore one will



650 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA find articles with flawed trials in even the most highly cited journals.1 Journals started as the diaries of researchers. These were then passed around colleagues so that the information could be shared. The development of trials and the assessment of bias are subsequent to that descriptive era. What we are now faced with are seemingly complex long papers with obscure statistical analyses. The scenario is no different to a medical student facing their first anatomy class. By reading around the subject and by practice, one is able to quickly identify key anatomical structures. The same is true of reading scientific articles. One can learn certain aspects of trial design very quickly to be able to assess whether an appropriate design was used for a certain research question. The journal Evidence Based Medicine critically appraises the literature from over 100 journals. However, it only includes a few that are likely to change the way you think about a problem or change the way you practise. Sadly for the other articles, we must appraise them ourselves. The need to assess the methodology of individual trials is extremely important. In large systematic reviews of the same clinical outcome with the same therapy, researchers have compared the results of those trials that were randomized and those that were not randomized. There are some large differences in certain studies that are more than one would expect by chance. One cannot always say that randomized trials show less effect than nonrandomized trials.2 However, it is not enough to accept that a trial is believable at face value. There is now overwhelming evidence that trial methodology has a major influence on the results of therapeutic trials.2, 3 Concealment of randomization, masking and randomization itself are three main components of randomized trial design. The numbers of patients involved in the study and their follow up are the other key features. If there is not concealment of randomization, this may exaggerate the efficacy of the treatment by as much as 30 percent more than trials where there is adequate concealment. Surprisingly, perhaps of lesser importance is blinding (masking) – the doctor looking after the patient, or the patient themselves, knowing whether they are giving (being given) the experimental or placebo treatment. If there is no blinding, the results may exaggerate the effectiveness of the treatment by 15 percent. If less than 80 percent of patients are followed up, one cannot tell what is happening and the results become meaningless. To assess a study there are checklists (see the Figures in the Appendix to this chapter), some of which have now been incorporated into computer programmes such as that found at the Centre for Evidence-Based Medicine website (www.cebm.net) that enable health professionals to appraise the quality of the evidence for themselves. Systematic reviews therefore need to assess the individual validity of the trials that are included. If they do not assess whether there was concealed randomization,



for example, then you, the reader, may be given incorrect information. Assessing the quality of a randomized controlled trial may be carried out using a Jadad score or other systems.4 Despite the Quorom statement5 setting out how systematic reviews should be presented, their quality may be poor.6 For systematic reviews, effective transparent searches and defined inclusion and exclusion criteria are important. The homogeneity (similarity) of the studies is also critical in determining whether a meta-analysis of the data is appropriate. The meta-analysis (combination of the results of the individual trials) is usually only appropriate when there is clinical and methodological homogeneity (that is, the methods and clinical outcomes and starting points are similar). In addition, the results must be put into context of present formats of care. The need to understand how trials may come to the ‘wrong’ conclusion is as important as understating the treatment itself. Clearly, all this may be daunting to a busy clinician so there are sensible short cuts. Always start by looking at evidence that has already been appraised such as that in Evidence Based Medicine and Clinical Evidence. The essential part of effective clinical practice is to continue to ask questions. We need to make sure that what we do is based on the soundest evidence possible. If the only evidence is experiential and anecdotal then that is perfectly satisfactory. However, if there is a systematic review of randomized control trials that all point to an alternative treatment being more effective then that should only be ignored at your patients’ peril.



TIPS FOR PRACTISING EVIDENCE-BASED HEALTH CARE Tip 1: Ask questions Try making one question per patient. Select one question because:  There is likely to be an answer.  The question has arisen more than once or is important; – use a sticky label or write down the patient’s name; – record briefly the problem for example ‘chronic obstructive pulmonary disease’; – write down the question, for example ‘are steroids an effective treatment (mortality and morbidity)?’; – put them in your pocket and look at them at the end of the week.



Tip 2: Searching  Search one question regularly: every two weeks, every month or every quarter!



Chapter 52 Critical appraisal skills



 Conduct your search in a logical order: – clinical evidence; – journal EBM; – Cochrane; – Medline. Often you will find too few articles, or that the articles you do find are not in your library or will take a long time to obtain. Sometimes you will find too many articles and a systematic review is needed. Unless you have time and the question is desperately important, pass and move on to the next question – let someone else answer this one! Appraise only the article(s) that answers your question, have the highest level of evidence and are readily available.



Tip 3: Appraisal  Look for letters about the article in subsequent issues of the journal.  Appraise with others until confident.  Appraise using worksheets or using software, for example CATmaker.  Mark (highlight) on the printed article where you found the important data.  Get someone else to check it for you.



Tip 4: Share your knowledge  Try sharing uncertainty with your colleagues.  Discuss your questions with colleagues (maybe they have already answered them).  Find fault with the article(s) rather than your colleagues.



BASICS OF CRITICAL APPRAISAL Read the abstract Assuming you have three or four articles to appraise, how should you start? First read the abstract. Read it briefly trying to gain an overview firstly of the methods used and secondly the results. For example, a parent brings in a child referred by their GP for recurrent epistaxis but is worried about surgery (describing cautery in dramatic terms) and wonders whether a cream used by a friend’s child might help (Naseptin). You make a note of this and later formulate the question ‘In children with recurrent epistaxis, is Naseptin effective, compared with placebo, at reducing attacks of epistaxis?’ and find this paper.7 The abstract is available on line. It takes a minute or two to read through the abstract: Epistaxis is common in children. Trials show antiseptic cream is as effective as cautery, but it is not known whether either is better than no treatment. We wished



] 651



to know the efficacy of cream in children with recurrent epistaxis. The design was a single-blind, prospective, randomized controlled trial set in the otolaryngology clinic in a children’s hospital. The participants were 103 children referred by their general practitioner for recurrent epistaxis. Excluded were those with suspected tumours, bleeding disorders or allergies to constituents of the cream. Referral letters were randomized to treatment and no treatment groups. Treatment was antiseptic cream to the nose twice daily for four weeks, which was prescribed by the general practitioner before clinic attendance. All children were given an appointment for eight weeks after randomization. The main outcome measures were the proportion of children in each group with no epistaxis in the four weeks preceding clinic review. Complete data were available for 88 (85 percent) of the children. Of the treatment group, 26/47 (55 percent) had no epistaxis in the four weeks before the clinic appointment. Of the controls, 12/41 (29 percent) had no epistaxis over the four weeks. This is a relative risk reduction of 47 percent for persistent bleeding (95 percent CI 9–69 percent) and an absolute risk reduction of 26 percent (95 percent CI 12–40 percent), giving a number needed to treat of 3.8 (95 percent CI 2.5–8.5). We conclude that antiseptic cream is an effective treatment for recurrent epistaxis in children.



The format of any abstract should be similar to this. If there is a complex description of the trial that is confusing, ask yourself why? Is it because the authors are hiding an overall negative result or is it really a necessary peculiar design? So, in this study there were 103 patients with epistaxis in a single blind randomized controlled trial with various outcomes described and a promising result overall.



Does the paper answer the question you are asking? This is a qualitative judgement that you must make on reading the abstract. If it is the highest level of evidence (systematic review) and it deals specifically with your question then you are home and dry (apart from having to critically appraise the review!). Often though this is not the case, in which case you must look for individual randomized controlled trials that address the specific outcomes relevant to your patient.



Find answers to the validity questions using a checklist for the relevant study It is very much easier to read an article when you know what you are looking for. The same occurs when an experienced doctor finds it easier to filter the enormous



652 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA number of questions and signs they can ask or examine for, and homes in on specific areas they think are relevant. This ability to put the patient’s symptoms quickly into context is the key to skilful practice. In exactly the same way, learning what to look for in a research article is key. When teaching medical students, we give exactly the same list of questions that are shown below. This gives them focus and they will work through an article in ten minutes answering all the validity questions. The first place to check for validity questions about the methods employed is, not surprisingly, in the methods section of the paper. However, you will frequently find certain elements of the methodology may be mentioned in the introduction, results or even the discussion. So if you cannot find out, for example, the rate of follow up in the methods, then scan the rest of the text for the answer.



What sort of research design is the trial? You may have to resort to checking the methods section of the paper to obtain this. This was a randomized control trial with single blinding.



Assess the level of bias At this stage one has all the methodology neatly summarized. For example, in this paper there were some problems with follow up. They had data available on 88 of the 103 patients randomized, but 15 of these were assessed by telephone. This introduces an element of bias that may reduce the validity. If this was less than 80 percent then really the trial is so severely compromised that one cannot make any interpretation of the result. If there is not concealed randomization (see under Concealed randomization) then the results may overestimate the effectiveness by 30 percent. So, assessing bias is the art of appraisal. In the end you have to decide whether the effect of the treatment demonstrated is likely, taking into consideration the level of bias, to translate into benefit in your patient.



Translate the results into something clinically meaningful If the article talks about risk reduction, how does that translate into my patient? A frequent term used now is numbers needed to treat. This puts the effect into the context of the patient and lets them decide whether the benefit is worth the trouble or potential side effect.



How can I implement this? Does this help my practice either by reinforcing what I know and making me more confident or does it change



my practice. If it is the latter, how am I going to go about implementing that change?



The numbers The most frequent fear expressed is that of the statistics. How can I tell whether these were the correct statistics to be used? I tackle this obliquely. First, ask yourself what is likely to be a beneficial outcome for your patients. Is a reduction in mortality of 1 percent enough for this invasive procedure, or should it be 5 percent? This is by far the most important feature of the results. If the trial was designed to establish a benefit that you regard as clinically insignificant then you can stop reading! Let us assume you settle for a 5 percent reduction in mortality, for example. The results of the trial say that this was achieved. The statistical test is there to show you that this difference was very unlikely to have occurred by chance in two ways. First, that the difference observed in this trial was unlikely to have happened by chance (p value) and, second, if this trial were repeated 100 times, 95 percent of the trials would have shown a similar result (confidence interval). How they determine the p value (i.e. with a Student t test or a chi-square is beyond the scope of this article but a quick tip is to check the letters written in response to the article because if the article is in a major journal, statisticians will be quick to point out the incorrect use of statistics). In our experience, there is much less use of incorrect statistical tests than there is of poor trial methodology. To summarize, check that the difference is clinically significant and did not occur by chance.



Summary Critical appraisal is a logical way of assessing the likely truth of a piece of research, evaluating the results in terms of individual patient care, and then assessing the whole and determining whether it is useful for your practice. It combines some knowledge of trial design, a small amount of basic arithmetic and some qualitative judgements.



FORMING ANSWERABLE QUESTIONS The reason we do not usually ask questions is they are so difficult to answer. This was clearly illustrated in the Hitchhiker’s Guide to the Galaxy when the super computer gave the answer ‘42’ to the Ultimate Question of Life, the Universe, and Everything.8 The people asking the question were horrified and angry. The computer calmly suggested that, instead of panicking, they should go back and consider the question and that he would of course help them do this. Clearly, this is an extreme example of where the question has been so badly formed that the answer is meaningless.



Chapter 52 Critical appraisal skills



I see a patient with diabetes. Their control had been haphazard and they had had repeated HbA1C’s performed over the last year. Their control seems better now but the HBA1c test is expensive to perform. I wonder how effective is blood glucose at monitoring diabetes. I do a search for ‘blood glucose’ and ‘diabetes mellitus’ and turn up 25,000 articles. I went wrong in two places. You may have spotted that my search was unstructured; the search terms were so vague that I was unable to generate a manageable list of relevant items. However, this arose partly because of the unfocused nature of my question. I want to know how ‘effective’ the test is. What do I mean by that? If it is negative does that rule out poor control – or if it is positive (high) does that mean they have poor control. In my case it is the former that really is of interest. We could have gone to a textbook and read thoroughly about diabetes, but we do not have the time, and textbooks are often many years out of date. There are many questions that arise from any consultation and being human we can only deal with those that are most important. In this case it may have been ‘diabetic control’. However if I was a medical student and wanted just to know more about diabetes, then a textbook is an excellent place to start. For understanding more about therapy in general, a resource such as Clinical Evidence9 is extremely valuable (Clinical Evidence cites all the evidence from which therapeautic recommendations are made). This form of enquiry is termed a ‘background question’ or one that helps the person understand the problem in general. This in contrast to what is being discussed here, which are termed ‘foreground questions’ or decision-making questions. As a student, one asks mainly background questions but as one’s experience with a condition increases and the clinical need for decision-making increases, the proportion of background questions fall as the foreground questions rise (Figure 52.1).



Background questions



Foreground questions



] 653



My question now is: ‘In diabetics, is fasting blood glucose (compared with HBA1c) effective at excluding poor diabetic control?10 (a fasting blood glucose o7.8 rules out poor glycaemic control). The query is becoming better but the first part is still vague. Perhaps I should enter something about noninsulin dependent diabetics. By structuring a question the answer may be found more efficiently.11 It is therefore important to try to break the question down into several parts:    



patient and problem; intervention (or diagnostic test); comparison intervention (optional); outcomes (PICO).



Patient and problem The first part is to identify the problem or the patient. Some health care problems are not always about patients! For example, an administrator may want to know whether having acute medical beds in a temporary holding ward next to the accident and emergency department is any better than having conventional acute medical wards in the hospital. Make sure at this stage in the question you are describing the problem or patients that you see. However, if one is too specific at this stage you may miss some important evidence and there is a balance to be struck between obtaining evidence about exactly your group of patients and obtaining all the evidence about all groups of patients.



Intervention The intervention is equally important. It may in fact be a postponement of an action, such as an operation. In patients with abdominal pain lasting less than 12 hours, does an additional 24-hour delay before referral to hospital alter outcome? Most interventions are more straightforward, such as types of dressings, drug therapies or counselling. Alternatively, they can be about the provision of differing environmental factors, such as the de´cor of waiting rooms or dealing with the way in which information is given to patients, i.e. positively or negatively.12 The intervention can be a diagnostic test. However, specify whether you are trying to detect or exclude a disease.



Comparison intervention Experience with the condition



Figure 52.1



Foreground and background questions.



Sometimes, there is a comparison of the intervention. For example, one might seek papers comparing the use of an active compound compared with either a placebo or



654 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA another active drug. Considering whether you are looking for comparative studies will help when searching for that evidence. You may be comparing one diagnostic test against another.



Outcomes Outcome measures are particularly important when considering the question. It is worth spending some time working out exactly what it is you want. In serious diseases it is often easy to concentrate on the mortality and miss the important aspects of morbidity. For example, the use of toxic chemotherapies for cancer may affect both aspects. This is also the time when it is important to know what outcome is relevant to the patient. In Menie`re’s, deafness and tinnitus may occur but the patient may only really be interested in controlling one problem as a priority. The search should primarily focus on that outcome.



Type of question Once we have created a question, it is helpful to think about what type of question you are asking, as this will affect where you look for the answer and what type of research you can expect to provide the answer. For example, although randomized controlled trials (RCTs) are a very valid type of study, it would be unethical to perform an RCT on harm (would you volunteer for the exposure arm of such a study?). Questions about harm should look for different types of research, such as case–control studies or cohort studies.



APPRAISING DIAGNOSIS ARTICLES Diagnoses are usually made through interpreting and combining information from a combination of the clinical history, examination findings and the results of investigations, such as blood tests and x-rays. An important component of evidence-based health care is working out how useful these different pieces of information are in making a diagnosis. For example, if you test a patient’s urine and find protein in it, how likely is it that this represents kidney disease? A test result needs to be interpreted in the light of how accurate you think the test is, and what you already know about the patient. In making a diagnosis you are consciously (or subconsciously) weighing up probabilities. Each question you ask or each observation changes those probabilities. For example, if a patient limps in then there is a probability of arthritis being the cause of the limp. If the patient is 70 years of age, the probability is higher than if the patient is 30. The age of the patient is a diagnostic test, either increasing or decreasing the probability. It is



essential that when contemplating a test that the person requesting the test has some idea of the probability of the diagnosis. They will also need to decide whether the test is helping them exclude or include the diagnosis. Finally, they need to know what is the likely probability of the diagnosis after a positive or a negative result. What is meant by test accuracy? Accuracy is a description of how well the test diagnoses and excludes the disease when the result is positive or negative (see under What are the results). Sensitivity reflects the ability of a test to identify people who have disease and specificity reflects the ability of the test to identify ‘normality’, or people who do not have disease. So this means that tests with very high sensitivities and specificities can be extremely useful at either excluding or including a diagnosis. In primary care some of our time is spent seeing the worried well. These are patients who present with symptoms caused by a mild or self-limiting disease but are naturally worried that it may be something more serious. In these cases we want to use tests that mainly exclude disease. In hospital we are seeing people who are ill and we quickly want to make a diagnosis (including disease). To see an example of how high sensitivity and specificity values of a test can be used in clinical practice, let us consider how general practitioners deal with patients with a sore throat. We see many sore throats in general practice and it would be useful to know whether a new ‘rapid test’ could be useful in helping diagnose bacterial sore throats.13 I selected this particular paper for illustrative purposes rather than it being the ‘best’ level of evidence. This paper was about better diagnosis of throat infections in general practice. Four features of patients were determined: 1. 2. 3. 4.



fever Z381C; lack of cough; tonsillar exudates; anterior cervical lymphadenopathy.



Five hundred and fifty-eight patients older than 11 years presenting with sore throat were swabbed and examined for these features. They all also had a rapid streptococcal antigen detection test for detecting group-A B-haemolytic streptococcus (GABHS). The next step was to go through the paper checking for validity.



Is the study valid?  Did the authors answer the question? The question is often found at the end of the introduction section of the paper. Some authors leave you guessing as to what the original question might have been. If the paper does not have a clear question, it cannot give any clear answer, so you can save yourself time by



Chapter 52 Critical appraisal skills































moving on to the next paper. You should also consider whether the research question is relevant to your clinical question. What were the characteristics of the groups? This would contain some description about the numbers of patients, their race and gender, age and any other features that are pertinent to the study. Is it clear how the test was carried out? To be able to apply the results of the study to your own clinical practice, you need to be confident that the test is performed in the same way in your setting, as it was in the study. What was the test undertaken? How was it done and by whom? Was it a multi-level test (i.e. a serum level) with different thresholds? Is the test result reproducible? This is essentially asking whether you obtain the same result if different people carry out the test, or if the test is carried out at different times on the same person. Was the reference standard (gold standard) appropriate? To start with, we need to find out how the study found out what the ‘truth’ really was about patients. That is to say, how did the investigators know whether or not someone really had a disease? To do this, they will have needed some reference standard test (or series of tests) which they know ‘always’ tells the truth. Were the reference standard and the diagnostic test interpreted blind and independently of each other? If the study investigators know the result of the reference standard test, this might influence their interpretation of the diagnostic test and vice versa. Was the reference standard applied to all patients? Ideally, both the test being evaluated and the reference standard should be carried out on all patients in the study. There may be a temptation, for example, if the test under investigation proves positive, not to bother administering the reference standard test. In many cases, the reference standard test may be invasive and may expose the patient to some risk and/or discomfort. While it would usually be ethical to use such a test on a patient in whom one had grounds to suspect that it might be positive, it would not be ethical if one thought that the test would be negative (i.e. if the diagnostic test being evaluated had been negative). Therefore, when reading the paper, you need to find out whether the reference standard was applied to all patients, and if it was not look at what steps the investigators took to find out what the ‘truth’ was in patients who did not have the reference test. Was the test evaluated on an appropriate spectrum of patients? Another complication is that a test may perform differently depending upon the sort of patients on whom it is carried out. For example, the more severe is a disease, the easier it tends to be to detect. Thus, you might find that testing for bacterial sore throats might be ‘better’ at detecting infection



] 655



when evaluated in patients who attend an ENT clinic, than in those attending family practice. These may be because people with chronic disease who attend an ENT clinic will be more symptomatic and have more bacteria in general, making it easier to detect. This problem is referred to as spectrum bias. A test is going to perform better in terms of detecting people with disease if it is used to identify it in people in whom the disease is more severe, or advanced. Similarly, the test will produce more false–positive results if it is carried out on patients with other diseases that might mimic the disease that is being tested for. The issue to consider when appraising a paper is whether the test was evaluated on the typical sort of patients in whom the test would be carried out in real life. The optimal study method for the assessment of a diagnostic test is a cohort study evaluating the test in a group of patients thought to be at risk of having the disease. A less valid alternative method that has been used is a case–control study. The latter method studies patients known to have the disease and control patients without the disease. The performance of the test in the two groups is compared. This sort of study has been shown to overestimate the effectiveness of the diagnostic tests.14



What are the results? In an ideal world, a positive test would mean that someone has disease, and a negative test would mean they do not have disease. Unfortunately, this is rarely the case. When a test is carried out, there are four possible outcomes: 1. The test can correctly detect disease that is present (a true positive result). 2. The test can detect disease when it is really absent (a false–positive result). 3. The test can correctly identify that someone does not have the disease (a true negative result). 4. The test can identify someone as being free of a disease when it is really present (a false–negative result). These possible outcomes are illustrated in Table 52.1. Given that a test may potentially mislead us if we obtain a false–positive or a false–negative result, we need to have some way of characterizing how accurate the test really is (Table 52.2). What does this mean? The high specificity means that there were very few people (16) out of the 375 people who did not have GABHS. So that means (using some reverse logic) that if the rapid test was positive (going across the row), by inference they almost certainly have the disease. We can simplify this to say that in tests with high specificities, where the result is positive, it rules the diagnosis in. The mnemonic for this is SpPIn.



656 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA Table 52.1



Possible outcomes of a diagnostic test.



Test result



Positive Negative Totals



Table 52.3 Potential results of rapid test screening for GABHS in patients with sore throat in an imaginary low prevalence setting.



‘Truth’ Disease present



Disease absent



Totals



a True positive c False negative a1c



b False positive d True negative b1d



a1b c1d a1b1c1d



Sensitivity = a/(a 1 c) Positive predictive value = a/(a 1 b). Specificity = d/(b 1 d) Negative predictive value = d/(c 1 d). Prevalence = (a 1 c)/(a 1 b 1 c 1 d). Accuracy = a 1 d/a 1 b 1 c 1 d.



Table 52.2 Results of rapid test screening for group-A Bhaemolytic streptococcus (GABHS) in patients with sore throat. ‘Truth’



Positive Negative Totals



GABHS1ve



GABHS  ve



Totals



119 64 211



16 359 375



135 423 558



Sensitivity = 119/211 = 65%. Specificity = 359/375 = 96%. Positive predictive value = 119/135 = 88%. Negative predictive value = 359/423 = 85%. Prevalence = 211/558 = 33%. Accuracy = 478/558 = 85%.



In tests with high sensitivities and the test is negative then it rules out the disease. So the mnemonic for this is SnNOut. Clearly it is not 100 percent accurate at excluding (that would only happen if the sensitivity was 100 percent). How you take that finding and put it into practice depends on the severity of the disease as well as the other symptoms and signs that indicate the pretest probability of the disease. In reality, patients who are well where you are trying to exclude disease need tests with high sensitivities (primary care). Patients who are ill need tests with high specificities to determine the cause of the illness. This is an oversimplification of the process of diagnosis but is helpful in terms of remembering what tests do. Often, of course, we are faced with tests whose sensitivity and specificity are not so high that we can use the test to rule in or rule out a disorder. For such cases, we need a measure of a particular test result’s ability to predict the presence or absence of disease. We might think that this is provided by the positive and negative predictive values (for positive and negative results, respectively). However, in practice we find that these values can change depending on the prevalence of the target disorder amongst the test population: if the



‘Truth’



Positive Negative Totals



GABHS1ve



GABHS  ve



Totals



12 6 18



16 359 375



28 365 393



Sensitivity = 67%. Specificity = 96%. Positive predictive value = 43%. Negative predictive value = 98%. Prevalence = 4.5%.



disorder is more common, the positive predictive value will be higher and the negative predictive value lower. For example, if the results are as follows with a much lower prevalence (Table 52.3). The sensitivity remains similar and the specificity is the same but the positive predictive value drops considerably. Now the test only has a 50:50 chance of detecting someone with disease. You can see how the prevalence will sometimes significantly alter the effectiveness of a test. It is important to always consider the setting before recommending tests that perform well in an alternative health care setting. However, since sensitivity and specificity are not affected by disease prevalence, we can combine them to create a combined measure of the efficacy of a particular test result. This can be thought of as, for a given test result, the likelihood that a patient with the disorder would yield that test results compared to the likelihood that a patient without the disorder would yield that same test result. This is the likelihood ratio (LR). As a rule of thumb, for positive test results, an LR of ten or above shows that the test result is good at ruling in disease; for negative test results, an LR of 0.1 or less shows that the test result is good at ruling out disease. Two additional benefits of the LR are that they can be used to generate specific post-test probabilities for your patient and that, where we have multiple independent tests, the LRs can be multiplied together to yield a much more powerful test. The prevalence of GABHS was 33 percent. If the test is positive it is very likely they have GABHS, but what about if it is negative? What I can do though is work out the likelihood ratios: Likelihood ratio ðþve resultÞ Sensitivity 0:65 ¼ ¼ 16:25 ¼ ð100%  specificityÞ ð1  0:96Þ Likelihood ratio ðve resultÞ ð100%  sensitivityÞ ð1  0:65Þ ¼ ¼ 0:36 ¼ specificity 0:96



Chapter 52 Critical appraisal skills



Figure 52.2 shows the nomogram for likelihood ratios. To use the nomogram I have drawn a line from the pretest probability of 33 percent through 16 (on the likelihood ratio line) to obtain a post-test probability of approximately 90 percent if the result is positive and through 0.36 (roughly) to obtain a post-test probability of approximately 12 percent if the result is negative. But what if our patient has a lower pretest probability? For example, this might be a patient with only a 20 percent pretest probability. Using a ruler, see what their post-test probability would be. The nomogram lets you quickly see the implications of a positive or negative result in terms of post-test probability.



0.1



] 657



99



0.2



0.5



1



95



1000



90



500 2



Will the results help my patient?



200



80



100



ARE THE RESULTS RELEVANT TO YOUR PATIENT?



50



70



20



60



10



50



5



40



2



30



5



So far, we have considered how to interpret results of tests for your patients. However, it may be that you feel that the results of studies carried out to assess the accuracy of the test are not generalizeable (or relevant) to your setting, or you may consider that it would not be helpful to your patient to perform the test.



10



20



GENERALIZABILITY TO YOUR SETTING



One aspect to consider is whether the assessment(s) that have been carried out of test validity are applicable to your setting. We have already seen how predictive value is dependent upon prevalence, so the predictive value of a test in one setting is usually not the same in another setting. Sensitivity and specificity (and likelihood ratios) are not dependent on prevalence, but they can vary according to the type of patients on which the test is carried out. The sensitivity of a test will depend upon the severity of disease in the population being tested. The more advanced or severe the disease, the more likely the test is to identify it. The specificity of a test will depend upon the prevalence of other diseases in the population that might lead to false–positive results. The more that other diseases are present, the more likely a false–positive result. A second aspect to consider is whether the test is carried out the same way in your setting as it was in the study. Some diagnostic tests depend upon the skill of the person carrying out the test, and the skill of the people interpreting the test result. Tests may be carried out in different ways, and it is important to know that it is carried out the same way in your local hospital or laboratory as it was in the study which evaluated the test.



WILL IT HELP YOUR PATIENT?



One way to think about whether or not to perform the test is whether it will influence the management of the



%



%



1 20



30



0.5



40



0.2



50



0.1



10



60



0.05 5



70



0.02 0.01



80



0.005



2



0.002 90



0.001



95



1



0.5



0.2



99 Pretest probability



Likelihood ratio



Figure 52.2 Nomogram for likelihood ratios.



0.1 Post-test probability



658 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA patient or groups of patients you are likely to see. If you think that a disease is unlikely in a particular patient, it may be that if you carried out the test and it was positive, you would still think that the disease was unlikely, so it would not influence your management. Conversely, if you think that a disease is likely in a given patient, it may be that a negative test result would not stop you treating the patient, since you would still think that the disease is likely whatever the test result. Putting this in terms of preand post-test probabilities, in the first instance the pretest probability of disease is so low, that even with a positive test result, the post-test probability will be less than 50 percent. In the second instance, the pretest probability of disease is so high that even with a negative test result, the post-test probability will be more than 50 percent. In practice, knowledge about the accuracy of diagnostic tests will perhaps be most useful when planning protocols or guidelines to manage particular groups of patients.



Summary An evidence-based approach to deciding whether a test is effective for your patient involves the following steps:  frame the clinical question;  search for evidence concerning the accuracy of the test;  assess the methods used to determine the accuracy of the test;  find out the likelihood ratios for the test;  estimate the pretest probability of disease in your patient;  apply the likelihood ratios to this pretest probability using the nomogram to determine what the post-test probability would be for different possible test results;  decide whether or not to perform the test on the basis of your assessment of whether it will influence the care of the patient, and the patient’s attitude to different possible outcomes.



APPRAISING THERAPY ARTICLES When we are required to answer a question about therapy, the identification of information from an article is not straightforward. The layout is frequently different depending on the journal you are reading. One is often relying on the editorial board of the journal, as well as the people who reviewed the article, to have ensured the quality of the article. There may be confusing statistics that make interpretation of the data seem difficult. These problems seem to be very daunting when one is trying to identify information that may affect the care of a patient. These following questions will help you identify the real validity of the therapy article.



 Is the study valid? Did the authors answer the question? For this example I shall be using the Kubba paper7 on Naseptin and epistaxis. The abstract of this article is shown under Read the abstract. Kubba et al.7 have answered their question about the effectiveness of Naseptin in epistaxis. They state clearly how they did this and what they used to determine ‘effective’.  What were the characteristics of the groups? This would contain some description about the numbers of patients, their race and gender, age and any other features that are pertinent to the study. This is usually found in Table 1 in most papers but in this case is described in the protocol part of the subjects and methods section.  Are the comparison groups similar? To determine whether bias has occurred in the selection procedure, certain potentially relevant characteristics of the two populations should be displayed in tabular form. Usually these include age, sex, duration of illness and other demographic and functional characteristics. This is covered in the first paragraph of the results section of this paper. There were more boys than girls, but the age and duration of history were the same in both groups. Is it likely that girls and boys would differ in response? The aim is not to find if there are statistical differences between the two groups but to find whether thare are clinically significant differences. You as a clinician reader should judge whether you think the differences are clinically significant.  What was the treatment? In this case it was either Naseptin or nothing. They describe clearly the reasons for not using a placebo.



Equal treatment The groups in the trial were treated equally throughout apart from the experimental intervention. It should be clear from the article that, for example, there were no cointerventions that were applied to one group but not the other, or more frequent or detailed assessments on one group compared to the other.



Placebo control Patients do better if they think they are receiving a treatment than if they do not, even if the treatment is an inactive substance: the placebo effect is a widely accepted potential bias in trials. A placebo is an inactive treatment that is given so that the patient does not know whether he or she has been given the active treatment; the control group in a trial will often receive a placebo treatment so that this effect is equal in both groups. Patients in both



Chapter 52 Critical appraisal skills



groups of the trial should not know whether or not they are receiving the active therapy. However there is a twist in the tail – one can have a placebo-controlled trial where the patients are having active treatments in both arms of the study in addition to the new treatment or the placebo. So the term ‘placebo-controlled trial’ does not exclude the use of active treatments (as long as both groups get the same).



Follow up There are three major aspects to assessing the follow up of trials: 1. Did so many patients drop out of the trial that its results are in doubt? 2. Was the study long enough to allow outcomes to become manifest? 3. Were patients analysed in the groups to which they were originally assigned (intention-to-treat)?



Drop-out rates A good clinical trial requires the complete follow up of patients in both the control and experimental groups. If less than 80 percent of patients are adequately followed up then the results may be invalid. In this study they had problems and had to phone subjects as they did not attend the outpatient clinic. They talk about this weakness of the study in the discussion. Does it invalidate the study?



Length of study The length of the study is critical in determining the clinical significance of the results. For a disease of short duration, such as an acute infectious disease, the period of the study need only be long enough to cover the course of that infection. Where the disease is more progressive and lengthy then the duration of the study needs to reflect that. The question that the research is trying to answer will also be helpful in deciding the appropriate length of the study.



] 659



different outcomes, even when the study medicine they have been prescribed is a placebo. The correct form of analysis, in which patients are analysed in the groups to which they were assigned, is called ‘intention-to-treat’ analysis.



Were the groups randomized? The major reason for randomization is to prevent bias. The largest potential cause of bias in research about effectiveness of an intervention, is if the patients allocated to the treatment group are different from those allocated to the control group in a way that might influence outcome. To reduce this bias as much as possible, the decision as to which treatment a patient receives should be determined by random allocation. Why is this important? There could be any number of confounding variables that affect a given situation, some of which we know about, but also others that our knowledge of human physiology has not yet uncovered. Randomization is important because it spreads all confounding variables evenly amongst the study groups, even the ones we do not know about.



CONCEALED RANDOMIZATION



As a supplementary point, clinicians who are recruiting patients into a trial may consciously or unconsciously distort the balance between groups if they know the treatments that will be given to patients. For this reason, it is preferable that the randomization list be concealed from the clinicians. The amount of impact the lack of concealment may have is larger than if the trial was not blinded. If there is inadequate concealment, the size of the effect may be overemphasized by as much as 30 percent.



BLINDING



It is also important that, if possible, both clinician and patient be ‘blind’ to whether or not they are receiving treatment. Lack of blinding may lead to exaggeration of the effectiveness of the therapy by as much as 20 percent.3



Outcome measures Intention-to-treat Because anything that happens after randomization can affect the chances that a patient in a trial has an event, it is important that all patients are analysed in the groups to which they were randomized. This is an essential prerequisite for valid evidence about therapy. For example, it has been repeatedly shown that patients who do and do not take their study medicine have very



To determine the impact of an intervention on patient health, one needs to define clearly the outcome that needs to be measured. This sounds straightforward enough but can take up a larger amount of the preparation time of the research. An outcome measure is any feature that is recorded to determine the progression of the disease or problem being studied. In this paper a comprehensive questionnaire was used to determine outcomes.



660 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA



What are the results?  How large is the treatment effect?  How precise is the estimate of the treatment effect? In any clinical therapeutic study there are three explanations for the observed effect: 1. the effect of the treatment. 2. chance variation between the two groups. 3. bias.



event rate (CER) is 0.71 (this number is also the risk of epistaxis in the control group, and corresponds to a percentage risk of 71 percent). In the experimental treatment group there were 21 children with epistaxis out of 47 children. The experimental event rate (EER) was therefore 0.44 (44 percent). Now that we have these numbers we can start working with them. RELATIVE RISK



Good research will explicitly endeavour to reduce the effects of chance and bias by using special designs, as we have seen.



NUMBERS NEEDED TO TREAT



The most common approach is to describe the effect of a drug in terms of the risk reduction. The risk is usually of continuing to have, for example, pain or an event. When evaluating a paper on a therapeutic manoeuvre, it is possible to establish its clinical effectiveness by determining the proportion of patients receiving treatment who gain benefit. From the paper the following results are drawn:



This simple statement obscures the fact that identifying the correct table in the paper and then obtaining the correct figures from it is often not as easy as it sounds. This is one of the parts of appraisal that it is often useful to share with others to ensure you have the correct data. From Table 52.4 it can be seen that 47 patients were assigned to the active treatment (or experimental) group and 41 to the watchful waiting (or control) group. There were a total of 50 children with epistaxis; 21 occurred in the experimental group and 29 in the control group.



One way of combining the event rates is to use relative risk (RR). Here, we make a ratio of the two risks to show which group has a higher or lower risk. The relative risk is calculated by dividing the risk of having epistaxis in the treatment group by the risk of the control group = 0.44/ 0.71 = 0.62. This means that the risk of epistaxis continuing with Naseptin was 0.62 (or 62 percent) of the risk without Naspetin. Relative risk reduction It is also possible to show the extent of the benefit compared to the original risk by calculating the relative risk reduction (RRR). This is similar to relative risk, except that we take the difference in risk between the two groups and divide that by the control group’s risk: RRR ¼



CER  EER 0:71  0:44 ¼ ¼ 0:38 CER 0:71



RRRs are often presented as percentages. The RRR means that the treatment reduced the risk of continued nosebleeds by 38 percent relative to that occurring in this control population (the RRR given by the authors is for persistent bleeding whereas I have calculated for cessation of bleeding). This is a large risk reduction in a population that are all bleeding. But what if it was an RRR for primary stroke reduction where the events are far fewer? So the RRR has to be judged against the prevalence of the outcome we want to prevent.



EVENT RATES



It is important to set some standard methods of describing this data. In therapy, we start by converting the numbers of patients into event rates. In the control group, 29 out of 41 patients had epistaxis. This may be presented as an event rate of (29/41) = 0.71. So the control Table 52.4 Results of a trial of Naseptin against watchful waiting in stopping epistaxis in children.



Active treatment (Naseptin) Watchful waiting Total



Epistaxis



No epistaxis for four weeks



Total



21



26



47



29 50



12 38



41 88



Absolute risk reduction The absolute risk is that 71 percent of individuals in the placebo group had nose bleeds and that 44 percent of the treatment group had nosebleeds. The difference (absolute risk reduction (ARR)) between these gives us another view of the effectiveness of the treatment. The ARR of this therapy is 0.27 (CER–EER = 0.71 – 0.44). ARR = CER – EER Consider 100 people given this treatment. We would expect 27 less children to have epistaxis if this group were given Naseptin than if they were not. We can convert this into how many people we would need to treat to prevent one child continuing to have epistaxis. This number is the reciprocal of the ARR. In this example 1/0.27 = 3.7. Therefore, to prevent a child continuing to have epistaxis we need to treat four people. This is the number needed



Chapter 52 Critical appraisal skills



to treat (NNT). It is more useful than relative measures because, as well as measuring the benefit of a therapy, it also represents the baseline risk and the absolute benefit you can expect from the therapy. NNT = 1/ARR Sometimes, papers will present their results in the form of NNTs, as in this one which is very helpful, but often they do not, and prefer to present risk ratios (perhaps not surprisingly, since these tend to make beneficial results look more spectacular) or similar measures. You will often be required to calculate NNTs from the data in the study. Fortunately, the arithmetic is fairly straightforward. Some other examples of NNTs are included in Bandolier (http:// www.jr2.ox.ac.uk/bandolier/band50/b50-8.html). This is a small study and like the relative risk there is a confidence interval around the NNT (to calculate this15 it is easier to use software (www.ebm.net). Also, there is the ‘believability’ issue of a relatively small trial. This is where the clinical judgement comes in – would you use this treatment based on 100 patients?



Will the results help my patient?  Can the results be applied to my patient?  Were all clinically relevant outcomes considered?  Are the benefits worth the harms and costs? Given that your patient was not in the trial and may not even have been eligible for it (due to age, sex, comorbidity, disease severity or for a host of other sociodemographic, biologic or clinical reasons), how can you extrapolate from the external evidence to your individual patient? (We use the term ‘extrapolate’ here, although this might more accurately be described as ‘particularizing’ the results of a trial (which looks at a generic population) to your particular patient.) This requires that you apply some of your knowledge of human biology and clinical experience to answer the question: ‘Is my patient so different from those in the trial that its results cannot help me make my treatment decision?’



HOW MUCH BENEFIT CAN YOU EXPECT FOR YOUR PATIENT?



Given that you have decided that your patient is sufficiently similar to those in the study for the effect to be applicable, the next questions is how much? Here is one way of going about this. f Method First estimate, as a decimal fraction, your patient’s risk compared to the control group from the study. Thus, if your patient is twice as susceptible as those in the trial, f = 0.2; if half as susceptible, f = 0.5. As long as the treatment produces a constant RRR across a range of susceptibilities, the NNT for your patient is simply the



] 661



trial’s reported NNT divided by f. (This is a big assumption and we are only beginning to learn when assuming a constant RRR is appropriate (for numerous medical treatments such as antihypertensive drugs) and inappropriate (for some operations such as carotid endarterectomy, where the RRR rises with increasing susceptibility). So, in the above example, we might assign an f of 0.7 (70 percent chance of having epistaxis to the control group of subjects), which would yield an NNT of (3.7/0.7) = 5. INCORPORATING YOUR PATIENT’S VALUES AND PREFERENCES



What does your patient think? Have you talked to them about their personal preferences, concerns and expectations? Here, we must try to determine whether the outcome of the therapy actually serves the values and preferences of the patient. Many factors can come into play, not least the difficulty of representing them. Sometimes, the situation may be very clear-cut or demand immediate action. For example, in a patient having a heart attack, the value of survival and the preference for a simple, low-risk intervention such as aspirin, given the efficacy of this regimen, usually makes this decision quickly agreed and acted upon. In other cases, the answer may take weeks and several visits to sort out, such as in choosing between radiation and adjuvant chemotherapy for stage II carcinoma of the breast. These questions are perhaps the most important, yet have the least written about them in this chapter. That is because they are also the most complex and depend to a great extent on the clinical situation.



Summary An evidence-based approach to deciding whether a therapy is effective for your patient involves the following steps. 1. Frame the clinical question. 2. Search for evidence concerning the accuracy of the test. 3. Assess the methods used to carry out the trial of the therapy. 4. Determine the NNT of the therapy. 5. Decide whether the NNT can apply to your patient, and estimate a particularized NNT. 6. Incorporate your patient’s values and preferences in to deciding on a course of action.



APPRAISING SYSTEMATIC REVIEWS What happens when there is more than one randomized control trial on the same treatment or diagnostic test? In



662 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA this case you want to look for a paper that has combined these trials together. What you need here is a systematic review, a publication that collects all of the evidence in a particular area and summarizes it. More formally, a systematic review is ‘a review of a clearly formulated question that uses systematic and explicit methods to identify, select and critically appraise relevant research, and to collect and analyse data from studies that are included in the review. Statistical methods may or may not be used to analyse and summarise the results of the included studies.’ This section will look at how you appraise studies that search for all research on a specific question and in some way combine it. You may be familiar with the narrative reviews traditionally found in many journals. Normally someone, usually an expert, looks at the evidence in a certain area. Mulrow16 argues that traditional reviews do not routinely use systematic methods to identify, assess and synthesize information. Thus, normally there is no methods section for the actual conduct of the review. The reader then has no way of knowing whether the review is based on a systematic review of the evidence, or on a collection of papers that the author has found in a less systematic way and thus the evidence presented may not be complete. There is a need for systematic reviews of the evidence. This needs to be undertaken in just as rigorous a way as any primary piece of research, with a clear question and explicit methods for all stages of the process, such that another person could replicate the review. Three key features of such a review are: 1. a strenuous effort to locate all original reports on the topic of interest; 2. critical evaluation of the reports; 3. conclusions are drawn based on a synthesis of studies which meet preset quality criteria. When synthesizing results, a meta-analysis may be undertaken. This is ‘the use of statistical techniques in a systematic review to integrate the results of the included studies’, which means that the authors have attempted to synthesize the different results into one overall statistic.



Where can you find systematic reviews? Systematic reviews have been published in a variety of journals, and in addition the Cochrane Collaboration provides an important resource in this area. The Cochrane Collaboration is an international collaboration that is committed to ‘preparing, maintaining and disseminating systematic reviews of the effects of health care’. It promotes and publishes the Cochrane Library (http://www.cochrane.org). This contains the Cochrane Database of Systematic Reviews. These reviews have a prespecified protocol, where every stage of the process of undertaking the review is made explicit.



Sections covered in a protocol include background, objectives, criteria for considering studies for the review (including types of participants, types of interventions, types of outcome measure, types of study), search strategy for identification of studies, study selection, methodological quality assessment, data extraction, meta-analysis, the comparisons to be made, subgroup analysis and sensitivity analysis to be undertaken. The point to note here is that all these aspects of the protocol are set out before the review is undertaken. Many of the systematic reviews so far completed are based on evidence of effectiveness of an intervention from RCTs.



Is the review valid In this section we will focus on systematic reviews of therapy. Of course, systematic reviews exist for other types of research.17  What databases and other sources did the authors of this review search? The paper should give a comprehensive account of the sources consulted in the search for relevant papers, and the search strategy used to find them. Since standard databases fail to correctly index up to half of published trials, and negative trials are less likely to be published in the first place, the search strategy should include hand searching of journals and searching for unpublished literature. Other questions to ask yourself about the search strategy are: Have any obvious databases been missed? Did the authors check the reference lists of articles and of textbooks? Did they contact experts (to obtain their list of references checked for completeness and to attempt to find out about ongoing or unpublished research)? Search terms used are also important – has an obvious medical subject heading (MeSH) (a standard thesaurus for medical indexing) term been missed or is there another term they have not used? For example, when searching for papers looking at care of the elderly, use of the term ‘senior’ in the North American literature may be overlooked. Searching should include languages other than English. Unpublished data are useful to include because studies with significant results are more likely to be published than studies without significant results, known as publication bias. For example, research which is carried out by non-English language teams is more likely to be published in English if it shows a strong result in favour of an intervention than if it shows a negative or indifferent result.  How the studies were selected – eligibility criteria? You should look for a statement of how the trial’s validity was assessed, using criteria such as those in the therapy guides. It is particularly reassuring when two or more investigators applied these criteria



Chapter 52 Critical appraisal skills



] 663



independently and achieved good agreement in their results. You need to know what criteria were used to select the research. This should include who the study participants were, what was done and what outcomes were assessed. A point to consider is that the narrower the inclusion criteria, the less generalizeable are the results. However, this needs to be balanced with using very broad inclusion criteria, when heterogeneity becomes an issue. Such studies may cover a wide range of patients and/or interventions and/or outcomes and the justification for then combining these differing groups comes into question. The importance of a clear statement of inclusion criteria is that studies should be selected on the basis of these criteria (that is, any study that matches these criteria is included) rather than selecting the study on the basis of the results.  How were the data abstracted? Data are usually obtained by two individuals working separately (independently) and then compared later. Often the data are not available within the paper and there may be a statement about obtaining the raw data from the authors.  Is there a description of the quality of each trial? A table of the studies meeting the inclusion criteria should include some data on the validity checks itemized in the section on randomized control trials.4  Were the results consistent from study to study? Although we might expect some variation from study to study, we would be concerned if some trials confidently concluded a beneficial effect of the therapy, while others confidently concluded harmful or no effects. Unless this heterogeneity can be explained to your satisfaction (such as by differences in patients, dosage or duration of treatment), this should lead you to be very cautious about believing any overall conclusion about efficacy from the review.



which might lead you to suggest it does not make sense to combine the data.



Statistical heterogeneity exists when there is greater variation between the results than is likely due to chance. The chi-square statistic on which they are based has on average a value equal to its degrees of freedom. Values much larger than the degrees of freedom suggest a smaller p value and thus significant statistical heterogeneity. This means greater variation exists between the studies than is likely by chance alone. One might then conclude that the studies are so different that it makes no sense to combine them. Since this test of heterogeneity is described as having a low power, a nonsignificant test cannot be interpreted as evidence of homogeneity. The aim of the authors should be to clearly describe the influences of specific clinical differences between the studies, rather than purely relying on an overall statistical test of heterogeneity. Clinical and statistical heterogeneity need to be distinguished. Clinical heterogeneity refers to differences between settings, patients, techniques and outcomes,



Odds are another way of describing risk. The odds of an event are the probability of it occurring compared to the probability of it not occurring. We cannot do a randomized control trial to assess whether gastrooesophageal reflux disease (GERD) may be associated with laryngeal neoplasm. However, we can do a case–control study looking at people with and without cancer.18 We do not know the incidence of laryngeal cancer in the population from this study as they took cases with cancer and compared them to controls, but we can work out the odds (Table 52.5). For GERD, the odds of having cancer is 731/1315 = 0.555. This clearly is not the risk in the general population! The odds of having laryngeal cancer in non-GERD patients was 0.237. The odds ratio was 0.555/0.237 = 2.3. So patients with GERD in this study had 2.3 times the odds of having laryngeal carcinoma. This was unadjusted for the other major risk factor of smoking. In large studies, odds ratios approximate risk ratio.



What are the results? Terms that you will probably come across when looking at systematic reviews include odds ratios, relative risks, weighted mean differences and fixed and random effects, amongst others. BINARY OR CONTINUOUS DATA



The type of data will dictate what you see when you look at a meta-analysis:  binary data (for example, an event rate: something that happens or it does not, such as myocardial infarction, stroke, improved/not improved) is usually combined using odds ratios;  continuous data (for example, numbers of days, peak expiratory flow rate) are combined using differences in mean values for treatment and control groups (weighted mean differences (WMD)) when units of measurement are the same (for example, all using the same anxiety scale), or standardized mean differences when units of measurement differ (for example, using a variety of anxiety scales, where one numerical value could mean very different things, depending on the scale used). Here, the difference in means is divided by the pooled standard deviation. Thus when you look at a meta-analysis, you will see odds ratios or relative risks and/or weighted mean differences or standardized mean differences, depending on the outcomes measures that have been used in the included studies. ODDS RATIO



664 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA vary to reflect the weight that particular study is given, with larger boxes representing higher weighting. Lower weighted studies are usually those with smaller samples and large confidence intervals. The overall or summary effect (from combining or pooling the studies) is usually depicted as a diamond. Since patients may respond differently to treatment, this pooling may need to be supplanted by summarizing evidence along multiple covariates of interest. Thus, although one overall number is appealing in its simplicity, it may oversimplify a more complex situation. An odds ratio of one on a linear scale means there is no difference between the experimental and control group. This value of one is the line down the middle of the figure (sometimes with the wording ‘favours treatment’ on one side and ‘favours control’ on the other). If the confidence interval does not cross the line, it means that there is a 95 percent chance that there is a true difference between the groups. That is, that there is a significant difference on that particular outcome between the intervention and control groups. If the confidence interval does cross this vertical line labelled ‘1’ it means that any difference in that outcome between the treatments could have occurred by chance (that is, there are no statistically significant differences). It may be that a very wide confidence interval is due to a very small sample size, and it would reduce with more people in the study. A 95 percent confidence interval covers likely results from a set of similar studies. It therefore provides a more realistic view of what will happen in practice than a point estimate, because it takes potential variability into account.



Table 52.5 Prevalence of gastrooesophageal reflux disease (GERD) in a case control study of patients with and without laryngeal cancer.



Laryngeal cancer No laryngeal cancer Total



GERD



No GERD



731 1315 2046



7497 31,597 39,094



Clearly, as in randomized controlled trials with risk reduction, unless we know the baseline risk it is impossible to put odds ratios into context. A number needed to treat can be derived from an odds ratio but remember that many assumptions have been made in that process and treat them with some caution. Many systematic reviews use odds ratios to represent differences between treatment and control groups. Relative risk and absolute risk reduction may also be used to express these sorts of differences, as we have already seen. WHAT DO ODDS RATIOS USED IN A META-ANALYSIS LOOK LIKE?



The square box or ‘blob’ on the Forrest Plot (which is sometimes known colloquially as a ‘blobogram’) (Figure 52.3) is the individual study effect (technically referred to as the point estimate), with its associated confidence interval as lines either side of that box (technically known as the interval estimate). Sometimes the size of the box may Review: Antibiotics for acute maxillary sinusitis Comparison: 06 Macrolide/cephalosporin versus Amoxicillin-clavulanate Outcome: 01 Clinical cure Study Mac/Ceph Amox-clav n/N n/N



Peto odds ratio 95% Cl



Weight (%)



Peto odds ratio 95% Cl



77/157



78/160



14.4



1.01 [0.65, 1.57]



Clement 1998



90/165



47/89



10.5



1.07 [0.64, 1.80]



Dubois 1993



85/246



86/251



20.5



1.01 [0.70, 1.47]



Gehanno 1996



86/145



77/139



12.6



1.17 [0.73, 1.88]



Klapan 1999



44/50



35/50



3.0



2.93 [1.12, 7.63]



Olmo 1994



24/25



14/22



1.3



7.74 [1.83, 32.67]



Pessey 1996



77/87



72/82



3.2



1.07 [0.42, 2.71]



Russel 1999



85/158



78/161



14.5



1.24 [0.80, 1.92]



Sterkers 1997



111/152



114/149



10.3



0.83 [0.49, 1.40]



Stevrer 2000



170/198



171/189



7.4



0.64 [0.35, 1.19]



Sydnor 1992



6/59



9/54



2.4



0.57 [0.19, 1.69]



100.0



1.06 [0.90, 1.25]



Camacho 1992



Total (95% Cl) 1442 1346 Total events: 855 (Mac/Ceph). 781 (Amox-clay) Test for heterogeneity chi-square = 17.00 df = 10 p = 0.07 I = 41.2% Test for overall effect Z = 0.71 p = 0.5 0.1 0.2 0.5 Favours Amox-clau



Figure 52.3



A Forrest plot. Redrawn from Ref. 19, with permission.



1



2 5 10 Favours Mac/ceph



Chapter 52 Critical appraisal skills



It is worth noting that a lack of evidence for an intervention is not the same as evidence of a lack of effect.



HOW PRECISE ARE THE RESULTS?



This is asking whether there are any confidence intervals used, so the reader can see the range of values around the effect size, where in one is 95 percent confidence that the ‘true’ effect would lie. The Cochrane handbook (1998), section 6.1, defines precision as ‘a measure of the likelihood of random errors. It is reflected in the confidence interval around the estimate of effect from each study y more precise results are given more weight’. Thus the more precise the results, the narrower (smaller) the confidence interval. NNTs can be calculated from the odds ratios (see Appendix). However, these should be treated with caution as they can oversimplify. WILL THE RESULTS HELP ME IN CARING FOR MY PATIENTS?



 Can the results be applied to my patient care? How similar are the patients you care for to those included in the review? The wider the type of patients included in the review, the more likely you are to feel the results may apply across a range of different patients. However, it is sometimes a difficult decision to know whether, for example, something that works on men will be equally effective on women, or something that works in younger people will work in the same way in older people. Subgroup analysis may be one way of addressing this concern, although there are concerns with this approach.  Were all clinically important outcomes considered? What is clinically important can depend on your perspective. Sometimes, whilst factors such as bone density may be included in a review to assess the effectiveness of certain orthopaedic procedures, it may be some other outcome, such as ability to get to the shops, that needs to be considered as this sort of outcome may be crucial from the patient’s perspective.  Are the benefits worth the harms and costs? This may cover a range of factors, such as benefit of early treatment following a positive cervical smear versus the harms of high anxiety. Or it may be the benefits of a chemotherapeutic agent versus its unpleasant side effects. An economic evaluation may be included. What is sometimes difficult using a set of criteria like this, is that you end up with a variety of no’s or can’t tell’s in answer to the questions you pose. It is sometimes difficult to decide when a review is ‘too bad’ to give you any confidence in the findings. Perhaps one pragmatic approach would be that when you have several ‘no’s’ or ‘can’t tell’s’ you are more wary about the validity and robustness of the review’s findings. What this does illustrate is that, like undertaking a systematic review, critically appraising these reviews is not an exact science,



] 665



but there are many subjective decisions along the way. Just like undertaking a systematic review, making explicit your decisions in the critical appraisal is therefore very important. Although a systematic review can provide you with high quality evidence, clinical experience and patient preferences are an important part of evidence-based medicine. It may be that even high quality evidence does not apply to a particular patient. Similarly, Cochrane reviews do not include recommendations, as reviewers cannot know the local situation or the patient, and this is the province of the health professional caring for the patient.



KEY POINTS Appraising diagnosis articles  Criteria for validity of diagnosis trials: was there a reference standard, was it appropriate and was it applied blind, independently and to all patients?  Appropriate spectrum of patients; clear and reproducible description of test, follow up, intention-to-treat.  Outcome measures.  Measures of the accuracy of diagnostic tests: sensitivity, specificity, positive predictive value negative predictive value, likelihood ratios.  Applying the results to your patient: pre- and post-test probability, SpPin and SnNout. Appraising therapy articles  Criteria for validity of therapy trials: randomization, blinding, placebo control, equal treatment, follow up, intention-to-treat, outcome measures.  Measures of the importance of the results of therapy trials: risk, event rates, NNTs. Appraising systematic reviews  Criteria for validity of systematic reviews: search for relevant trials, assessment of included trials’ quality, heterogeneity.  Measures of the importance of the results: risk, event rates, NNTs, odd ratios.



APPENDIX Levels of evidence Table 52.6 provides a breakdown of the types of research that answers questions on therapy, prognosis, diagnosis, differential diagnosis or prevalence and economic studies.



Table 52.6 Level



Levels of evidence. Therapy/prevention, aetiology/ harm



Prognosis



Diagnosis



Differential diagnosis/symptom prevalence study



Economic and decision analyses



1a



SR (with homogeneity)a of RCTs



SR (with homogeneity)a of level 1 economic studies



Individual RCT (with narrow Confidence interval)c



SR (with homogeneity)a of Level 1 diagnostic studies; CDRb with 1b studies from different clinical centres Validatingk cohort study with goodh reference standards; or CDRb tested within one clinical centre



SR (with homogeneity)a of prospective cohort studies



1b



SR (with homogeneity)a of inception cohort studies; CDRb validated in different populations Individual inception cohort study with 480% followup; CDRb validated in a single population



Prospective cohort study with good follow-upc



1c



All or noned



All or none case-series



Absolute SpPins and SnNoutsg



All or none case-series



2a



SR (with homogeneity)a of cohort studies



SR (with homogeneity)a of Level 42 diagnostic studies



SR (with homogeneity)a of 2b and better studies



2b



Individual cohort study (including low quality RCT; e.g., o80% follow-up)



SR (with homogeneity)a of either retrospective cohort studies or untreated control groups in RCTs Retrospective cohort study or follow-up of untreated control patients in an RCT; Derivation of CDRb or validated on split-samplef only



Analysis based on clinically sensible costs or alternatives; systematic review(s) of the evidence; and including multi-way sensitivity analyses Absolute better-value or worsevalue analysesj SR (with homogeneity)a of level 42 economic studies



Exploratoryk cohort study with goodi reference standards; CDRb after derivation, or validated only on splitsamplef or databases



Retrospective cohort study, or poor follow-up



2c



‘Outcomes’ research; Ecological studies SR (with homogeneity)a of case–control studies Individual case–control study



3a 3b



‘Outcomes’ research



Ecological studies SR (with homogeneity)a of 3b and better studies Nonconsecutive study; or without consistently applied reference standards



SR (with homogeneity)a of 3b and better studies Nonconsecutive cohort study, or very limited population



Analysis based on clinically sensible costs or alternatives; limited review(s) of the evidence, or single studies; and including multi-way sensitivity analyses Audit or outcomes research SR (with homogeneity)a of 3b and better studies Analysis based on limited alternatives or costs, poor quality estimates of data, but including sensitivity analyses incorporating clinically sensible variations



4



Case-series (and poor quality cohort and case–control studies)e Expert opinion without explicit critical appraisal, or based on physiology, bench research or ‘first principles’



5



Case-series (and poor quality prognostic cohort studies)l Expert opinion without explicit critical appraisal, or based on physiology, bench research or ‘first principles’



Case–control study, poor or nonindependent reference standard Expert opinion without explicit critical appraisal, or based on physiology, bench research or ‘first principles’



Case-series or superseded reference standards



Analysis with no sensitivity analysis



Expert opinion without explicit critical appraisal, or based on physiology, bench research or ‘first principles’



Expert opinion without explicit critical appraisal, or based on economic theory or ‘first principles’



The levels of evidence used throughout this book are as stated in the prelims on the How to use this book page. These are very similar to the five levels given in this table, except that Levels 3 and 4 of the system described in this table are equivalent to Level 3, and Level 5 (expert opinion) is equivalent to Level 4 in the system used in this book. Produced by Bob Phillips, Chris Ball, Dave Sackett, Doug Badenoch, Sharon Straus, Brian Haynes and Martin Dawes since November 1998. Users can add a minus sign ‘  ’ to denote the level of that fails to provide a conclusive answer because of:



 EITHER a single result with a wide confidence interval (such that, for example, an ARR in an RCT is not statistically significant but whose confidence intervals fail to exclude clinically important benefit or harm).



 OR a systematic review with troublesome (and statistically significant) heterogeneity.  Such evidence is inconclusive, and therefore can only generate Grade D recommendations. a



By homogeneity we mean a systematic review that is free of worrisome variations (heterogeneity) in the directions and degrees of results between individual studies. Not all systematic reviews with statistically significant heterogeneity need be worrisome, and not all worrisome heterogeneity need be statistically significant. As noted above, studies displaying worrisome heterogeneity should be tagged with a ‘–’ at the end of their designated level. b Clinical decision rule. (These are algorithms or scoring systems which lead to a prognostic estimation or a diagnostic category.) c Good follow-up in a differential diagnosis study is 480%, with adequate time for alternative diagnoses to emerge (e.g. one to six months acute, one to five years chronic). d Met when all patients died before the Rx became available, but some now survive on it; or when some patients died before the Rx became available, but none now die on it. e By poor quality cohort study we mean one that failed to clearly define comparison groups and/or failed to measure exposures and outcomes in the same (preferably blinded), objective way in both exposed and nonexposed individuals and/or failed to identify or appropriately control known confounders and/or failed to carry out a sufficiently long and complete follow-up of patients. By poor quality case–control study we mean one that failed to clearly define comparison groups and/or failed to measure exposures and outcomes in the same (preferably blinded), objective way in both cases and controls and/or failed to identify or appropriately control known confounders. f Split-sample validation is achieved by collecting all the information in a single tranche, then artificially dividing this into ‘derivation’ and ‘validation’ samples. g An ‘Absolute SpPin’ is a diagnostic finding whose Specificity is so high that a Positive result rules-in the diagnosis. An ‘Absolute SnNout’ is a diagnostic finding whose Sensitivity is so high that a Negative result rulesout the diagnosis. h Good, better, bad and worse refer to the comparisons between treatments in terms of their clinical risks and benefits. i Good reference standards are independent of the test, and applied blindly or objectively to applied to all patients. Poor reference standards are haphazardly applied, but still independent of the test. Use of a nonindependent reference standard (where the ‘test’ is included in the ‘reference’, or where the ‘testing’ affects the ‘reference’) implies a level 4 study. j Better-value treatments are clearly as good but cheaper, or better at the same or reduced cost. Worse-value treatments are as good and more expensive, or worse and the equally or more expensive. k Validating studies test the quality of a specific diagnostic test, based on prior evidence. An exploratory study collects information and trawls the data (e.g. using a regression analysis) to find which factors are ‘significant’. l By poor quality prognostic cohort study we mean one in which sampling was biased in favour of patients who already had the target outcome, or the measurement of outcomes was accomplished in o80% of study patients, or outcomes were determined in an unblinded, nonobjective way, or there was no correction for confounding factors.



668 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA Table 52.7



Grades of recommendation.



A Consistent level 1 studies B Consistent level 2 or 3 studies or extrapolations from level 1 studies C Level 4 studies or extrapolations from level 2 or 3 studies D Level 5 evidence or troublingly inconsistent or inconclusive studies of any level ‘Extrapolations’ are where data are used in a situation that has potentially clinically important differences than the original study situation.



For each type of research a level of evidence is shown. This helps the reader identify the likely validity of the research. It does not mean that all systematic reviews are ‘better’ than randomized controlled trials. It should be used more as a pointer for searching and then a guide to the likely level of evidence. If one has a therapeutic question then looking for a systematic review first is logical. If there is no review then one next looks for a randomized controlled trial, etc. If there is no trial evidence then consensus statements becomes the highest level of evidence answering that question. This just makes the grade of recommendation D rather than A if there was a systematic review. The grades of recommendations vary from country to country and organization to organization and should only be considered as a guide (Table 52.7).



Worksheets The worksheets shown in Figures 52.4, 52.5 and 52.6 can be used to appraise a randomized controlled trial, a diagnostic study or a systematic review. They are aide memoirs about the items to look for in a paper when appraising the validity, identifying the results and assessing the applicability. They can be used in journal clubs or in other settings when an important new paper is being discussed. The data from these checklists can be entered into software to make your own electronic database of critical appraisals and saved as web pages or text documents.



REFERENCES 1. Kraaijenhagen RA, Haverkamp D, Koopman MM, Prandoni P, Piovella F, Buller HR. Travel and risk of venous thrombosis. Lancet. 2000; 356: 1492–3. 2. Ioannidis JP, Haidich AB, Pappa M, Pantazis N, Kokori SI, Tektonidou MG et al. Comparison of evidence of treatment effects in randomized and nonrandomized studies. Journal of the American Medical Association. 2001; 286: 821–30. 3. Juni P, Altman DG, Egger M. Systematic reviews in health care: Assessing the quality of controlled clinical trials. British Medical Journal. 2001; 323: 42–6.



4. Moher D, Jadad AR, Nichol G, Penman M, Tugwell P, Walsh S. Assessing the quality of randomized controlled trials: an annotated bibliography of scales and checklists. Controlled Clinical Trials. 1995; 16: 62–73. 5. Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Onkologie. 2000; 23: 597–602. 6. Shea B, Moher D, Graham I, Pham B, Tugwell P. A comparison of the quality of Cochrane reviews and systematic reviews published in paper-based journals. Evaluation and the Health Professions. 2002; 25: 116–29. 7. Kubba H, MacAndie C, Botma M, Robison J, O’Donnell M, Robertson G et al. A prospective, single-blind, randomized controlled trial of antiseptic cream for recurrent epistaxis in childhood. Clinical Otolaryngology. 2001; 26: 465–8. 8. Adams D. Life, the universe, and everything. The hitchhiker’s guide to the galaxy, Vol. 3. Isbn: ISBN- 0-33026738-8. London: Millenium (Orion), 1994. 9. BMJ Publishing Group Limited. Clinical evidence. www.clinicalevidence.com/ 10. Bouma M, Dekker JH, van Eijk JT, Schellevis FG, Kriegsman DM, Heine RJ. Metabolic control and morbidity of type 2 diabetic patients in a general practice network. Family Practice. 1999; 16: 402–6. 11. Booth A, O’Rourke AJ, Ford NJ. Structuring the pre-search reference interview: a useful technique for handling clinical questions. Bulletin of the Medical Library Association. 2000; 88: 239–46. 12. Thomas KB. General practice consultations: is there any point in being positive? British Medical Journal (Clinical Research Ed.). 1987; 294: 1200–2. 13. Dagnelie CF, Bartelink ML, van der Graaf Y, Goessens W, de Melker RA. Towards a better diagnosis of throat infections (with group A beta-haemolytic streptococcus) in general practice. British Journal of General Practice. 1998; 48: 959–62. 14. Lijmer JG, Mol BW, Heisterkamp S, Bonsel JG, Prins MH, van der Meulen JH et al. Empirical evidence of designrelated bias in studies of diagnostic tests. Journal of the American Medical Association. 1999; 282: 1061–6. 15. Bender R. Calculating confidence intervals for the number needed to treat. Controlled Clinical Trials. 2001; 22: 102–10. 16. Mulrow CD. The medical review article: state of the science. Annals of Internal Medicine. 1987; 106: 485–8. 17. Oxman AD, Cook DJ, Guyatt GH. Users’ guides to the medical literature. VI. How to use an overview. EvidenceBased Medicine Working Group. Journal of the American Medical Association. 1994; 272: 1367–71. 18. El-Serag HB, Hepworth EJ, Lee P, Sonnenberg A. Gastroesophageal reflux disease is a risk factor for laryngeal and pharyngeal cancer. American Journal of Gastroenterology. 2001; 96: 2013–8. 19. Williams Jr. JW, Aguilar C, Cornell J, Chiquette E, Dolor RJ, Makela M et al. Antibiotics for acute maxillary sinusitis. Cochrane Database of Systematic Reviews. 2003; 2: CD000243.



Chapter 52 Critical appraisal skills



] 669



The answer: Are the results of this single preventive or therapeutic trial valid? PICO QUESTION Search terms used Databases searched to find this article Citation of article Why did you select this article(s)? Did the authors answer the question? 1. What were the characteristics of the patients? 2. Were the groups similar at the start of the trial? 3. Aside from the experimental treatment, were the groups treated equally? 4. What was the treatment and what was it compared against (placebo?) 5a. Was the assignment of patients to treatments randomized? 5b. and was the randomization list concealed 6a. Were all patients who entered the trial followed up at its conclusion? 6b. and were they analyzed in the groups to which they were randomized? (Intention to treat) 7. Were patients and clinicians kept ‘blind’ to which treatment was being received? 8. Was the length of study appropriate? Are the valid results of this randomized trial important? YOUR CALCULATIONS:



Control group rate (CER)



Experimental Group Rate (EER)



Absolute risk reduction (ARR)



Number needed to treat (NNT)



Relative risk reduction (RRR)



CER-EER



1/ARR



ARR/CER



95% Confidence Interval (CI) on an NNT = 1/(limits on the CI of its ARR) = +/−1.96



CER × (1−CER) No. of control pts.



+



EER × (1−EER)



=



No. of exp. pts.



Can you apply this valid, important evidence about a treatment in caring for your patient? Do these results apply to your patient? Is your patient so different from those in the trial that its results can not help you? How great would the potential benefit of therapy actually be for your individual patient? Method I: f



Risk of the outcome in your patient, relative to patients in the trial. expressed as a decimal: ______ NNT/F = _____/_____ = _________ (NNT for patients like yours)



Are your patient’s values and preferences satisfied by the regimen and its consequences? Do your patient and you have a clear assessment of their values and preferences? Are they met by this regimen and its consequences?



Figure 52.4 Randomized control trial worksheet.



670 ] PART 10 INTERPRETATION AND MANAGEMENT OF DATA The Answer:



Are the results of this diagnostic study valid? PICO QUESTION Search terms used Databases searched to find this article Citation of article Why did you select this article(s) Did the authors answer the question? 1. What were the characteristics of the patients? 2a. New diagnostic test description. Is it clear how the test was carried out? 2b. Is the test reproducible in your setting? 3a. Gold standard description. 3b. Was there an independent, blind comparison with a reference (‘gold’) standard of diagnosis? 4. Was the reference standard applied regardless of the diagnostic test result? 6. Was the diagnostic test evaluated in an appropriate spectrum of patients (like those in whom it would be used in practice)? YOUR CALCULATIONS:



Diagnostic test result



Target disorder Present



Absent



Positive



a



b



a+b



Negative



c



d



c+d



Totals



a+c



b + d1770



a+b+c+d



Totals



Sensitivity = a /(a + c) = Specificity = d/(b + d) = Likelihood ratio for a positive test result = LR+ = sens/(1−spec) = Likelihood ratio for a negative test result = LR− = (1−sens)/spec = Positive predictive value = a /(a + b) = Negative predictive value = d/(c + d) = Pretest probability (prevalence) = (a + c)/(a + b + c + d) = Pretest-odds = prevalence/(1−prevalence) = Post-test odds = Pretest odds x likelihood ratio = Post-test probability = post-test odds/(post-test odds + 1) = Can you apply this valid, important evidence about a diagnostic test in caring for your patient? Is the diagnostic test available, affordable, accurate and precise in your setting? Can you generate a clinically sensible estimate of your patient’s pretest probability (from practice data, from personal experience, from the report itself, or from clinical speculation) Will the resulting post-test probabilities affect your management and help your patient? (Could it move you across a test-treatment threshold?; Would your patient be a willing partner in carrying it out?) Would the consequences of the test help your patient? Additional notes:



Figure 52.5 Diagnosis worksheet.



Chapter 52 Critical appraisal skills



The Answer:



Are the results of this systematic review of therapy valid? PICO QUESTION Search terms used Databases searched to find this article Citation of article Why did you select this article(s) Did the authors answer the question? 1. What databases and other sources did the authors of this review search? 2. What were their eligibility criteria (inclusion and exclusion) for papers in this study? Do these seem appropriate? 3. Was their independent data extraction of the results by the reviewers (then compared later) 4. Is there a description of the quality of each trial included? 5. Were the results consistent from study to study? (Homogeneous) Are the valid results of this systematic review important? What was the odds ratio? What was the control group rate (PEER) Can you apply this valid, important evidence from a systematic review in caring for your patient? Do these results apply to your patient? Is your patient so different from those in the overview that its results can not help you? How great would the potential benefit of therapy actually be for your individual patient? To calculate the NNT for any OR and PEER NNT =



1 − (PEER * ( 1 − OR)) (1 − PEER) * PEER * (1 − OR)



Or use calculator from www.cebm.net Are your patient’s values and preferences satisfied by the regimen and its consequences? Do your patient and you have a clear assessment of their values and preferences? Are they met by this regimen and its consequences? Should you believe apparent qualitative differences in the efficacy of therapy in some subgroups of patients? Only if you can say ‘yes’ to all of the following: Do they really make biologic and clinical sense? Is the qualitative difference both clinically (beneficial for some but useless or harmful for others) and statistically significant? Was this difference hypothesized before the study began (rather than the product of dredging the data), and has it been confirmed in other, independent studies? Was this one of just a few subgroup analyses carried out in this study? Additional notes:



Figure 52.6 Systematic review (of therapy) worksheet.



] 671



This page intentionally left blank



PART



11



RECENT ADVANCES IN TECHNOLOGY EDITED BY MARTIN J BURTON



53 Functional magnetic resonance imaging: Principles and illustrative applications for otolaryngology Paul M Matthews



675



54 Positron emission tomography and integrated PET/computed tomography Wai Lup Wong



684



55 Image-guided surgery, 3D planning and reconstruction Ghassan Alusi and Michael Gleeson



701



56 Ultrasound in ear, nose and throat practice Keshthra Satchithananda and Paul S Sidhu



711



57 Interventional techniques James V Byrne



731



58 Laser principles in otolaryngology, head and neck surgery Brian JG Bingham



742



59 Electrophysiology and monitoring Patrick R Axon and David M Baguley



748



60 Optical coherence tomography Mariah Hahn and Brett E Bouma



755



61 Contact endoscopy Mario Andrea and Oscar Dias



762



This page intentionally left blank



53 Functional magnetic resonance imaging: Principles and illustrative applications for otolaryngology PAUL M MATTHEWS



Introduction The physiological basis for blood oxygenation leveldependent functional magnetic resonance imaging Practical details Illustrative applications of functional magnetic resonance imaging



675 677 678



Conclusions Key points Best clinical practice Current deficiencies and suggestions for future research References



681 681 682 682 682



679



SEARCH STRATEGY The data in this chapter are supported by a PubMed search using the key words functional brain imaging and hearing, restricted to humans.



INTRODUCTION Functional magnetic resonance imaging (fMRI) has become one of the most versatile tools in cognitive neuroscience. Using blood oxygenation level-dependent (BOLD) fMRI, grey matter regions active during sensory, motor and cognitive activity can be defined based on the accompanying local increase in blood flow and oxygenation. Increasing numbers of applications to basic questions in auditory physiology have been reported. Now, there is the exciting potential for clinical use of the methodology.



An overview of functional brain imaging methods Functional brain imaging includes any of a broad range of techniques by which physiological changes accompanying brain activity can be followed. For example, neuronal depolarization can be monitored using



electroencephalography (EEG) or magnetoencephalography (MEG). The energy-requiring metabolic changes in neurons and glia that accompany neurotransmitter release and reuptake and receptor occupancy can be measured using positron emission tomography (PET). Currently, the most commonly used functional brain imaging technique is fMRI. Functional magnetic resonance imaging is sensitive to the increase in local blood flow with neuronal activation. These changes in blood flow can be measured directly using perfusion MRI.1 However, as the increase in total oxygen delivery exceeds the increase in oxygen utilization, another fMRI method, BOLD fMRI, can also be used. In BOLD fMRI the imaging contrast arises as a consequence of the higher ratio of oxyto deoxy-haemoglobin in local capillaries and draining venules with brain activity.2 While the neuronal events thus are only indirectly observed, the method is widely exploited because of its sensitivity, relatively low cost and ease of examination and the lack of recognized risks. The different functional brain imaging methods have different temporal and spatial resolutions. Methods that



676 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



Figure 53.1 Different forms of contrast can be generated using MRI, identifying structural and functional differences across normal or pathological brains. (a) T1-weighted contrast in the brain of a patient with multiple sclerosis, showing brain tissue as brighter signal and cerebrospinal fluid (CSF) (or areas of tissue matrix destruction) darker. (b) T2-weighted contrast in the brain of a patient with multiple sclerosis. Here CSF is bright as are regions of inflammation with its associated oedema. (c) A contrast-enhanced T1-weighted image from the brain of a patient with multiple sclerosis. A region of acute inflammation with loss of an intact blood-brain barrier allows leak of the gadolinium-DTPA contrast agent into the brain parenchyma, enhancing the signal intensity locally (arrow). (d) BOLD fMRI contrast defines the hand region of the primary sensorimotor cortex during a simple hand movement. The colours represent the statistical significance of signal enhancement during the active vs. the rest phases.



map the transient brain electrical dipoles generated by neuronal depolarization (e.g. EEG) or the associated magnetic dipoles (MEG) define the underlying cortical neuronal events in real time (10–100 msec), but offer relatively poor spatial resolution (on the order of a centimetre or more). PET methods require relatively long acquisition times (tens of seconds) and give moderately high spatial resolution (5–15 mm). Functional magnetic resonance imaging provides information on the increases in blood flow accompanying neuronal activation with potentially rather higher spatial resolution (approximately 1–10 mm), but has a temporal resolution limited by the haemodynamic changes, which are made slower (several seconds) than the neuronal response (milliseconds).



Generation of image contrast in functional magnetic resonance imaging Let us consider the problem of developing contrast between different tissues in greater detail. How is a difference in signal intensity (image contrast) generated between regions of brain that are ‘active’ and those that are at ‘rest’ in fMRI? Once the nuclear spins in the tissue are excited by a radiofrequency pulse, they can return (or ‘relax’) to a low energy state by emission of energy. It is this energy, the frequency of which is made to be spatially variable, that is detected in MRI. The rate at which the emissions decay is determined by interactions with the environment. As local (i.e. on a molecular and cellular level) environments for water molecules vary across a heterogeneous tissue such as the brain, this so-called ‘relaxation rate’ also varies. The MRI image can be made sensitive to this by



varying the interpulse delay or ‘TR’ in a pulse sequence. A ‘T1-weighted’ clinical scan generates image contrast using this principle (Figure 53.1). However, while useful for differentiating grey from white matter, this alone does not distinguish ‘active’ from ‘resting’ brain. Some functional information can be introduced by scanning after injection of an MRI contrast agent (e.g. gadolinium DTPA), which can define vessels or regions around vessels that become more permeable to small molecules with inflammation, infarction or tumours by changing the local with relaxation time. However, while the first functional imaging was performed in this way,1 it remains a cumbersome technique for most applications. In tissue, huge numbers of spins are being observed simultaneously (in the brain for example, there are more than 4  1019 water protons/mm3). Even in a small tissue region, each of the nuclei has a different relation to other molecules and experiences a slightly different, varying local magnetic field. These shifting fields allow an exchange of energy between the nuclei spins. This leads to a faster loss of the net signal from all of the nuclei together. This is described by the so-called ‘spin–spin’ or T2 relaxation time. The T2 is an intrinsic property of nuclei in a particular chemical environment that accounts for the contrast (generated by changing the imaging parameter ‘TE’) in a ‘T2-weighted’ MRI image. While T2 differences are useful for defining differences in, for example, water content between different brain regions (Figure 53.1), they also tell us nothing directly about brain function. However, a related phenomenon provides the answer. The rate of decay of signal is faster if there are local (i.e. on a molecular scale) magnetic field variations that the molecules can diffuse through over the time course of a single TE. As molecules move through different



Chapter 53 Functional magnetic resonance imaging: Principles and illustrative applications for otolaryngology



magnetic fields, their resonance frequencies change (although only slightly), lowering the coherence of the nuclear spins. This leads to more rapid decay of the net signal. In the presence of local magnetic field inhomogeneities the rate of signal decay is expressed by the T2 relaxation time. In regions of rapidly changing local magnetic fields (e.g. in tissue adjacent to a blood vessel filled predominantly with paramagnetic deoxyhaemoglobin), the T2 can be substantially shorter than the T2. This mechanism determines the dramatic signal contrast provided by blood breakdown products in tissue for clinical MRI. It is also the mechanism underlying generation of BOLD fMRI images (Figure 53.1).



THE PHYSIOLOGICAL BASIS FOR BLOOD OXYGENATION LEVEL-DEPENDENT FUNCTIONAL MAGNETIC RESONANCE IMAGING Mechanisms of haemodynamic change The locally increased blood flow with brain activation is a consequence of increased energy utilization at the synapse.3, 4 Several processes contribute to this, including increased energy utilization in adjacent astrocytes with the uptake of the excitatory neurotransmitter, glutamate, depolarization of the post-synaptic membrane and subsequent spike discharges.5 However, there is not a simple relationship between increased energy utilization and increased blood flow or BOLD fMRI contrast. ‘Metabolic’ functional imaging techniques, such as BOLD fMRI, can identify activation-related changes in grey matter (where the synapses are found), but not in the white matter. It is important to emphasize that results should be interpreted primarily in terms of synaptic activity, rather than neuronal activity directly.6 Because of the lack of a fixed relationship between synaptic energy utilization and neuronal firing, as well as the large number of potentially interacting processes generating the signal, caution should be taken in trying to relate the magnitude of the BOLD response quantitatively to neuronal discharge rate.



Coupling of haemodynamic changes to neuronal activation Multiple mechanisms contribute to the control of blood flow to the brain and the precise way in which blood flow is coupled to increased metabolic demands with synaptic activity. Nitric oxide (primarily from neuronal nitric oxide synthase) may be the most important chemical signal responsible for local increases in perfusion with neuronal activation and the cerebral vasodilatory response to hypercapnia, but a variety of factors would also



] 677



contribute, including K 1 release with neuronal depolarization and H 1 and adenosine release. Locally generated eicosanoids (e.g. prostacyclin)7 and circulating factors (e.g. norepinephrine, serotonin) can modulate blood flow regulation, as can some drugs in common use (e.g. theophylline or scopolamine, inhibit the haemodynamic response to neuronal activation).8, 9 Neuronal-haemodynamic coupling can change with age,10 although the changes after maturity are modest.11



Perfusion functional magnetic resonance imaging The most direct approach to measuring functional activation with MRI would be to directly define the changes in perfusion (or the associated small increase in local increase in blood volume) with the accompanying neuronal activation. The first functional MRI studies did just this using a paramagnetic MRI contrast agent to follow changes in local blood volume with increased activation.1 While technically challenging, a more attractive approach to measuring functional activation changes from direct perfusion measurements is to use noninvasive arterial ‘spin-tagging’ (Figure 53.2).12, 13, 14 Arterial spintagging methods use radiofrequency pulses from the MRI system itself to transiently ‘label’ flowing blood without the need for an exogenous contrast agent. Unfortunately, the sensitivity of this approach is low, which limits its applicability at present.



Figure 53.2 Both perfusion MRI and BOLD fMRI can be used to define regions active during brain activity. In this study, the subject alternately moved the right hand in flexion-extension or kept it at rest. The hemodynamic response with brain activation in the left primary sensorimotor cortex is shown using: (a) BOLD fMRI; and (b) arterial spin tagging (ASL) perfusion imaging. Similar regions are active, although the perfusion method is less sensitive (giving a smaller activation region. The green target shows the mean and standard deviation (in the plane of the image) of activation centres of gravity with respect to the large sulcal draining vein (blue) identified by MR venography in the same experiment. Images were prepared by Dr. T. Tjandra, FMRIB Centre, Oxford.



678 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



How is blood oxygenation level-dependent contrast generated? The most robust and widely used fMRI method is BOLD fMRI. In BOLD fMRI the image signal contrast arises from changes in T2 generated with differences in local ‘magnetic susceptibility’, an index of the extent to which an applied magnetic field is distorted as it interacts with a material, as blood oxygenation changes. Normal blood can be considered simply as a concentrated solution of haemoglobin (10–15 g haemoglobin/100 cm3). When bound to oxygen, haemoglobin is diamagnetic, while deoxygenated haemoglobin is paramagnetic. Magnetic flux is reduced in diamagnetic materials, i.e. the applied magnetic field is repelled. Paramagnetic materials, in contrast, have an increased magnetic flux, i.e. the applied magnetic field is attracted into the material. A change in haemoglobin oxygenation therefore leads to changes in the local distortions of a magnetic field applied to it. ‘Gradient echo’ MR images of a cat brain show signal loss around blood vessels when the animal is made hypoxic, for example.2 More local changes in blood oxygenation, from the increase in the ratio of oxy- to deoxy-haemoglobin specifically in a region of increased blood flow with brain activation, give rise to small local increases in signal intensity.



PRACTICAL DETAILS Design of functional magnetic resonance imaging studies Methods such as PET provide an absolute measure of tissue metabolism. In contrast, BOLD fMRI can be used only for determining relative signal intensity changes associated with different cognitive states during a single imaging session. The most time-efficient approach for comparing brain responses in different states during the imaging experiment is the ‘block’ design.15 This design uses relatively long, alternating periods (e.g. 30 s), during each of which a discrete cognitive state is maintained. In the simplest form, there may only be two such states, a ‘rest’ and an ‘active’ state (although ‘rest’ is defined only with respect to the specific activity being considered). These are alternated through the experiment in order to ensure that signal variations from small changes in scanner sensitivity, patient movement or changes in attention have a similar impact on the signal responses associated with each of the different states. However, it can become difficult to control a cognitive state precisely for the relatively long periods of each block. A ‘rest’ state is rarely a true rest, as mental events are difficult to constrain in a subject who is not engaged in a specific task. Some types of stimuli (particularly sensory stimuli) may show rapid habituation (a particular



problem in studies of olfaction, for example).16 Some cognitive tasks simply may not be amenable to a block design. For example, an ‘oddball’ paradigm (in which the reaction to an unexpected stimulus is probed) cannot be adapted to a classical ‘block’ design directly. An alternative approach is so-called single-event fMRI,17 in which data are acquired while discrete stimuli or responses are given. Results from many trials are then averaged together to give a measurable response. This is a potentially powerful approach, as it allows considerable flexibility for study of responses to individual or periodically presented stimuli, but it demands relative longer acquisition times than the block design. A related approach is to present stimuli in a periodic fashion and then to map responses in terms of their phase relative to that of the stimulus presentation. This was used first for study of the visual cortex, which is organized functionally in multiple retinotopic maps arranged according to increasing eccentricity and polar angle. Because it is effectively a difference mapping procedure,18 the edge contrast between adjacent functional regions is emphasized.19 Tonotopy in the auditory system has been mapped using a similar strategy.20



Analysis of functional magnetic resonance imaging studies The raw BOLD fMRI data can be acquired over periods as short as a few minutes. If the approach to analysis is well defined, near ‘real-time’ viewing of final statistical maps of activation is possible (although in research applications exploratory analyses may take substantial periods of time!). The basic problem in analysis of functional imaging experiments is to identify voxels that show signal changes varying with the changing brain states of interest across the serially acquired images. This is a challenging problem for fMRI data because the signal changes are small (leading to potential false-negatives or type II errors) and the number of voxels simultaneously interrogated across the imaged volume is very large (giving potential false-positives or type I errors). One of the potentially most significant artifacts for fMRI that distinguishes it from other functional imaging techniques is its extreme susceptibility to motion from movements either of the whole head or even the brain alone (e.g. with the respiratory or cardiac cycles).21 Effects of small head movements can be reduced by mathematically realigning the brain volumes after the data are first acquired, but larger movements can degrade image quality irrecoverably. This may be a particular confound in studies of patients who may be less comfortable in the magnet environment. After optimization of the signal-to-noise by digital filtering, analysis of the time course by comparison of the signal of both models of the ‘baseline’ and ‘active’ states must be performed. There are many valid ways of



Chapter 53 Functional magnetic resonance imaging: Principles and illustrative applications for otolaryngology



] 679



Figure 53.3 New forms of representation of brain activation patterns can make the spatial relationships easier to appreciate. All three images here reflect visual cortex activation associated with watching a flashing checkboard. A ‘model-free’ analysis using independent components analysis was performed, which distinguishes between the primary visual cortex (orange) and the motion-sensitive, MT cortex (blue). (a) Inflated brain representation; (b) flattened brain representation; (c) spherical brain representation. Images were prepared by Drs S. Smith and C. Beckmann (FMRIB Centre, Oxford, UK) using MELODIC (www.fmrib.ox.ac.uk/fsl) for analysis and the Freesurfer (Athinoula Martinos Imaging Centre, Massachusetts General Hospital NMR Centre, Boston) for representation.



performing statistical comparisons between signals in images acquired during different brain states. A simple approach is to generate a map of the t-statistic for changes on a voxel-by-voxel basis and use this to identify voxels with significance levels exceeding a chosen significance threshold (e.g. t43, which might correspond in a particular case to po0.01). A related approach is to correlate the time course of signal change in each voxel with a model time course based on the expected response, which again can be used to generate a t-statistic change. Valid statistical inference from the image time course data is complicated by the fact that large numbers of voxels are being assessed simultaneously for changes. The significance level must be corrected for the number of truly independent comparisons that are being made. For example, if 10,000 independent voxels are being tested and a threshold for significance of change in each of the individual voxels of p = 0.05 is chosen, then 500 (= 0.05  10,000) would be ‘active’ by chance alone! The significance threshold therefore must be made more stringent in proportion to the number of independent comparisons. Fortunately, because observations of immediately neighbouring regions are not really independent, the correction factor is reduced. There are now a number of software packages produced by academic centres that include a full set of tools for analysis of fMRI data (e.g. FEAT (www.fmri.ox.ac.uk/fsl), SPM (www.fil.ion.ucl.ac.uk/spm/) or AFNI (www.cc.nih. gov/cip/ip_packages/ip_packages.html#afni)).



common to many members of the group. This can be used to define the ‘normal’ pattern of behaviour as a basis for detecting pathological patterns in patients, for example. Averaging between brains is possible with the use of a ‘common’ brain space, a canonical brain shape that can be related easily to the variable anatomy of individuals. The most well-known common brain space was developed by Talairach as a simple geometric parcellation of a single brain according to major anatomical landmarks.22 This has been extended into a probabilistic framework.23 More abstract representations of spatial variations in activation across the brain are possible using, for example, flattened representations, produced by computationally unfolding the cortex (Figure 53.3).24, 25, 26 Other representations using standard geometric shapes for which coordinate systems (which allow relative positions of activations to be described precisely) are derived readily and that demand less distortion (e.g. a sphere) are also possible.27



Neuroanatomical interpretation of functional data for groups of subjects: Use of a common brain space



Imaging the normal auditory system



A strength of fMRI is that it is sensitive enough to allow observations to be made even in individual subjects. Nonetheless, averaging results from studies of many subjects is often useful for identifying areas of activation



ILLUSTRATIVE APPLICATIONS OF FUNCTIONAL MAGNETIC RESONANCE IMAGING Applications of fMRI are growing rapidly and potentially have relevance to several aspects of otolaryngology. Here, applications to selected issues in hearing research are described to illustrate the range of problems that can be studied using the technique.



Functional magnetic resonance imaging has already contributed important information concerning the organization of the human auditory cortex, directly demonstrating parallels with the functional organization of the auditory brain in animals that previously were only assumed.28 For example, tonotopic representations in the



680 ] PART 11 RECENT ADVANCES IN TECHNOLOGY human primary auditory cortex comparable to those identified electrophysiologically in animals have been demonstrated using fMRI.29 Differences between humans and other species have also begun to be defined. Animal studies, for example, have not demonstrated evidence for differences in cortical organization reflecting the sound level of the stimulus, while evidence for this level of representation of ‘sound space’ has been found in humans.28 One of the most striking areas of difference between humans and animals lies in the relative hemispheric specialization for auditory signal processing. Microelectrode studies in animals have consistently demonstrated a simple contralateral dominance for processing, similar with monoaural stimulation to either of the two ears. In contrast, recent fMRI studies have provided evidence for left hemispheric dominance for this process in humans even for pure tones (Devlin et al., unpublished observations) (Figure 53.4a and b). This lateralization for pure tone processing may be central to understanding the specialization of the left hemisphere for language. In auditory language processing, a left hemisphere dominance has long been recognized. This can be made clear with functional imaging by appropriate choice of contrasts of auditory tasks during the imaging study. Processing associated specifically with language intelligibility is strongly left-lateralized, for example, but this is made apparent only when an unintelligible control (‘rest’) stimulus of identical auditory complexity is contrasted (Figure 53.4).30 Development of function in the brain may also be studied with these techniques. While the demands for prolonged stationary periods in the unusual environment of the scanner probably preclude direct application of the technique to very young children while awake, responses in primary sensory systems, including audition, can be studied even under conditions of anaesthesia.31



Figure 53.4 Functional imaging of audition. A ‘sparsesampling’ protocol was used for analysis of auditory and language cortex activation by fMRI. Intelligible vs. unintelligible speech matched for acoustic characteristics were used. Both images show the left hemisphere for a group average activation map. (a) Activation primarily in the region of Heschl’s gyrus with acoustic stimulation. (b) Activation primarily in more posterior, language-related regions with presentation of intelligible speech. Images courtesy of Dr. C. Narain and Dr. S. Scott (FMRIB Centre, Oxford).



As described earlier in this chapter, one of the general problems with the use of fMRI to study auditory processing is the noise of the imaging itself. Use of paradigms such as ‘sparse sampling’ (in which the stimuli are presented during quiet periods, with only intermittent sampling) avoid this problem, but are highly timeinefficient. An intriguing alternative approach that can be used with deaf patients is electrical stimulation via cochlear implants. This strategy poses technical challenges because of magnetic susceptibility image artifacts from the metal and safety concerns regarding the potential for current generation in the wires with gradient switching. However, with proper shielding of the stimulation cables, any induced currents can be reduced to levels below the acoustic perceptual threshold for cochlear implant subjects. In a recent report, stimulation of different intracochlear electrodes was associated with distinct cortical activations, consistent with the known tonotopical organization of the auditory cortex.32



Defining functional relations between brain regions An important aspect of fMRI studies is the ability to define potentially interacting regions across broad regions of the brain. This ability to discriminate anatomically specific processing regions sets it apart from electrophysiological methods such as evoked potential studies. Functional magnetic resonance imaging has recently been used to probe the ways in which the brain processes identity and spatial information in sound.33 In one study, participants had to match two sounds either for pitch or spatial location. Relative to location, pitch processing generated greater activation in the auditory cortex and the inferior frontal gyrus. Conversely, the task of recognizing location generated greater activation in posterior temporal cortex, parietal cortex and the superior frontal sulcus. The specialized auditory processing streams in the human brain implied by these differences are analogous to the ‘what’ and ‘where’ segregation of visual information processing, suggesting common modes of information organization. Another general area of interest catalyzed by functional brain imaging has been understanding the functions of the cerebellum, classically thought of simply as a motor centre. Data now show that the cerebellum is involved in a broad range of tasks (perceptual, as well as motor) that demand representations of temporal information. Temporal sequencing plays an important role in auditory perception. Consistent with this hypothesis, bilateral injury to the cerebellar hemispheres can compromise the ability to make distinctions between words whose differences are determined by the closure time, a purely temporal measure of the lengths of pauses within words. To test for cerebellar involvement with this function, subjects were studied by fMRI to contrast patterns of



Chapter 53 Functional magnetic resonance imaging: Principles and illustrative applications for otolaryngology



brain activation during word discriminations made possible by variations in relative timing or the sound of central consonants.34 The timing categorization selectively activated the right cerebellar cortex (Crus I) and the prefrontal lobe (anterior to Broca’s area), while the sound categorization preferentially activated the supratemporal plane of the dominant hemisphere. This suggests that representation of the temporal structure of speech sound sequences is the crucial aspect of cerebellar involvement in cognitive tasks involving auditory comprehension.



Imaging plasticity Short-term plasticity or ‘learning’ has become a particularly fruitful area of fMRI-based research. One example of this is in responses to an ‘oddball’ tone.35 Brain responses elicited to pure tones of 950 Hz (standard) with occasional deviant tones of 952, 954 and 958 Hz were measured before and one week after subjects had been trained in these frequency discriminations. Frequency discrimination improved after the training session for some subjects (T 1 ), but not for others (T  ). Haemodynamic responses in the auditory cortex comprising the planum temporale, planum polare and sulcus temporalis superior significantly decreased during training, but only for the T 1 group. The results suggest that there is a short-term, plastic reorganization of cortical representations specifically for the trained frequencies. Auditory reafferentation by cochlear implants (CI) offers a unique opportunity to study longer-term auditory plastic changes taking place at up to the supraor polymodal organizational level.36 One functional magnetic resonance imaging study investigated the impact of early auditory deprivation and use of a visuospatial language (American sign language (ASL)) on the organization of neural systems important in visual motion processing by comparing hearing controls with deaf and hearing native signers.37 Participants monitored images of moving dots under different conditions of spatial and featural attention. Recruitment of the motionselective brain visual-processing area, known as MT in hearing controls, was greater when attention was directed centrally and when the task was to detect motion features, confirming previous reports that the motion network is selectively modulated by different aspects of attention. More importantly, marked differences in the recruitment of motion-related areas as a function of early experience were observed. First, the lateralization of MT was found to shift toward the left hemisphere in early signers, suggesting that early exposure to ASL is associated with a greater reliance on the left MT. Second, whereas the two hearing populations displayed more MT activation under central than peripheral attention, the opposite pattern was observed in deaf signers, indicating enhanced recruitment of MT during peripheral attention after early deafness. Third, deaf signers (but not subjects with



] 681



normal hearing), displayed increased activation of the posterior parietal cortex, supporting the view that parietal functions are modified after early auditory deprivation. Finally, attention to motion lead to enhanced recruitment of the posterior superior temporal sulcus only in deaf signers, establishing for the first time in humans that this polymodal area is modified after early sensory deprivation.



Imaging drug responses The ability to monitor changing responses of subjects over several examinations makes fMRI potentially attractive for studying the effects of drugs. Cholinergic blockade is known to attenuate conditioning-related neuronal responses in the human auditory cortex. One recent study investigated the effects of cholinergic enhancement on experience-dependent responses in the auditory cortex.38 The cholinesterase inhibitor physostigmine, or a placebo control, were continuously infused into healthy young volunteers, during a conditioning protocol conducted while fMRI data were being acquired. Volunteers were presented with two tones, one of which (CS 1 ) was paired with an electrical shock, while the other was always presented without the shock (CS  ). Conditioning-related fMRI activations were found in the left auditory cortex during placebo administration, but not with physostigmine. Analysis showed that this absence of conditioning-related activation with physostigmine administration was due to enhanced responses to CS  with physostigmine relative to placebo. The study report concluded that increases in cholinergic activity in the auditory cortex enhances processing of behaviourally irrelevant stimuli and that attenuates differential conditioning-related cortical activations.



CONCLUSIONS Functional magnetic resonance imaging is already a major tool for cognitive neuroscience research. Studies can be performed on single subjects and in all current high-field MRI systems. With a growing appreciation for interpretation of the data, fMRI can also be expected to provide new clinical strategies for establishing diagnosis and prognosis and in treatment monitoring.



KEY POINTS  Functional magnetic resonance imaging is currently the most commonly used functional brain imaging technique.  With blood oxygenation level dependent fMRI, image contrast arises as a consequence



682 ] PART 11 RECENT ADVANCES IN TECHNOLOGY







 







of the higher ratio of oxy- to deoxyhaemoglobin in local capillaries and draining venules with brain activity. Functional magnetic resonance imaging thus has a spatial (approximately 1–10 mm) and temporal resolution determined by the haemodynamic changes accompanying neuronal activity. Results should be interpreted primarily in terms of synaptic activity, rather than neuronal firing directly. Coupling between the neuronal and haemodynamic changes is complex and BOLD fMRI provides a measure of only relative synaptic activity. Applications of fMRI are growing rapidly and potentially have relevance to several aspects of otolaryngology.



accurate models for the haemodynamic response. The better this is defined, the greater the potential sensitivity of the method. Several research groups are working to develop approaches that make no assumptions about the timing or nature of the haemodynamic response. Successful work in this area would allow brain activity to be defined independent of information regarding stimuli or behaviour. These developments and rapidly growing experience are fuelling an increasingly confident application of the technique to complex clinical problems, such as those related to understanding decision-making, motivation and emotion, as well as more sensitively studying changes in brain functions over time.



REFERENCES



Best clinical practice A rigorous evaluation of fMRI as a tool for clinical decision-making has not been performed. While encouraging preliminary evidence has been presented in applications such as localization of functional tissue and lateralization of language and memory activity in preoperative surgical planning, there is little information with which to judge the validity and informativeness of data outside specific research contexts. However, the techniques are promising in several areas and active, clinically based research to test the possible utility in decision-making should be encouraged. At this point, fMRI may be best regarded as experimental in a clinical context and the data derived from it should not be used independently to provide primary information for clinical decision-making.



Current deficiencies and suggestions for future research There are several areas of current research that should provide the new clinical imaging strategies of the future. Some of these can be highlighted here to give a vision of what is to come. MRI is highly versatile and can be used to generate contrast relevant to brain function based on phenomenon other than perfusion or BOLD. It may be possible to detect local distortions in MRI signal induced by the transient magnetic fields generated with neuronal depolarization, for example. New approaches to analysis of fMRI are addressing the problem of developing more



1. Belliveau JW, Kennedy Jr. DN, McKinstry RC, Buchbinder BR, Weisskoff RM, Cohen MS et al. Functional mapping of the human visual cortex by magnetic resonance imaging. Science. 1991; 254: 716–9. 2. Ogawa S, Lee TM, Kay AR, Tank DW. Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proceedings of the National Academy of Sciences of the United States of America. 1990; 87: 9868–72. 3. Duncan GE, Stumpf WE. Brain activity patterns: assessment by high resolution autoradiographic imaging of radiolabeled 2-deoxyglucose and glucose uptake. Progress in Neurobiology. 1991; 37: 365–82. 4. Duncan GE, Stumpf WE, Pilgrim C. Cerebral metabolic mapping at the cellular level with dry-mount autoradiography of [3H]2-deoxyglucose. Brain Research. 1987; 401: 43–9. 5. Attwell D, Laughlin SB. An energy budget for signaling in the grey matter of the brain. Journal of Cerebral Blood Flow and Metabolism. 2001; 21: 1133–45. 6. Logothetis NK, Pauls J, Augath M, Trinath T, Oeltermann A. Neurophysiological investigation of the basis of the fMRI signal. Nature. 2001; 412: 150–7. 7. Kuschinsky W. Coupling of function metabolism and blood flow in the brain. Neurosurgical Review. 1991; 14: 163–8. 8. Dirnagl U, Niwa K, Lindauer U, Villringer A. Coupling of cerebral blood flow to neuronal activation: role of adenosine and nitric oxide. American Journal of Physiology. 1994; 267: H296–H301. 9. Ogawa S, Menon RS, Tank DW, Kim SG, Merkle H, Ellermann JM et al. Functional brain mapping by blood oxygenation level-dependent contrast magnetic resonance imaging. A comparison of signal characteristics with a biophysical model. Biophysical Journal. 1993; 64: 803–12. 10. Meek JH, Firbank M, Elwell CE, Atkinson J, Braddick O, Wyatt JS. Regional hemodynamic responses to visual



Chapter 53 Functional magnetic resonance imaging: Principles and illustrative applications for otolaryngology



11.



12.



13.



14.



15.



16.



17.



18.



19.



20. 21.



22.



23.



stimulation in awake infants. Pediatric Research. 1998; 43: 840–3. Hock C, Muller-Spahn F, Schuh-Hofer S, Hofmann M, Dirnagl U, Villringer A. Age dependency of changes in cerebral hemoglobin oxygenation during brain activation: A near-infrared spectroscopy study. Journal of Cerebral Blood Flow and Metabolism. 1995; 15: 1103–8. Detre JA, Zhang W, Roberts DA, Silva AC, Williams DS, Grandis DJ et al. Tissue specific perfusion imaging using arterial spin labeling. NMR in Biomedicine. 1994; 7: 75–82. Wong EC, Buxton RB, Frank LR. Implementation of quantitative perfusion imaging techniques for functional brain mapping using pulsed arterial spin labeling. NMR in Biomedicine. 1997; 10: 237–49. Kim SG, Tsekos NV, Ashe J. Multi-slice perfusion-based functional MRI using the FAIR technique: Comparison of CBF and BOLD effects. NMR in Biomedicine. 1997; 10: 191–6. Friston KJ, Zarahn E, Josephs O, Henson RN, Dale AM. Stochastic designs in event-related fMRI. NeuroImage. 1999; 10: 607–19. Sobel N, Prabhakaran V, Desmond JE, Glover GH, Goode RL, Sullivan EV et al. Sniffing and smelling: Separate subsystems in the human olfactory cortex. Nature. 1998; 392: 282–6. Buckner RL, Bandettini PA, O’Craven KM, Savoy RL, Petersen SE, Raichle ME et al. Detection of cortical activation during averaged single trials of a cognitive task using functional magnetic resonance imaging. Proceedings of the National Academy of Sciences of the United States of America. 1996; 93: 14878–83. Grinvald A, Slovin H, Vanzetta I. Non-invasive visualization of cortical columns by fMRI. Nature Neuroscience. 2000; 3: 105–7. Sereno MI, Dale AM, Reppas JB, Kwong KK, Belliveau JW, Brady TJ et al. Functional MRI reveals borders of multiple visual areas in humans. Science. 1995; 268: 889–93. Servos P, Zacks J, Rumelhart DE, Glover GH. Somatotopy of the human arm using fMRI. Neuroreport. 1998; 9: 605–9. Liepert J, Tegenthoff M, Malin JP. Changes of cortical motor area size during immobilization. Electroencephalogr. Clinical Neurophysiology. 1995; 97: 382–6. Talairach J, Tournoux P. Coplanar stereotactic atlas of the human brain: 3-dimensional system, and approach to cerebral imaging. Stuttgart, New York: George Thieme Verlag, 1988. Collins DL, Neelin P, Peters TM, Evans AC. Automatic 3D intersubject registration of MR volumetric data in standardized Talairach space. Journal of Computer Assisted Tomography. 1994; 18: 192–205.



] 683



24. Van Essen DC, Drury HA. Structural and functional analyses of human cerebral cortex using a surface-based atlas. Journal of Neuroscience. 1997; 17: 7079–102. 25. Van Essen DC, Drury HA, Joshi S, Miller MI. Functional and structural mapping of human cerebral cortex: solutions are in the surfaces. Proceedings of the National Academy of Sciences of the United States of America. 1998; 95: 788–95. 26. Dale AM, Fischl B, Sereno MI. Cortical surface-based analysis. I. Segmentation and surface reconstruction. NeuroImage. 1999; 9: 179–94. 27. Fischl B, Sereno MI, Dale AM. Cortical surface-based analysis. II: Inflation, flattening, and a surface-based coordinate system. NeuroImage. 1999; 9: 195–207. 28. Palmer AR, Summerfield AQ. Microelectrode and neuroimaging studies of central auditory function. British Medical Bulletin. 2002; 63: 95–105. 29. Guimaraes AR, Melcher JR, Talavage TM, Baker JR, Ledden P, Rosen BR et al. Imaging subcortical auditory activity in humans. Human Brain Mapping. 1998; 6: 33–41. 30. Scott SK, Blank CC, Rosen S, Wise RJ. Identification of a pathway for intelligible speech in the left temporal lobe. Brain. 2000; 123: 2400–6. 31. Altman NR, Bernal B. Brain activation in sedated children: auditory and visual functional MR imaging. Radiology. 2001; 221: 56–63. 32. Lazeyras F, Boex C, Sigrist A, Seghier ML, Cosendai G, Terrier F et al. Functional MRI of auditory cortex activated by multisite electrical stimulation of the cochlea. NeuroImage. 2002; 17: 1010–7. 33. Alain C, Arnott SR, Hevenor S, Graham S, Grady CL. ‘What’ and ‘where’ in the human auditory system. Proceedings of the National Academy of Sciences of the United States of America. 2001; 98: 12301–6. 34. Mathiak K, Hertrich I, Grodd W, Ackermann H. Cerebellum and speech perception: a functional magnetic resonance imaging study. Journal of Cognitive Neuroscience. 2002; 14: 902–12. 35. Jancke L, Gaab N, Wustenberg T, Scheich H, Heinze HJ. Short-term functional plasticity in the human auditory cortex: an fMRI study. Brain research. Cognitive Brain Research. 2001; 12: 479–85. 36. Giraud AL, Truy E, Frackowiak R. Imaging plasticity in cochlear implant patients. Audiology and Neuro-otology. 2001; 6: 381–93. 37. Bavelier D, Brozinsky C, Tomann A, Mitchell T, Neville H, Liu G. Impact of early deafness and early exposure to sign language on the cerebral organization for motion processing. Journal of Neuroscience. 2001; 21: 8931–42. 38. Thiel CM, Bentley P, Dolan RJ. Effects of cholinergic enhancement on conditioning-related responses in human auditory cortex. European Journal of Neuroscience. 2002; 16: 2199–206.



54 Positron emission tomography and integrated PET/ computed tomography WAI LUP WONG



Principles of PET Squamous cell carcinoma Occult primary tumours Other malignant tumours Indeterminate pulmonary lesions Alternative imaging methods for FDG uptake Radiotracers other than FDG Nononcological applications



684 686 691 692 694 694 696 696



Integrated PET/CT Key points Best clinical practice Deficiencies in current knowledge and areas for future research Acknowledgements References



696 696 696 697 697 697



SEARCH STRATEGY This chapter is supported by literature obtained by electronic searches of four biomedical databases (Medline, Embase, Cochrane Library, PubMed) using combinations of key words including PET, PET-CT, integrated PET/CT, FDG, positron emission tomography, tomography emission computed, fluoro-deoxyglucose F18, head and neck, cancer, thyroid cancer, staging, nodal metastases, occult primary, recurrence, surveillance, coincidence imaging and gamma camera. Studies were also sought from hand searches of head and neck surgery and other relevant journals, and scanning the reference lists of relevant articles. In this discussion, PET refers to a full ring bismuth germinate PET system.



PRINCIPLES OF PET Computed tomography (CT), magnetic resonance (MR) and ultrasound produce anatomical images, whereas positron emission tomography (PET) provides a means of identifying pathology based on altered tissue metabolism. This functional imaging technique relies on a radioactive molecule (radiotracer) that decays with positron emission. The radiotracer is given intravenously to the patient and is taken into cells. The cell recognizes it as being ‘foreign’ and as a consequence it is trapped early in its metabolic pathway. Malignant cells trap more radiotracer compared with nonmalignant cells. The local radiotracer concentration can be measured in vivo since



these unstable radiotracers decay by positron emission. Positrons travel a short distance in tissue before colliding with electrons. When they collide, the annihilation reaction results in two photons also known as gamma rays of 511 kilo electron volts (keV) each emitted at approximately 1801 to each other. The photons are detected by opposing detectors. A computer reassembles these signals into images that represent radiotracer uptake in the part of the body scanned (Figure 54.1a,b). One of the main strengths of PET is the ability to perform quantitative studies. Absolute quantification is not often carried out in clinical practice as it is complicated to obtain and can demand direct arterial blood sampling. The standardized uptake value (SUV),



Chapter 54 Positron emission tomography and integrated PET/computed tomography



] 685



Patient



Gamma ray pairs used to reconstruct images



Detector ring (PET Scanner)



Two gamma rays emitted 180° apart



Positron annihilates with an electron in tissue



Positron path



Positron emitting isotope (a) 2-[18F ] fluoro-2-deoxy-D-glucose FDG



HO



CH2OH O



HO



FDG-6-phosphate



Glycolysis Block



Fructose-1, 6-diphosphate 18F



OH



CITRIC ACID CYCLE



(b)



CT



PET



FDG



(c)



Figure 54.1 (a) The principles of PET imaging. Projection FDG-PET shows a patient with a primary bronchogenic carcinoma with extensive bony and cutaneous deposits. (b) Schematic representation of FDG metabolism. (c) Schematic representation of scanning with integrated PET/CT. The radiation burden is equivalent to two whole body diagnostic CTs. The CT is usually obtained using a low-voltage technique with no intravenous contrast given, but it is of diagnostic quality for detecting pulmonary metastases. Courtesy of Dr Bal Sanghera, Mount Vernon Hospital.



686 ] PART 11 RECENT ADVANCES IN TECHNOLOGY which provides a semiquantitative index of radiotracer uptake is widely used in clinical PET: Tracer uptake ðMBq=mLÞ Administered activity ðMBqÞ=ðpatient weight ðkgÞ  1000Þ



PET images lack anatomical detail, which can be overcome by combining PET with CT/MR using software techniques (registration). With 2-[18F] fluoro-2-deoxy-Dglucose (FDG), these techniques are most robust in the cranium, extracranial head and neck and pelvis.1 Depending on the radiotracer used, different aspects of tissue metabolism can be measured, such as distribution of blood flow, oxygen utilization and protein synthesis. The overwhelming majority of clinical studies are in conjunction with an analogue of glucose, 2-[18F] fluoro2-deoxy-D-glucose, which reflects glucose metabolism. Cancer cells have a greater avidity for glucose than normal cells. Otto Warburg and colleagues2 made this observation in the 1920s. FDG can be used to exploit the differences in glucose metabolism between cancer cells and normal cells. 2-[18F] fluoro-2-deoxy-D-glucose–positron emission tomography (FDG-PET) has been used effectively for imaging of a variety of malignancies including breast, lung, colorectal, oesophageal cancer, brain tumours, malignant melanoma and lymphoma.3 Seifert and others demonstrated that head and neck cancers take up FDG, an observation which was subsequently confirmed by other studies.4



SQUAMOUS CELL CARCINOMA Primary disease STAGING



FDG-PET can detect the majority of clinically visible primary tumours.1, 5 The ability of FDG-PET to detect clinically visible lesions is of limited clinical use, but can be useful for more precisely delineating the primary tumour where submucosal extension of disease is a feature, such as in post-cricoid carcinoma and tracheal carcinoma (Figure 54.2). Drawing from the cumulative results of three studies which included 2004 nodes correlated with resection specimens, FDG-PET had a sensitivity of 90 percent (200/222), specificity 94 percent (1666/1782), positive predictive value (PPV) 63 percent (200/316), negative predictive value (NPV) 99 percent (1666/1688).6, 5, 7 With regard to ‘neck sides’, from the summed data of 280 neck dissections, FDG-PET had a sensitivity of 82 percent (96/117), specificity 80 percent (130/163), PPV 74 percent (96/129), NPV 86 percent (130/151) and accuracy 81 percent (226/280).5, 7, 8, 9, 10, 11 Encouragingly,



Figure 54.2 A 59-year-old man with a post-cricoid SCC. MR suggested more extensive disease than suspected on clinical assessment. (a) Sagittal T1-weighted MR scan demonstrates an intermediate signal intensity mass in the post-cricoid region (arrowhead). (b) FDG-PET sagittal image shows the post-cricoid SCC (arrowheads). The extent of disease seen on FDG-PET was in agreement with the MR. B, physiological uptake in brain; UB, normal activity in bladder as FDG is excreted via the kidneys through the bladder.



Chapter 54 Positron emission tomography and integrated PET/computed tomography



FDG-PET detected contralateral neck metastases, but it is unclear how many of these were unsuspected prior to FDG-PET.6, 5, 7, 8, 12 All studies show a high NPV but this may be overestimated because not all patients had radical neck dissections. Preliminary results suggest that FDG-PET is at least as effective as conventional imaging, if not more so, at staging the neck.6, 5, 7, 8, 9, 13, 14 However, the additional information gained from FDG-PET compared with other imaging is unclear. In a prospective study which included 100 neck sides, colour-coded duplex ultrasonography (CCDU) was as accurate as FDG-PET for detecting nodal disease, with similar sensitivity and specificity.14 In accessible areas, namely levels I–VI, CCDU especially when combined with fine needle aspiration cytological examination (FNAC) may well prove to be as accurate as FDG-PET.15 However, CCDU will not detect retropharyngeal and mediastinal nodal disease. What is clear from the literature is that FDG-PET cannot consistently detect subclinical disease and so does not obviate elective treatment of the neck.5, 7, 11, 12 It can have a role to play in those patients with equivocal nodal disease following conventional assessment. In patients with high risk of neck metastases, a positive scan will be highly indicative of neck nodal disease and the detection of contralateral or bilateral nodal disease can be particularly useful in this group of patients (Figure 54.3). Conversely, a patient with low risk of nodal metastases and a negative FDG-PET is most unlikely to have neck nodal disease (Figure 54.4). In patients with advanced disease in the head and neck (stage III/IV), FDG-PET is potentially useful for detecting occult mediastinal nodal disease.16 Distant metastatic disease in patients presenting with head and neck cancer is unusual and limits the use of FDG-PET as a screening test. However, it may have a role in those patients with a high risk of distant metastases, e.g. locally advanced nasopharyngeal cancer.17



Detection of synchronous malignancies The cumulative incidence of a second malignancy following head and neck cancer is up to 15 percent. Fewer than half will have a detectable synchronous primary malignancy and almost half will be detected by a thorough clinical assessment and a further number by chest CT. So, the pick-up rate of FDG-PET at the time of primary diagnosis is likely to be 2–3 percent, substantiated by one study of 56 patients where it detected one occult lung cancer, a case finding yield of 2 percent.18 FDG-PET has the advantage over other imaging modalities as it can scan the entire body for disease, which is often not possible with other techniques. The low rate of synchronous malignancy perhaps does not justify routine screening.



] 687



OTHER COMMENTS



Pilot studies suggest that pretreatment FDG-PET can be useful as an independent predictor of treatment response and also subsequent prognosis.19, 20, 21, 22, 23 If these results can be substantiated, it will allow the clinician to identify patients where a more aggressive treatment approach should be considered. [***/**]



Recurrent/residual disease PRIMARY SITE



A recent survey of the literature identified 10 relevant studies where FDG-PET was used to assess recurrent/ residual disease and where there was pathological confirmation of the positive FDG-PET findings and follow-up of at least six months for negative FDG-PET results. The sensitivity of FDG-PET varied between 71 and 100 percent, the specificity between 85 and 100 percent, PPV between 64 and 100 percent, NPV between 60 and 100 percent and accuracy between 78 and 100 percent.13, 14, 24, 25, 26, 27, 28, 29, 30, 31 The varying results are a reflection not only of the small number of patients in each study, 8 to 43 patients, but also the various different clinical situations where FDG-PET was used. The challenging clinical issue is the detection of recurrent disease when it is still amenable to curative salvage treatment. Prospective studies that have focussed on patients who become symptomatic during follow-up show FDG-PET to be effective and superior to CT/MR in this area.13, 14, 24, 29, 30 Cumulatively, these studies included 215 patients, 218 FDG-PET scans and an analysis of 226 anatomical sites. FDG-PET distinguished between recurrent disease and treatment sequelae at the primary site with an overall sensitivity of 88 percent (102/116), specificity 83 percent (91/110), PPV 85 percent (102/121), NPV 87 percent (91/105) and accuracy 85 percent (193/226).13, 14, 24, 25, 26, 27, 28, 29, 30, 31 In a group of 38 patients where FDG-PET was used to differentiate between post-irradiation laryngeal recurrent cancer from laryngeal oedema, it achieved a correct diagnosis in 79 percent and was superior to CT (61 percent) and clinical assessment (43 percent).30 FDG-PET is potentially useful as a surveillance tool for patients with high risk of relapse. In one study, in 10/10 patients FDG-PET accurately detected the presence of recurrent disease despite negative or equivocal MR and indeterminate clinical examinations.25 In a prospective study of 30 advanced head and neck cancer (stage III/1V) patients with FDG-PET performed 2 and 10 months after completion of therapy, FDG-PET detected all 16 recurrences which occurred during the first year posttreatment. In five of these 16 patients, FDG-PET was the only technique which detected the recurrences, four at the primary site and one patient with lung metastases.32



688 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



Figure 54.3 A patient referred for assessment of supraglottic laryngeal squamous cell carcinoma with equivocal nodal disease on conventional evaluation. Clinical assessment suggested contralateral nodal disease not confirmed by MR and ultrasound-guided fine needle aspiration cytology examinations. FDG with integrated PET/CT demonstrated (a) the primary site (arrow); (b) ipsilateral neck nodal disease (arrow); and also (c) contralateral neck nodal disease (arrows). (d) The contralateral neck nodal disease, which corresponded to a cluster of three subcentimetre nodes on CT, was subsequently confirmed by neck dissection.



Following radiotherapy, the minimum interval that should be allowed before performing FDG-PET for evaluation of recurrent/residual disease has yet to be determined. In a group of 25 radiotherapy (RT) patients who were followed up between 4 and 11 months and where all failures were confirmed by pathological correlation, there were fewer false-negative results when FDG-PET was performed four months after RT compared with one month following RT.33 In a series



of 44 consecutive patients there were no false-negative results, but a significant number of false-positive results when FDG-PET was performed 12 weeks or less following various treatments for recurrent/residual disease.34 This is in contrast to our own experience of 11 patients with a high risk of recurrence where FDG-PET was performed eight weeks post-RT as part of a surveillance study. Of the nine negative studies, there was one possible false-negative result and no



Chapter 54 Positron emission tomography and integrated PET/computed tomography



] 689



Figure 54.4 A 33-year-old woman presented with a right level II node due to SCC. Conventional assessment including EUA with multiple biopsies and CT/MR did not identify the primary site. FDG with integrated PET/CT (a) shows the tonsillar SCC (arrow); (b) confirms the level II nodal mass (arrowhead) and also shows contralateral neck nodal disease (arrow); (c) corresponds to a node of normal size, short axis diameter of 8 mm, on the CT component of integrated PET/CT (arrow).



false-positive results; both patients with positive FDGPET scans relapsed. Recurrent/residual disease as a group show higher FDG uptake compared with sequelae of treatment. Semiquantitative analysis, however, does not improve discrimination between active disease and post-treatment changes as there is considerable overlap of SUVs between these two groups.24, 35 In clinical practice, FDG PET and now FDG with integrated PET/CT is routinely considered eight weeks or more following RT. A definitively positive scan should be taken as highly suspicious of disease as long as there is no infection. A negative scan is highly likely to represent absence of active disease, but still demands careful



surveillance as microscopic foci of active disease cannot be excluded. An equivocal scan requires biopsy in areas of FDG activity or an early repeat FDG-PET. FDG-PET is more accurate than conventional methods for assessment of recurrent/residual disease, but falsepositive and false-negative results occur and so must be used in the context of the overall assessment of the patient.24, 25, 26, 27, 28, 29, 31, 34, 35 A positive scan can influence a more aggressive biopsy approach and a negative scan a more conservative approach. In some instances it can be used to select sites for further biopsies in biopsy-negative patients with suspected recurrent/residual disease, thus increasing the yield of positive biopsies and avoiding unnecessary deep biopsies (Figure 54.5).



690 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



Figure 54.5 A 62-year-old man, nine months after radical RT with curative intent for a T3 N1 laryngeal SCC. EUA, including multiple biopsies, did not show active disease. (a) Sagittal T1-weighted MR scan demonstrating an intermediate signal intensity mass in the supraglottic larynx. (b) Axial T2-weighted MR scan shows the mass to be predominately of low signal. (c) FDG-PET axial and (d) sagittal images shows intensive uptake in the laryngeal oedema consistent with recurrent disease (arrow), confirmed by further biopsies.



Furthermore, as in patients with primary disease, FDGPET will detect other unexpected cancers in some patients (Figure 54.6). [***/**]



Post-treatment neck assessment Two retrospective reviews, with 34 patients in one study and 36 patients in the other, found FDG-PET had a sensitivity of 100 and 93 percent, specificity of 89 and 77 percent, PPV 70 and 72 percent, NPV 100 and 94 percent, respectively. Both studies suggested FDG-PET to be superior to CT/MR.26, 27 These results are in agreement with a prospective study where FDG-PET identified recurrent nodal disease in all eight patients seen with suspected nodal recurrence who subsequently had histologically proven disease. In this series, the



sensitivity and specificity of FDG-PET was 100 percent, compared with CT/MR (75 and 80 percent, respectively) and clinical evaluation (100 and 60 percent, respectively).13 These results suggest that FDG-PET is a potentially useful tool for assessing recurrent disease in the post-treatment neck and is an area that deserves further attention. [***/**] Preliminary results show that FDG PET is accurate for detection of residual neck disease when performed at least eight weeks after chemoradiotherapy.36, 37, 38, 39 There is considerable interest in the use of FDG with integrated PET/CT for the detection of residual disease in the neck in patients with advanced (N2 and N3) nodal metastases following chemoradiotherapy. A multicentre UK trial has recently been funded by the department of health to test the hypothesis that FDG with integrated PET/CT guided watch and wait is a valid alternative to the current



Chapter 54 Positron emission tomography and integrated PET/computed tomography



] 691



Figure 54.6 A 63-year-old man with clinical suspicion of recurrence at the surgical bed at the primary site. Direct inspection including biopsies, CT and MR did not provide a definitive diagnosis. FDG with integrated PET/CT demonstrated (a) recurrent disease within the surgical bed (arrows); (b) and (c) uptake in the rectum suspicious of rectal carcinoma (arrowheads), confirmed by subsequent colonoscopy and biopsies.



practice of planned neck dissection in this group of patients. Results from this study are awaited with interest.



OCCULT PRIMARY TUMOURS The cornerstone for investigating this group of patients is with examination under anaesthesia (EUA), including multiple biopsies and CT/MR. As such, FDG-PET can only be justified if it can detect more primary sites following these investigations. The various studies which have investigated FDG-PET in this setting have shown, without exception, that FDG-PET can detect more primary sites



compared with conventional assessment.40, 41, 42, 43, 44, 45, 46 In a study of 27 patients which included 21 squamous cell carcinomas (SCC) and undifferentiated carcinomas, FDG-PET showed 28 percent more histologically proven primary sites.45 Our own experience is similar with FDGPET detecting 29 percent more pathologically confirmed primary sites.46 In four other studies, with 7–14 patients with occult SCC or undifferentiated carcinoma in each study, FDG-PET detected between 8 and 42 percent more histologically verified primary sites compared with conventional work up.40, 41, 42, 43, 44 Some of the occult primaries detected were lung cancers in patients who did not have chest CT; it is unclear how many of these lung



692 ] PART 11 RECENT ADVANCES IN TECHNOLOGY cancers would have been detected by CT. Excluding the lung cancers, FDG-PET detected up to 20 percent more primary sites.40, 41, 42, 43, 44 Well-documented false-positive results are unusual in the literature: there were 15 (9 percent) clinically or histologically verified cases from a cumulative total of 173 head and neck sites (nasopharynx, 2; tonsil, 5; tongue base, 3; submandibular gland, 1; oral cavity, 1; thyroid, 2; epiglottis, 1) and a further five (2 percent) in the lungs where no tumour was identified at subsequent bronchoscopy or CT.27, 40, 41, 42, 43, 44, 45, 46, 47, 48 One study is at variance with a false-positive result of 46 percent (6/13);43 the authors suggest that it was possible that the biopsy was in error and that the true PPV of the technique was underestimated. In one study, all 17 patients including the three patients where FDG-PET showed a primary site which was not subsequently confirmed, were treated on the basis of the FDG-PET result. No other primary head and neck tumour developed during subsequent follow-up in any of the patients.46 This result supports the speculation that FDG-PET may be more specific than the current figures indicate. Further clarification may be available with long-term follow-up of larger groups of patients. Inflammatory pulmonary lesions including granulomas and abscesses and recent biopsy can result in falsepositive FDG-PET.40, 42 Dental caries and benign thyroid and salivary gland nodules are recognized causes of intense FDG uptake and so do not usually cause a diagnostic dilemma. We have found that tonsillectomy can cause potential confusion, not on the side of surgery but on the contralateral side, where normal tonsillar uptake in the remaining tonsil can mimic a malignancy. Considering all the studies with false-negative results reported, there was a cumulative total of over 127 patients and only 10 (8 percent) false-negative results confirmed at histology or subsequent follow-up: tongue carcinoma, 3, including 2 base of tongue carcinomas; tonsillar carcinoma, 2; nasopharyngeal cancer, 1; lung cancer, 2; mesothelioma, 1; and gastric carcinoma, 1.27, 40, 41, 42, 43, 44, 45, 46 The ability to detect cancers less than 1 cm in diameter with FDG-PET is currently unclear. Using attenuation corrected scanning, FDG-PET showed a 4-mm base of tongue SCC in one study.40 Even in retrospect, a further study failed to show a 7-mm pharyngeal wall SCC with attenuation correction scanning.46 Most studies report that in the majority of patients with an occult primary tumour after staging which includes FDG-PET, no head and neck primary tumour will emerge after treatment during subsequent followup.40, 41, 42, 43, 44, 45, 46, 47, 48 FDG-PET can influence the therapeutic plan further in up to 59 percent of patients.41, 42, 44, 45, 46, 47, 48 It can alter management by not only identifying the primary site, but also other occult sites of disease including bony metastases, mediastinal nodal disease and synchronous carcinomas at other sites.40, 45



Biopsies carried out after FDG-PET can improve the number of occult primaries detected compared with EUA with speculative biopsies.40, 41, 42, 45 [***/**]



OTHER MALIGNANT TUMOURS FDG-PET is of limited value for the preoperative assessment of thyroid nodules as both malignant and benign thyroid lesions can be equally avid for FDG. However, it can have a role in the assessment of the posttreatment patient. In a study of 24 patients with treated differentiated papillary and follicular thyroid cancer, the metastases had an alternating pattern of uptake with 131iodine and FDG. Either some 131-iodine uptake combined with low FDG trapping or no 131-iodine uptake combined with high FDG trapping.49 The hypothesis was that persistent iodine metabolism is consistent with better cell differentiation, while the loss of this ability together with increased glucose metabolism is consistent with dedifferentiation.49 Several reports have demonstrated the value of FDG-PET in treated papillary and follicular thyroid cancer patients who have suspected recurrent disease with a high thyroglobulin levels and negative 131iodine scan.49, 50, 51, 52, 53 In a retrospective review, FDGPET was found to be useful in 64 percent (39/61) of patients who had an elevated thyroglobulin and negative 131-iodine scan with a PPV of 83 percent (34/41) and an NPV of 25 percent (5/20). Furthermore, true positive FDG-PET findings correlated positively with increasing thyroglobulin levels.52 In a study of 18 patients with elevated thyroglobulin levels FDG-PET localized occult disease in 71 percent (12/17) of patients who subsequently had confirmation of disease, mainly unsuspected neck and chest nodal disease; FDG-PET was of much more limited value in patients with normal thyroglobulin levels.53 The overall accuracy of FDG-PET in this setting is unclear. In all published studies, many lesions were not verified histopathologically and in most studies the follow-up period after FDG-PET was unclear. The incremental improvement afforded by FDG-PET over other imaging is also uncertain. Various studies show FDG-PET has an advantage over 131-iodine in its ability to detect more nodal disease in the neck and in the mediastinum.53, 54, 55, 56 With regard to the neck, however, none of these studies systematically compared neck MR and ultrasound with FDG-PET. In one survey, minimal cervical nodal disease present in three patients not detected by 131-iodine and FDG-PET was shown on ultrasound.53 In most of the patients where FDG-PET detected mediastinal nodal disease, a comparison with CT (chest) was not made. In the literature, FDG-PET demonstrated lung metastases which were not iodine avid, but these same studies showed that CT was superior to FDG-PET in this area (Figure 54.7).53, 55



Chapter 54 Positron emission tomography and integrated PET/computed tomography



] 693



Figure 54.7 A 38-year-old man who presented with elevated thyroglobulin after total thyroidectomy and 131-iodine ablation for papillary thyroid cancer. (a) Whole body 123-iodine scan showed no active disease. (b) FDG with integrated PET/CT shows the nodal recurrence corresponding to (c) normal sized nodes on CT (arrow). Surgery and histological examination confirmed retropharyngeal nodal disease.



694 ] PART 11 RECENT ADVANCES IN TECHNOLOGY Thallium and hexakis (2-methoxyisobutyl-isonitirile) technetium-99m (MIBI) have been proven to be useful radiopharmaceuticals as alternatives to 131-iodine in the follow-up of differentiated thyroid cancer. A retrospective multicentre study evaluated the use of 131-iodine, FDG and MIBI in the follow-up of 222 such patients. FDG was more sensitive than iodine 131 and MIBI, 75, 50 and 53 percent respectively with no compromise in specificity, 90, 99 and 92 percent, respectively.51 FDG-PET can detect metastases from Hurthle cell carcinoma and anaplastic carcinoma.53, 57 Its clinical value in these tumours is currently unclear. With medullary thyroid cancer, preliminary results suggest FDG-PET to be an accurate technique for detecting metastases and recurrent disease. The largest study in the literature, a German multicentre retrospective survey, reviewed 100 FDG-PET studies performed on 85 patients. The results of the FDG-PET were compared with 46 indium-111 pentetreotide (SMS) studies, 33 pentavalent technetium dimercaptosuccinate acid (DMSA) scans, 8 hexakis (2-methoxyisobutylisonitirile) technetium-99m (MIBI) scans, 64 CT and 37 MR. One hundred and eighty-one lesions were identified by at least one technique; FDG-PET detected the majority of lesions with a lesion detection probability of 68 percent compared with 25 percent for SMS, 29 percent for DMSA, 6 percent for MIBI, 53 percent for CT and 58 percent for MR. In the 55 lesions confirmed histologically, FDG-PET had the highest overall accuracy with a sensitivity of 78 percent and specificity of 79 percent; MR had a slightly higher sensitivity of 82 percent, but a specificity of 67 percent and MIBI had a specificity of 100 percent, but a sensitivity of only 25 percent.58 In one published report of postoperative parathyroid carcinoma, FDG-PET provided more accurate information compared with other imaging.59 FDG-PET cannot reliably distinguish between malignant and benign major salivary gland tumours as both can be equally FDG avid. It can, however, be useful for distinguishing active disease from sequelae of treatment. FDG-PET is effective for detecting lymphoma except low-grade non-Hodgkin’s lymphoma and extranodal marginal zone B-cell lymphoma, which are not consistently FDG avid.60, 61 Head and neck malignant melanomas, sarcomas and paraganglionomas have been shown to be FDG avid.62, 63 FDG-PET can be useful in these rare head and neck tumours for detecting occult sites of disease, especially when more disease is suspected than can be accounted for by conventional assessment, provided the known sites of disease are FDG avid (Figure 54.8). [***/**]



INDETERMINATE PULMONARY LESIONS Cancer patients with indeterminate pulmonary lesions following full clinical and radiological assessment can



Figure 54.8 A 49-year-old man presented with left neck nodes due to high-grade non-Hodgkin’s lymphoma (NHL). He had systemic symptoms but conventional assessment showed disease that was localized to the left neck. (a) FDG-PET coronal slice demonstrates disease in the neck (arrowheads) and also in the mediastinum (hollow arrowhead) and axilla (arrow). The patient’s management was changed from localized RT to the neck to chemotherapy. (b) Following chemotherapy, the FDG abnormalities in the neck, axilla and mediastinum resolved. Increased uptake is seen along an infected Hickmann line (arrowheads).



present a diagnostic dilemma to the surgeon, and head and neck cancer patients are no exception. Accurate characterization of the pulmonary abnormality can alter treatment plan. FDG-PET has been shown to be an accurate noninvasive test for the diagnosis of pulmonary lesions (Figure 54.9). A metanalysis of the data from 40 studies showed FDG-PET has a sensitivity of 96.8 percent and specificity of 77.8 percent.64 Abscesses and granulomas including those due to sarcoid, tuberculosis (TB), anthracite and fungus (asperilloma, blastomycosis, cocidiomycosis, cryptococcisis, histolplasmosis) can mimic malignant lesions.65 False-negative FDG-PET results are very unusual, although bronchoalveolar carcinoma, highly differentiated neuroendocrine carcinoma and adenocarcinoma, especially within a scar, have been reported to cause confusion.65 Lesions smaller than 1 cm across can be detected by FDG-PET, but falsenegative results do occur and so a negative scan must be interpreted with caution.65 [***/**]



ALTERNATIVE IMAGING METHODS FOR FDG UPTAKE Owing to the high costs and limited availability of full ring bismuth germinate PET systems, alternative methods of imaging 511 keV photons of positron emitters have been sought. To this end, partial ring bismuth germinate systems, such as the ECAT advanced rotating tomograph (ART), scanners with six positron-sensitive sodium



Chapter 54 Positron emission tomography and integrated PET/computed tomography



] 695



Figure 54.9 A 54-year-old man with clinical suspicion of recurrence of oropharyngeal cancer. CT and MR did not provide a definitive diagnosis. (a) FDG with integrated PET/CT demonstrated recurrent disease in the post-treatment neck (arrow); and (b), (c) and (d) uptake in the right hilum of the lung suspicious of a bronchogenic carcinoma (hollow arrowhead), confirmed by subsequent bronchoscopy and biopsies.



iodide detectors and dual-headed gamma cameras operating in coincidence have been developed. These systems are less sensitive compared with full ring dedicated PET systems. Nevertheless, their performance may be adequate in the specific clinical settings. Stokkel and colleagues, in a retrospective review of 54 consecutive SCC patients, found FDG dual-head PET a valuable tool for staging the neck.66 In a prospective study of 48 patients with suspected recurrent laryngeal or hypopharyngeal cancer, the results obtained with FDG dual-headed PET was comparable with those from studies using full ring PET.67 Similar results were obtained in one study using a sodium iodide detector system and a partial ring PET in another; in both, FDG was superior to CT/MR.68, 69



In three studies, non-full ring PET systems detected 28–73 percent more occult head and neck primary tumours compared with conventional imaging. It is unclear how many of the occult malignancies would have been detected by EUA and multiple biopsies, as FDG was performed prior to EUA in all patients studied.69, 70, 71 With regards to thyroid cancer, several pilot studies have shown the feasibility of FDG dual-head PET for detecting recurrent disease in patients with elevated thyroglobulin and negative 131-iodine scan.72, 73 Despite these encouraging preliminary reports for the use of non-full ring PET for the assessment of head and neck cancer patients, there is now very limited interest in the use of this technology with the advent of combined PET-CT scanners.



696 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



RADIOTRACERS OTHER THAN FDG Carotid artery resection may occasionally offer a chance of cure in patients with advanced head and neck cancer. [15-O] H2O PET has been used a rapid quantitative means of determining cerebral blood flow prior to carotid artery resection in this group of patients.74 Beyond clinical studies, PET provides the unique opportunity to assess specific biological characteristics of head and neck cancer with minimal disturbance of tumour physiology. Hypoxic areas in head and neck tumours have been successfully evaluated with PET using radiolabelled nitroimidazole compounds including fluorine-18 fluoromisonidazole (FMISO) and fluorine18 fluoroerythronitroimidazone (FETNIM).75, 76 Thymidine labelled with 11-carbon, [methyl-11C] thymidine and L-[methyl-11C] methionine (11-C methionine) have been developed to enable the measurement of the proliferative capacity of head and neck tumours.77, 78 Chao and others demonstrated the feasibility of using Cu(II)-diacetyl-bis-N4-methylthiosemicarbozone (Cu-ATSM), a hypoxic marker, for planning intensity modulated radiotherapy.79



KEY POINTS  The use of FDG-PET in head and neck cancer is supported by data from observational and nonrandomized studies. There are no randomized controlled trials available.  The body of evidence shows that FDG-PET can contribute to more accurate assessment of patients with head and neck cancer, including improved staging of the neck, detection of recurrent/residual disease, detection of the occult primary and detection of occult disease in treated differentiated thyroid cancer.  Beyond FDG, PET provides the unique opportunity to assess specific biological characteristics of head and neck cancer with minimal disturbance of tumour physiology.  Integrated PET/CT should improve the accuracy of FDG PET.



NONONCOLOGICAL APPLICATIONS FDG-PET is potentially a useful tool in the assessment of cochlea implants and their pattern on stimulation of the central nervous system. A pilot study using FDG-PET and [15-O] H2O demonstrated that the primary auditory cortex does not develop normally in prelingually deaf subjects, but that in post-lingually deaf subjects this is not so and the primary auditory cortex may be reactivated by cochlear implant after many years of deafness.80



INTEGRATED PET/CT Positron emission tomography is limited by the imprecise localization of radiotracer. Software image fusion is labour-intensive and is usually unsuccessful unless data are acquired prospectively. Furthermore, accurate software registration is generally only valid for FDG-PET. The advantages of integrated PET/CT include superior localization of lesions and better distinction between physiological uptake and pathology. With PET/CT, CT is acquired followed by PET. Using CT for attenuation correction also means shorter scanning times; 30 minutes compared with 60 minutes with standard FDG PET. PET/ CT should augment the information provided by PET; it will more accurately localize the occult primary in the orophyarnx. In patients with suspected recurrence of thyroid cancer, the CT component will detect pulmonary metastases not identified on FDG and in other patients more precise anatomical localization of active disease within treatment sequelae will influence the treatment plan.81, 82, 83



Best clinical practice FDG-PET, and now FDG with integrated PET/CT, is not necessary for all patients with head and neck cancer. For most, clinical assessment complemented by conventional imaging will give all the necessary information for planning treatment. However, FDG with integrated PET/CT can be useful for a selected minority. With the present state of knowledge, the following are recommended as current indications of FDG with integrated PET/CT: [ SCC patients with – equivocal nodal disease following conventional assessment, especially if the nodes in question are not accessible to ultrasonography or ultrasonography with FNA has not provided a conclusive result; – suspicion of recurrent/residual disease and equivocal clinical or radiological assessment. [ Patients with occult primary tumours: FDG with integrated PET/CT should be considered prior to EUA and after full clinical assessment, laryngoscopy/ fibreoptic endoscopy and CT/MR fail to show a primary tumour. This investigation algorithm provides the best opportunity for detecting the occult primary, whilst minimizing misleading results. [ Post-treatment papillary and follicular thyroid cancer patients with elevated thyroglobulin and negative 131-I scan. [ Patients with clinical suspicion of more disease than conventional assessment demonstrates. [ Patients where resectability is in doubt.



Chapter 54 Positron emission tomography and integrated PET/computed tomography



[ Patients with indeterminate lung lesions which are not accessible to precutaneous biopsy or where a pneumothorax would be particularly hazardous. [Grade C/D]



Deficiencies in current knowledge and areas for future research



$ $ $ $ $



Large prospective comparative studies of FDG with integrated PET/CT with state-of-the-art ultrasound (neck) for patients referred for assessment of primary disease and equivocal necks following conventional assessment. Prospective studies evaluating FDG with integrated PET/CT as an early predictor of response to treatment and as an independent prognostic factor for head and neck SCC. Large prospective studies evaluating the value of FDG with integrated PET/CT as a surveillance tool in head and neck SCC patients with high risk of recurrence. Prospective comparative studies including MR and state-of-the-art neck ultrasound, CT (chest) and FDG with integrated PET/CT for the assessment of patients with treated differentiated thyroid cancer, elevated thyroglobulin and negative whole body 131-I scan. Integrated PET/CT studies with hypoxic and cell proliferation markers for evaluating the biology of malignant head and neck tumours.



ACKNOWLEDGEMENTS I am very grateful to my clinical oncology and head and neck surgery colleagues with whom I work closely, especially Professor Michele Saunders, Mr Roy Farrell, Dr Katherine Lemon, Dr Kate Goodchild. I also gratefully acknowledge Dr Jane Chambers for her useful critical comments.



REFERENCES 1. Wong WL, Hussain K, Chevretton E, Hawkes DJ, Baddeley H, Maisey M et al. Validation and clinical application of computer-combined computed tomography and positron emission tomography with 2-(18F)fluoro-2-deoxy-dglucose head and neck images. American Journal of Surgery. 1996; 172: 628–32. 2. Warburg O, Wind F, Negelein G. Uber den stoffwechsel von tumoren im korper. Klinische Wochenschrift. 1926; 5: 829–32.



] 697



3. Hustinx R, Benard F, Alavi A. Whole body FDG-PET imaging in the management of patients with cancer. Seminars in Nuclear Medicine. 2002; 32: 35–46. 4. Seifert E, Schadel A, Haberkorn U, Strauss LG. Evaluating the effectiveness of chemotherapy in patients with head and neck tumours using positron emission tomography. HNO. 1992; 40: 90–3. 5. Braams JW, Pruim J, Freling NJM, Nikkels PGJ, Rodenburg JLN, Boering G et al. Detection of lymph node metastases of squamous cell cancer of the head and neck with FDGPET and MR. Journal of Nuclear Medicine. 1995; 36: 211–6. 6. Adams S, Baum RP, Stuckensen T, Bitter K, Hor G. Prospective comparision of 18-F-FDG with conventional imaging modalities (CT, MRI, US) in lymph node staging of head and neck cancer. European Journal of Nuclear Medicine. 1998; 25: 1255–60. 7. Laubenbacher C, Saumweber D, Wagner-Manslau C, Kau RJ, Herz M, Avril N et al. Comparision of fluorine-18fluorodeoxyglucose PET, MR and endoscopy for head and neck squamous cell carcinoma. Journal of Nuclear Medicine. 1995; 36: 1747–57. 8. Kau RJ, Alexiou C, Laubenbacher C, Werner M, Schwaiger M, Arnold W. Lymph node detection of head and neck squamous cell carcinomas by positron emission tomography with fluorodeoxyglucose F18 in a routine clinical setting. Archives of Otoloaryngology – Head and Neck Surgery. 1999; 125: 1322–8. 9. McGuirt WF, Williams III DW, Keyes JW, Greven KM, Watson Jr. NE, Geisinger KM et al. A comparative diagnostic study of head and neck metastases using positron emission tomography. Laryngoscope. 1995; 105: 373–5. 10. Myers LL, Wax MK, Nabi H, Simpson GT, Lamonica D. Positron emission tomography in the evaluation of the N0 neck. Laryngoscope. 1998; 108: 232–6. 11. Stoeckli SJ, Steinert H, Pflaltz M, Schmid S. Is there a role for positron emission tomography with 18Ffluorodeoxyglucose in the initial staging of nodal negative oral and oropharyngeal squamous cell carcinoma. Head and Neck. 2002; 24: 345–9. 12. Nowak B, di Martino E, Janicke S, Cremerius U, Adam G, Zimmy M et al. Diagnostik maligner Kopf-Hals-Tumouren durch F-18-FDG PET im vergleich zu CT/MRT. Nuklearmedizin. 1999; 38: 312–8. 13. Wong WL, Chevretton E, McGurk M, Hussain K, Davis J, Beaney R et al. A prospective study of FDG PET imaging for the assessment of head and neck squamous cell carcinoma. Clinical Otolaryngology. 1977; 22: 209–14. 14. di Martino E, Nowak B, Hassan HA, Hausman R, Adam G, Buell U et al. Diagnosis and staging of head and neck cancer – a comparision of modern imaging modalities (positron emission tomography, computed tomography, colour coded duplex sonography) with pan endoscopic and histopathological findings. Archives of Otolaryngology – Head and Neck Surgery. 2000; 126: 1457–61.



698 ] PART 11 RECENT ADVANCES IN TECHNOLOGY 15. Brouwer J, de Bree Remco, Comans EFJ, Castelijns JA, Hoekstra OS, Leemans CE. Positron emission tomography using [18F] flurodeoxyglucose (FDG-PET) in the clinically negative neck: is it likely to be superior? European Archives of Oto-rhino-laryngology. 2004; 261: 479–83. 16. Teknos TN, Rosenthal EL, Lee D, Taylor R, Marn CS. Positron emission tomography in the evaluation of stage III and IV head and neck cancer. Head and Neck. 2001; 23: 1056–60. 17. Chang J T-C, Chan S-C, Yen T-C, Liao C-T, Lin C-Y, Lin K-J et al. Nasopharyngeal carcinoma by (18)-Ffluorodeoxyglucose positron emission tomography. International Journal of Radiation Oncology, Biology, Physics. 2005; 62: 501–7. 18. Keyes JW, Chen MYM, Watson Jr. NE, Greven KM, McGuirt WF, Williams III DW. FDG PET evaluation of head and neck cancer: value of imaging the thorax. Head and Neck. 2000; 22: 105–10. 19. Brun E, Ohlsson T, Erlandsson K, Kjellen E, Sandell A, Tennvall J et al. Early prediction of treatment outcome in head and neck cancer with 2-18FDG PET. Acta Oncologica. 1997; 36: 741–7. 20. Halfpenny W, Hain SF, Biassoni L, Maisey MN, Sherman JA, McGurk M. FDG PE. A possible prognostic factor in head and neck cancer. British Journal of Cancer. 2002; 86: 512–6. 21. Minn H, Lapela M, Klemi PJ, Grenman R, Leskinen S, Lindholm P et al. Prediction of survival with fluorine-18flurodeoxyglucose and PET in head and neck cancer. Journal of Nuclear Medicine. 1997; 38: 1907–11. 22. Rege S, Safa AA, Chaikin L, Hoh C, Juillard G, Withers R. Positron emission tomography: An independent indicator of radiocurability in head and neck carcinomas. American Journal of Clinical Oncology. 2000; 23: 164–9. 23. Allal AS, Dugluerov P, Allaoua M, Haenggeli C-A, El Ghazi EA, Lehmann E et al. Standardized uptake value of 2-[18F] Fluoro-2-Deoxy-D-glucose in predicting outcome in head and neck carcinomas treated with radiotherapy with and without chemotherapy. Journal of Clinical Oncology. 2002; 20: 1398–404. 24. Anzai Y, Carroll WR, Quint DJ, Bradford CR, Minoshima S, Gregory T et al. Recurrence of head and neck cancer after surgery or irradiation: Prospective comparison of 2-deoxy2-(F-18)fluoro-d-glucose PET and MR imaging diagnoses. Radiology. 1996; 200: 135–41. 25. Bailet JW, Sercarz JA, Abemayor E, Anzai Y, Lufkin RB, Hoh CK. The use of positron emission tomography for early detection of recurrent head and neck squamous cell carcinoma in postradiotherapy patients. Laryngoscope. 1995; 105: 135–9. 26. Fischbein NJ, Aassar OS, Caputo GR, Kaplan MJ, Singer MI, Price DC et al. Clinical utility of positron emission tomography with 18F-fluorodeoxyglucose in detecting residual/recurent squamous cell carcinoma of the head and neck. AJNR American Journal of Neuroradiology. 1998; 19: 1189–96.



27. Hanasono MM, Kunda LD, Segall GM, Ku GH, Terris DJ. Uses and limitations of FDG positron emission tomography in patients with head and neck cancer. Laryngoscope. 1999; 109: 880–5. 28. Kim HJ, Boyd J, Dunphy F, Lowe AV. F-18 FDG PET scan after radiotherapy for early-stage larynx cancer. Clinical Nuclear Medicine. 1998; 23: 750–2. 29. Lapela M, Grenman R, Kurki T, Joensuu H, Leskinen S, Lindholm P et al. Head and neck cancer: Detection of recurrence with PET and 2-(F-18)Fluoro-2-deoxy-dglucose. Radiology. 1995; 197: 205–11. 30. McGuirt WF, Greven KM, Keyes JW, Williams III DW, Watson N. Laryngeal radionecrosis versus recurrent cancer: A clinical approach. Annals of Otology, Rhinology, and Laryngology. 1998; 107: 293–6. 31. Li P, Shuang H, Mozley PD, Denittis A, Yeh D, Machtay M et al. Evaluation of recurrent squamous cell carcinoma of the head and neck with FDG positron emission tomography. Clinical Nuclear Medicine. 2001; 26: 131–5. 32. Lowe VJ, Boyd JH, Dunphy FR, Kim H, Dunleavy T, Collins BT et al. Surveillance for recurrent head and neck cancer using positron emission tomography. Journal of Clinical Oncology. 2000; 18: 651–8. 33. Greven KM, Williams III DW, Keyes JW, McGuirt WF, Watson NE, Randall ME et al. Positron emission tomography of patients with head and neck carcinoma before and after high dose irradiation. Cancer. 1994; 74: 1355–9. 34. Lonneux M, Lawson G, Ide C, Bausart R, Remacle M, Pauwels S. Positron emission tomography with fluorodeoxyglucose for suspected head and neck tumor recurrence in the symptomatic patient. Laryngoscope. 2000; 110: 1493–7. 35. Lapela M, Eigtved A, Jyrkkio S, Grenman R, Kurki T, Lindholm P et al. Experience in qualitative and quantitative FDG PET in follow-up of patients with suspected recurrence from head and neck cancer. European Journal of Cancer. 2000; 36: 858–67. 36. Yao M, Smith RB, Graham MM, Hoffman HT, Tan H, Funk GF et al. The role of post-radiation therapy FDG PET in prediction of necessity for post-radiation therapy neck dissection in locally advanced head and neck squamous cell carcinoma. International Journal of Radiation Oncology, Biology, Physics. 2005; 63: 991–9. 37. Porceddu SV, Jarmolowski E, Hicks RJ, Ware R, Weih L, Rischin D et al. Utility of PET for the detection of disease in residual neck nodes after (chemo)radiotherapy in head and neck cancer. Head and Neck. 2005; 27: 175–81. 38. McCullom DA, Burrell SC, Haddad RI, Norris CM, Tishler RB, Case MA et al. PET with 18F-FDG to predict pathological response after induction chemotherapy and definitive chemoradiotherapy in head and neck cancer. Head and Neck. 2004; 26: 890–6. 39. Rogers JW, Greven KM, McGuirt WF, Keyes Jr JW, Williams 3rd DW et al. Can post RT neck dissection be omitted for



Chapter 54 Positron emission tomography and integrated PET/computed tomography



40.



41.



42.



43.



44.



45.



46.



47.



48.



49.



50.



51.



patients with head and neck cancer who have a negative PET scan after definitive radiation therapy? International Journal of Radiation Oncology, Biology, Physics. 2004; 58: 694–7. Aassar OS, Fischbein NJ, Caputo GR, Kaplan MJ, Price DC, Singer MJ et al. Metastatic head and neck cancer: Role and usefulness of FDG PET in locating occult primary tumours. Radiology. 1999; 210: 177–81. Braams JW, Pruim J, Kole AC, Nikkels PJC, Vaalburg W, Vermey J et al. Detection of unknown primary head and neck tumors by positron emission tomography. International Journal Oral Maxillofacial Surgery. 1997; 26: 112–5. Safa AA, Luu MT, Rege S, Brown CV, Mandelkern MA, Wang MB et al. The role of positron emission tomography in occult primary head and neck cancers. The Cancer Journal from Scientific American. 1999; 5: 214–8. Greven KM, Keyes JW, Williams III DW, McGuirt WF, Joyce III WT. Occult primary tumours of the head and neck. Cancer. 1999; 86: 114–8. Lassen U, Daugaard G, Eigtved A, Damgaard K, Friberg L. 18F-FDG whole body positron emission tomography (PET) in patients with unknown primary tumours (UPT). European Journal of Cancer. 1999; 35: 1076–82. Jungehulsing M, Scheidhauer K, Damm M, Pietrzyk U, Eckel H, Schicha H et al. 2[18F]-fluoro-2-deoxy-D-glucose positron emission tomography is a sensitive tool for the detection of occult primary cancer (cancer of unknown syndrome) with head and neck lymph node manifestation. Otolaryngology – Head Neck Surgery. 2000; 123: 294–301. Wong WL, Saunders M. The impact of FDG PET on the management of occult primary head and neck tumours. Clinical Oncology. 2003; 15: 461–6. Bohuslavizki KH, Klutmann S, Sonnemann U, Thomas J, Kroger S, Werner JA et al. F-18 FDG-PET zur detektion des okkulten primartumors bei patienten mit lymphnotedmetastsen der halsregion. Laryngo-RhineOtologie. 1999; 78: 445–9. Schipper JH, Schrader M, Arweiler D, Muller S, Scuik J. Die positronenemissonstomographie zur primatumoursuche bei halslyphknotenmetastasen mit unbekanntem primartumor. HNO. 1996; 44: 254–7. Feine U, Lietzenmayer R, Hanke JP, Wohrle H, MullerSchauenburg W. 18FDG-Ganzkorper-PET bei differenzierten schilddrusenkarzinomen. Nuclear Medicine. 1995; 34: 127–34. Conti PS, Durski JM, Bacqai F, Grafton ST, Singer PA. Imaging of locally recurrent and metastatic thyroid cancer with positron emission tomography. Thyroid. 1999; 9: 797–804. Grunwald F, Kalicke T, Feine U, Lietzenmayer R, Scheidhauer K, Dietlein M et al. Fluorine-18 fluorodeoxyglucose positron emission tomography in thyroid cancer:results of a multicentre study. European Journal of Nuclear Medicine. 1999; 26: 1547–52.



] 699



52. Schluter B, Bohuslavizki KH, Beyer W, Plotkin M, Buchert R, Clausen M. Impact of FDG PET on patients with differentiated thyroid cancer who present with elevated thyroglobulin and negative 131I scan. Journal of Nuclear Medicine. 2001; 42: 71–6. 53. Wang W, Macapinlac H, Larson SM, Yeh SDJ, Akhurst T, Finn RD et al. (18F)-2-Fluoro-2-deoxy-D-glucose positron emission tomography localizes residual thyroid cancer in patients with negative diagnostic 131I whole body scans and elevated serum thyroglobulin levels. The Journal of Clinical Endocrinology and Metabolism. 1999; 84: 2291–302. 54. Chung J-K, So Y, Lee JS, Choi CW, Lim SM, Lee DS et al. Value of FDG PET in papillary thyroid carcinoma with negative 131I whole-body scan. Journal of Nuclear Medicine. 1999; 40: 986–92. 55. Dietlein M, Scheidhauer K, Voth E, Theissen P, Schicha H. Fluorine-18 fluorodeoxyglucose positron emission tomography and iodine-131 whole-body scintigraphy in the follow-up of differentiated thyroid cancer. European Journal of Nuclear Medicine. 1997; 24: 1342–8. 56. Javad H, McDougall IR, Segall GM. Evaluation of suspected recurrent thyroid carcinoma with [18F]fluorodeoxyglucose positron emission tomography. Nuclear Medicine Communications. 1998; 19: 547–54. 57. Lind P, Kumnig G, Matschnig S, Heinisch M, Gallowitsch HJ, Mikosch P et al. The role of F-18FDG PET in thyroid cancer. Acta Medica Austriaca. 2000; 27: 38–41. 58. Diehl M, Risse JH, Brandt-Mainz K, Dietlein M, Bohuslavizki KH, Matheja P et al. Fluorine-18 fluorodeoxyglucose positron emission tomography in medullary thyroid cancer: results of a multicentre study. European Journal of Nuclear Medicine. 2001; 28: 1671–6. 59. Neumann DR, Esselstyn CB, Kim EY. Recurrent post operative parathyroid carcinoma: FDG PET and sestamibi SPECT findings. Journal of Nuclear Medicine. 1996; 37: 2000–1. 60. Walsh RM, Wong WL, Chevretton EB, Beaney RP. The use of PET-18FDG imaging in the clinical evaluation of head and neck lymphoma. Clinical Oncology. 1996; 8: 51–4. 61. Hoffman M, Kletter K, Becherer A, Jaeger U, Chott A, Raderer M. 18F-fluorodeoxyglucose positron emission tomography (18F FDG PET) for staging and follow-up of marginal zone B-cell lymphoma. Oncology. 2003; 64: 336–40. 62. Goerres GW, Stoeckli SJ, von Schulthess GK. FDG PET for mucosal malignant melanoma of the head and neck. Laryngoscope. 2002; 112: 381–5. 63. Wittekindt C, Theissen P, Jungehulsing M, Brochhagen HG. FDG-PET imaging of malignant paraganglioma of the neck. Annals of Otology, Rhinology, and Laryngology. 1999; 108: 909–12. 64. Gould MK, Maclean CC, Kushner WG, Ware G, Rydzak CE, Douglas K. Accuracy of positron emission tomography for the diagnosis of pulmonary nodules and mass lesions: a meta-analysis. Journal of the American Medical Association. 2001; 285: 914–24.



700 ] PART 11 RECENT ADVANCES IN TECHNOLOGY 65. Wong WL, Campbell H, Saunders M. Positron emission tomography (PET) - evaluation of indeterminate pulmonary lesions. Clinical Oncology. 2002; 14: 123–8. 66. Stokkel MPM, ten Broek F-W, Hordijk G-J, Koole R, van Rijk PP. Preoperative evaluation of patients with primary head and neck cancer using dual-head 18 fluorodeoxyglucose positron emission tomography. Annals of Surgery. 2000; 231: 229–34. 67. Stokkel MPM, Terhaard CHJ, Hordijk G-J, van Rijk PP. The detection of local recurrent head and neck cancer with fluorine-18 fluorodeoxyglucose dual-head positron emission tomography. European Journal of Nuclear Medicine. 1999; 26: 767–73. 68. Farber LA, Benard F, Machtay N, Smith RJ, Weber RS, Weinstein GS et al. Detection of recurrent head and neck squamous cell carcinomas after radiation therapy with 2-18F-fluoro-2-deoxy-D-glucose positron emission tomography. Laryngoscope. 1999; 109: 970–5. 69. Kresnik E, Mikosch P, Gallowitsch HJ, Kogler D, Wieser S, Heinisch M et al. Evaluation of head and neck cancer with 18 F-FDG PET: a comparison with conventional methods. European Journal of Nuclear Medicine. 2001; 28: 816–21. 70. Mukherji SK, Drane WE, Mancuso AA, Parsons JT, Mendenhall WM, Stringer S. Occult primary tumors of the head and neck: Detection with 2-(F-18)fluoro-2-deoxy-Dglucose SPECT. Radiology. 1996; 199: 761–6. 71. Stokkel MPM, Terhaard CH, Hordijk GJ, van Rijk PP. The detection of unknown primary tumors in patients with cervical metastases by dual-head positron emission tomography. Oral Oncology. 1999; 35: 390–4. 72. Alnafisi NS, Driedger AA, Coates G, Moote DJ, Raphael SJ. FDG PET of recurrent or metastatic 131I-negative papillary thyroid carcinoma. Journal of Nuclear Medicine. 2000; 41: 1010–15. 73. Muros MA, Llamas-Elvira JM, Ramirez-Navarro A, Gomez MJA, Rodriguez-Fernandez A, Muros T et al. Utility of fluorine-18-fluorodeoxyglucose positron emission tomography in differentiated thyroid carcinoma with negative radioiodine scans and elevated serum thyroglobulin levels. American Journal of Surgery. 2000; 179: 457–61. 74. Okamoto Y, Inugami A, Matsuzaki Z, Yokomizo M, Konno A, Togawa K et al. Carotid artery resection for head and neck cancer. Surgery. 1996; 120: 54–9.



75. Lehtio K, Oikonen V, Gronroos T, Eskola O, Kalliokoski K, Bergman J et al. Imaging of blood flow and hypoxia in head and neck cancer: initial evaluation with [15O]H2O and [18F]fluoroerythronitroimidazole. Nuclear Medicine. 2001; 42: 1643–52. 76. Rasey JS, Koh Wui-Jin J, Evans ML, Peterson LM, Lewellen TK, Graham MM et al. Quantifying regional hypoxia in human tumours with positron emission tomography of [18F]fluoromisonidazole: a pretherapy study of 37 patients. International Journal of Radiation Oncology, Biology, Physics. 1996; 36: 417–28. 77. Minn H, Clavo AC, Grenman R, Wahl RL. In vitro comparision of cell proliferation kinetic and uptake of tritiated fluorodeoxyglucose and L-Methionine in squamous cell carcinoma of the head and neck. Journal of Nuclear Medicine. 1995; 36: 252–8. 78. Van Eijkeren ME, Thierens H, Seuntjens J, Goethals P, Lemahieu I, Strijckmans K. Kinetics of [methyl11C]thymidine in patients with squamous cell carcinoma of the head and neck. Acta Oncologica. 1996; 35: 737–41. 79. Chao KSC, Bosch WR, Mutic S, Lewis JS, Dehdashti F, Mintun MA et al. A novel approach to overcome hypoxic tumour resistance: Cu-ATSM-guided intensity modulated radiation therapy. International Journal of Radiation Oncology, Biology, Physics. 2001; 49: 1171–82. 80. Okazawa H, Naito Y, Yonekura Y, Sadato N, Hirano S, Nishizawa S et al. Cochlear implant efficiency in the preand post-lingually deaf subjects. A study with H2(15)O and PET. Brain. 1996; 119: 1297–306. 81. Schoder H, Yeung HWD, Gonen M, Kraus D, Larson SM. Head and neck cancer: Clinical usefulness and accuracy of PET/CT image fusion. Radiology. 2004; 231: 65–72. 82. Branstetter BF, Blodgett TM, Zimmer LA, Snyderman CH, Johnson JT, Raman S et al. Head and neck malignancy: is PET/CT more accurate than PET or CT alone? Radiology. 2005; 235: 580–6. 83. Syed R, Bomanji JB, Nagabhushan N, Hughes S, Kayani I, Groves A et al. Impact of combined 18F-FDG PET/CT in head and neck tumours. British Journal of Cancer. 2005; 92: 1046–50.



55 Image-guided surgery, 3D planning and reconstruction GHASSAN ALUSI AND MICHAEL GLEESON



Introduction Background and overview Image reconstruction Visualization of image data Surgical planning Surgical simulation



701 701 703 703 704 704



Image-guided surgery Key points Deficiencies in current knowledge and the areas for future research References



705 709 709 709



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words image-guided surgery, endoscopic sinus surgery and skull base surgery.



INTRODUCTION



BACKGROUND AND OVERVIEW



Image-guided surgery (IGS) has developed as a result of advances in computer science, digital scanning and image processing. It is a relatively new technology that provides surgeons with information before and during an operative intervention. IGS is an important part of modern surgical practice in disciplines that include neurosurgery, otorhinolaryngology, maxillo-facial and orthopaedic surgery. Synonyms for this technology include computeraided surgery, navigational surgery and computer-guided surgery. With this technology, magnetic resonance (MR), computed tomography (CT) or combined image data sets are used to create three-dimensional (3D) reconstructions of the operative volume. These 3D reconstructions can be used to plan surgery (surgical planning), practise a certain surgical procedure (surgical simulation) or to navigate during the surgical procedure (image-guided surgery). Computer-aided surgery is the umbrella term that is sometimes used to encompass all these processes.



It has been possible to generate 3D models of a specific anatomical region from two-dimensional (2D) CT data sets for some time (Figure 55.1).1 At the outset, these 3D models were used to make measurements, drive numerically controlled milling machines to create physical models and to view a patient’s anatomy and disease process from every aspect.2 The extraction of specific structural detail (segmentation) from the data became possible with the development of sophisticated imageprocessing software and the advantages of preprocessing data in this way became apparent very quickly. It was then only a matter of time before this technology was used to facilitate live surgery. In the operating theatre, images were presented to the surgeon as 2D slices within the three orthogonal planes of space, or as 3D reconstructions displayed on a 2D computer screen. Surgical tools were developed to enable the localization of anatomical features or to confirm the precise position of a pointer within the operative field.



702 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



Figure 55.1 3D reconstruction of CT image data showing the involvement of the skull base by an extensive chordoma (arrowed).



Surgeons appreciated that this information was particularly useful in those situations where the disease process or previous surgery has distorted the normal anatomy.3 Since then, alternative methods of image display have been developed that include image overlays on the surgical field visible through the operating microscope.4 These computer-generated images may be vector representations of previously segmented and preprocessed data. Alternatively, part of the 3D data set can be superimposed on the patient’s anatomy, thus providing the surgeon with an extraordinary ‘x-ray’ type vision (Figure 55.2).5 Augmented reality is an allied area of research that has yet to fulfil its true potential, but will surely replace conventional IGS in the not too distant future.4 Preoperative imaging serves two distinct purposes. First, it is a diagnostic tool. Second, it is used to assess the extent of disease. By careful scrutiny of the images, the surgeon becomes familiar with the anatomy and surgical needs of his patient. The amount of preoperative imaging for complex cases has increased significantly over recent years. Surgeons no longer have to rely so much on their anatomical acumen and intuition in difficult situations. Nowadays, complex and thorough preoperative image processing removes some of the guess work in surgery and utilizes to the full all the information that is acquired. It is hoped that at the least this makes surgery safer and the training of future generations of surgeons better.



Figure 55.2 Vestibular schwannoma. (a) MR scan; (b) intraoperative view through the microscope with the segmented image data overlain on the operative scene.



Surgeons nowadays plan their approach using data obtained from various imaging modalities. Maxillofacial surgeons were among the first to make use of computers for surgical planning.6 They used 3D reconstruction of CT scans of the skull to drive a numerically controlled milling machine that produced prostheses for patients, precise replicas of their normal tissues. The outcome was better fitting prostheses and surgery that took a fraction of the time required by more traditional methods where the prostheses were hand-made during the operative process.6 Orthopaedic surgeons currently use 3D data sets to drive robots designed to perform accurate drilling of the femur or tibia during hip and knee replacements. Robots controlled by computers using patient specific 3D data sets are extremely accurate. The result is better fitting prostheses and fewer failures and morbidity for patients.7 Image-guided surgery, when used by trainee surgeons in endoscopic sinus surgery, improves surgical accuracy and reduces the risk of major intracranial or intraorbital



Chapter 55 Image-guided surgery, 3D planning and reconstruction



complications that might be caused by inexperience. It enhances surgical efficiency and accelerates the learning curve by reducing operative time while maintaining greater than 90 percent accuracy in identifying critical anatomical landmarks.8 In the sections that follow, we discuss the principles and elements of IGS so that the user can better understand the technology and jargon that accompanies it.



IMAGE RECONSTRUCTION The creation of a 3D data set from 2D sections is performed by software using highly complex algorithms that are specific to the particular scanner and scanning protocol employed. Powerful computer workstations are required to run these software programs. This process is automated and requires little clinician input. It is important to consider and understand the various aspects of 3D reconstruction as it has a direct effect on accuracy and visualization.



The concept of slices Conventional CT scan converts one-dimensional x-ray data into a 2D slice. This is referred to as a single image section and may be acquired along axial, coronal or other planes by changing the gantry angle. Tilting the gantry tilt is extremely useful as sections can be acquired that avoid predictable artifacts, for example those caused by dental restorations. Subtle changes to the gantry angle can also help show certain anatomical features better. Each individual section has a width known as a slice thickness, a factor that pertains to the width of the x-ray beam detector window used. Smaller width beams (windows) produce higher resolution 2D images. Slices are spaced along the area to be scanned with an inter-slice distance that varies depending on the scanning protocol being used and on the size of the anatomical structure that is being imaged. Spiral and multi-slice CT scans are faster and expose the patient to less radiation than their traditional counterparts. They collect x-ray data in a helical fashion that is reconstructed later into 2D sections or indeed 3D volumetric voxel data sets. The resulting data set from a spiral CT scan may be reformatted along the sagital, coronal or axial planes, as well as emulating any gantry tilt.



] 703



gives a high definition picture, whereas in a large field this gives a low resolution. By convention, the image resolution is stated by the number of pixels in the x and y axis only and the size of the image field will not be actually defined. Each pixel has a value ranging, in the case of the CT scan, from –1000 to 3096. This is often referred to as Hounsfield units (HU). CT scans are calibrated such that the value of 0 equates to the density of water. By convention, high pixel values are displayed as white. The lower the value, the lower the density of the tissue and the darker the pixel would appear. Pixels are 2D, but it is possible to convert these picture elements into a 3D block. This 3D block is referred to as a voxel or a volumetric picture element. Voxels can be a cube or cuboid depending on the inter-slice distance.



Volume averaging If a structure falls partially within a pixel, the true value of that structure will then be assigned a value less than the normal value of the structure. In the case of bone, which has a value of say 1000 HU, the value of a pixel that includes a part of that part of the bone will be much less, perhaps 500. This is referred to as partial volume averaging and is a major problem in delineating where the edge of a structure should lie within a given image. Higher resolution images would overcome that problem to some extent. This is not always possible as higher x-ray dosages would be required and an order of magnitude increase in the complexity of the hardware. Algorithms have been developed to overcome this problem and define the edge of a structure with a greater degree of confidence. Each one has its own unique problems and for clinical use in the area of neurosurgery and otorhinolaryngology the accuracy difference is submillimetric and can be ignored.



VISUALIZATION OF IMAGE DATA Many factors have to be considered when deciding which visualization technique to use. The clinical purpose for which the data is to be used is the most significant and the image data are rendered for that purpose. Rendering is the process of generating computer images which represent 3D anatomy with some degree of tissue transparency if that is appropriate.



Surface rendering Pixels and voxels Two-dimensional slice sections are composed of pixels which are the smallest picture elements. The more pixels in a certain distance, the better the resolution. Currently, the highest resolution available for both CT and MRI is 512  512 pixels per image. In a field that is 30  30 cm, this



Optical surface scanners produce structured 3D coordinate point sets that relate to an anatomical surface.9 Triplets of data points are generally grouped as the vertices of adjacent triangles that interconnect to make up the entire surface. These triangular surface patches are known as facets or polygons.



704 ] PART 11 RECENT ADVANCES IN TECHNOLOGY The rendering of surfaces from volume data requires preprocessing. The properties of voxels containing anatomical surfaces must first be decided and voxels with these properties can then be processed. Usually the voxels are considered to lie on a surface if they are connected and all have the same associated property within a given range of values. The derived surfaces are called isosurfaces and are selected to correspond to the surfaces of anatomical structures or to surfaces of equal functional activity. Geometric primitives are also derived from volumetric data by processes such as contour tracing, surface extraction or boundary following. Alternatively, voxels belonging to anatomical parts may be isolated from the full data set by applying thresholds to the data values associated with them. This extraction of tissue topology has become known as ‘segmentation’. The derived geometric primitives (such as polygon meshes or contours) are rendered for display using conventional computer graphics techniques. Figure 55.3b shows the rendering of a facial surface that is made up of a large set of triangular facets that are also shown in Figure 55.3a. This type of image can be produced using data from an optical scan of the face. Volume rendering can be undertaken for notional surfaces between distinct tissue types. The surface normals are derived directly from the voxel values neighbouring the boundary. These derived surface normals are used for calculating the final image using algorithms. A technique of this sort was used to produce the image depicted in Figure 55.3c.



Volume rendering Volume rendering has become the most commonly used method with which to visualize 3D medical image data. The basic idea is that 3D volumes are composed of voxels and that these are analogues of their 2D counterparts, the pixels. Fine details throughout a volume of interest are displayed, enabling a more direct understanding of visualized data with fewer artifacts. This visualization technique works by projecting each voxel on to a viewing plane with a value related to the physical property represented in the voxel array. For example, a voxel containing bone with a high x-ray absorption coefficient might be projected with a high value. The most advanced systems allow the operator to construct a look-up table which relates the physical value associated with the voxel to the value it contributes to the image at the chosen viewing plane. A pixel in the viewing plane usually receives contributions from many voxels and the operator may control the manner in which these contributions are composed. For example, the operator may choose to display only the maximum contribution from any voxel along a ray. This produces an image known as the maximum intensity projection (MIP). On



the other hand, any individual voxel-associated value may be assigned a maximum opacity value to produce the same images as would be produced by surface rendering. Figure 55.4a and b show two volume-rendered images with different look-up tables. These demonstrate how different anatomical structures may be made visible through others which are being rendered transparent. Major blood vessels may, for example, be effectively rendered visible within the anatomical data set. Generally, a volume-rendered image appears different from that of a surface rendered image in that anatomical structures are presented as having some degree of transparency. For some clinical procedures, such as imageguided biopsy or transcutaneous thermal ablation, transparency may greatly enhance depth perception and thus increase the accuracy of the procedure. It is apparent that surfaces are not explicitly rendered, but human perception reconstructs images that are perceived to be in the correct spatial relationship. The transparency which volume rendering offers also enables the placement of surgical instruments within 3D structures with great accuracy.10



SURGICAL PLANNING Surgical planning is usually carried out in the surgeon’s mind, drawing on past experience, knowledge of anatomy and anatomical variation. Some individuals are able to do this very well, others are not. In endoscopic sinus surgery, the surgeon must have the ability to think in 3D, having studied the information available from 2D CT slices. Surgeons may or may not be able to communicate their impressions to other surgeons. Furthermore, their mental image cannot be reviewed, audited or measured by other surgeons or indeed by the surgeon themselves when assessing patients at a later date. Planning using reconstructed 3D data enables the surgeon to assess the patient’s anatomy objectively, to communicate this with other surgeons and to review these films at a later date. The data can be segmented and manipulated to familiarize the surgeon with specific features pertinent to that patient and the procedure to be performed.



SURGICAL SIMULATION The major driving force behind the development of surgical simulators was the morbidity that accompanied the introduction of new techniques such as endoscopic ‘key hole’ surgery. These techniques were conceptually different from traditional surgery. Surgeons had to operate using a 2D image on a television monitor and had to develop new hand–eye coordination skills not previously required for open surgery. Training surgeons in these new skills and the assessment of their competence was entirely different from the ‘apprentice–master’ approach that had been the mainstay of traditional surgical teaching.



Chapter 55 Image-guided surgery, 3D planning and reconstruction



] 705



Figure 55.3 (a) Wire mesh; (b) polygon-rendered; and (c) volume surface.



Three-dimensional reconstructions of image data can be used for surgical simulation. Simulators that create a virtual surgical environment or a set of tasks that enhance hand–eye coordination skills seem to be the answer.11, 12 Surgical simulators are now being developed and employed in other surgical specialties, for example skull base surgery. They are being used by trainees and experienced surgeons who might find it useful to practise a specific difficult surgical technique before performing the actual operation. Surgeons are able to ‘virtually’ manipulate, move and drill the mastoid process, to



perform laparoscopic or endoscopic sinus procedures. The sensations and resistance encountered during real surgery may be simulated using positive feedback mechanisms in the form of a passive robotic arm.



IMAGE-GUIDED SURGERY A thorough knowledge of anatomy is essential for all types of surgery, particularly those using microscopes and endoscopes. Preoperative imaging alerts the surgeon to



706 ] PART 11 RECENT ADVANCES IN TECHNOLOGY Both processes are potentially subject to significant mathematical error and it is vital that how this might happen is fully appreciated.



Registration



Figure 55.4 Volume-rendered data with different look-up tables



variations of anatomy that are either inherent in the patient or caused by a disease process or previous surgery. As stated before, image guidance offers the surgeon the ability to use the patient’s image data intraoperatively to determine position, distance to vital organs and other anatomical features that might be hidden from direct vision. Structures in the vicinity of the skull base are either embedded or closely adherent to bone, making this region ideal for image guidance as CT has great spatial accuracy. Two fundamental processes are required for intraoperative image guidance, registration and tracking. Registration is the process that relates the patient in the operating theatre to preoperatively acquired image data sets. Tracking is the mechanism of following the position of the patient or an instrument within the operative field.



Once the patient is secured on the operating table, registration of the Cartesian coordinates of the CT scan to that patient in the theatre is achieved by one of two methods. The first is by locating anatomical landmarks visible both on the patient and image data using a probe that is visible to the tracking device. For example, using navigation software, the surgeon would point to prespecified landmarks, such as the tragus, outer canthus and nasion. This is usually sufficient to achieve a registration accuracy or error of 3–4 mm. This can then be improved by entering a random sample of points from around the operative region (typically 40–100 points) using the probe. The position of the tip of the probe is identified by the tracking device and the coordinates are fed back to the navigation software. In this way, accuracy/errors of 2–3 mm can be achieved. Other methods of registration use masks13 and laser scanning tools (Figure 55.5).14 These speed up the acquisition of surface points. With all these methods, calculations made in real time indicate the point accuracy. It is then up to the surgeon to reject or accept that point in order to improve the overall registration accuracy. Some types of IGS systems make use of fiducial markers that are applied to the patient before scanning on the day of surgery. These markers are clearly visible in both MR and CT data sets and, of course, on the patient when in theatre. They help to improve the accuracy of registration. No matter whether skin fiducials or anatomical landmarks are used for registration, these reference points should describe or be immediately adjacent to the surgical field. Failure to pay attention to this results in navigational inaccuracies. It is important to realize that the registration error is only a measure of the accuracy of correlation between selected points in the virtual data set and the fiducial markers or anatomical landmarks identified on the patient. It is not synonymous with target error, which is the error that could be expected if a probe was placed on a random point of interest within the surgical field. The target error is influenced by the registration error and is unlikely to be much worse if the target is within the volume described by the fiducial markers. However, if the point is outside that volume, the target error will be proportionately greater.



Tracking Sensors that provide dynamic positional information are known as ‘tracking devices’. They are employed in a multitude of everyday situations that range from virtual reality games to missile guidance and electronic tagging of



Chapter 55 Image-guided surgery, 3D planning and reconstruction



] 707



very precise, consistently accurate, fast enough to provide more than 25 readings per second, be insensitive to changes in air temperature, unaffected by metal objects and able to track two objects simultaneously. Several tracking systems have been employed over the years. Some of the earliest devices were mechanical arms fitted with potentiometers at every joint. These were fast and accurate but were cumbersome, had a restricted range of movement and hindered the movement of the tracked object to which it had to be attached. Systems based on magnetic field distribution are effective and relatively cheap. However, their disadvantage is that tracking accuracy can be both easily and significantly affected by the presence of metal objects. Infrared light sensors are the most commonly used image-guided surgical systems. They are referred to as either active or passive devices. Active devices sense infrared light from light-emitting diodes (LED) attached to the patient or location probe. Passive tracking devices detect infrared light reflected from metallic balls attached to the patient or probe. In the latter case, the infrared light source is situated on the sensing device itself (Figures 55.6 and 55.7). In order to detect changes in the patient’s head position, haloes or arches with lightemitting diodes or reflective spheres can be fitted to a Mayfield clamp or the patient’s head. Systems based on light-emitting diodes are reliable if the light path is not impeded.16 Within the operating theatre an accuracy of 2–5 mm can usually be achieved.4 In general, optical tracking systems are expensive. Cheaper variations do exist, but they lack high degrees of accuracy. Inertial trackers have also been developed that are small and accurate enough for virtual reality applications. These devices generally provide one rate of change of rotational measurement only; hence, a number would be



Figure 55.5 (a) A laser registration being undertaken before endoscopic sinus surgery. (b) The laser beam is shone on to the skin around the operative site. (c) A band on to which fiducial balls have been mounted is attached to the patient’s head.



prisoners.15 Although these measurement acquisition systems may be similar, each has different tracking properties that are optimal for a particular application. Without some form of tracking device, it would not be possible to move the patient once they had been registered. In surgical practice, tracking systems must be



Figure 55.6 A patient with their head held in a Mayfield clamp being prepared for image-guided surgery. An arc with metallic ball fiducials has been attached to the clamp. The arc describes the operative volume. A laser light is being centred on the field to find the optimal position for the tracking device.



708 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



Figure 55.7 An infrared optical tracking system. In the centre of the gantry is a laser light that is used to position the arm during the set-up procedure.



required. They are also not accurate for slow position changes. Systems based on standard ultrasound signals have the potential to achieve much greater accuracy, but are susceptible to changes in temperature and air currents. Other drawbacks include long lag times and interference from echoes and other noises in the environment.



Clinical applications Image-guided surgery offers most promise in the areas of skull base and endoscopic sinus surgery. Skull base



surgeons have found this technology particularly useful for preoperative surgical planning, design of bone flaps, identification of important structures and for finding small tumours in obscure parts. For example, in Figure 55.8, the guidance system had been used to maximize the access for correction of basilar invagination in a patient with osteogenesis imperfecta. The normal anatomical arrangements and landmarks were not present as a result of the disease process. In rhinology, IGS systems are already in use in difficult revision endoscopic sinus surgery where there are no recognizable anatomical landmarks (Figure 55.9).17 In particular, surgeons have found the guidance systems useful to localize the frontal recess during Draf type 2 and 3 procedures, where the floor of the frontal sinus is to be opened.18 With the help of these systems, the need to perform open, external procedures will become less. Extended applications of endoscopic sinus surgery include trans-sphenoidal, trans-nasal endoscopic hypophysectomy where conventionally x-ray image intensifiers have been used to localize the sella.19, 20, 21 In the field of otology, IGS has been used to aid the surgeon in locating the facial nerve, identifying and localizing lesions of the petrous apex22 and tumours, particularly in the internal auditory meatus. These include meningiomas and vestibular schwannomas.23, 24 It is possible that when these systems gain in acceptance, are more accurate and involve shorter set-up times, they may become mainstream intraoperative equipment in all types of mastoid surgery. Their role then would be to aid the surgeon in localizing important atomical structures such as the facial nerve, dura, brain and jugular bulb, as well as diseases, such as cholesteatoma and acoustic schwannoma.



Figure 55.8 Intraoperative screen view during correction of basilar invagination. The cross lines of the cursor are centred over the displaced odontoid peg.



Chapter 55 Image-guided surgery, 3D planning and reconstruction



$ $



$ Figure 55.9 Functional endoscopic surgery being undertaken using image guidance.



] 709



At present, augmented reality still has to prove its position in surgical practice. Several research groups are producing encouraging results and future reliance on these types of systems is likely to replace traditional IGS systems. Robots are already being used for drilling in the field of orthopaedics and to manoeuvre and control cameras in laparoscopic surgery. These are far more accurate than humans and are likely to enter other fields of surgery where there is a reliance on power instrumentation, including drills and debriders. In the future, we are likely to use IGS systems to audit surgical procedures as they already have the inherent capacity to store all information relating to intraoperative navigation. This will be an invaluable and, perhaps, incriminating operative record.



KEY POINTS  IGS aids the surgeon in diagnosing the extent of disease, planning surgery and providing intraoperative positional information.  Image-guided surgical techniques may reduce complications in endoscopic sinus surgery.  IGS techniques offer greater accuracy and confidence in structure and disease localization.  Major applications are skull base and revision nasal surgery where there is complex or distorted anatomy.



REFERENCES



 Deficiencies in current knowledge and the areas for future research



$



$ $



Image-guided and augmented reality surgery relies heavily on technology research in the areas of computer hardware (processors, graphics and tracking) and software for segmentation and registration. Developments in these fields are driven by a variety of customer-driven demands. It is vital that clinicians make these demands and provide ideas for the computer scientists. The most important requirements are for accurate, robust, fail-safe systems that will give clinicians greater confidence. These systems must be userfriendly and ergonomically efficient. In the future, we are likely to see more developments in the areas of surgical simulation and planning. Surgical training should benefit from simulators that will enable the trainee to develop their hand–eye coordination skills prior to applying them in clinical practice.







1. Herman GT, Liu HK. Three-dimensional display of human organs from computed tomograms. Computer Graphics and Image Processing. 1979; 9: 1–21. 2. Linney AD, Tan AC, Richards R, Gardner J, Grindrod S, Lees WR. The use of three dimension data on human body for diagnosis and surgical planning. Revista di Neuroradiologia. 1992; 5: 483–8. 3. Olson G, Citardi MJ. Image-guided functional endoscopic sinus surgery. Otolaryngology – Head and Neck Surgery. 2000; 123: 188–94. 4. Edwards PJ, Hawkes DJ, Hill DL, Jewell D, Spink R, Strong A et al. Augmentation of reality using an operating microscope for otolaryngology and neurosurgical guidance. Journal of Image Guided Surgery. 1995; 1: 172–8. 5. Lapeer RJ, Tan AC, Alusi G, Linney AD. Computer-assisted surgery and planning using augmented reality (CASSPAR): Visualisation and calibration. In: 2003 IEEE/ACM International Symposium on Mixed and Augmented Reality. Washington: IEEE Computer Society, 2005: 272–3. 6. Vannier MW, Marsh JL, Warren JO. Three-dimensional computer graphics for craniofacial surgical planning and evaluation. ACM SIGGRAPH Computer Graphics Quarterly. 1983; 17: 263–73. 7. Kienzle III TC, Stulberg SD, Peshkin MA, Quaid A, Lea J, Goswami A et al. A computer-assisted total knee replacement surgical system using a calibrated robot. In: Taylor RH, Lavalle´e S, Burdea GS, Mo¨sges R (eds). Computer integrated surgery. Cambridge, MA: MIT Press, 1996: 409–16. 8. Casiano RR, Numa Jr. WA. Efficacy of computed tomographic image-guided endoscopic sinus surgery in residency training programs. Laryngoscope. 2000; 110: 1277–82.



710 ] PART 11 RECENT ADVANCES IN TECHNOLOGY 9. Cline HE, Dumoulin CL, Hart Jr. HR, Lorensen WE, Ludke S. 3D reconstruction of the brain from magnetic resonance images using a connectivity algorithm. Magnetic Resonance Imaging. 1987; 5: 345–52. 10. Jolesz FA, Kikinis R. Intraoperative imaging revolutionizes therapy. Diagnostic Imaging. 1995; 17: 62–8. 11. Caversaccio M, Eichenberger A, Hausler R. Virtual simulator as a training tool for endonasal surgery. American Journal of Rhinology. 2003; 17: 283–90. 12. Bockholt U, Muller W, Voss G, Ecke U, Klimek L. Real-time simulation of tissue deformation for the nasal endoscopy simulator (NES). Computer Aided-Surgery. 1999; 4: 281–5. 13. Albritton FD, Kingdom TT, DelGaudio JM. Malleable registration mask: application of a novel registration method in image guided sinus surgery. American Journal of Rhinology. 2001; 15: 219–24. 14. Raabe A, Krishnan R, Wolff R, Hermann E, Zimmermann M, Seifert V. Laser surface scanning for patient registration in intracranial image-guided surgery. Neurosurgery. 2002; 50: 797–801; Discussion 802–3. 15. Watanabe K, Kobayashi K, Munekata F. Multiple sensor fusion for navigation systems vehicle navigation and information systems. In: Vehicle Navigation & Information Systems Conference, 1994 Proceedings. Washington: IEEE Computer Society, 1994: 575–8. 16. Azuma R, Bishop G. Improving static and dynamic registration in an optical see-through hmd. In: Proceedings of the 21st Annual Conference on Computer



 17.  18.



19.



20.



21.



22.



23.



24.



Graphics and Interactive Techniques. New York: ACM Press, 1994: 197–204. Metson R. Image-guided sinus surgery: lessons learned from the first 1000 cases. Otolaryngology – Head and Neck Surgery. 2003; 128: 8–13. Neumann Jr. AM, Pasquale-Niebles K, Bhuta T, Sillers MJ. Image-guided transnasal endoscopic surgery of the paranasal sinuses and anterior skull base. American Journal of Rhinology. 1999; 13: 449–54. Otori N, Haruna S, Yoshiyuki M, Moriyama H. Endoscopic endonasal surgery with image-guidance. Nippon Jibiinkoka Gakkai Kaiho. 2000; 103: 1–6. Kajiwara K, Nishizaki T, Ohmoto Y, Nomura S, Suzuki M. Image-guided transsphenoidal surgery for pituitary lesions using Mehrkoordinaten Manipulator (MKM) navigation system. Minimally Invasive Neurosurgery. 2003; 46: 78–81. Sandeman D, Moufid A. Interactive image-guided pituitary surgery. An experience of 101 procedures. Neurochirurgie. 1998; 44: 331–8. Van-Havenbergh T, Koekelkoren E, De-Ridder D, Van-DeHeyning P, Verlooy J. Image guided surgery for petrous apex lesions. Acta Neurochirurgica. 2003; 145: 737–42; Discussion 742. Sargent EW, Bucholz RD. Middle cranial fossa surgery with image-guided instrumentation. Otolaryngology – Head and Neck Surgery. 1997; 117: 131–4. Klimek L, Mosges R, Schlondorff G, Mann W. Development of computer-aided surgery for otorhinolaryngology. Computer-Aided Surgery. 1998; 3: 194–201.



56 Ultrasound in ear, nose and throat practice KESHTHRA SATCHITHANANDA AND PAUL S SIDHU



Introduction Basic physics and equipment design Programme software technology advances Transducer selection Doppler ultrasound Microbubble contrast agents Advantages of ultrasound Disadvantages of ultrasound Examination technique Ultrasound appearances of common abnormalities



711 711 713 713 714 714 715 715 715 715



Reliability of ultrasound examination of thyroid nodules Cystic lesions Soft tissue neoplasms Conclusion Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



721 724 725 727 727 727 727 727



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words ultrasound or imaging and neck, head and neck neoplasm, thyroid diseases, parathyroid masses and salivary gland disorders.



INTRODUCTION



BASIC PHYSICS AND EQUIPMENT DESIGN



Ultrasound has achieved wide clinical use over the past half century. It is a crucial imaging modality in medical practice and now accounts for almost a quarter of all imaging studies carried out in the world. The equipment used has evolved from large expensive Bmode gantry systems with coarse, static, bistable displays, which only demonstrated organ boundaries, to convenient handheld machines with high resolution, real-time imaging with colour Doppler facility making use of digital technology. This chapter will discuss some of the basic physics and principles of both the hardware and software of the ultrasound machine and of real-time B-mode imaging. An overview of the benefits and pitfalls of the more important applications in ear, nose and throat (ENT) practice will be covered, with an emphasis on new advances and future developments.



Ultrasound is defined as sound with a frequency above the upper limit of normal hearing (over 20 kHz). In medical practice, much higher ultrasound frequencies are used (2.5–30 MHz) to produce an image based on the detection and display of sound waves reflected from interfaces within the body. Ultrasound produces tomographic images similar to computed tomography (CT) and magnetic resonance imaging (MRI); however, unlike CT where x-rays are transmitted through the patient to produce an image on the other side of the patient; ultrasound images are formed from reflected sound waves. Sound waves are generated in short bursts by the transducer (or probe) and the sound energy that is reflected back is collected at the point of origin (the transducer) during the relatively long intervals between the sequential generation of sound waves. The ultrasound image is based on the back



712 ] PART 11 RECENT ADVANCES IN TECHNOLOGY scattering of sound energy by interfaces of tissues with different physical properties through interactions governed by acoustic physics; the amplitude of the reflected echoes is used to generate the contrast of the different tissues to form the image. Coaxial cable



Transducer design B-mode (brightness mode) real-time imaging is most commonly used in medical practice. The images are formed line by line from the information within the transducer that both transmits and receives ultrasound signals from the body’s tissues. A typical transducer comprises five layers (Figure 56.1): 1. a protective layer; 2. mechanical lens (‘Hanafy’ lens) that allows you to focus the beam in a particular elevation plane; 3. matching layers that overcome the large acoustic impedance mismatch between the ultrasound crystal and the patient’s tissues (the acoustic impedance of a material is a measure of how efficiently the sound waves are conducted through the tissue); 4. an active piezoelectric material to produce the sound waves. In practice, this is usually lead zirconate titanate (PZT); 5. a backing block to stop the crystal from freely resonating and thus producing poor echoes that have inadequate axial resolution and are not useful in producing a diagnostic image.



Plastic housing



Electrical leads



Rear electrode



Backing layer PZT plate Matching layer Lens Front electrode



PHYSICS OF ULTRASOUND



Sound is mechanical energy travelling through matter as a wave producing alternating areas of compression and rarefaction.1 Sound obeys the rules of basic acoustic physics as summarized in the following equation: Velocity ðcÞ ¼ frequency ðf Þ  wavelength ðlÞ In the clinical setting, the ultrasound transducer generates sound waves in short bursts or pulses of energy that are transmitted into the body where they are propagated through tissues. These acoustic pressure waves may behave either as a transverse wave (travel in a direction perpendicular to the direction of the displaced particle) or as a longitudinal wave (in the same direction as the particle movement). In tissues and fluids, in medical practice, sound behaves in a longitudinal fashion. Propagation velocity The speed at which the pressure waves travel through different tissues varies greatly and is affected by the physical properties of the tissues, with this velocity largely



Figure 56.1 Transducer design. Schematic drawing illustrating the design of an ultrasound transducer.



dependent on the resistance of the medium to compression, as well as the tissue stiffness or elasticity. Increasing the stiffness or decreasing the density of the material will increase the propagation velocity. In the body, the propagation velocity of sound is assumed to be constant, and is quoted at 1540 m/s. This value is the average of measurements for all the tissues and is applicable to most soft tissues of the body. However, it should be noted that some tissues, such as aerated lung and fat, have a much lower velocity than 1540 m/s, whilst others, such as bone, have a much greater velocity. In practice, this property of propagation velocity is critical as it is used to evaluate the distance of a reflecting surface from the transducer. By measuring the time it takes for an echo to return to the probe and knowing the velocity with which the wave travelled, the distance it traversed (which is the depth of the reflecting tissue interface from the probe) can be calculated.



Chapter 56 Ultrasound in ear, nose and throat practice



Acoustic impedance Ultrasound images are formed by detecting and displaying backscattered (reflected) sound waves from a reflecting surface within tissue. Sound travelling in any homogenous medium will not have such a reflecting interface and will consequently appear anechoic (black) as is seen with a simple cyst. At the junction between different tissues with different physical properties, an acoustic interface will be present. In the body, there are many of these interfaces, each of which will produce varying amounts of reflected sound that is detected by the transducer. The strength of the returning sound wave echo is determined by the difference in acoustic impedances of the materials forming the interface. The acoustic impedance is defined by the following formula: Acoustic impedance ðZÞ ¼ density ðpÞ  propagation velocity ðcÞ



Interfaces with large differences in acoustic impedances such as soft tissue/air or soft tissue/bone, reflect almost all the incident sound waves, producing returning echoes of higher amplitude, which will form better images. In tissues where there are smaller differences in acoustic impedances, such as muscle and fat, only part of the incident sound wave is reflected and the remainder is allowed to continue on through the material; the returning sound wave echoes will be of smaller strength. It is important to note that both the propagation velocity and acoustic impedance are determined by the physical properties of the tissues and are independent of the frequency of the waves.



Image generation The foundation of modern ultrasound imaging is with real-time, grey scale, B-mode display where the intensity/ brightness of each pixel of the image represents the different amplitudes of the returning sound wave echoes. New digital technology with the consequent improvements in the software of the modern ultrasound machine has allowed improvement of image quality. To create a two-dimensional (2D) image, multiple ultrasound pulses are sent down a series of successive scan lines, building up a 2D representation of echoes originating from the object being insonnated. The image is normally displayed on a black background, with the greatest amplitude signals appearing as white and no signal as black. Signals of intermediate intensities are displayed as varying shades of grey. In modern ultrasound machines, a digital memory of 512  512 or 512  640 pixels is used to store values corresponding to the values of the returning echoes, as well as their position in the patient. The digitally stored image can then be displayed on a video monitor by converting the digital signal to an analogue format. The dynamic range of the information stored in the digital



] 713



memory is much larger than that which can be seen on a video monitor or be appreciated by the human eye. This information needs to be compressed, a technique called ‘post-processing’, the purpose being to optimally view the information that is likely to be of diagnostic value and selectively eliminate less useful information stored in the memory. One important recent advance is the use of application specific integrated circuits (ASIC), which has led to improving the complexity, speed and reliability of the machines whilst reducing the cost and machine size with lower power consumption. Manufacturers have also developed new techniques to optimize the images by allowing the machine to manipulate the echo signals and therefore reducing the operator-dependent nature of ultrasound. Real-time ultrasound gives the impression of a moving image by generating a series of 2D images at a rate of 15–60 frames per second, permitting the real-time ultrasound assessment of both anatomy and motion. This innovation has made ultrasound a dynamic examination technique.



PROGRAMME SOFTWARE TECHNOLOGY ADVANCES Advances in digital technology2 have allowed electronic and software technology improvement at all stages of the ultrasound imaging systems. Spatial compounding is an example of such a development, improving the quality of the final displayed image. With spatial compounding, the beam is electronically steered so that overlapping scans of an object are acquired at different angles. These entire scan lines are averaged to produce a real-time compound image with better spatial resolution than conventional Bmode imaging. A further example of post-processing to improve imaging is ‘coded excitation’. This technique uses the fact that the transmitted pulse is digitally encoded and thus can be modified by changing its frequency or amplitude and the receiver is set to optimally receive this ‘code’ in the returning echoes, allowing weak signals to be distinguished from background noise. This allows imaging at higher frequency (as used in imaging the superficial structures in the neck) with improved spatial and contrast resolution at greater depth. Other recent advances, less applicable to imaging of neck structures, include ‘extended field-of-view’ methods that produce a panoramic image to expand diagnostic capabilities particularly in musculoskeletal ultrasound3 and tissue harmonic imaging, particularly useful in abdominal imaging.4



TRANSDUCER SELECTION When choosing the optimal transducer for a given application, both the requirements for spatial resolution



714 ] PART 11 RECENT ADVANCES IN TECHNOLOGY and the distance of a target from the transducer need to be considered. In general, the higher the frequency of the probe, the less the depth of tissue penetration, but the better the spatial resolution. To optimize the imaging, a probe of the highest frequency is chosen that allows penetration to the depth of interest. The imaging of superficial structures requires a higher frequency probe; in the neck, most of the structures of interest are superficial and probes of greater than 7.5 MHz are used. Occasionally, in the larger individual, deeper penetration is required, usually a 5 MHz probe is used. High frequency linear array probes are ideal for imaging superficial structures as the near-field resolution is better; this allows axial resolution of 0.5 mm or less and lateral resolution of 1.0 mm or less and provide a rectangular image format that is well suited to this application.5 To evaluate the parotid gland area or the floor of the mouth, a curved array or sector transducer may be better suited. Curved array probes make a smaller footprint contact with the skin surface, but give a larger field-of-view at depth as it provides a sector display format.6



DOPPLER ULTRASOUND Doppler ultrasound7 allows the imaging of blood flow in vessels; an important adjunct to grey scale imaging and has allowed greater diagnostic confidence in image interpretation. The Doppler effect is defined as a change in the apparent frequency of a wave as a result of relative movement between the observer and the source. The observer in this context is the ultrasound transducer and the source is the moving red blood cells within a vessel. The change in frequency (Doppler shift frequency) that is detected can be used to assess movement. The Doppler shift frequency is proportional to the ultrasound frequency, the velocity of the moving target and the angle of insonation of the target. These parameters are summarized in the Doppler equation: Df ¼ 2



ft V cos f c



where Df is the Doppler shift frequency; ft is the frequency of sound emitted from the transducer; V is the velocity of the target towards the transducer; c is the velocity of sound in the medium; and f is the angle of insonation (the angle between the axis of flow and the incident ultrasound beam). In B-mode imaging, the pulse sound wave echoes are transmitted and the backscattered signal contains information regarding phase and frequency. Precise timing allows the depth from which the echo originates to be measured. However, the additional information contained from these echoes can be used to evaluate moving targets, such as blood flow. The Doppler systems that are in use are of two kinds: continuous wave and pulsed wave Doppler systems.



Continuous-wave Doppler systems use two transducers mounted within the same housing: one transmits sound continuously and one receives the reflected echoes continuously. The traces obtained by this method can be then analysed (spectral analysis). The Doppler shift frequencies are displayed graphically and it is possible to visualize the time-varying waveform of blood flow, measure objective parameters, such as velocity, pulsatility index (PI) and resistance index (RI). The PI is the peakto-peak value divided by the mean of the peak values throughout the cardiac cycle. Pulsatility index ¼



Peak systolic velocity  End diastolic velocity Mean velocity



The RI is the peak systole minus the end diastole velocity divided by the peak systole velocity. Resistance index ¼



Peak systolic velocity  End diastolic velocity Peak systolic velocity



The RI gives an indication of the degree of resistance to blood flow in the distal circulation. The disadvantage of a continuous wave Doppler system is that the range or depth of the moving structure cannot be calculated. Pulsed-wave Doppler systems overcome this drawback by transmitting short pulses of sound waves and then leaving long intervals between to receive the returning echoes, using a single crystal. The depth of the returning Doppler signal may be calculated by timing the returning signal, with the knowledge of the speed of the sound waves in the tissues. Duplex scanners combine colour Doppler acquisition equipment with a conventional grey-scale imaging system, allowing visualization of the vessel under interrogation and placing the sample gate over the region of interest to obtain a spectral Doppler signal. In modern machines, the switch from 2D imaging to Doppler mode is instantaneous and many machines allow continued Bmode imaging, whilst acquiring the Doppler information, a practical advantage.



MICROBUBBLE CONTRAST AGENTS Ultrasound contrast medium8 is a relatively new technique and is gaining increasing acceptance in improving diagnostic accuracy. Microbubbles are less than 10 mm in diameter, cross capillary membranes and are safe, effective echo enhancers. The agent remains in the vascular compartment and is stable for several minutes after an intravenous bolus. Tissue harmonics is used to image ultrasound contrast agents by tuning the receiver to listen to a band of frequencies that allows the contrast signal to be resolved from the background, allowing visualization of vessels as small as 100 mm, enough to characterize tumour vascularity.9 Tumour vascularity can be further enhanced using 3D displays to demonstrate the vascular anatomy.



Chapter 56 Ultrasound in ear, nose and throat practice



ADVANTAGES OF ULTRASOUND The benefits of ultrasound imaging are well recognized. The key points to be noted are:  Ultrasound does not use ionizing radiation making it ideal for paediatric patients, as well as for patients requiring multiple examinations, as in oncology practice.  Ultrasound is readily available and is relatively cheap.  Ultrasound allows good spatial resolution and is the ‘gold standard’ in determining whether a focal lesion is cystic or solid, surpassing even CT or MRI.  Ultrasound is well tolerated by patients and is a ‘patient friendly’ technique.  Ultrasound provides diagnostic information with a high degree of accuracy, allowing for more efficient use of CT and MRI when choosing how, if necessary, to image the patient further.  An increasingly important role is to guide interventional procedures such as fine needle aspirations, biopsies and drainage catheter insertions accurately and in real time.



DISADVANTAGES OF ULTRASOUND The disadvantages of ultrasound are few but are very important to consider when making clinical management decisions. The primary disadvantage is that the technique is operator dependent. There should be confidence in your operator and he/she should work in a multidisciplinary environment to ensure the highest standards. Though the cost of this examination is relatively low, the initial capital costs need to be considered; the equipment has to be high quality. The size and position of the structures and diseases affecting the neck require highresolution linear array and curvilinear probes with good supporting software to produce images of diagnostic quality. The facility to use colour Doppler is also highly desirable. The images obtained in ultrasound are susceptible to artifact that may cause mistaken diagnoses. The operator has to be aware how to identify, reduce or eliminate such faults from the image. Despite these disadvantages, ultrasound is a valuable tool in the diagnosis and treatment of diseases of the head and neck and is accepted as the first imaging modality after clinical examination.



EXAMINATION TECHNIQUE As many conditions that affect the head and neck occur bilaterally and often in an asymmetrical way, it is imperative that both sides of the neck are examined with a systematic and comprehensive approach.



] 715



It is best to examine the patient supine and to image the right side of the neck with the patient’s head turned to the left and vice versa. The areas that should be covered are the thyroid gland, paravascular spaces, cervical triangles, supraclavicular fossae and the major salivary glands. All these areas need to be examined in at least two orthogonal planes. Colour Doppler ultrasound should also be used to assess if there is involvement of the great vessels or if there is increased vascularity in or around abnormalities.



ULTRASOUND APPEARANCES OF COMMON ABNORMALITIES Lymph nodes Both surgeons and oncologists use the regional classification of lymph nodes into seven levels adapted by the American Joint Committee on Cancer. Importantly, the retropharyngeal lymph nodes cannot be seen with ultrasound (Table 56.1). This group of lymph nodes is important in pharyngeal cancers and in patients who have previously undergone a neck dissection. Ultrasound is commonly used for the evaluation of patients with enlarged cervical lymph nodes. It has been suggested that, of the numerous lymph nodes in the body, over a third are sited in the cervical region.10 Although normal lymph nodes may be visualized by ultrasound in healthy subjects, they are often not seen due to both their small size and their similar echo-texture to the surrounding structures.11 When apparent, lymph nodes are invariably reactive (hyperplastic), inflammatory or neoplastic. Reactive lymph nodes are typically elongated, oval and have smooth margins, predominantly hypoechoic with an eccentric echogenic fatty hilum (Figure 56.2). The upper limit of size of a reactive lymph node in the upper jugular chain is 10 mm in the short axis.12 However, this carries the caveat that in the paediatric population, larger hyperplastic nodes may be encountered.13 Table 56.1 Regional classification of lymph nodes into seven levels, adapted by the American Joint Committee on Cancer. Level Level Level Level Level Level



I II III IV V



Level VI Level VII



Submental and submandibular nodes Upper internal jugular chain nodes Middle internal jugular chain nodes Lower internal jugular chain nodes Spinal accessory chain nodes and transverse cervical chain nodes Anterior cervical nodes Upper mediastinal nodes



See also Chapter 137, Surgical anatomy of the neck, Figure 137.5.



716 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



Figure 56.3 Squamous cell carcinoma lymph node metastasis. An enlarged low reflective mass with an irregular border (long arrow), displacing the surrounding soft tissues, anterior to the carotid artery (short arrow), in keeping with a squamous cell metastasis from a primary laryngeal tumour. Figure 56.2 Reactive lymph node. (a) An oval-shaped, lowreflective lymph node with an echogenic hilum (arrow), in keeping with a reactive lymph node. (b) Florid colour Doppler flow to the central hilum consistent with a benign reactive lymph node.



Inflammatory nodes are enlarged (410 mm), round or oval in shape with well-defined borders and in which the hilum is not readily identified. Colour Doppler ultrasound shows hilar vascularization with smooth branching of the intranodal vessels. Ultrasound may assess for the presence of an abscess formation (hypoechoic or anechoic area with no colour Doppler flow) within a mass of inflammatory nodes that may require surgical intervention.6 Up to 90 percent of cervical lymph node metastases are from squamous cell carcinomas (Figure 56.3) and less commonly from a primary tumour in the thyroid, breast, lung, stomach or from a melanoma.5 Malignant lymph nodes lose the normal ultrasound architecture, have an eccentric hilum, disordered colour Doppler signal and become more rounded (Figure 56.4). Assessment of cervical lymph nodes, the number, level in the neck and the presence of extranodal spread are important prognostic indicators and determine therapeutic options. Most tumours from the upper gastrointestinal tract have a predictable pattern of metastases to the neck depending on the pattern of tumour extent at the primary site. It has been suggested that, if the primary tumour crosses the midline then cervical lymph nodes of both



Figure 56.4 Adenocarcinoma lymph node. Colour Doppler images of a low-reflective, irregular, rounded lymph node with an eccentric echogenic hilum (arrow), infiltrated by adenocarcinoma.



sides of the neck are at risk of metastases.14 [**] The reported prevalence of cervical lymphadenopathy in patients with nasopharyngeal cancer at the time of diagnosis is up to 87 percent.15 Traditionally, staging



Chapter 56 Ultrasound in ear, nose and throat practice



was performed by palpation which has a sensitivity and specificity of 60–70 percent, but importantly the rate of occult metastases, undetectable with palpation, with a T1–3 oropharyngeal/supraglottic tumour is 20–50 percent.16, 17 Ultrasound imaging has a sensitivity of 84 percent, a specificity of 68 percent and an accuracy of 76 percent in detecting abnormal nodes, improved further with the combination of ultrasound-guided fine-needle aspiration cytology (FNAC) to 97 percent sensitivity and 93 percent specificity.18 [***] The important morphological ultrasound characteristics of lymph nodes to ascertain the presence of malignant change are summarized in Table 56.2.19 [***]



] 717



carcinoma of the thyroid, lymph nodes are hyperechoic (87 percent) and homogenous (81 percent). Peripheral echogenic foci are seen in 68 percent of metastatic nodes from papillary carcinoma of the thyroid and have been histologically correlated to psammoma bodies; a characteristic and important feature to recognize.24 With the newer high-resolution transducers, lymphomatous nodes have a reticular echo pattern, but on the older machines these nodes had a ‘pseudocystic’ appearance of low reflectivity with posterior enhancement (Figure 56.5).25



CYSTIC/NECROTIC NODES



The ‘roundness index’ (longitudinal to transverse diameter ratio) of less than 1.5 indicates metastases in 71 percent of nodes. If the ratio is greater than 2.0, then 84 percent of such nodes are inflammatory or reactive.20



The heterogeneous appearance of nodes with cystic spaces, representing areas of necrosis, is seen most commonly with squamous cell carcinoma, but this is a nonspecific finding seen with melanoma, papillary thyroid carcinomas and with tuberculous nodes.26 Cystic necrosis appears as echo-lucent areas within the lymph nodes (Figure 56.6).



LYMPH NODE SIZE



CALCIFICATION



Optimal size is difficult to define as the size of lymph nodes varies with the site in the neck, varying from 5 to 30 mm and the presence of micrometastases does not always cause enlargement.12, 21 Traditionally, the maximum short axis diameter is quoted in such situations and a cut off of 10 mm is widely accepted. However, when all three lymph node diameters are assessed, the minimal axial diameter is the best criteria for enlargement.12 The most acceptable size of minimum axial diameter is 7 mm in the upper internal jugular chain and 6 mm for the rest of the neck.22 It is important to note that the lower the limit criteria for nodal size, the higher the sensitivity in differentiating malignant from benign nodes, but this is at the cost of lower specificity. Conversely, the higher the limit criteria for nodal size, the lower the sensitivity but with increased specificity.



Intranodal calcification is unusual in metastatic nodes, except in papillary cell thyroid carcinoma, where calcification is reported in 69 percent.24 Calcification is also found rarely in irradiated or post-chemotherapy lymph nodes.26



LYMPH NODE SHAPE



ECHOGENIC HILUM



The presence of an echogenic (fatty) hilum is associated with benign disease. Normal lymph nodes have an echogenic hilum in 75–100 percent, whereas it is reported that there is an absent hilum in 96 percent of metastatic lymph nodes,27 63 percent of lymphomatous lymph nodes28 and 83 percent of tuberculous lymph nodes.23



LYMPH NODE ECHOGENICITY



Nodes containing metastases are usually hypoechoic when compared with adjacent muscle.23 In metastatic papillary Table 56.2 Ultrasound morphology of malignant infiltrated lymph nodes. Morphology Round shape Short axis diameter of 8 mm or more Heterogeneous echo-texture Cystic or necrotic areas No echogenic (fatty) hilum Ill-defined margins Invasion of surrounding tissues



Figure 56.5 Lymphomatous lymph node. An enlarged, rounded, low reflective lymph node that maintains a central echogenic hilum (arrow) infiltrated with lymphoma.



718 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



Thyroid gland



Figure 56.6 Tuberculous lymph nodes. Collection of lymph nodes, with necrosis occurring (arrows) in the posterior lymph nodes, manifest as low reflective fluid. The anteriorly located lymph node has the appearance of a reactive lymph node.



However, as an echogenic hilum is found in both benign and malignant nodes, an echogenic hilum alone should not be used to assess cervical lymph nodes.29



VASCULAR PATTERN



Normal and reactive lymph nodes examined with colour Doppler are either avascular or demonstrate hilar vessels with regular branching, whereas metastatic lymph nodes demonstrate tortuous vessels with an aberrant pattern, a peripheral vascular pattern or a mixture of both.30 The accuracy of differentiating metastases from benign reactive lymph nodes using these criteria is 88 percent (sensitivity 89 percent, specificity 87 percent).30 Tuberculous lymph nodes may have a variable vascular pattern on colour Doppler imaging and simulate both benign and malignant conditions.31 Vessels displaced to one side of the lymph node, a relatively common feature in tuberculous lymph nodes (81 percent), is not seen in normal lymph nodes and is uncommon in other pathologies; a useful feature in classifying such lymph nodes.31 Using spectral Doppler ultrasound, comparing the highest PI and RI within suspect lymph nodes, malignant nodes showed low resistance flow. Malignant lymph nodes have a PI Z1.1 or an RI Z0.7, with the accuracy of differentiating malignant from benign lymph nodes estimated at 76 and 73 percent, respectively.30 Ultrasound microbubble contrast medium has also been used to help evaluate the nature of cervical lymph nodes, where smaller vessels in the lymph nodes are more obvious and the typical pattern of vascular distribution can be ascertained with greater clarity.32 A high degree of accuracy (99 percent) and a change of diagnosis in 14 percent have been reported with microbubble contrast-enhanced imaging of lymph nodes, with a particularly useful role in smaller or borderline sized lymph nodes.32 [***]



The thyroid gland is a superficial structure and is favourably sited for ultrasound examination. Highfrequency B-mode and colour Doppler ultrasound is widely used in both the diagnosis and follow-up of patients with thyroid diseases. Examination technique consists of obtaining longitudinal and transverse images of the lower half of the neck from the midline. The thyroid gland is made up of two lobes connected medially by the isthmus, which has a transverse course. A third (pyramidal lobe) arises from the isthmus in 10–40 percent of normal individuals. The parenchyma has a homogenous echo texture, which is of higher reflectivity than adjacent muscle. The thyroid gland is initially assessed for size, which varies with body habitus. Thin subjects have glands that measure 70–80 mm in length and 7–10 mm in depth. Obese patients have glands measuring less than 50 mm in length with an anteroposterior diameter of up to 20 mm.33 Thyroid volumes have also been used to assess size and are calculated from the right anteroposterior (RAP) diameters and left anteroposterior (LAP) lobes using the following formula: Volume of gland ¼ ð6:91  RAPÞ 1ð3:05  LAPÞ  3:48 Thyroid volume for males is 19.6  4.7 mL and for females is 17.5  3.2 mL.34 Volumetric studies can be easily performed using 3D ultrasound, but unfortunately this technique is not widely available.



THYROID DISEASE



For ultrasound imaging, thyroid disorders may be broadly considered in two groups: nodular and diffuse. Almost all diffuse thyroid disease and the majority (90 percent) of thyroid nodules are benign, with thyroid malignancy a rare entity accounting for only 1 percent of all cancers.33



THYROID NODULES



There are five major roles of ultrasound in the assessment of nodular diseases: 1. detection of focal masses; 2. differentiation of multinodular goiter/hyperplasia from other nodular diseases; 3. to document the extent of a known thyroid malignancy; 4. follow up to look for residual, recurrent or metastatic carcinoma; 5. guidance for FNAC or fine needle aspiration for biopsy (FNAB).



Chapter 56 Ultrasound in ear, nose and throat practice



GOITROUS NODULES



as follows:



Multinodular goiters or hyperplasia is the most common abnormality encountered (80–85 percent).33 Hyperplasia is recognized as either a diffuse or nodular form. The nodules are separated by normal parenchyma and are mostly isoechoic or hyperechoic with well-defined margins. The adjacent great vessels may be displaced posterior or laterally, but the walls are never involved (Figure 56.7). Less than 5 percent of thyroid nodules are hypoechoic, cystic areas are seen in 60–70 percent of cases secondary to haemorrhage or collections of colloid substance.33 Colloid substance gives a typical ‘comet tail’ artifact within the nodules.35 Up to a quarter of goitrous thyroid nodules, particularly in the elderly, have areas of macrocalcification, seen as curvilinear, annular or dystrophic areas of high echoes with posterior acoustic shadowing. Colour Doppler ultrasound shows these thyroid nodules to have less vascularity than the surrounding normal parenchyma.36



    



ADENOMATOUS AND MALIGNANT NODULES



Adenomas and thyroid carcinomas account for most of the nongoitrous nodules, with adenomas accounting for 5–10 percent of all thyroid focal lesions. Thyroid adenomas have a variable appearance on ultrasound, appearing as isoechoic, hypoechoic or hyperechoic masses.13, 33 Thyroid adenomas have a thick, echo-poor halo, thought to be a fibrous capsule (Figure 56.8). The blood supply is often a regular, ‘spoke-and-wheel’ type of appearance on colour Doppler ultrasound.33 Malignant neoplasm of the thyroid are rare (two to three cases per 100,000 individuals) and may be classified



Figure 56.7 Multinodular goitre. A large multinodular goitre, in a transverse view, at the level of the thyroid isthmus and trachea (arrow), demonstrating two large nodules (short arrows).



] 719



papillary cell carcinoma; follicular cell carcinoma; anaplastic cell carcinoma; medullary cell carcinoma; lymphoma.



Papillary cell carcinoma Papillary carcinoma is the most common thyroid neoplasm and makes up 60–70 percent of all thyroid malignancies.33 On ultrasound, the characteristic appearance is that of a hypoechoic, solid nodule, with up to a third demonstrating cystic changes with detectable blood supply within the intracystic septa (Figure 56.9). Punctuate areas of microcalcification are seen in 85–90 percent of lesions; a highly reliable feature in diagnosing papillary carcinoma.37 Microcalcifications are only seen with the appropriate high frequency imaging, appearing as highly echogenic areas without the posterior acoustic shadowing normally seen with calcification. Colour Doppler studies demonstrate increased vascularity with a chaotic arrangement of blood vessels in 90 percent of lesions, due to arteriovenous shunting, as well as tortuous vessels within the malignant lesion.38 Regional lymphadenopathy may be present at initial diagnosis in 50 percent of patients.39 Recurrent lymphadenopathy may develop even after a total thyroidectomy, with the lymph nodes demonstrating microcalcification, cystic change and chaotic hypervascularity, similar in appearance to the primary tumour.37 [****] Follicular cell carcinoma Follicular cell carcinoma accounts for 5–15 percent of all thyroid malignancies, is seen in older patients and often develops in pre-existing thyroid adenomas; approximately 60–70 percent of follicular cell carcinomas are associated with a hyperplastic or adenomatous thyroid nodule.33 On ultrasound, follicular cell carcinomas are solid,



Figure 56.8 Solitary thyroid adenoma. A transverse section through the thyroid gland at the level of the thyroid isthmus, with a left lobe solitary adenoma present as a wellcircumscribed isoechoic lesion (arrow).



720 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



Figure 56.9 Papillary cell carcinoma. A solitary nodule (long arrows) in the right lobe of the thyroid demonstrating an irregular margin and punctate areas of calcification (short arrow).



homogenous, hyperechoic or isoechoic nodules with a thick irregular capsule and tortuous perinodular and intranodular vessels often with extracapsular spread. Unfortunately, the presence of extracapsular spread, a crucial feature, may be difficult to demonstrate, the capsular and vascular invasion only being detectable on histology. Anaplastic carcinoma Anaplastic carcinomas account for 5–10 percent of all thyroid cancers and, are aggressive tumours seen mostly in the elderly patient. Anaplastic carcinomas are diffusely hypoechoic with areas of necrosis in 78 percent and dense amorphous calcification in 58 percent of lesions.40 The boundaries of the tumour nodules are irregular due to early invasion of adjacent structures, with nodal or distant metastases seen in 80 percent.33 [****/*] Medullary cell carcinoma Medullary cell carcinoma, an uncommon tumour, is familial in 20 percent with an association with multiple endocrine neoplasia (MEN IIa) syndrome. In MEN IIa, the medullary cell carcinomas are often multicentric, bilateral or both (Figure 56.10). On ultrasound, medullary cell carcinoma is similar to papillary cell carcinoma being hypoechoic with irregular margins, demonstrating microcalcifications and irregular vascularity.41 An important distinguishing feature is the frequent presence of metastatic lymphadenopathy in medullary cell carcinoma.



Figure 56.10 Medullary cell carcinoma. Multifocal wellcircumscribed, low reflective areas within the left lobe of the thyroid (arrows), with no colour Doppler signal present.



Primary thyroid lymphoma Primary thyroid lymphoma is a rare tumour, making up 4 percent of all thyroid tumours, mostly of the nonHodgkin’s type and in 70–80 percent of cases arise from a gland with pre-existing chronic thyroiditis.33 Primary thyroid lymphoma appears as a hypoechoic, lobulated, vascular mass lesion with large areas of cystic necrosis and encasement of nearby large neck vessels.42 The adjacent thyroid parenchyma is usually heterogeneous as a consequence of the underlying chronic thyroiditis. [**]



DIFFUSE THYROID DISEASE



Several pathological processes of the thyroid involve the gland in a diffuse manner: acute suppurative thyroiditis, subacute granulomatous (De Quervain’s) thyroiditis, chronic lymphocytic thyroiditis (Hashimoto’s disease), colloid diffuse goiter and Graves’ disease. Colloid diffuse goiter and Graves’ disease are the most common cause of thyrotoxicosis. Ultrasound has a limited role in the diagnosis in these disorders, as assessment is based on clinical and biochemical findings.33 In Graves’ disease, the gland is lobulated and enlarged with a reduction in size used as an indicator of therapeutic success.43 The parenchyma is more homogenous than a diffuse goiter due to the presence of a large number of intraparenchymal vessels which are seen on colour Doppler ultrasound as turbulent flow within arteriovenous shunts; the ‘thyroid inferno’ or ‘thyroid storm’ (Figure 56.11).44 In the normal thyroid gland, Doppler ultrasound of the intrathyroid arteries demonstrates a peak systolic velocity (PSV) of 20–40 cm/s and an end diastolic velocity (EDV) of 10–15 cm/s. However, in thyroid disease, the PSV is



Chapter 56 Ultrasound in ear, nose and throat practice



] 721



Figure 56.11 Thyrotoxicosis. (a) Transverse image through the thyroid gland demonstrating normal parenchymal reflectivity in a patient with thyrotoxicosis. (b) The florid colour Doppler signal in the thyroid gland of the same patient is demonstrated, termed a ‘thyroid storm’. 33, 43



elevated to 50–120 cm/s due to increased flow rates. There is no correlation between gland hyperfunction and vascularity, measured flow or velocity rates.33 [***] In Graves’ disease, hypoechoic parenchymal echotexture with a high flow rate in the thyroid arteries is suggestive of post-treatment relapse.43 During the course of medical treatment, Doppler studies of the vessels demonstrate a marked fall in flow velocities measured in the superior and inferior thyroid arteries, proportional to the decrease in free circulating thyroid hormone. Subacute (De Quervain’s) thyroiditis is a clinical diagnosis, but on ultrasound the appearances may be characteristic (Figure 56.12).33 Initially, the gland enlarges with ill-defined, irregular margins. Subsequently, the parenchyma becomes hypoechoic and colour Doppler ultrasound demonstrates either normal or decreased flow. As the disease progresses, the gland recovers at different rates and assumes a ‘pseudonodular’ configuration. Hypoechoic areas may enlarge at follow-up ultrasound, necessitating further medical treatment, where ultrasound assumes an important role.33 Chronic autoimmune thyroiditis is visualized as an enlarged gland with lobulated margins, internal fibrous septa and micronodules. The micronodules are seen as



Figure 56.12 Thyroditis. (a) An enlarged thyroid gland with scattered ‘pockets’ of low reflectivity. (b) Colour Doppler ultrasound demonstrates a florid increase in vascularity with the inflammatory process.



hypoechoic areas measuring 1.0–6.5 mm, dispersed throughout the entire gland.45 On histology, micronodules represent massive infiltration of plasma cells and lymphocytes. The presence of micronodules is highly suggestive of chronic autoimmune thyroiditis, with a positive predictive value estimated at 94.7 percent.45 Colour Doppler ultrasound demonstrates increased vascularity similar to the ‘thyroid inferno’ described in Graves’ disease, but the flow velocities are within normal limits both before and during treatment.33 The end stage of chronic autoimmune thyroiditis is gland atrophy; ultrasounds demonstrating a small gland with ill-defined margins, a heterogeneous echo-texture and absent colour Doppler flow.33



RELIABILITY OF ULTRASOUND EXAMINATION OF THYROID NODULES The finding of a thyroid nodule during an ultrasound examination requires the examiner to try and determine if



722 ] PART 11 RECENT ADVANCES IN TECHNOLOGY the nodule is malignant or benign.36 The reliability of Bmode ultrasound imaging with colour Doppler in differentiating benign from malignant nodules based on morphological and vascular features is reported to have a sensitivity of 63–87 percent, specificity of 61–91 percent with an overall accuracy of 80–94 percent.33 Early work with ultrasound microbubble agents has been encouraging; in carcinomas, there is a significantly reduced measured arrival time of the contrast agent in comparison to hyperplastic or adenomatous nodules, but no significant difference in arrival time of the contrast agents in benign nodules.46 FNAC is the most accurate way of distinguishing benign from malignant thyroid nodules with a reported sensitivity of 65–98 percent, specificity of 72–100 percent, false negative rates of 1–11 percent and false positive rates of 1–8 percent.47 [****] Ultrasound is used to guide such procedures to increase the diagnostic yield of samples.



Parathyroid glands Normal parathyroid glands are not usually visible due to their small size and similar echo-texture to adjacent thyroid tissue. Occasionally, normal parathyroid glands are seen as either hypoechoic or hyperechoic oval areas adjacent to the lower or upper poles of the thyroid gland.48 Consequently, the parathyroid glands are usually only imaged when a patient has biochemical evidence of hyperparathyroidism. Primary hyperparathyroidism may be secondary to either an adenoma (where a single gland is involved), hyperplasia (where all four glands are involved) or rarely a carcinoma. Adenomas are the main cause of primary hyperparathyroidism (80 percent of cases). Secondary hyperparathyroidism is usually a consequence of chronic hypocalcaemia in patients with chronic renal impairment. Parathyroid adenomas and hyperplastic glands are oval in shape, measuring 7–15 mm in length and are usually hypoechoic in comparison to adjacent thyroid tissue (Figure 56.13).49 Parathyroid adenomas have an echogenic capsule and are usually vascular on colour Doppler ultrasound with prominent diastolic flow demonstrated on spectral Doppler ultrasound. Areas of calcification are rare in parathyroid adenomas, but more common in hyperplasia and carcinoma.49 Intrathyroidal parathyroid glands are seen in 1 percent of cases and mimic thyroid nodules as they appear as hypoechoic areas with well-defined margins.33 Solitary parathyroid cysts are rare, often occurring below the level of the inferior thyroid margin. The cystic fluid in functioning lesions contains high levels of parathyroid hormone.50 The sensitivity of ultrasound in localizing parathyroid glands in primary hyperparathyroidism is between 70 and 82 percent.51, 52 However, in persistent or recurrent hyperparathyroidism, the sensitivity falls and ranges from 36 to 63 percent.53



Figure 56.13 Parathyroid adenoma. (a) Low reflective ovoid area at the inferior aspect of the left lobe of the thyroid gland (between cursors). (b) Magnified image of a parathyroid adenoma demonstrating increased vascularity on colour Doppler ultrasound.



Salivary gland disease Both the parotid and submandibular glands are superficial and well sited for ultrasound examination. The parotid glands are examined, initially in the axial plane in order to identify anatomical landmarks before completing the examination in the coronal plane. The submandibular glands are best seen in the coronal plane with paramandibular images as useful adjuncts. The normal submandibular gland parenchymal echo-texture is of a homogenous, highly reflective nature, well demarcated from surrounding tissues. The parotid gland is generally



Chapter 56 Ultrasound in ear, nose and throat practice



] 723



of higher echogenicity than the submandibular gland. The parotid (Stensen’s) and submandibular (Wharton’s) ducts are only seen on ultrasound when dilated.6



SALIVARY GLAND TUMOURS



Salivary gland tumours account for 3 percent of all head and neck tumours.54, 55 Ultrasound accurately differentiates salivary gland tumours from other lesions outside the gland in 98 percent of cases.54, 56 Tumours that arise in the parotid gland are benign pleomorphic adenomas in 80 percent of cases. The accepted treatment for pleomorphic adenomas is parotidectomy, as there is a 40–50 percent recurrence rate if they are only enucleated.54, 57 Malignant transformation is reported in 3–10 percent of cases, with the risk increasing with time.55 Pleomorphic adenomas are typically hypoechoic, slightly heterogeneous solid mass lesions with a smooth lobulated outline (Figure 56.14). Unusually, unlike other soft tissue masses, pleomorphic adenomas exhibit through transmission of sound in a similar fashion to simple cysts.56 Ultrasound delineates pleomorphic adenomas in the parotid in 95 percent of cases and further cross-sectional imaging is advised in only in a minority of cases.55 [**/*] The second most common tumour in the parotid gland is a Warthin’s tumour (adenolymphoma), accounting for 2–24 percent of parotid tumours.54 Adenolymphomas are markedly hypoechoic due to the increased mucus content of cysts and contain discrete septations, suggested as characteristic of adenolymphomas.55, 58 Parotid carcinomas are rare and are hypoechoic, heterogeneous lesions with irregular margins.6 When differentiating malignancy from benign lesions using irregular margins as a marker for malignancy, ultrasound has an accuracy of 94 percent in lesions over 1.0 cm. However, using this as the only criterion is limited, as up to 28 percent of malignant tumours have well-demarcated borders.56 [*]



Figure 56.14 Parotid gland pleomorphic adenoma. (a) A well-circumscribed low reflective abnormality in the right parotid gland (long arrow) demonstrating through transmission of sound (short arrow), normally a feature of a cystic structure. (b) Colour Doppler ultrasound confirms lesion vascularity and the solid nature of the pleiomorphic adenoma.



SIALOLITHIASIS



The majority of salivary gland calculi (80 percent) occur in the submandibular gland.56 Calculi larger than 2 mm may be detected on ultrasound, with a typical appearance of an echogenic rim with posterior acoustic shadowing (Figure 56.15). Calculi impacted at the ostium are not well seen on ultrasound, but often main salivary duct dilatation as well as intraglandular duct dilatation are present, with gland enlargement and a heterogeneous, hypoechoic appearance (Figure 56.16).59 Both intraand extraductal calculi may be treated with ultrasound lithotripsy but symptomatic intraparenchymal calculi require surgical removal. Ultrasound is useful in defining the location of calculi in relation to the gland parenchyma, which has important implications for clinical management.55 [*]



Figure 56.15 Submandibular gland calculus. Linear area of high reflectivity (arrow) and posterior acoustic shadowing present in the submandibular gland, in keeping with a calculus.



724 ] PART 11 RECENT ADVANCES IN TECHNOLOGY differentiate solid from cystic abnormalities.61 Necrotic lymph node metastases from squamous cell carcinoma of the head and neck or papillary carcinoma of the thyroid may appear cystic and caution is needed in interpretation. Any lymph node is likely to demonstrate increased vascular flow with colour Doppler ultrasound, which may be used as a discriminator.5 [*]



Thryoglossal cysts



Figure 56.16 Submandibular gland duct dilatation. Dilated main duct in the submandibular gland.



INFLAMMATORY AND AUTOIMMUNE DISORDERS



Acute inflammation causes swollen salivary glands and this enlargement may be documented on ultrasound and compared to the contralateral side if unaffected, or to established normal measurements. Inflammation causes the gland to appear heterogeneous and hypoechoic on ultrasound; the main role of ultrasound is to exclude salivary duct dilatation.56 Ultrasound may also identify any complications, such as abscess formation, and image guidance could then be used to aspirate fluid for microbiological analysis. Acute suppurative sialadenitis is a painful condition occurring in postoperative, dehydrated and debilitated patients, seen almost exclusively in the parotid gland. Ultrasound may demonstrate an enlarged, well-defined gland with either a hypoechoic or hyperechoic parenchymal echo-texture. The role of ultrasound in this situation is to detect the presence and extent of any abscess formation.59 In chronic inflammation, one or both parotid glands appear heterogeneous with a small number of hypoechoic lesions, measuring up to 10 mm in diameter, often with associated reactive lymph nodes with their characteristic echogenic hilum present.56 In Sjo¨gren’s syndrome, there is bilateral gland enlargement, which on ultrasound may appear of normal or hyperechoic texture with enlargement in the early stages.59 As the disease progresses, the gland becomes atrophic with multicystic changes and a reticular pattern. The degree of parenchymal damage correlates well with glandular vascularity on colour Doppler ultrasound imaging.59 As there is an increased risk of developing lymphoma, ultrasound may be used in the follow up of these patients.60 [**]



CYSTIC LESIONS Soft tissue cystic lesions occur with a frequency of 1–10 percent in the neck, with ultrasound able to clearly



Thyroglossal cysts occur close to the midline over the anterior aspect of the neck, anywhere between the foramen caecum to the pyramidal lobe of the thyroid gland. Thyroglossal cysts appear as anechoic areas with posterior acoustic enhancement, with the typical appearance of a cyst (Figure 56.17).62, 63 The majority of thyroglossal cysts occur in the region of the hyoid bone and tend to communicate with the bone. This latter finding should be specifically sought on ultrasound, as this will allow the corpus of the hyoid bone to be removed with the cyst at the time of surgery.56 It is possible on ultrasound imaging to establish if the cyst is multilocular or if a fistula is present.62



Branchial cysts Branchial cysts most commonly arise from the first or second branchial clefts and are found in the region of the parotid glands or anterolateral to the carotid bifurcation. Branchial cysts vary in appearance on ultrasound from anechoic to heterogeneous, depending on the amount of cellular debris or cholesterol crystals in the cysts (Figure 56.18).64 There is no vascularity on colour Doppler ultrasound and if there is colour Doppler signal, the possibility of a branchiogenic carcinoma should be raised.6



Dysontogenetic cysts Epidermoid, dermoid cysts and teratomas are included in the dysontogenetic cyst group, occurring in the midline from the floor of the mouth or tongue.65 These lesions are usually hypoechoic, but may have high reflective echoes due to the presence of hair, sebum and fluid.66 The presence of calcification is considered characteristic of teratomas.6



Laryngoceles Laryngoceles are dilatations of the saccule of Morgagni containing mucus, saliva and air, and are classified relative to their position to the larynx. Laryngoceles are seen as anechoic lesions if examined thorough an uncalcified thyroid cartilage.67



Chapter 56 Ultrasound in ear, nose and throat practice



Figure 56.17 Thyroglossal cyst. A thick-walled, bilobulated thyroglossal cyst is present in the soft tissues above the hyoid bone (long arrows). Echogenic debris, manifest as reflective areas, is present within the cyst, which demonstrates characteristic posterior acoustic enhancement (short arrow).



] 725



Figure 56.18 Branchial cyst. A cystic abnormality present in the soft tissues of the lateral aspect of the neck, with an area (arrow) of echogenic debris present.



Ranulas Ranulas are mucous retention cysts of the minor sublingual salivary glands and can extend over the posterolateral aspect of the mylohyoid muscle. Ranulas are recognized as hypoechoic lesions with smooth, welldefined margins with no colour Doppler signal.6



SOFT TISSUE NEOPLASMS Ultrasound is often the only technique used for imaging non-nodal cervical lesions. CT and MRI are used as adjuncts in patients where the lesion is not completely delineated or a specific diagnosis is not made by ultrasound. [*]



Lipomas Lipomas are common soft tissue masses that are oval, elongated lesions with varying echogenicity depending on the amount of fat and fibrous tissue within the lesion (Figure 56.19).68 The greater the fat content, the more hypoechoic the lesion appears, with the echogenic manifestation due to the presence of multiple fat–water interfaces within the lesion acting as reflectors, thus



Figure 56.19 Lipoma. A well-circumscribed isoechoic abnormality (arrows) in the soft tissues of the neck causing displacement of the surrounding tissue planes with no increase in colour Doppler signal.



increasing the number and intensity of reflected echoes.69 In the rare cases where there is doubt, CT or MRI can confirm the presence of fat.



Paragangliomas Paragangliomas are tumours arising from the autonomic nervous tissue, commonly the carotid body or vagal ganglia (Figure 56.20). Paragangliomas appear as hypoechoic solid tumours and lie between the bifurcation



726 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



Figure 56.20 Carotid body tumour. A low reflective abnormality (between arrows) with evidence of displacement of the internal carotid artery (ICA) and the external carotid artery (ECA).



of the internal and external carotid arteries. Colour Doppler ultrasound demonstrates a markedly vascular lesion with a low resistance flow, resultant to arteriovenous shunts.70 Ultrasound is additionally useful in the follow up of these patients to facilitate early detection of recurrence.6



Neurogenic tumours Neurofibromas and schwannomas, when they are present in the neck, are most commonly sited in the posterior triangle. Neurogenic tumours are well-defined, hypoechoic, fusiform lesions with a number of central cystic areas in continuity with a nerve and exhibit posterior acoustic enhancement.71 With neurofibromas, the nerve enters the centre of the lesion, whereas in schwannomas the mass is eccentric to the nerve and this may be demonstrated on ultrasound. Shwannomas have a variable vascularity but are generally hypervascular, whereas neurofibromas are less vascular which may be an important distinguishing factor between these two entities.71



Vascular malformations Haemangiomas are the most common tumour in infants and up to 60 percent will occur in the head and neck area.13 Vascular-based lesions may be seen in the prenatal period, are usually evident in the first two years of life and are classified into two broad categories: haemangiomas and other vascular malformations. Vascular malformations



are further divided depending on flow characteristics and the presence of anomalous channels.72 It is important to distinguish between haemangiomas and other vascular malformations as they are treated differentially and have different outcomes. A haemangioma does not regress spontaneously. Haemangiomas are a clinical diagnosis and the role of imaging is to identify deep malformations or if they present as an atypical soft tissue mass. Haemangiomas have a variable ultrasound appearance depending on the number of ‘cystic’ vascular spaces present. Small lesions are usually echogenic but often the larger lesions are hypoechoic, compressible with increased colour Doppler flow.73 Lesions in the proliferative phases have increased colour Doppler flow with increased vessel density (45 vessels/cm2) and a low resistance pattern on spectral Doppler ultrasound.72, 74 Congenital venous malformations occur in the head and neck in 40 percent of cases, but are often not diagnosed at birth.13, 72 Grey-scale imaging can aid the diagnosis of a venous malformation. The majority of venous malformations are heterogeneous and hypoechoic relative to the subcutaneous tissue and phleboliths are seen in 16 percent.73 On spectral Doppler ultrasound, venous malformations tend to have a low velocity monophasic flow pattern and mixed vascular lesions tend to have a low velocity, biphasic flow pattern.73 If the flow velocity is so low to be below the level that can be detected with normal colour Doppler imaging, then the administration of ultrasound microbubble contrast agent may help.13 Arteriovenous malformations (AVM) have a heterogeneous appearance due to visualization of the feeding vessels; colour and spectral Doppler analysis show high vessel density with high systolic flow from numerous sites of arteriovenous shunting. Distinct from haemangiomas, an AVM always demonstrates pulsatile flow in the veins caused by arterialization of the veins.72



Lymphangiomas Lymphangiomas are thought to originate from sequestered lymphatic sacs that do not communicate with peripheral draining channels. Cystic hygromas, usually diagnosed on antenatal ultrasound examination, are a form of lymphangioma. Seventy-five percent of all lymphangiomas are located in the neck, usually in the posterior triangle, with up to 10 percent extending down to the mediastinum.75 Four types are classified on histological criteria: cystic hygroma, cavernous lymphangiomas, capillary lymphangiomas and vasculolymphatic malformations. These four types cannot be differentiated by ultrasound as all are multilocular, predominantly cystic masses with both septa and solid components. Ultrasound may also be used to guide therapeutic sclerosing injection in the management of macrocystic lymphangiomas.76



Chapter 56 Ultrasound in ear, nose and throat practice



Ectopic thymic tissue Ectopic thymic tissue is seen in the neck in 21–42 percent of infants and resembles a lymphoma, appearing as an echogenic mass with septa that are isoechoic with normal thymus tissue. Increased vascularity on colour Doppler ultrasound allows differentiation of ectopic thymic tissue from a lipoma.5



Best clinical practice [ A linear high frequency transducer will allow the



[



Foreign bodies Ultrasound is useful in localizing nonradio-opaque foreign bodies, especially in the neck to guide the operator to a suture granuloma with its typical appearance of a hypoechoic lesion with a central curvilinear echogenic line.6



CONCLUSION Ultrasound is a safe, widely available imaging technique that is patient-friendly and extensively used in most aspects of clinical medicine. In neck imaging, ultrasound use has become more prevalent with the development of real-time, high frequency ultrasound transducers, digital imaging technology coupled with the inherent anatomical advantage of the superficial structures of the head and neck. Ultrasound is currently the imaging modality of choice after clinical examination in ENT practice. Ongoing improvements to ultrasound technology will lead to increased diagnostic accuracy.



KEY POINTS  High frequency, high resolution ultrasound provides excellent visualization of structures of the neck, allowing guided interventional procedures.  Ultrasound readily identifies enlarged lymph nodes and may usefully divide the appearances into benign or malignant.  Ultrasound detects nodular thyroid disease, but is nonspecific in establishing the type of nodule.  Ultrasound often identifies papillary carcinoma of the thyroid.  Ultrasound is not sensitive, but specific for the detection of parathyroid adenomas.  Ultrasound is a reliable method of examination for salivary gland disease.  Ultrasound should always be the first-line imaging modality in assessing disease of the soft tissues and glandular structures of the neck.



] 727



[ [ [ [



optimal resolution of the structures of the neck and will guide needle placement in the lesion for tissue sampling. Ultrasound of lymph nodes using B-mode, colour and spectral Doppler ultrasound allows accurate delineation of lymph nodes into benign and malignant types, but fine needle aspiration is always required. Ultrasound in the delineation of thyroid nodules is not specific except in papillary cell carcinoma where some ultrasound features are characteristic. Ultrasound defines an enlarged thyroid and, with the addition of colour Doppler ultrasound, identifies the appearances of thyroiditis and Graves’ disease. Parathyroid adenomas have a characteristic lowreflective ovoid appearance on ultrasound, with colour Doppler flow present in the lesion. Ultrasound of obstructive salivary gland disease is accurate and will often identify the calculus.



Deficiencies in current knowledge and areas for future research



$ $ $ $



Ultrasound probe technology and image postprocessing will continue to improve, which will allow better visualization of neck structures in difficult patients. The use of microbubble contrast agents will allow physiological aspects of tumour circulation to be better understood and aid tissue characterization, using transit time and blood vessel distribution. Better needle visualization will allow better needle guidance for cytological analysis. An improvement in the identification of benign versus malignant thyroid nodules is desirable.



REFERENCES 1. Merritt CR. Physics of ultrasound. In: Rumack CM, Wilson SR, Charboneau JW (eds). Diagnostic ultrasound. St Louis: Mosby, 1998: 3–33. 2. Harvey CJ, Pilcher JM, Eckersley RJ, Blomley MJK, Cosgrove DO. Advances in ultrasound. Clinical Radiology. 2002; 57: 157–177. 3. Weng L, Tirumalai AP, Lowery CM, Nock LF, Gustafson DE, Von Behren PL et al. US extended field-of-view imaging technology. Radiology. 1997; 203: 877–80.



728 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



 



4. Whittingham TA. Tissue harmonic imaging. European Radiology. 1999; 9: S323–6. 5. Koischwitz D, Gritzmann N. Ultrasound of the neck. Radiologic Clinics of North America. 2000; 38: 1029–45. 6. Gritzmann N, Hollerweger A, Macheiner P, Rettenbacher T. Sonography of soft tissue masses of the neck. Journal of Clinical Ultrasound. 2002; 30: 356–73. 7. Merritt CRB. Doppler US: The basics. Radiographics. 1991; 11: 109–19. 8. Harvey CJ, Blomley MJK, Eckersley RJ, Cosgrove DO. Developments in ultrasound contrast media. European Radiology. 2001; 11: 675–89. 9. Ferrara KW, Merritt CR, Burns PN, Foster FS, Mattrey RF, Wickline SA. Evaluation of tumor angiogenesis with US: Imaging, Doppler contrast agents. Academic Radiology. 2000; 7: 824–39. 10. Castelijns JA, van den Brekel MWM. Imaging of lymphadenopathy in the neck. European Radiology. 2002; 12: 727–38. 11. Ying M, Ahuja A. Sonography of neck lymph nodes. Part I. Normal lymph nodes. Clinical Radiology. 2003; 58: 351–8. 12. van den Brekel MWM, Stel HV, Castelijns JA, Nauta JJP, vander Waal I, Valk J et al. Cervical lymph node metastasis: assessment of radiologic criteria. Radiology. 1990; 177: 379–84. 13. Toma P, Rossi UG. Paediatric ultrasound. II. Other applications. European Radiology. 2001; 11: 2369–98. 14. Wakisaka M, Mori H, Fuwa N, Matsumoto A. MR analysis of nasopharyngeal carcinoma: Correlation of the pattern of tumor extent at the primary site with the distribution of metastasized cervical lymph nodes. Preliminary results. European Radiology. 2000; 10: 970–7. 15. Khoury GG, Paterson IC. Nasopharyngeal carcinoma: A review of cases treated by radiotherapy and chemotherapy. Clinical Radiology. 1987; 38: 17–20. 16. Shingaki S, Kobyashi T, Suzuki I, Kohno M, Nakajima T. Surgical treatment of stage I and II oral squamous cell carcinomas: analysis of causes of failure. British Journal of Oral and Maxillofacial Surgery. 1995; 33: 304–8. 17. Levendag P, Sessions R, Vikram B, Strong EW, Shah JP, Spiro R et al. The problem of neck relapse in early stage supraglottic larynx cancer. Cancer. 1989; 63: 345–8. 18. Stuckensen T, Kovacs AF, Adams S, Baum RP. Staging of the neck in patients with oral cavity squamous cell carcinomas: a prospective comparison of PET, ultrasound, CT and MRI. Journal of Cranio-Maxillo-Facial surgery. 2000; 28: 319–24. 19. Vassallo P, Wernecke K, Roos N, Peters PE. Differentiation of benign from malignant superficial lymphadenopathy: The role of high resolution US. Radiology. 1992; 183: 215–20. 20. Tohnosu N, Onoda S, Isono K. Ultrasonography evaluation of cervical lymph node metastases in esophageal cancer with special reference to the relationship between the short to long axis (S/L) and the cancer content. Journal of Clinical Ultrasound. 1989; 17: 101–6.



 21. 22.



23.



24.



25.



26.



 27.



28.



29.



30.



31.



32.



 33. 34.



Hajek PC, Salomonowitz E, Turk R, Tscholakoff D, Kumpan W, Czemberik H. Lymph nodes of the neck: Evaluation with US. Radiology. 1986; 158: 739–42. van den Brekel MW, Castelijns JA, Snow GB. The size of lymph nodes in the neck on sonograms as a radiologic criterion for metastasis: how reliable is it? American Journal of Neuroradiology. 1998; 19: 695–700. Ying M, Ahuja AT, Evans R, King W, Metreweli C. Cervical lymphadenopathy: Sonographic differentiation between tuberculous and nodal metastases from non-head and neck carcinomas. Journal of Clinical Ultrasound. 1998; 26: 383–9. Ahuja A, Chow L, Chick W, King W, Metreweli C. Metastatic cervical nodes in papillary carcinoma of the thyroid: Ultrasound and histological correlation. Clinical Radiology. 1995; 50: 229–31. Ahuja AT, Ying M, Yuen HY, Metreweli C. ‘Pseudocystic’ appearance of non-Hodgkin’s lymphomatous nodes: An infrequent finding with high-resolution transducers. Clinical Radiology. 2001; 56: 111–5. Som PM. Lymph nodes of the neck. Radiology. 1987; 165: 593–600. Solbiati L, Rizzatto G, Bellotti E, Montali G, Ciotti V, Croce F. High-resolution sonography of cervical lymph nodes in head and neck cancer: Criteria for differentiation of reactive versus malignant nodes. Radiology. 1988; 169P: 113. Ahuja A, Ying M, Yang WT, Evans R, King W, Metreweli C. The use of sonography in differentiating cervical lymphomatous lymph nodes from cervical metastatic lymph nodes. Clinical Radiology. 1996; 51: 186–90. Ahuja AT, Ying M, Evans R, King W, Metreweli C. The application of ultrasound criteria for malignancy in differentiating tuberculous cervical adenitis from metastatic nasopharyngeal carcinoma. Clinical Radiology. 1995; 50: 391–5. Wu CH, Chang YL, Hsu WC, Ko JY, Sheen TS, Hsieh FJ. Usefulness of Doppler spectral analysis and power Doppler sonography in the differentiation of cervical lymphadenopathies. American Journal of Roentgenology. 1998; 171: 503–9. Ahuja A, Ying M, Yuen YH, Metreweli C. Power Doppler sonography to differentiate tuberculous cervical lymphadenopathy from nasopharyngeal carcinoma. American Journal of Neuroradiology. 2001; 22: 735–40. Moritz JD, Ludwig A, Oestmann JW. Contrast-enhanced color Doppler sonography for evaluation of enlarged cervical lymph nodes in head and neck tumors. American Journal of Roentgenology. 2000; 174: 1279–84. Solbiati L, Osti V, Cova L, Tonolini M. Ultrasound of thyroid, parathyroid glands and neck lymph nodes. European Radiology. 2001; 11: 2411–24. Hegedus L, Perrild H, Poulsen LR, Andersen JR, Holm B, Schnohr P et al. The determination of thyroid volume by ultrasound and its relationship to body weight, age and sex in normal subjects. Journal of Clinical Endocrinology and Metabolism. 1983; 56: 260–3.



Chapter 56 Ultrasound in ear, nose and throat practice











35. Ahuja A, Chick W, King W, Metreweli C. Clinical significance of the comet-tail artifact in thyroid ultrasound. Journal of Clinical Ultrasound. 1996; 24: 129–33. 36. Solbiati L, Osti V, Cova L, Tonolini M. Thyroid nodules: Which sonographic criteria for differentiation between benign and malignant lesions? British Medical Ultrasound Society Bulletin. 2001; 9: 11–8. 37. Ahuja AT, Chow L, Chick W, King W, Metreweli C. Metastatic cervical node in papillary carcinoma of the thyroid: ultrasound and histological correlation. Clinical Radiology. 1995; 50: 229–31. 38. Frates MC, Benson CB, Doubilet PM, Cibas ES, Marqusee E. Can color Doppler sonography aid in the prediction of malignancy of thyroid nodules? Journal of Ultrasound in Medicien. 2003; 22: 127–33. 39. McConahey WM, Hay ID, Woolner LB, van Heerden JA, Taylor WF. Papillary thyroid cancer treated at the Mayo Clinic, 1946 through 1970: Initial manifestations, pathologic findings, therapy, and outcome. Mayo Clinic Proceedings. 1986; 61: 978–96. 40. Takashima S, Morimoyo S, Ikezoe J, Takai S, Kobayashi T, Koyama H et al. CT evaluation of anaplastic thyroid carcinoma. American Journal of Roentgenology. 1990; 154: 1079–85. 41. Gorman B, Charboneau JW, James EM, Reading CC, Wold LE, Grant CS et al. Medullary thyroid carcinoma: Role of high-resolution US. Radiology. 1987; 162: 147–50. 42. Takashima S, Morimoto S, Ikezoe J, Arisawa J, Hamada S, Ikeda H et al. Primary thyroid lymphoma: comparison of CT and US assessment. Radiology. 1989; 171: 439–43. 43. Castagnone D, Rivolta R, Rescalli S, Tozzi R, Cantalamessa L. Color Doppler sonography in Grave’s disease: Value in assessing activity of disease and predicting outcome. American Journal of Roentgenology. 1996; 166: 203–7. 44. Ralls PW, Mayekawa DS, Lee KP, Radin DR, Boswell WD, Halls JM. Color-flow Doppler sonography in Graves disease: ‘thyroid inferno’. American Journal of Roentgenology. 1988; 150: 781–4. 45. Yeh HC, Futterweit W, Gilbert P. Micronodulation: Ultrasonographic sign of Hashimoto thyroiditis. Journal of Ultrasound in Medicine. 1996; 15: 813–9. 46. Spiezia S, Farina R, Cerbone G, Assanti AP, Iovino V, Siciliani M et al. Analysis of color Doppler signal intensity variation after Levovist injection. A new approach to the diagnosis of thyroid nodules. Journal of Ultrasound in Medicine. 2001; 20: 223–31. 47. Gharib H, Goellner JR. Fine-needle aspiration biopsy of the thyroid: An appraisal. Annals of Internal Medicine. 1993; 118: 282–9. 48. Simeone JF, Mueller PR, Ferrucci JT, vanSonnenberg E, Wang CA, Hall DA et al. High-resolution real-time sonography of the parathyroid. Radiology. 1981; 141: 745–51. 49. Reeder SB, Desser TS, Weigel RJ, Jeffrey RB. Sonography in primary hyperparathyroidism. Journal of Ultrasound in Medicine. 2002; 21: 539–52.







] 729



50. Forston JK, Patel VG, Henderson VJ. Parathyroid cysts: A case report and review of the literature. Laryngoscope. 2001; 111: 1726–8. 51. Attie JN, Khan A, Rumancik WM, Moskowitz GW, Hirsch MA, Herman PG. Preoperative localization of parathyroid adenomas. American Journal of Surgery. 1988; 156: 323–6. 52. Weinberger MS, Robbins KT. Diagnostic localization studies for primary hyperparathyroidism. A suggested algorithm. Archives of Otolaryngology – Head and Neck Surgery. 1994; 120: 1187–9. 53. Rodriquez JM, Tezelman S, Siperstein AE, Duh QV, Higgins C, Morita E et al. Localization procedures in patients with persistent or recurrent hyperparathyroidism. Archives of Surgery. 1994; 129: 870–5. 54. Freling NJM. Imaging of salivary gland disease. Seminars in Roentgenology. 2000; 35: 12–20. 55. Wittich GR, Scheible FW, Hajek PC. Ultrasonography of the salivary glands. Radiologic Clinics of North America. 1985; 23: 29–37. 56. Gritzmann N. Sonography of the salivary glands. America Journal of Roentgenology. 1989; 153: 161–6. 57. Koral K, Seyre J, Bhuta S, Abemayor E, Lufkin R. Recurrent pleomorphic adenoma of the parotid gland in pediatric and adult patients: Value of multiple lesions as a diagnostic indicator. America Journal of Roentgenology. 2003; 180: 1171–4. 58. Howlett DC, Kesse KW, Huges DV, Sallomi DF. The role of imaging in the evaluation of parotid disease. Clinical Radiology. 2002; 57: 692–701. 59. Ching ASC, Ahuja A. High-resolution sonography of the submandibular space: anatomy and abnormalities. America Journal of Roentgenology. 2002; 179: 703–8. 60. MAkula E, Pokorny G, Kiss M, Voros E, Kovacs L, Kovacs A et al. The place of magnetic resonance and ultrasonographic examinations of the parotid gland in the diagnosis and follow-up of primary Sjogrens’s syndrome. Rheumatology. 2000; 39: 97–104. 61. Koeller KK, Alamo L, Adair CF, Smirniotopoulos JG. Congenital cystic masses of the neck: radiologicpathologic correlation. Radiographics. 1999; 19: 121–46. 62. Ahuja AT, King AD, Metreweli C. Sonographic evaluation of thyroglossal duct cysts in children. Clinical Radiology. 2000; 55: 770–4. 63. Ahuja AT, King AD, King W, Metreweli C. Thyroglossal duct cysts: Sonographic appearances in adults. American Journal of Neuroradiology. 1999; 20: 579–82. 64. Badami JP, Athey PA. Sonography in the diagnosis of branchial cysts. America Journal of Roentgenology. 1981; 137: 1245–8. 65. Smirniotopoulos JG, Chiechi MV. Teratomas, dermoids, and epidermoids of the head and neck. Radiographics. 1995; 15: 1437–55. 66. Turetschek K, Hospodka H, Steiner E. Case report: Epidermoid cyst of the floor of the mouth: Diagnostic imaging by sonography, computed tomography and



730 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



67.



68.



 69. 70.



71.



 72.



magnetic resonance imaging. British Journal of Radiology. 1995; 68: 205–97. Baatenburg de Jong RJ, Rongen RJ, Lameris JS, Knegt P, Verwoerd CD. Ultrasound in the diagnosis of laryngoceles. ORL, Journal for Oto-Rhino-Laryngology and its Related Specialties. 1993; 55: 290–3. Fornage BD. Sonographic appearances of superficial soft tissue lipomas. Journal of Clinical Ultrasound. 1991; 19: 215–20. Behan M, Kazam E. The echographic characteristics of fatty tissues and tumors. Radiology. 1978; 129: 143–51. Derchi L, Serafini G, Rabbia C, De Albertis P, Solbiati L, Candiani F et al. Carotid body tumors: US evaluation. Radiology. 1992; 182: 457–9. King AD, Ahuja AT, King W, Metreweli C. Sonography of peripheral nerve tumors of the neck. America Journal of Roentgenology. 1997; 169: 1695–8. Dubois J, Garel L. Imaging and therapeutic approach of hemangiomas and vascular malformations in the



73.



74.



75.



76.



pediatric age group. Pediatric Radiology. 1999; 29: 879–93. Trop I, Dubois J, Guibaud L, Grignon A, Patriquin H, McCuaig C et al. Soft-tissue venous malformations in pediatric and young adult patients: Diagnosis with Doppler US. Radiology. 1999; 212: 841–5. Dubois J, Patriquin HB, Garel L, Powell J, Filiatrault D, David M et al. Soft-tissue hemangiomas in infants and children: diagnosis using Doppler sonography. American Journal of Roentgenology. 1998; 171: 247–52. Zadvinskis DP, Benson MT, Kerr HH, Mancuso AA, Cacciarelli AA, Madrazo BL et al. Congenital malformations of the cervico-thoracic lymphatic system: Embryology and pathogenesis. Radiographics. 1992; 12: 1173–89. Dubois J, Garel L, Abela A, Laberge L, Yazbeck S. Lymphangiomas in children: Percutaneous sclerotherapy with an alcoholic solution of Zein. Radiology. 1997; 204: 651–4.



57 Interventional techniques JAMES V BYRNE



Introduction Indications Tools used in interventional neuroradiology techniques Application of embolization techniques Embolization for meningiomas and paragangliomas



731 732 732 734 735



Embolization for epistaxis Key points Deficiencies in current knowledge and areas for future research References



739 740 740 740



SEARCH STRATEGY AND EVIDENCE-BASE The data in this chapter are supported by a Medline search using the key words interventional neuroradiology, endovascular therapy, embolization, meningioma, paraganglioma, epistaxis and tumour embolization. The evidence base for most INR treatments is lacking. Research has largely involved reports of single institute case audits. Large-scale trials are difficult because relatively small patient numbers are treated at each hospital and multicentre cooperation is needed. This problem is currently being addressed by various specialist interest groups who have organized case registries and a large multicentre trial to study the efficacy of embolization alone for treatment of meningioma is ongoing.



INTRODUCTION What is interventional neuroradiology and what is its role in the management of patients with pathologies of the head and neck? This chapter is intended to answer these two questions by describing the current (and future) capabilities of the discipline. Interventional radiology originated in surgery, but its complete dependence on accurate in vivo corporal imaging places it in radiology. In 1930, Brooks1 closed an arteriovenous fistula of the cavernous carotid by introducing a muscle embolus via an arteriotomy in the proximal carotid artery. This desperate operation was performed in order to avoid the extensive dissection required to expose and directly repair the fistula. The need of imaging to monitor such embolization was obvious, but the capacity to do so developed slowly after the first description of cerebral angiography by Egas Moniz in 1927,2 and the technologies needed to navigate catheters through blood vessels and to deposit



emboli at specific sites did not become available for another 30 years. In 1964, Lusenhop and Valasquez3 described an operation to embolize a caroticocavernous fistula using an endovascular catheter under x-ray control and the current form of most procedures was born. Interventions for head and neck pathologies are usually practised within the subspecialty of neuroradiology or, more specifically, interventional neuroradiology (INR). The term ‘most procedures’ is used because endovascular embolization is the most common technique. It is used to treat arteriovenous shunts, devascularize tumours and close large arteries prior to surgical resections. However, also included in the INR repertoire are percutaneous techniques for embolization of vascular malformations and tumours, as well as image-guided biopsy and the focussed endovascular delivery of chemotherapy agents. The technologically-driven evolution of the discipline will almost certainly produce new techniques and applications to replace more invasive



732 ] PART 11 RECENT ADVANCES IN TECHNOLOGY surgery in the future and it remains to be seen, into which other branch(es) of medicine (or rather surgery) it extends.



INDICATIONS Detection of head and neck lesions is by magnetic resonance (MR) or computed tomography (CT) scanning and the possible role of INR should be considered before invasive diagnostic x-ray angiography is performed. Combining embolization or functional testing with diagnostic x-ray angiography means that the indications for INR techniques should be considered early in treatment planning. Preoperative treatment planning of patients with tumours generally involves obtaining biopsy tissue and mapping the lesion’s blood supply, as well as its extent and relationship to adjacent normal structures. Information about invasion of large arteries and veins may be vital. Major vessels are at risk, either because of tumour spread or because their sacrifice is a necessary component of the surgical approach. In these situations, preoperative testing of blood supply by temporary vessel occlusion is indicated and endovascular ligation may be appropriate (see Coils and balloons). Embolization of highly vascular tumours is indicated prior to surgical resection of these tumours in order to reduce intraoperative blood loss and operation times. It may also be indicated to relieve patients of symptoms by inducing tumour shrinkage. Embolization is targeted to occlude the tumour vascular bed, rather than feeding arteries, since occlusion of proximal arteries may fail to reduce blood supply because blood flow from collateral arteries can still be recruited. It is, therefore, most effective in tumours with a rich vascular bed. In most instances, this can be inferred from the degree of contrast enhancement evident on CT or MR scanning, but the uptake of radiographic contrast within a tissue reflects the degree of extravascular leakage of these agents and does not necessarily imply high vascularity, so occasionally tumours showing pronounced enhancement are not suitable for embolization. The decision to perform embolization therefore depends on clinical and anatomical factors. These should be reviewed in consultation between otolaryngologist and Table 57.1



interventional neuroradiologist. The principle guiding the decision to recommend preoperative embolization is that the addition of the INR procedure to the patient’s treatment plan should not add to its overall risk. The risk/ benefit analysis may require information that can only be obtained at the time of x-ray angiography. For example, embolization via arteries with intracranial territories of supply is obviously more likely to be complicated by stroke than treatments performed entirely in extracranial vessels. Therefore, a team approach is needed to ensure that the risks of complications caused by embolization or preoperative functional testing are acceptable and are likely to improve the results of subsequent surgery. These issues, as they apply to specific tumour types, and the use of embolization in the management of epistaxis, will be discussed below. The purpose and indications for INR techniques are summarized in Table 57.1.



TOOLS USED IN INTERVENTIONAL NEURORADIOLOGY TECHNIQUES Endovascular catheters Transarterial or transvenous catheterizations are performed for embolization, temporary vessel closures for dynamic testing or intraarterial drug delivery. The catheters used are available in a range of sizes with shaped tips. They are used to select and inject radiographic contrast media in cranial vessels for angiography or as guiding catheters for smaller catheters used for the section of small vessels (i.e. microcatheters). Microcatheters are generally 0.02 inches (0.508 mm) or smaller in diameter and used to inject embolization materials. They are constructed with flexible distal ends for navigation through tortuous vessels and stiffer proximal sections to allow the operator to advance them by pushing. Two basic types of microcatheters are used: flowdirected and over-the-wire catheters. The former have extremely light and supple distal ends so that they are carried by antegrade blood flow to the target position. The latter are pushed over a guide wire to the objective. Flow-directed catheters work best when blood flow is increased and are therefore used to treat arteriovenous shunts. Their size limits their use to the injection of



Indications for INR techniques.



Procedure



Technique



Purpose/Benefit



Biopsy Functional testing



Percutaneous needle Balloon occlusion



Embolization of small vessels Embolization of large vessels Chemotherapy



Particles or liquid embolics Balloon or coils Superselective injection



Appropriate treatment selection Assessment of collateral blood supply prior to surgical or endovascular ligations Devascularization of tumour to stop haemorrhage To aid surgical exposure/dissection To maximize drug delivery



Chapter 57 Interventional techniques



liquids or very small particles. Over-the-wire catheters are used when blood flow is slower. The combination of guide wire and a stiffer catheter makes navigation easier and their larger lumen allows the delivery of both particles and coils. They are usually used in the treatment of head and neck tumours.



Materials used for embolization



] 733



proximal arteries and achieve less reliable tumour devascularization.6 The effect of PVA or gelatin particles on tissues is to cause an initial acute inflammatory response, which is followed by a chronic foreign body reaction within weeks.5, 7 Devascularization causes ischaemia and necrosis within hours, which in turn causes tumours to swell and may exacerbate symptoms, particularly pain. Preoperative embolization is therefore best performed the day before or immediately prior to operation.



PARTICLES



Embolic agents can be categorized on various criteria. Traditionally, they are divided according to whether the resulting devascularization is permanent or temporary. Temporary agents being autologous blood clot, gelfoam powder and microfibrillar collagen, whilst polyvinyl alcohol particles (PVA) are considered permanent agents (see Table 57.2). However, embolization using PVA is liable to be temporary because, though the particles are not degraded, their effect may be circumvented by development of new collateral vessels.4, 5 In practice, the development of new vessels is less likely if particles can be deposited in vessels within the tumour, rather than in proximal arteries. To ensure the deep penetration of injected particles, they are engineered to mix well with radiographic contrast so that they can be injected as an even suspension. PVA or Trisacryl gelatin microspheres are currently used in the author’s department. These come in a range of sizes varying between 40 and 800 microns in diameter. The choice of particle size is important because, though smaller particles will penetrate to smaller vessels, they are more likely to pass through arteriovenous shunts and reach the lungs. They are also more likely to cause ischaemic damage to normal tissues. On the other hand, larger particles occlude Table 57.2



Embolic agents.



Agent



Embolization



Target vessels



Autologous blood clot Gelfoam Polyvinyl alcohol particles Trisacryl gelatin particles Coils Balloons Alcohol (ethanol)



Temporary Temporary Permanent



Large arteries Large arteries Tumour vessels



Permanent



Tumour vessels



Permanent Permanent Permanent



Sodium tetradecyl sulphate Cyanoacrylate adhesives Ethylene vinyl alcohol copolymer



Permanent



Large arteries/veins Large arteries Malformations or tumour vessels Malformations or tumour vessels Malformations or tumour vessels Malformations or tumour vessels



COILS AND BALLOONS



Coils are made from steel, tungsten or platinium. Because of problems of corrosion, tungsten is generally no longer used and platinium, though expensive, is preferred because it is inert and not liable to magnetization effects during MRI. A new development is to enhance their effect by adding a thrombogenic coating. The materials used to coat coils include collagen, bioabsorbable polymers and dacron fibres.8 Balloons are made of latex or silicone. They are either fixed to a catheter tip for temporary test occlusions or can be detached for vessel occlusion. They are inflated by injecting fluid; usually radiographic contrast media. Coils and balloons are used to occlude large vessels. The delivery of coils is easier to control than balloons and they can be positioned very accurately. There are various methods for controlling the detachment of coils from a delivery wire. These include couplings released by mechanical or hydrostatic pressure, electrolysis or thermal heating.9 Detachable balloons are released mechanically from a delivery microcatheter, but their detachment is less reliable than the newer methods developed for coils. In many centres they have been superseded by coils. Coils or balloons are used for large artery occlusion (i.e. endovascular ligations) after a satisfactory temporary balloon occlusion test. High blood flow fistulas are occluded with coils or balloons positioned at the point of transition between artery and vein. Coils are also commonly used to embolize large veins or dural sinuses in order to close dural arteriovenous malformations (AVM). They are less often used in tumour embolizations.



LIQUID EMBOLIC AGENTS



Permanent Permanent



Various liquids are used for embolization including sclerosants, adhesives and soluble plastics. They have the advantage over coils and particles of being injectable via small lumen catheters or needles. Most commonly used are quick-setting adhesives and polymers. The latter are used for percutaneous puncture of small vessels within tumours or facial vascular malformations. Examples of liquid agents are the tissue adhesive n-butyl-2-cyanoacrylate (NBCA) and the copolymer ethylene vinyl alcohol. There are various commercial formulations



734 ] PART 11 RECENT ADVANCES IN TECHNOLOGY available. NBCA thickens and sets on exposure to blood and copolymers set after injection because of solvent dispersal. Both agents are liable to stick the delivery catheter to vessels and require considerable expertise to deliver safely, but they are generally regarded as producing the most permanent form of embolization. They are therefore used for definitive or palliative treatments and in situations where immediate surgical resection is not planned. Sclerosants, such as absolute alcohol, are used to treat facial vascular malformations, but replacement agents are being tested because of toxicity problems.



APPLICATION OF EMBOLIZATION TECHNIQUES



particles or NBCA are used.13 Though technically relatively straightforward, safe embolization of skull base lesions should only be undertaken after adequate training. Safe embolization demands a sound knowledge of vascular anatomy and the potential sites of spontaneous external to internal carotid artery or vertebral artery anastomoses, if unintended cerebral or spinal migration of emboli is to be avoided.14



PERCUTANEOUS EMBOLIZATION



This technique involves the injection of a liquid embolic agent or sclerosant directly into tumour vessels (Figure 57.1). The objective is to devascularize tumours before surgical resection or for palliation of symptoms by



Tumour embolization TRANSARTERIAL



Embolization of tumours of the head and neck is most often performed to reduce blood loss and facilitate surgical resection or for palliation (Table 57.3). It is rarely curative. Preoperative embolization is performed for vascular tumours such as meningioma.7, 9, 10 haemangioblastoma, juvenile nasopharyngeal angiofibroma,11 schwannoma and paraganglioma.12 Functional evaluations by temporary balloon occlusion of major arteries at risk during surgery can also be performed. Preoperative embolization is largely reserved for extraaxial intracranial tumours and only rarely are intraaxial central nervous system (CNS) tumours treated. Preoperative transarterial embolization is usually performed with PVA particles, sized to occlude the pathological circulation. Liquid agents, including alcohol, can be delivered transarterially by catheter, but are more frequently injected percutaneously (see below). Serious procedural complications occur in about 4 percent of patients and these include cranial nerve palsies (including blindness), stroke due to unrecognized spread of emboli and induced bleeding due to vessel perforation or tissue necrosis (see Embolization for epistaxis). Complicating cranial nerve palsies are more common when very small Table 57.3 Head, neck and skull base tumours treated by embolization. Commonly treated tumours



Meningioma Paraganglioma Juvenile angiofibroma Haemangiopericytoma



Less frequently treated tumours Schwannoma Carcinoid Alveolar sarcoma Thyroid carcinoma Granular cell myoblastoma Capillary haemangioma Esthesioneuroblastoma Neurinoma



Figure 57.1 Percutaneous embolization of facial venous angioma. (a) A coronal T1W MRI showing the left-sided lesion just below the ramus of the mandible (arrows). In (b), a lateral plain radiograph, a needle has been inserted into the angioma and radiographic contrast injected prior to embolization using ethanol. Because the venous space empties slowly, a sclerosant can be used to obliterate the lesion.



Chapter 57 Interventional techniques



causing tumour involution. Preoperative embolization is usually performed following percutaneous puncture, but perioperative needling can be performed. The technique was first used to devascularize hypervascular juvenile angiofibromas,15 but has also been used to treat haemangiopericytoma and paraganglioma.16 The technique is particularly attractive for the management of recurrent lesions and has been reported to achieve total or near-total tumour devascularization.17 INTRAARTERIAL CHEMOTHERAPY



Delivery of drugs via selective catheterization of a feeding artery aims to maximize the dose the tumour receives by allowing otherwise systemically toxic dose levels to be administered. The amount of drug delivered to a tumour is proportional to its rate of plasma clearance within the tissue and inversely related to tumour plasma flow.18 Selective catheterization allows the delivery rate to be fine tuned to the prevailing rate of blood flow. This benefit has been used in various studies of local chemotherapy for head and neck cancers, specifically advanced squamous cell carcinomas.19 Combination therapies with radiotherapy, for example, RADPLAT, have been effective at inducing remission, and Kumar and Robbins20 recently reviewed their use. The systemic toxic effects of cisplatin can be mitigated by the simultaneous intravenous administration of its competitive antagonist thiosulphate. Despite logistical problems associated with catheterization and long infusion times, preliminary trials have shown the technique to be highly effective at causing tumour shrinkage.21



] 735



can be performed by simultaneously lowering the systemic blood pressure or administering a vasodilator drug, such as acetazolamide.22 All these techniques are designed to improve the reliability of the test but the lack of a consensus protocol testifies to the fact that there remains a risk of the collateral blood flow being inadequate, despite a normal test. It is generally accepted that preliminary testing reduces the risk of delayed stroke, but the data supporting this conclusion come largely from reports of its use in aneurysm patients.23 Amongst patients undergoing ligation for skull base tumours, rates for complications range between 5 and 20 percent.24 The additional risks faced by cancer patients were documented in a single institution report which found their complication rate to be 10 percent, whilst that of aneurysm patients was only 3 percent.25 Following a satisfactory period of temporary occlusion, the artery is permanently occluded by detaching the balloon used to perform temporary occlusion, or replacing it with embolization coils. Some practitioners prefer the latter because the detachment systems for coils are more reliable. It is obviously important that permanent occlusion is performed at the same arterial level (i.e. at the same place) as the test occlusion.



EMBOLIZATION FOR MENINGIOMAS AND PARAGANGLIOMAS Two commonly treated tumour types will be described in detail to illustrate the goals of INR in the management of patients with tumour of the skull base and neck.



Temporary and permanent large artery occlusions



Meningioma



Temporary endovascular occlusion of large arteries prior to head and neck surgery allows angiographic and neurological testing of the awake patient to determine the consequences of vessel sacrifice. Rarely is large vessel sacrifice indicated for cancer palliation since the proximal occlusion of arteries is a relatively ineffective method of causing tumour necrosis and/or regression. Large vessel occlusions may be required to manage tumour haemorrhage, as described under Embolization for epistaxis. There are many described protocols for temporary artery occlusion and dynamic testing. Only two elements are common to all: (1) the procedure is performed under local anaesthesia so that the patient is accessible for neurological examination and (2) anticoagulants are given. Test occlusion involves inflating a balloon in the target artery for 20–30 minutes and assessing the effect on cerebral blood flow. The adequacy of collateral blood flow to the territory of the occluded vessel can be demonstrated by angiography, Doppler ultrasound, xenon CT, SPECT and PET scanning. Additional provocative testing



These typically benign tumours originate from arachnoid cap cells found in arachnoid granulations. They therefore arise in continuity with dura. Though generally intracranial, they are found at extracranial sites by extension from a dural origin or rarely occur entirely extracranially (presumably from ectopic dural rests). Thus tumours may involve the skull base, orbit and cervical spine or the upper neck. They occur in middle age, affect women twice as commonly as men and are linked to a genetic deficit on chromosome 22. Multiple lesions occur in patients with neurofibromatosis II and meningiomatosis.26, 27 Associations have been reported with exposure to ionizing radiation28 and hormonal influences due to the presence of androgen and other hormone receptors.27 They may be locally invasive and the WHO classification of histological findings recognizes three grades (grade 1, benign; grade 2, atypical; grade 3, anaplastic), which are independent of the traditional histopathological descriptions of meningothelial, fibrous, transitional, syncytial and psammomatous subtypes. They are usually highly vascular tumours



736 ] PART 11 RECENT ADVANCES IN TECHNOLOGY with the transitional subtype being most vascular and the psammomatous subtype least vascular. Clinical presentation varies according to location and imaging by CT scan or MRI is usually adequate for diagnosis and preoperative assessment. Tumour calcification and skull hyperostosis is easier to recognize on CT than MRI and suggest a less vascular subtype. Tumour enhancement after intravenous contrast media administration is typical and is more avid in vascular tumours, but such enhancement does not imply that a particular lesion is suitable for embolization. Treatment is by resection and radiotherapy is reserved for more aggressive histological types. Preoperative embolization in indicated for vascular tumours and for the palliation of inoperable newly diagnosed or recurrent tumours. Intraarterial angiography is required to assess the extent of tumour vascularity and therefore the potential of embolization. In a minority of tumours, dense calcification on CT or lack of enhancement after contrast media administration imply a less vascular tumour and embolization will not contribute to management, but the majority warrant angiography and possible embolization. The vascular supply to meningioma is typically arranged in a radial pattern of dilated feeding arteries of decreasing size and a delayed venous phase of contrast passage through the tumour bed. Arterial supply may be from external carotid artery branches or from transpial internal carotid artery branches or a mixture of the two (see Figure 57.2). The endovascular therapist therefore needs a detailed preembolization angiogram to evaluate the feasibility (and risks) associated with superselective catheterization and injection of particles. In practice, treatment and catheter angiography are usually combined in the same session and scheduled as part of the patient’s overall treatment plan. Embolization is indicated as an adjuvant to tumour resection in order to reduce operative blood loss and facilitate surgery. Since the typical blood supply is by arteries and arterioles of gradually reducing size, small particles (for example, 150 micron) are used first to obstruct intratumour vessels and larger particles injected subsequently to obstruct larger feeding arteries (see Figure 57.3). If embolization is performed preoperatively, this technique will cause acute tumour infarction and swelling should be anticipated. Operation is therefore best performed within hours of embolization or if delayed then steps must be taken to ensure that worsening neurological symptoms or signs are quickly detected and treated. Rarely, tumour necrosis causes bleeding and lifethreatening expansion of tumours. This complication occurs in about 1–2 percent of instances. Revascularization of tumour after particulate embolization, due to growth of new vessels and collateral routes of blood supply, can also be anticipated within as short a period as two to three weeks. Another indication for meningioma embolization is to arrest or impede tumour growth when lesions are



inoperable and radiotherapy inappropriate. In theory, particles delivered into the tumour bed only will cause tissue ischaemia/necrosis without stimulating growth of collateral blood vessels. In this situation, very small particles (50 microns) are injected slowly and no attempt is made to obstruct larger feeding arteries. Tumour devascularization and shrinkage is monitored by MRI enhanced by intravenous gadolinium administration; induced failure of enhancement implying devascularization. The efficacy of preoperative embolization is difficult to evaluate objectively. Criteria used in reported studies include comparisons of blood transfusion volumes between tumour resection operations performed with or without preoperative embolization, lengths of hospital stay and clinical outcomes. A reduced need for blood transfusion was reported by Teasdale et al.29 and Oka et al.,30 shorter lengths of hospital admission by Dean et al.31 and Oka et al.30 and improved clinical results by Oka et al.30 when adjuvant embolization was performed. A reduction in neurological complications after surgery was reported by Oka et al.,30 but this benefit was only evident for patients with large tumours (greater than 6 cm in diameter). The conduct and evaluation of such studies is difficult because of the various sizes and sites of tumours studied. Several specific complications associated with meningioma embolization have been reported. Intratumoural haemorrhage resulting from devascularization may occur in the 24–36 hours after embolization. In the author’s practice, this complication has occurred once in 40 treatments over the last five years. Selective catheterization risks causing vessel damage or rupture. The latter is more likely to occur at sharp bends in vessels. Such bends occur in the course of the middle meningeal artery (a common route for meningioma embolization) and rupture of this artery has been reported to cause bleeding or delayed arteriovenous shunting.32, 33 The incidence of all procedural complications is about 4 percent, though complications resulting in permanent neurological deficits occur in less than half of affected patients.



Paraganglioma These benign but locally invasive tumours arise from paraganglionic chemoreceptor cells of neural crest origin. About 50 percent occur in the temporal bone, arising from either the cochlear promontory (i.e. typanicum) or the jugular blub (i.e. jugulare), 35 percent in the carotid body, 12 percent in the region of the high cervical vagus and the rest at various sites of the head and neck.34 Though relatively rare tumours, they are usually targets for embolization because of their highly vascular nature. Most occur sporadically though familial cases occur with autosomal dominant inheritancy. In approximately 10 percent of patients, tumours are multifocal and



Chapter 57 Interventional techniques



] 737



Figure 57.2 Parasellar meningioma. This tumour is supplied by meningeal branches arising from both internal and external carotid arteries. (a) and (b) are lateral views from internal and external carotid injections, respectively. Preoperative embolization of the external carotid supply only was performed. (c) The tumour circulation following selective catheterization of the accessory meningeal artery. The arrow marks the tip of the microcatheter. (d) The effect of embolization on the external carotid angiogram (compare with panel b). (e) and (f) Axial T1W MRI of the tumour before and after embolization of the external carotid blood supply. Note that enhancement occurs in only the medial portion of the tumour on (f) (arrows).



738 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



Figure 57.3 Temporal paraganglioma. Lateral angiograms following selective injections into the ascending pharyngeal artery before (a), (b) and after (c), (d) embolization with particles. Images (b) and (d) show the bony details of the digitally subtracted angiograms (a) and (c). The tumour circulation (arrows in (a) and (b)) has been occluded.



up to 5 percent of tumours secrete catacholamines. Symptoms are related to tumour location and typically consist of mass, tinnitus (tympanicum), cranial nerve palsy (jugulare, vagale) and pain. The role of imaging by plannar scanning is to support the clinical diagnosis and



usually includes both CT and MRI. The former (performed with high resolution parameters) will best demonstrate bony involvement and MRI will best demonstrate the soft tissue character of tumours. Tumour vascularity and extent can be assessed from scans



Chapter 57 Interventional techniques



performed after intravenous contrast administration and evidence of involvement of the carotid artery or intracranial invasion should be sought. Intraarterial angiography is usually undertaken as part of a preembolization evaluation rather than for surgical planning.35 The typical angioarchitecture of this tumour type makes angiography a very specific diagnostic test and it should be considered prior to biopsy. On intraarterial angiography, feeding arteries are found to be enlarged and there is an early and intense blush in the tumour bed. The intratumoural angioarchitecture comprises centripetally orientated arterioles estimated to be 90 mm in diameter at the periphery and 300–600 mm in the centre of the tumour with arteriovenous shunts (see Figure 57.3).36 To further complicate embolization, the majority of tumours show a multicompartment pattern of blood supply with arterial and venous supply confined to a haemodynamic unit so that separate injections of embolic agent have to be made into each feeding artery. The technique for embolization has evolved with the development of embolization particles small enough to penetrate the smaller peripheral arteries (see Figure 57.4). The presence of intratumour shunts may justify the use of a percutaneous (or intraoperative) direct puncture technique with injections of a liquid adhesive agent (see Tools used in interventional neuroradiology techniques, above) since spread of emboli to the lungs is less likely. However, arteriovenous (AV) shunting is more rare in paraganglioma than within the vascular bed of some other tumour types (for example, nasopharyngeal angiofibroma).37 Embolization is usually performed as an adjuvant to surgical resection. In rare instances, when surgery and/or radiotherapy are considered inappropriate, embolization alone may provide symptomatic relief by stabilizing tumour growth.38 The benefits of preoperative embolization have been demonstrated in several single institution reports.39, 40 The efficacy of preoperative embolization has



Figure 57.4 Histological section showing embolization particles (arrows) within tumour vessels.



] 739



been studied by several groups. Murphy and Brackman compared two groups, totalling 35 patients, and found preoperative embolization reduced the volume of operative blood loss and the lengths of procedure times, but not the length of bed stay.41 Tikkakoshi et al.42 reported that preoperative embolization improved operating conditions with subjective benefits to surgical results if performed by superselective catheterization and with effective devascularization of the tumour vascular bed.



EMBOLIZATION FOR EPISTAXIS Transarterial embolization using particles, coils or balloons has been used to control intractable idiopathic epistaxis. For effective embolization, superselective catheterization of the sphenopalatine artery is performed after lateralization of the bleeding site. This goal is frustrated if prior surgical ligation of the internal maxillary artery has been performed or if the site of bleeding cannot be established. In these situations, particles may need to be injected via both internal maxillary and facial arteries. There is no consensus on the most effective timing of embolization after the onset of spontaneous epistaxis, and different definitions of intractable epistaxis make it difficult to compare published reports. The vast majority of patients admitted to hospital with epistaxis respond to packing, balloon tamponade or local cautery and only a small minority of patients will need embolization.43 Whether embolization is substantially safer or more effective than internal maxillary artery (IMA) ligation is also debated. In a recent review of the literature comparing embolization with IMA ligation, Cullen and Tami44 found that embolization was generally reported to be more reliable, but this was not the case in their institution, where outcomes were similar. However, amongst their patients the procedural complication rate was higher after IMA ligation. If the risks are no greater, then it seems logical to employ embolization first, since IMA ligation limits its subsequent use. For patients with epistaxis secondary to a vascular malformation or nasal tumour, particulate embolization is indicated to stop acute haemorrhage. In patients with Osler–Weber–Rendu disease, multiple sessions may be required to induce remission and embolization is rarely curative.45 In patients with epistaxis due to the rare internal carotid aneurysm or pseudoaneurysm that erodes the sphenoid bone and bleeds intranasally, emergency embolization using coils or balloons can be life saving. Embolization also has a small but important role in the management of patients bleeding as a result of head and neck cancers elsewhere. Haemorrhage may occur from large or small vessels and may be associated with radiation-induced tissue necrosis. A variety of embolization techniques including particles, coils and balloons may be needed to control such bleeding.46



740 ] PART 11 RECENT ADVANCES IN TECHNOLOGY probably open new areas of research. The in vivo response of lesions will need further study.



KEY POINTS  Pre-operative embolization should be considered for all vascular tumours.  Tumour vascularity can only be inferred from CT and MR scans. Intraarterial angiography is still required to identify tumours suitable for embolization.  Functional testing of vessels ‘at risk’ during surgery is performed under local anaesthesia by temporary inflation of an endovascular balloon.  Most embolizations are performed transarterially, but direct percutaneous injections of tumour or vascular malformation vessels or transvenous injections are also possible.  Embolization alone is used for palliation and management of inoperable tumours.  Embolization should be considered early rather than late in the management of intractable epistaxis.



Deficiencies in current knowledge and areas for future research Technical



$



$ $



Development of non-x-ray imaging. Fluoroscopy using magnetic resonance imaging is now feasible. This avoids the potential hazards of ionizing radiation but requires the development of new catheters and embolic agents. Attempts to develop liquid agents that remain stable in high strength magnetic fields are proving difficult. A great deal of investment is being made to develop ways of monitoring minimally invasive surgical techniques without x-ray fluoroscopy. Current liquid embolic agents are adhesive and nonadhesive materials that remain coherent during delivery and are permanent need to be developed. Improvements in reliability of data from functional tests are needed. A great deal of research has been directed at the problem of falsely reassuring temporary vessel occlusion tests. A completely reliable protocol is still needed.



Biological



$ $



The evolution of embolized tissue has been studied, but we do not know much about the mechanisms that cause secondary haemorrhage in some tumours. The potential benefits of local delivery for gene therapy and future chemotherapy agents will



REFERENCES







1. Northfield DWC. Chapter 13. In: The surgery of the central nervous system. Oxford: Blackwell Scientific Publication, 1973: 395–6. 2. Miniz E, Lima A, Caldas P. Angiographie en serie de la circulation de la tete. Revue Neurologique. 1934; 1: 4. 3. Luessenhop AJ, Velasquez AC. Observations on the tolerance of the intracranial arteries to catheterization. Journal of Neurosurgery. 1964; 21: 85–91. 4. Quisling RG, Mickel JP, Ballinger W. Small particle polyvinyl alcohol embolisation of cranial lesions with minimal arteriolar-capillary barriers. Surgical Neurology. 1986; 25: 243–52. 5. Hamada J, Kai Y, Nagashiro S, Hashimoto N, Iwata H, Ushio Y. Embolization with porous beads. II: Clinical trial. American Journal of Neuroradiology. 1996; 17: 1900–6. 6. Latchlow RE. Preoperative intracranial meningioma embolisation: Technical considerations affecting the riskto-benefit ratio. American Journal of Neuroradiology. 1993; 14: 583–6. 7. Kerber CW, Bank WP, Horton JA. Polyvinyl alcohol foam: Prepackaged emboli for therapeutic embolisation. American Journal of Neuroradiology. 1978; 130: 1193–4. 8. Sellar R. Endovascular techniques. In: Byrne JV (ed.). A Textbook of interventional neuroradiology: Theory and practice. Oxford University Press, 2002: 53–4. 9. Guglielmi G, Vinuela F, Sepetka I, Macellari V. Electrothrombosis of saccular aneurysms via endovascular approach. Part 1: Electrochemical basis, technique and experimental results. Journal of Neurosurgery. 1991; 75: 1–7. 10. Manelfe C, Lasjaunias P, Ruscalleda J. Preoperative embolizaton of intracranial meningioma. American Journal of Neuroradiology. 1986; 7: 963–72. 11. Davis KR. Debrun. Embolization of juvenile nasopharyngeal angiofibroma. Seminars in Interventional Radiology. 1987; 4: 309–20. 12. Lacour P, Doyon D, Manelfe C, Picard L, Salisachs P, Schwaab G. Treatment of chemodectomas by arterial embolization. Journal of Neuroradiology. 1975; 2: 275–87. 13. Lasjaunias P, Berenstein A. Chapter 4. In: Surgical neuroangiography: endovascular treatment of craniofacial lesions. Berlin: Springer-Verlag, 1987: 90–99. 14. Valavanis A. Preoperative embolization of the head and neck: Indications, patient selection, goals, and precautions. American Journal of Neuroradiology. 1986; 7: 943–52. 15. Tranbahuy P, Borsik M, Herman P, Wassel M, Casasco A. Direct intracranial embolisation of juvenile angiofibroma. American Journal of Otolaryngology. 1994; 15: 429–35.



Chapter 57 Interventional techniques















16. George B, Casasco A, Deffrennes D, Houdart E. Intratumoral embolisation of intracranial and extracranial tumours: technical note. Neurosurgery. 1994; 35: 771–3. 17. Chaloupka JC, Mangla S, Huddle DC, Roth TC, Mitra S, Ross DA et al. Evolving experience with direct puncture therapeutic embolization for adjunctive and palliative management of head and neck hypervascular neoplasms. Laryngoscope. 1999; 109: 1864–72. 18. Howell SB. Pharmakokinetic principles of regional chemotherapy. Contributions to Oncology. 1988; 29: 1–8. 19. Lee YY, Dimery IW, Von Tassel P, De Pena C, Blacklock JB, Goepfert H. Superselective intraarterial chemotherapy of advanced paranasal sinus tumours. Archives of Otolaryngology – Head and Neck Surgery. 1989; 115: 503–11. 20. Kumar P, Robbins KT. Treatment of advanced head and neck cancer with intra-arterial Cisplatin and concurrent radiation therapy: The RADPLAT protocol. Current Oncology Reports. 2001; 3: 56–65. 21. Kerber CW, Wong WH, Howell SB, Hanchett K, Robbins KT. An organ-preserving selective arterial chemotherapy strategy for head and neck cancer. American Journal of Neuroradiology. 1998; 19: 935–41. 22. Rogg J, Rutigliano H, Yonas H, Johnson DW, Pentheny S, Latchaw RE. The acetazolamide challenge: Imaging techniques designed to evaluate cerebral blood flow reserve. American Journal of Neuroradiology. 1989; 10: 830–910. 23. Byrne JV, Guglielmi G. Chapter 4. In: Endovascular treatment of intracranial aneurysms. Berlin Heidelberg, New York: Springer-Verlag, 1998: 120–2. 24. Standard SC, Ahuja A, Guterman LR, Chavis TD, Gibbons KJ, Barth AP et al. Balloon test occlusion of the internal carotid artery with hypotensive challenge. American Journal of Neuroradiology. 1995; 16: 1453–8. 25. Gonzalez CF, Moret J. Balloon occlusion of the carotid artery prior to surgery for neck tumours. American Journal of Neuroradiology. 1990; 11: 649–52. 26. Langford LA. Pathology of meningiomas. Journal of NeuroOncology. 1996; 29: 217–21. 27. Black PMcL. Meningiomas. Neurosurgery. 1993; 32: 643–57. 28. Bondy M, Ligon BL. Epidemiolgy and etiology of intracranial meningiomas: a review. Journal of NeuroOncology. 1996; 29: 197–206. 29. Teasdale E, Patterson J, McLellan D, MacPherson P. Subselective preoperative embolisation for meningiomas. A radiological and pathological assessment. Journal of Neurosurgery. 1984; 60: 506–11. 30. Oka H, Kurata A, Kawano N, Seregusa H, Ikuo K, Ohmomo T et al. Preoperator superselective embolisation of skullbase meningiomas: indications and limitations. Journal of Neuro-Oncology. 1998; 40: 61–71. 31. Dean BL, Fiom RA, Wallace RC, Khayata MH, Obuchowski NA, Hodak JA et al. Efficacy of endovascular treatment of



32.



33.



34.



35.



36.



 37.



38.



39.



40.



 41. 42.



43.



44.



45.



46.



] 741



meningiomas: evaluation with matched sample. American Journal of Neuroradiology. 1994; 15: 1675–80. Terada T, Nakai E, Tsumoto T, Itakura T. Iatrogenic arteriovenous fistula of the middle meningeal artery caused during embolisation for meningioma – case report. Neurologia Medico-Chirurgica. 1997; 37: 677–80. Barr JD, Mathis JM, Horton JA. Iatrogenic carotidcavernous fistula occurring after embolization of a cavernous sinus meningioma. American Journal of Neuroradiology. 1995; 16: 483–5. Zak FG, Lawson W. The paraganglionic chemoreceptor system. Physiology, pathology and clinical medicine. New York: Springer Verlag, 1982. Phelps PD, Cheesman AD. Imaging jugulotympanic glomus tumors. Archives of Otolaryngology – Head and Neck Surgery. 1990; 116: 940–5. Willis AG, Birrel JH. The structure of a carotid body tumour. Acta Anatomica (Basal). 1955; 25: 220–65. Schroth G, Haldemann AR, Mariani L, Remonda L, Raveh J. Preoperative embolization of paragangliomas and angiofibromas. Archives of Otolaryngology – Head and Neck Surgery. 1996; 122: 1320–5. Maier W, Marangos N, Laszig R. Paraganglioma as a systemic syndrome: pitfalls and strategies. Journal of Laryngology and Otology. 1999; 113: 978–82. Ogura JH, Spector GJ, Gado M. Glomus jugulare and vagale. Annals of Otology, Rhinology, and Laryngology. 1978; 87: 622–9. Simpson 2nd. GT, Konrad HR, Takahashi M, House J. Immediate postembolization excision of glomus jugulare tumors: advantages of new combined techniques. Archives of Otolaryngology – Head and Neck Surgery. 1979; 105: 639–43. Murphy T, Brackmann DE. Effects of preoperative embolization on glomus jugulare tumors. Laryngoscope. 1989; 99: 1244–7. Tikkakoshi T, Luotonen J, Leinonen S, Siniluoto T, Heikkila O, Paivansalo M et al. Preoperative embolisation in the management of neck paragangliomas. Laryngoscope. 1997; 107: 821–6. Pollice PA, Yoder MG. Epistaxis: A retrospective review of hospitalised patients. Otolaryngology and Head and Neck Surgery. 1997; 117: 49–53. Cullen MM, Tami T. A Comparison of internal maxillary artery ligation versus embolisation for refractory posterior epistaxis. Otolaryngology and Head and Neck Surgery. 1998; 118: 636–42. Hirsch JA, Choi I-S. Vascular Malformations of the head and neck. In: Byrne JV (ed). Interventional neuroradiology. Oxford: Oxford University Press, 2002: 197–212. Morrissey DD, Anderson PE, Nesbit GM, Barnwell SL, Everts EC, Cohen JL. Endovascular management of haemorrhage in patients with head and neck cancer. Archives of Otolaryngology – Head and Neck Surgery. 1997; 123: 15–9.



58 Laser principles in otolaryngology, head and neck surgery BRIAN JG BINGHAM



Introduction History Principles of laser action Laser light delivery devices Laser–tissue interaction Laser applications in otolaryngology



742 742 742 743 745 745



Laser safety Photodynamic therapy in otorhinolaryngology Key points Deficiencies in current knowledge and areas for future research References



745 745 746 747 747



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words head and neck cancer and photodynamic therapy.



INTRODUCTION Laser is the acronym for light amplification by stimulated emission of radiation. Surgical lasers are devices that amplify light and create coherent light beams ranging from the infrared to the ultraviolet parts of the spectrum.



HISTORY In 1917, Albert Einstein described the theory of stimulated emission which is the underlying process for laser action. The American physicists, Arthur Schawlow and Charles Townes, described the working principles of lasers in 1958.1 In 1960 Theodore Maiman demonstrated the first laser action in solid ruby2 and a year later Ali Javan built the first helium-neon gas laser.3 C. Kumar N. Patel introduced the carbon dioxide (CO2) gas laser in 1962. In 1972, in Boston, USA, Jako and Strong were the



first to pioneer the use of the CO2 laser in otolaryngology, head and neck surgery.4, 5, 6



PRINCIPLES OF LASER ACTION A laser empowers atoms to store and emit light in a coherent form. The electrons in the atoms of a laser medium are first pumped or energized to an excited state by an external energy source. These electrons are then stimulated by external photons to emit their stored energy in the form of photons. This process is ‘stimulated emission’. The photons emitted have a frequency characteristic to their atoms and travel in step with the stimulating photons. These photons now strike other excited atoms to release even more photons. These photons move back and forth between two parallel mirrors triggering further stimulated emission. This part of the process is known as ‘light amplification’. One mirror in the laser tube is partially silvered and it allows the exit or leak of the intense, collimated, monochromatic and coherent laser light.



Chapter 58 Laser principles in otolaryngology, head and neck surgery



Nature of laser light



] 743



 coherent: the photons or waves travel in step, or in phase with one another.  collimated: the laser light travels in one direction.  monochromatic: one wavelength or colour is in the visible spectrum.



output may be continuous wave or pulsed. A continuous wave laser operates with a constant intensity. A laser that operates with a continuous output for longer than 0.1 seconds is considered a continuous wave laser. A pulsed laser produces a single or train of pulses with each individual pulse less than 0.1 seconds. A Q-switch is an electro-optical component that facilitates the production of a very short (less than 1 microsecond) but high intensity pulse of laser energy.



Types of lasers



Basic laser tissue interaction



Depending on the type of laser medium used, lasers can be classified as solid state, gas, semiconductor, liquid or free electron. A pioneering example of a solid state laser is the ruby laser. The neodymium yttrium aluminium garnate (YAG) laser and the related (frequency doubled) potassium titanyl phosphate (KTP) are examples of solid state lasers in surgical practice. The carbon dioxide laser (CO2) is a very efficient gas laser that delivers laser light as an invisible continuous wave beam. The carbon dioxide surgical laser has been the most widely applied laser in oral and laryngological practice. The helium-neon laser is a gas laser with high frequency stability, good colour stability (red) and minimal beam spread. The helium-neon gas laser is often superimposed on an invisible laser beam, such as the CO2, to facilitate surgical targetting. Activation of chemicals in photodynamic therapy can be achieved with the helium-neon laser. The gallium arsenide laser is the most commonly used semiconductor laser. This laser is used in CD players and laser printers and has some surgical applications. The most common medium for a liquid laser is an inorganic dye contained in a glass vessel. This laser is pumped or energized by intense flash lamps or by a gas laser in a continuous wave mode. The frequency of a tuneable dye laser can be altered with the aid of a prism inside the laser cavity. Tuneable dye lasers can be well suited for treating pigmented cutaneous lesions. Lasers that utilize beams of electrons unattached to atoms and spiralling around magnetic field lines were initially developed in 1977 and are important research instruments. Free-electron lasers are tuneable and, in theory, could cover the electromagnetic spectrum from infrared to x-rays. Free-electron lasers could be capable of producing very high power radiation and may have medical applications in the future.



The reaction of laser energy with living tissue can be photoablative, photochemical, photomechanical or photothermal. Most lasers react with a combination of all these mechanisms although for a specific wavelength and delivery system one form of tissue reaction may predominate.



A beam of laser light is:



Patterns of laser output The configuration of the resonator cavity and the method in which an energy source is applied to the ‘active laser medium’ will determine the pattern of a laser output. The



 Photoablative reactions occur when molecular bonds are divided. The ruby laser, for example, can split the molecular bonds of tattoo ink with minimal local thermal damage. Macrophages remove the tattoo ink after the molecular bonds are broken.  Photochemical reactions occur when infrared, visible or ultraviolet laser light interacts with photosensitizers to produce chemical and physical reactions. This forms the basis for photodynamic therapy that is discussed later in this chapter.  A photomechanical effect occurs when the laser energy is pulsed to disrupt tissue or stones by the mechanism of shock waves. An example of this mechanism would be the use of the Holmium YAG laser to shatter ureteric and renal calculi.  The conversion of absorbed laser light into heat is a photothermal reaction. The tissue effect can be cutting, coagulation or vaporization depending on the laser wavelength and the laser delivery device.



LASER LIGHT DELIVERY DEVICES Many delivery devices are available to deliver laser energy to tissue in a safe and efficient manner. Some delivery devices are suited ideally for the wavelength that they transmit. Examples of delivery systems would be an articulated arm, a mirror lens system, micromanipulator, fibreoptic fibre, shaped tip fibreoptic fibres and robotized scanners. An articulated arm uses a system of hollow tubes and mirrors to direct the laser beam to the target area. Most articulated arms have a lens system to focus the emerging beam on the target tissue. The carbon dioxide laser wavelength is absorbed in fibreoptic material and hence this laser energy is delivered with an articulated arm system.



744 ] PART 11 RECENT ADVANCES IN TECHNOLOGY Micromanipulators and other focussing devices can be connected to microscopes. Micromanipulators and focussing devices will create an accurate and reproducible spot on target tissue. This facilitates very accurate multiple laser strikes which is important in otology and laryngology. Laser energy exits from a micromanipulator or focusing device in a similar style to the ‘laser beam weapon’ of science fiction. This means that inadvertent strikes beyond the target tissue or misdirection from a reflective surface are a significant operative risk. It also means that a ‘laser beam’ can be deliberately redirected by a mirror. An example of this would be to use a mirror to redirect CO2 laser energy on to adenoid tissue or on to lymphoid tissue on the base of the tongue. The bare fibreoptic fibre is the most common technique for delivering laser energy to tissue. The flexibility and diameter of fibreoptic fibre facilitates their use with both rigid and flexible endoscopes. This means that laser energy can be delivered to almost any tissue that can be seen with an endoscope. The tissue response to laser energy exiting from the tip of a fibre can be controlled by altering the distance of the fibre from the tissue. Changing the distance between the fibre and the tissue can produce all the tissue photothermal effects of coagulation, incision and vaporization. Laser light exits the tip of the fibre in what is called the ‘angle of divergence’. The light exiting from the tip of the fibre projects in a similar manner to water projecting from the nozzle of a hosepipe. The smallest light spot and, consequently, the highest concentration of energy are found at the fibre tip. As the distance from the fibre to tissue increases, the projected area of light decreases and the energy intensity reduces. A fibre ‘in contact’ with tissue will create an incision. The cut is similar to a saw rather than a knife. This means that some tissue is removed like the kerf of a saw as compared to being split with a scalpel. There is a degree of tactile control with a laser fibre ‘in contact’ mode. As the fibre is retracted and hovers above the tissue (2–4 mm), it is in the ‘near contact’ position and vaporizes tissue. As the laser fibre is retracted even further from tissue into the ‘noncontact’ mode, then the tissue effect is coagulation. A fibreoptic laser cable can be inserted through the biopsy channel of a fibreoptic endoscope. The end of the laser fibre must protrude beyond the end of the endoscope or heat damage to the endoscope can occur when the laser is fired. To simplify this problem one can perform a preliminary check of the length of fibre required to achieve a satisfactory distal position and then mark the laser fibre with a ‘steristrip’ at the entry port. A fibreoptic laser fibre can be inserted through a narrow metal tube to help direct the laser energy. In otology, for example, the laser fibre can be presented through a hollowed out needle. An aspiration channel can be added to the hollow metal fibre carrier to remove smoke from the surgical site. Figure 58.1 shows a



Figure 58.1 A 0.6-micron laser fibre inserted through a modified aural suction tube to create a laser delivery device. The natural torque of the fibre holds the laser fibre in place.



Figure 58.2 The pressure required to move or regress the laser fibre tip into the delivery device.



modified ear suction tube and fibre holder that is used for endoscopic intranasal surgery. In the design of these instruments, the curves of the fibre carrier should create sufficient resistance that when combined with the torque of the laser fibre it results in a fibre that is ‘held’ in position without any additional mechanism. In Figure 58.2, the 0.6-micron laser fibre can be set at the position of choice for the surgeon and will not move unless the surgeon applies additional pressure. The indent of the finger shows the pressure to overcome the self-retaining torque. A shaped tip laser is a laser fibre with a tip constructed of another material, such as metal or synthetic sapphire. The laser energy heats the tip and it is the conduction of heat from the tip that produces the tissue interaction. Different materials and shapes of tip will create different tissue reactions. Some tips become so hot that a secondary cooling mechanism is required. An example of this type of device in head and neck practice would be a sapphire tipped neodynium YAG fibre used to ablate tracheobronchial tumours.



Chapter 58 Laser principles in otolaryngology, head and neck surgery



Robotized scanners can facilitate laser treatment by reproduction of previous instrument settings or by being able to control precisely the distance a laser is placed or ‘hovered above’ the target tissue. A robotized scanner can ‘trace’ an area of treatment before delivery of laser energy to ensure accuracy or can apply a pattern to the area of laser treatment. At present, the greatest use for robotized scanners is in the cutaneous treatment of pigmented lesions and for skin resurfacing or recontouring.



LASER–TISSUE INTERACTION When laser light strikes tissue, it scatters until all the laser energy is either absorbed or reflected. The absorption of laser energy heats tissue. The volume of tissue affected is larger than the area shown by the laser spot size. The depth of penetration by a given wavelength of laser light is determined by the absorption and scattering of the type of tissue and the wavelength of the laser. The strength of the tissue effect is altered by changing the fluence. Fluence is defined as the laser power in watts, multiplied by the length of exposure in seconds, divided by the area (mm2) of the target tissue. This implies that when the distance between a fibreoptic tip and the tissue varies, the spot sizes becomes bigger or smaller and the fluence will accordingly increase or decrease. The heat in the tissue produces a series of changes in the tissue as the temperature rises. The changes are denaturation, coagulation, vaporization, carbonisation and incandescence. Cutting with a laser is narrow controlled vaporization! The heat produced by the absorption of laser light is partly dissipated into air but also produces a secondary thermal effect in surrounding tissue. This lateral thermal effect produces haemostasis by coagulation. The lateral haemostasis effect varies with wavelength of laser, the rate at which the energy is applied, fluence and the nature of the tissue.



LASER APPLICATIONS IN OTOLARYNGOLOGY The specific clinical applications of lasers are not described in this chapter, but the following list identifies particular lasers for disease-specific regions:  stapedectomy and tympanomastoid surgery: carbon dioxide laser, KTP laser, argon laser;  endonasal dacrocystorhinostomy: Holmium YAG laser, KTP laser, 810 nm laser;  nasal and sinus surgery, including tumour surgery, telangectasia destruction, antrostomy creation and turbinate reduction: Holmium YAG laser, KTP laser, Nd YAG laser, argon laser, 810 nm;  nasal polyp reduction: CO2 laser ‘swiftlase’ (oscillating device).



] 745



LASER SAFETY Each country applies different safety regulations throughout the world. Usually the Medical Devices Agency of a country or state will produce guidance for clinicians and hospitals. In the UK, the local Medical Physics Department will produce local laser rules (based on national guidance) and ensure training of all staff involved in the use of the laser. Training for medical and nursing staff should be mandatory. A laser company will typically provide a safety manual on their particular wavelength of laser. Staff should read the laser safety manual and in many countries produce signed documentation that the manual has been read. The principles of safety are as follows.  A laser beam may hit or damage objects outside the target area and cause a fire, tissue damage or eye/ visual damage.  The patient should be shielded by wet gauze or fireproof material from accidental strikes.  The anaesthetic tube and airway should be protected from accidental strikes that could produce a fire.  All personnel in the operating room should wear correct eye protection appropriate to the wavelength of the laser.  A warning sign plus locked doors should prevent unprotected and unprepared individuals from walking into the operating room when the laser is in use.  The key to switch on the laser should be held by a senior member of the operating team to ensure only properly qualified individuals use the laser.  The operating room and windows should be laser protected.  Endoscopic equipment should be blackened to reduce accidental reflective strikes of the laser.



PHOTODYNAMIC THERAPY IN OTORHINOLARYNGOLOGY Principles of photodynamic therapy Phototherapy is the use of light in the treatment of disease. An example would be the treatment of neonatal hyperbilirubinaemia with visible light. Photochemotherapy is a subset of phototherapy where a treatment requires the administration of a drug in addition to the application of light. An example would be treatment of a skin condition such as psoriasis with the combination of a furocoumarin sensitizer drug and UV-A (320–400 nm) light. Photodynamic therapy (PDT) is a subset of photochemotherapy where, in addition to an administered drug and the application of light, oxygen is required to complete the process. The administered drug is known



746 ] PART 11 RECENT ADVANCES IN TECHNOLOGY as a ‘photosensitizer’. This type of drug accumulates within a cell and reacts with light and oxygen to form ‘singlet’ oxygen. The ‘singlet’ oxygen damages the cell membranes and produces cell death. One of the attractions of PDT for oncology is that the photosensitizer drug is typically retained in tumour tissue for a longer period than in normal tissue. This improves the therapeutic effect while reducing toxicity to normal tissue. The use of PDT is not affected by prior radiotherapy, chemotherapy or surgery. To date, the principle shortcomings of PDT have been the limited depth of tissue penetration and extensive skin phototoxicity. This skin phototoxicity requires that the patient is not exposed to daylight during treatment or they develop extensive and painful skin reactions. In photodynamic therapy for head and neck cancer the photosensitizer is usually administered by intravenous infusion. The perfect photosensitizer drug would be absorbed selectively only by tumour tissue in the drug uptake period (also called the drug–light interval). The drug uptake period may last between 3 and 96 hours. The tumour is then irradiated with a measured light dose. Singlet oxygen is released to kill only tumour cells and thus destroy the target tumour. The tumour targeting arises firstly from the structure of the photosensitizer that leads to the selective uptake by a tumour and secondarily by the accurate application of an appropriate light wave. The diseases in which trials of photodynamic therapy have taken place include malignant tumours, such as squamous carcinoma of the oral cavity, squamous carcinoma of the nasopharynx, oesophageal carcinoma and metastatic squamous carcinoma of the neck. Nonmalignant disease such as inverted papilloma of the sinonasal cavity has also been targeted.



Photodynamic clinical reports in otorhinolaryngology Both palliative and curative therapy has been reported for oesophageal cancer. The patients receive a photosensitizer followed by light delivered via optical fibres through a flexible endoscope. A phase II trial7 randomized 218 patients to receive palliative oesophageal PDT (porfimer sodium sensitizer) versus Nd:YAG laser ablative therapy. Each arm gave an equivalent improvement in dysphagia but with fewer perforations in the PDT group (1 versus 7 percent, po0.05). [***] Overholt and Panjehpour8 reported the results of 55 patients with dysplasia or early carcinoma in Barrett’s mucosa of the lower end of the oesophagus treated with PDT. The patients received porfimer sensitizer followed at 48 hours with red light therapy. There was a good response at the six-month follow up. Forty-three patients with high-grade dysplasia/adenocarcinoma had endoscopic-proved ablation of their problem. Eleven of the 12



patients with low-grade dysplasia had no dysplasia on endoscopic review. Oesophageal stricture, however, developed in 53 percent of the patients. [***] There has been hope for PDT in the treatment of head and neck tumours (See Chapter 188, Nasopharyngeal carcinoma; Chapter 192, Oral cavity tumours including the lip; and Chapter 194, Tumours of the larynx). The endoscopic access to squamous tumours of the upper aerodigestive tract combined with the tendency to develop ‘field cancerization’ make these tumours good candidates for PDT. Gluckman9 treated a mixture of carcinoma in situ, early and advanced squamous cell carcinoma of the head and neck. Dihaematoporphyrin ether and light at 630 nm were used to treat these patients. The best results were obtained in the oral cavity and oropharynx where 11 out of 13 patients had a complete response and two out of 13 a partial response. Four of these tumours had recurrence within one year of treatment. Eight patients with advanced tumours were treated palliatively, but the results were no better that with standard regimes of the time. [**] Biel10 treated 25 patients with early squamous cell tumours of the larynx, obtaining a complete response in patients despite the fact that radiation therapy had previously been unsuccessful in 17 of them. [**] Biel11 also reported a summary of the results of a collection of studies on the treatment of early squamous cell carcinoma of the head and neck with PDT. There were a total of 217 patients. One hundred and ninety-four (89 percent) showed a complete response to treatment, 23 (10.6 percent) showed a partial response and no patient failed to show any response to the therapy. [**] Dilkes et al.12 had considerable experience with the use of PDT using the sensitizers m-TPHC, photofrin II and ALA for squamous carcinoma at multiple sites in the head and neck during the mid-1990s. The patients they treated included palliative, primary and adjunctive forms of treatment and, in nearly all cases, the authors identified a visible response to the effects of PDT. Their complications included local pain at the photosensitizer injection site and, in some patients where appropriate precautions had not occurred, post-treatment skin photosensitivity. [**] Lofgren et al.13 used PDT in five patients with circumscribed nasopharyngeal carcinoma. The drug was activated with laser light under topical anaesthesia and after four years, three of the patients has no evidence of disease. [**]



KEY POINTS  Light amplification by stimulated emission of radiation.  Laser light – collimated, coherent and monochromatic.  Laser tissue action – photoablative, photochemical and photomechanical.



Chapter 58 Laser principles in otolaryngology, head and neck surgery



 Laser ‘star wars’ beam of micromanipulator – hosepipe type jet of fibreoptic laser delivery.  Read and follow the safety regulations for your laser and operating room.  PDT requires administered drug, light and oxygen.  Skin phototoxicity and limited tissue penetration are the current PDT shortcomings.  PDT has a future place in head and neck cancer management.



Deficiencies in current knowledge and areas for future research All of the papers on photodynamic therapy for head and neck cancer suggest that photodynamic therapy has a significant place in the future management of this disease. The problem is in determining the exact type of photodynamic therapy associated with which photosensitizer and whether primary therapy, adjunctive therapy or palliation is the optimum course.



REFERENCES 1. Schawlow AL, Townes CH. Infrared and optical lasers. Physical Review. 1958; 112: 1940–9. 2. Maiman TH. Stimulated optical radiation in ruby. Nature. 1960; 187: 493–4. 3. Javan A, Bennett WR, Harriott DR. Population inversion and continuous laser oscillation in a gas discharge containing He-Ne mixtures. Physical Review Letters. 1961; 6: 106.



] 747



4. Jako GJ. Laser surgery of the vocal cords. An experimental study with carbon dioxide lasers on dogs. Laryngoscope. 1972; 82: 2204–16. 5. Strong MS, Jako GJ. Laser surgery in the larynx. Early clinical experience with continuous CO2 laser. Annals of Otology, Rhinology and Laryngology. 1972; 81: 791–8. 6. Strong MS, Jako GJ, Vaughan CW, Healy GB, Polanyi T. The use of C02 laser in otolaryngology: a progress report. Transactions. Section on Otolaryngology. American Academy of Ophthalmology and Otolaryngology. 1976; 82: 595–602. 7. Lightdale CJ, Heier SK, Marcon NE, McCaughan Jr. JS, Gerdes H, Overholt BF et al. Photodynamic therapy with porfimer sodium versus thermal ablation therapy with Nd:YAG laser for palliation of esophageal cancer: a multicenter trial. Gastrointestinal Endoscopy. 1995; 42: 507–12. 8. Overholt BF, Panjehpour M. Photodynamic therapy for Barrett’s oesophagus. Gastrointestinal Endoscopy Clinics of North America. 1997; 2: 207–20. 9. Gluckman JL. Hematoporphyrin photodynamic therapy: is there truly a future in head and neck oncology? Reflections on a 5 year experience. Laryngoscope. 1991; 101: 36–42. 10. Biel MA. Photodynamic therapy and the treatment of neoplastic diseases of the larynx. Laryngoscope. 1994; 104: 399–403. 11. Biel MA. Photodynamic therapy and the treatment of head and neck neoplasia. Laryngoscope. 1998; 108: 1259–68. 12. Dilkes MG, Alusi G, Djaezeri BJ. Treatment of head and neck cancer with PDT: Clinical experience. Review of Contemporary Pharmacotherapy. 1999; 10: 47–57. 13. Lofgren LA, Hallgren S, Nilsson E, Westerborn A, Nilsson C, Reizenstein J. Photodynamic therapy for recurrent nasopharyngeal cancer. Archives of Otolaryngology – Head and Neck Surgery. 1995; 121: 997–1002.



59 Electrophysiology and monitoring PATRICK R AXON AND DAVID M BAGULEY



Development of facial nerve monitoring Technique for continuous facial nerve monitoring Predicting postoperative facial function Difficulties of monitoring facial function Monitoring facial function for nonotological procedures Monitoring auditory function Techniques



748 749 749 750 750 750 751



Evidence of the efficacy of monitoring auditory function Key points Best clinical practice Deficiencies in current knowledge and areas of future research References



751 752 752 752 752



SEARCH STRATEGY The information presented in this chapter is supported by a Medline search using the key words intraoperative monitoring, facial nerve, cranial nerve and auditory evoked potentials.



DEVELOPMENT OF FACIAL NERVE MONITORING The first description of intraoperative cranial nerve stimulation was by Fedor Krause in 1898.1 During cochlear nerve section for tinnitus, he noted ‘y unipolar faradic stimulation of the (facial) nerve-trunk with the weakest possible current of the induction apparatus resulted in contractions of the right facial region, especially of the orbicularis oculi, as well as the branches supplying the nose and mouth’. Over the next threequarters of a century, a series of articles refined the technique, all relying on observing movement of the face in order to confirm the functional integrity of the facial nerve.2, 3, 4, 5, 6, 7 The evoked facial twitch was observed either by the anaesthetist or ancillary staff under the drapes using a flashlight or mirror. These techniques lacked both quantitative control of the stimulus and objective recording of the evoked responses. In 1979, Delgado and colleagues first described the use of evoked compound muscle action potentials (CMAP) to monitor facial nerve function in response to stimulating



the intracranial portion of the facial nerve.8 The introduction of facial electromyography (EMG) enabled not only facial nerve identification either by electrical stimulation or inadvertent manipulation, but also the possibility of mapping its course through the temporal bone and assessing changes in function during surgical resection of tumour from the nerve’s surface.9 Facial nerve monitoring has proved an invaluable aid during vestibular schwannoma surgery.10, 11, 12 The introduction of an auditory signal enabled instantaneous real-time auditory feedback to the surgeon during tumour dissection without information passing through an intermediary.13, 14 Three trials tested the hypothesis that facial nerve outcome improved when using intraoperative facial nerve monitoring. Harner and colleagues12 demonstrated the usefulness of facial nerve monitoring in 91 consecutive cases of vestibular Schwannoma resection via the suboccipital route. At one year, 78 percent of those patients who were monitored demonstrated facial function, compared with 65 percent in an unmonitored group, these data were not studied statistically. Niparko and colleagues15 described the results of 29 patients who



Chapter 59 Electrophysiology and monitoring



underwent translabyrinthine removal of vestibular Schwannoma and compared them with a similar group of 75 unmonitored patients. They demonstrated that monitoring was associated with a significant improvement of facial function at one year for tumours over 2 cm intracranial diameter. Kwartler and colleagues16 demonstrated that monitored patients with tumours over 2.5 cm had a significant improvement of facial function when compared with a matched unmonitored group. [**] The benefit of facial nerve monitoring during surgery for chronic middle ear disease is less certain. Facial nerve injury after otological surgery is rare in experienced hands and there are no randomized controlled trials examining its efficacy. Silverstein and others recommend that the facial nerve should be monitored during all general anaesthetic cases where the facial nerve is at risk.17, 18 This view is in contrast to most American otolaryngologists, although those that trained in the 1990s, those in an academic setting and those who perform more otology than other types of surgery are more likely to use monitoring techniques.19, 20 Facial nerve injury often occurs when there is an unexpected change to normal middle ear anatomy, precisely the time when monitoring is so valuable. The senior surgeon must take responsibility to ensure that the equipment is functioning normally. A simple technique for confirming equipment function is described under Technique for continuous facial nerve monitoring. Intraoperative facial nerve monitoring is no substitute for experience in the otological setting and should not replace good surgical practice, but if the operating team adopt the approach that all patients are monitored, the set up technique becomes routine and more reliable.



TECHNIQUE FOR CONTINUOUS FACIAL NERVE MONITORING The operating theatre is filled with electrical interference generated by the equipment surrounding the anaesthetized patient. Monitoring techniques have developed to minimize this interference and amplify only relevant information. Two sets of subdermal platinum or stainless steel needle recording electrodes are inserted into the upper and lower face. The amplifiers amplify the difference between the potentials recorded at each electrode. This arrangement has the advantage of common mode rejection; electrical interference from other sources is recorded by both electrodes equally and therefore does not create a potential difference between the two closely aligned electrodes. A number of commercial EMG cranial nerve monitoring systems are now available including the NIM-2 (Xomed Treace, Jacksonville, FL, USA) and the Neurosign (Magstim, Whitland, UK). They rely on recording facial muscle activity and delivering the information as a visual and audible representation of the CMAP response. The audible response is either presented as raw EMG activity or a



] 749



characteristic sound when EMG activity reaches a set threshold. All systems are isolated and self-contained electrical nerve stimulator and monitoring units. The electrodes are connected to a preamplifier pod, which is attached to the operating table. The recorded electrical signal is filtered through high- and low-pass filters and either rectified and displayed on a logarithmic bar chart or presented as a CMAP waveform. Different systems use different methods of presenting the same information to the surgeon. The logarithmic bar chart has a delayed response decay to enable calculation of rectified CMAP amplitude. Systems that present a CMAP waveform present it as visual and audible real-time information or utilize image capture strategies that also give waveform amplitude information. This allows the surgeon time to examine the waveform and size of the CMAP. Familiarity with the set up and function of a chosen monitoring system is essential. The senior surgeon must take responsibility and should check that the equipment is functioning normally. Tapping the skin overlying the two sets of subdermal electrodes will generate a recorded response on the monitor. This confirms that electrodes are connected to the preamplifier pod and in turn the preamplifier pod is connected to the monitor, which is switched on. The volume should be checked so that a response is audible over background theatre noise. Facial nerve stimulation is delivered as a short (0.1 ms) electrical pulse. The stimulating electrode is either monopolar or bipolar. The monopolar electrode is favoured because it is simple to use, but has the disadvantage of stimulating a larger area. The bipolar electrode requires careful positioning of both electrode tips on to the tissue surface; this can prove difficult in the tight confines of the temporal bone. The use of constant voltage stimuli has an advantage over constant current stimuli because it delivers a relatively reliable current to the nerve whatever the medium that surrounds the nerve.



PREDICTING POSTOPERATIVE FACIAL FUNCTION Recently, a number of studies have described an objective technique that correlates parameters of the evoked CMAP to eventual facial outcome.15, 21, 22, 23, 24 The test gives nondichotomous results and therefore a retrospective cut-off point is used to predict those patients who have a good prognosis. Results indicate that a low stimulation threshold, across the site of tumour dissection, is a valuable prognostic indicator of good long-term facial function. The technique, which is simple to perform, assesses the minimum current required to evoke a muscle response after tumour resection. The drawback to the described technique is that the majority of patients have some degree of facial function immediately after surgery. This group will almost certainly have good long-term outcome.25 It is



750 ] PART 11 RECENT ADVANCES IN TECHNOLOGY the small group of patients with poor facial function immediately after surgery that will benefit most from a sensitive and specific predictive test. Axon and Ramsden compared post-dissection minimal stimulation thresholds with immediate postoperative facial function for predicting long-term facial function in 184 patients undergoing vestibular Schwannoma surgery.25 Post-dissection stimulation thresholds demonstrated only a moderate relation to eventual outcome, which was of limited clinical value. The test criteria were then applied to patients with poor immediate postoperative facial function for predicting long-term outcome, the predictive accuracy fell, further reducing test validity. [**] Some studies compared the supramaximal CMAP to either facial nerve stimulation proximal and distal to tumour dissection or before and after tumour dissection. These techniques have been advocated as more accurate methods of analysing data, because they remove absolute amplitude comparisons and rely on comparison of ratios.26, 27 [**]



DIFFICULTIES OF MONITORING FACIAL FUNCTION All otological procedures rely on a facial muscle response, warning the surgeon that the facial nerve is near. A simple audible noise is all that is required. Recent monitoring systems have increased the amount of information available to the surgeon, stimulating the desire to expand monitoring techniques and so improve patient outcome. This information is superfluous to most procedures and of benefit to only a few. Facial CMAPs represent a complicated interplay between groups of muscle fibres depolarizing in response to facial nerve stimulation. The muscles of the face are very different to those found in the limbs. The facial motor units are small, often having only 25 muscle fibres supplied by each motoneuron compared with many thousands in more peripheral muscles. As a consequence, each muscle has a wide ill-defined motor end-plate zone. The muscles are also arranged in an almost haphazard arrangement, overlying each other and aligned in different directions. This makes meaningful electrophysiological recording difficult. Intrasubject variability (test–retest variability) is high and intersubject comparison almost impossible. The CMAP waveform is usually multiphasic instead of the wellrecognized biphasic responses recorded from peripheral muscles, a consequence of phase cancellation. Calculation of maximum amplitude or area under the waveform bears little relation to the number of motoneurons innervating the muscle fibres that create the response.



MONITORING FACIAL FUNCTION FOR NONOTOLOGICAL PROCEDURES The facial nerve is at risk of iatrogenic injury in the cerebellopontine angle and parotid. Intraoperative facial



nerve monitoring has been advocated during microvascular decompression and superficial parotidectomy.28, 29 Arguments for adopting its use for all surgical procedures are the same as those for otological surgery.



MONITORING AUDITORY FUNCTION The aim of monitoring the status of the auditory pathway during cerebello-pontine angle (CPA) surgery is the prevention of avoidable postoperative hearing deficit. The achievement of this laudable aim is fraught with difficulty. The cochlear nerve is sensitive to mechanical manipulation and easily damaged, as the intracranial section of the nerve is sheathed in central myelin and has no perineurium.30 Additionally, the cochlear nerve is intimately involved with pathologies, such as vestibular schwannoma, and hence at very considerable risk during the surgical removal of such lesions, even when every care is taken to preserve the nerve anatomically. The basic principle of intraoperative monitoring is that changes in recordable neuroelectric potentials occur whilst the injury is still reversible and before permanent deficits result.31 Recent research32 has demonstrated that this principle holds for changes in auditory brainstem responses (ABR) wave V amplitude (and to a lesser extent latency) in rat auditory nerves manipulated in a fashion analogous to that undergone in humans during vestibular schwannoma removal. It is therefore theoretically feasible that monitoring auditory function may inform the surgeon of reversible injury to the cochlear nerve. One prerequisite of monitoring auditory function during surgery is that sufficient preoperative hearing must remain such that meaningful recordings may be made in the operating theatre and change observed if and when it occurs. If ABR or electrocochleography (EcochG) are the monitoring techniques of choice, a further prerequisite is for preoperative recordings with that technique. Monitoring with ABR is not possible if waveforms are absent or of grossly abnormal morphology. Whilst the techniques and equipment utilized for auditory monitoring may seem familiar from the outpatient clinic, their use in monitoring is significantly different. Rather than comparing an individual’s data with group data, in monitoring one is comparing a sequence of waveforms over time and evaluating for change. The time allowed for eliciting each recording is limited and this will influence the choice of recording parameters. The decision whether to inform the surgeon of a change in recorded activity is both urgent and crucial. As such, the person undertaking monitoring should be familiar with the operating theatre environment and team, and be able to interpret and communicate observed and measured change to the surgeon confidently and concisely.33 Such skills are rare and this may prove to be a hindrance to the widespread adoption of these techniques.



Chapter 59 Electrophysiology and monitoring



TECHNIQUES Otoacoustic emissions Cane et al.34 undertook a feasibility study of the use of otoacoustic emissions (OAE) as a technique for monitoring the auditory pathway during surgical removal of a vestibular schwannoma. The experience was that OAE recordings were feasible despite the noise within the surgical environment, but the study did not indicate that OAE would be a useful indicator of early, reversible injury. Theoretically, OAE recording may furnish evidence of cochlear function, but as the site of surgery is the internal auditory canal (IAC), one would not be measuring activity deriving from the point of possible injury. OAE monitoring might therefore give information about ischaemic or noise injury to the cochlea, but it has not yet been demonstrated that this would be at a point where the injury was reversible. At the present time, however, the use of OAE in monitoring auditory function during surgery is rare.



Electrocochleography Similarly, electrocochleography records activity associated with cochlear function and the distal portion of the cochlear nerve, rather than the intracranial portion of the cochlear nerve, which is acutely at risk during surgery. As such, even strong advocates of EcochG in monitoring have seen it as adjunct to ABR, and not providing the surgeon with the required information when used alone. There is research evidence that responses to auditory stimulation recorded by EcochG persist after complete transection of the VIIth nerve in the rat,35 and so this technique has the potential to mislead if used in isolation. One potential advantage of the use of EcochG is the large amplitude of the activity recorded.36 This allows interpretation of the elicited waveform after fewer averages, so that decision-making may be more rapid.37 As with ECochG in the outpatient context, there is debate about the optimal site of the recording electrode, with advocates for extratympanic intrameatal placement36 and transtympanic placement.38 A further advantage is that EcochG recordings are less compromised by acoustic noise or by electrical artifact as may occur during cautery, and may serve some purpose whilst these are present.39



Auditory brainstem responses Auditory brainstem responses have a theoretical advantage over other techniques in that certain peaks within the waveform derive from anatomical structures at risk of injury during surgery. The generator of wave I is considered to be the distal portion of the cochlear nerve, the site also held to be the generator of the N1 component



] 751



of the compound action potential recorded by EcochG.30 Wave II is associated with the proximal section of the cochlear nerve,31 wave III with the lower pons (specifically the superior olivary complex)40 and wave V with the inferior colliculus.40 The generators of waves III and V are, therefore, unlikely to be directly challenged by surgical manipulation during hearing preservation surgery for space occupying lesions of the CPA, but the latencies of these easily identifiable waves are used during monitoring as change reflects latency change of waves I and II. Specific methodologies and parameters utilized in intraoperative ABR have been described in detail. The reader who intends to perform this procedure is directed to the comprehensive reviews by Moller31 and Martin and Mishler.30 Several points should be noted. There is considerable variability in the methods advocated and clinical trial evidence is not apparent. Second, this variability may account in part for the lack of a strong evidence base for the benefits of auditory monitoring. Further, the reviews cited underline the difference between utilizing ABR in the outpatient clinic and the operating theatre, the need for the technique not to inconvenience surgical techniques and the utility of clear and timely interpretation and communication of change in recordings to the surgeon.



Direct recordings from the cochlear nerve The need for rapid acquisition of responses may be met by recording directly from the cochlear nerve, the large amplitude of the elicited activity obviating the need for lengthy averaging.41 This technique was first developed to monitor cochlear nerve function during microvascular decompression surgery, but has been adapted for use during surgical removal of space occupying lesions of the CPA.37 The site of recording is usually the root entry zone, and both monopolar and bipolar electrodes have their advocates.42 Click stimuli similar to those used in ABR are used to elicit the response, which comprises a compound action potential. Changes in latency, amplitude and morphology should all be reported to the surgical team.31



Combining techniques Given that each of the techniques described above has technical challenges and shortcomings, many teams choose to utilize more than one, the most common combination being direct cochlear nerve recording and ABR.43



EVIDENCE OF THE EFFICACY OF MONITORING AUDITORY FUNCTION Authors have reported hearing preservation rates following vestibular schwannoma excision, but the heterogeneity



752 ] PART 11 RECENT ADVANCES IN TECHNOLOGY of pathological status, surgical technique, surgical experience and criteria for reporting of hearing preservation combine to frustrate analysis of the clinical utility of auditory monitoring. Observational evidence in this area abounds.44, 45, 46 The need for well-designed, hypothesisdriven clinical trials in this field is paramount. A similar dearth of evidence is found when determining the relative efficacy of techniques or combinations of techniques. Several studies have reported that direct cochlear nerve recordings are more effective in the maintenance of hearing function than ABR, to an extent that is statistically significant.47, 48 This effect is said to be due to the immediacy of direct recordings.49 [**] Despite the lack of evidence, a consensus is building that monitoring auditory function during CPA surgery is best practice,42 and that a combination of ABR and direct cochlear nerve recording is optimal,35 with the adjunctive use of EcochG if desired.38



$



REFERENCES



KEY POINTS  Monitoring facial nerve function is straightforward and neither hampers nor impedes surgery.  There is a good evidence that monitoring facial nerve function improves outcomes of facial function.  A consensus has not yet been reached on techniques for monitoring auditory function, but it looks likely that a combination of techniques will be optimal.



Best clinical practice [ Regarding monitoring of facial nerve function, there is evidence that these techniques offer benefit in improving facial nerve outcomes in surgery that may challenge the facial nerve. As such it is strongly indicated. [Grade B] [ The evidence base for monitoring auditory nerve function is less robust, but it should be regarded as best clinical practice [Grade D/C].



Deficiencies in current knowledge and areas of future research



$



Further work is required to develop an intraoperative technique that enables accurate assessment of clinical facial nerve function at any point during skull



base procedures. Only accurate estimation of motorneuron function will give the surgeon a true representation of immediate facial function and hopefully then enable development of a valid predictive technique. The speciality as a whole requires a good evidence base to support the use of intraoperative facial nerve monitoring as the standard for all otological procedures that place the facial nerve at risk.







1. Krause F (ed.). Surgery of the brain and spinal cord. New York: Rebman Company, 1912. 2. Frazier CH. Intracranial division of the auditory nerve for persistent aural vertigo. Surgery, Gynaecology and Obstetrics. 1912; 15: 524–9. 3. Olivecrona H. Acoustic tumors. Journal of Neurology, Neurosurgery, and Psychiatry. 1940; 3: 141–6. 4. Hullay J, Tomits GH. Experiences with total removal of tumours of the acoustic nerve. Journal of Neurosurgery. 1965; 22: 127–35. 5. Rand RW, Kurze TL. Facial nerve preservation by posterior fossa transmeatal microdissection in total removal of acoustic tumors. Journal of Neurology, Neurosurgery, and Psychiatry. 1965; 28: 311–6. 6. Poole JL. Suboccipital surgery for acoustic neurinomas: Advantages and disadvantages. Journal of Neurosurgery. 1966; 24: 483–92. 7. Albin MS, Babinski M, Maroon JC. Anesthetic management of posterior fossa surgery in the sitting position. Acta Anaesthesiologica Scandinavica. 1976; 20: 117–28. 8. Delgado TE, Buchheit WA, Rosenholtz HR. Intraoperative monitoring of facial muscle evoked responses obtained by intracranial stimulation of the facial nerve: a more accurate technique for facial nerve dissection. Neurosurgery. 1979; 4: 418–21. 9. Silverstein H, Willcox Jr. TO, Rosenberg SI, Seidman MD. Prediction of facial nerve function following acoustic neuroma resection using intraoperative facial nerve stimulation. Laryngoscope. 1994; 104: 539–44. 10. Moller AR, Janetta PJ. Preservation of facial function during removal of acoustic neuromas. Use of monopolar constant voltage stimulation and EMG. Journal of Neurosurgery. 1984; 61: 757–60. 11. Benecke JE, Calder HB, Chadwick G. Facial nerve monitoring during acoustic neuroma removal. Laryngoscope. 1987; 97: 697–700. 12. Harner SG, Daube JR, Beatty CW, Ebersold M. Intraoperative monitoring of the facial nerve. Laryngoscope. 1988; 98: 209–12. 13. Prass RL, Luders H. Acoustic (loudspeaker) facial electromyography (EMG) monitoring: I. Evoked



Chapter 59 Electrophysiology and monitoring



14.



15.



16.



17.



18. 19.



20.



21.



22.



23.



24.



 25. 26.



27.



28.



electromyographic (EMG) activity during acoustic neuroma resection. Neurosurgery. 1986; 19: 392–400. Prass RL, Kenney SE, Hardy RW, Hahn JF, Luders H. Acoustic (loudspeaker) facial EMG monitoring: II. Use of evoked EMG activity during acoustic neuroma resection. Otolaryngology and Head and Neck Surgery. 1987; 97: 541–51. Niparko JK, Kileny PR, Kemink JL. Neurophysiologic intraoperative monitoring: II. Facial nerve function. American Journal of Otology. 1989; 10: 55–61. Kwartler JA, Luxford WM, Atkins J, Shelton C. Facial nerve monitoring in acoustic tumor surgery. Otolaryngology and Head and Neck Surgery. 1991; 104: 814–7. Silverstein H, Smouha EE, Jones R. Routine intraoperative facial nerve monitoring during otologic surgery. American Journal of Otology. 1988; 9: 269–75. Noss RS, Lalwani AK, Yingling CD. Facial nerve monitoring in middle ear surgery. Laryngoscope. 2001; 111: 831–6. Greenberg JS, Manolidis S, Stewart MG, Kahn JB. Facial nerve monitoring in chronic ear surgery: US practice patterns. Otolaryngology and Head and Neck Surgery. 2002; 126: 108–14. Pensak ML, Willging JP, Keith RW. Intraoperative facial nerve monitoring in chronic ear surgery: A resident training program. American Journal of Otology. 1994; 15: 108–10. Selesnick SH, Carew JF, Victor JD, Heise CW, Levine J. Predictive value of facial nerve electrophysiologic stimulation thresholds in cerebellopontine-angle surgery. Laryngoscope. 1996; 106: 633–8. Prasad S, Hirsch BE, Kamerer DB, Durrant J, Sekhar LN. Facial nerve function following cerebellopontine angle surgery: Prognostic value of intraoperative thresholds. American Journal of Otology. 1993; 14: 330–3. Silverstein H, Willcox TO, Rosenberg SI, Seidman MD. Prediction of facial nerve function following acoustic neuroma resection using intraoperative facial nerve stimulation. Laryngoscope. 1994; 104: 539–44. Nissen AJ, Sikand A, Curto FS, Welsh JE, Gardi J. Value of intraoperative threshold stimulus in predicting postoperative facial nerve function after acoustic tumor resection. American Journal of Otology. 1997; 18: 249–51. Axon PR, Ramsden RT. Intraoperative EMG for predicting facial function in vestibular Schwannoma surgery. Laryngoscope. 1999; 109: 922–6. Axon PR, Ramsden RT. Assessment of real-time clinical facial function during vestibular Schwannoma surgery. Laryngoscope. 2000; 110: 1911–5. Goldbrunner RH, Schlake HP, Milewski C, Tonn JC, Helms J, Roosen K. Quantitative parameters of intraoperative electromyography predict facial nerve outcomes for vestibular schwannoma surgery. Neurosurgery. 2000; 46: 1140–6; discussion 1146–8. Mooj JJ, Mustafa MK, van Weerden TW. Hemifacial spasm: intraoperative electromyographic monitoring as a guide for microvascular decompression. Neurosurgery. 2001; 49: 1365–70; discussion 1370–1.



] 753



29. Lopez M, Ouer M, Leon X, Orus C, Recher K, Verges J. Usefulness of facial nerve monitoring during parotidectomy. Acta Otorrinolaringolo´gica Espan˜ola. 2001; 52: 418–21. 30. Martin WH, Mishler ET. Intraoperative monitoring of auditory evoked potentials and facial nerve electromyography. In: Katz J (ed.). Handbook of clinical audiology. Philadelphia: Lippincott Williams and Wilkins, 2001: 323–48. 31. Moller AR. Intraoperative neurophysiological monitoring. In: Roeser RJ, Valente M, Hosford-Dunn H (eds). Audiology diagnosis. New York: Thieme, 2000: 545–70. 32. Sekiya T, Shimamura N, Yagihashi A, Suzuki S. Axonal injury in auditory nerve observed in reversible latency changes of brainstem auditory evoked potentials (BAEP) during cerebellopontine angle manipulations in rats. Hearing Research. 2002; 173: 91–9. 33. Fisher RS, Raudzens P, Nunemacher M. Efficacy of intraoperative neurophysiological monitoring. Journal of Clinical Neurophysiology. 1995; 12: 97–109. 34. Cane MA, O’Donoghue GM, Lutman ME. The feasibility of using oto-acoustic emissions to monitor cochlear function during acoustic neuroma surgery. Scandinavian Audiology. 1992; 21: 173–6. 35. Rosahl SK, Tatagiba M, Gharabaghi A, Matthies C, Samii M. Acoustic evoked response following transection of the eighth nerve in the rat. Acta Neurochirurgica. 2000; 142: 1037–45. 36. Mullatti N, Coakham HB, Maw AR, Butler SR, Morgan MH. Intraoperative monitoring during surgery for acoustic neuroma: benefits of an extratympanic intrameatal electrode. Journal of Neurology, Neurosurgery, and Psychiatry. 1999; 66: 591–9. 37. Yingling CD. Intraoperative monitoring of cranial nerves in neurotologic surgery. In: Cummings CW, Fredickson JM, Harker LA, Krause CJ, Schuller DE, Richardson MA (eds). Otolaryngology head and neck surgery, 3rd edn. St Louis: Mosby, 1998: 3331–55. 38. Schlake HP, Goldbrunner R, Milewski C, Siebert M, Behr R, Riemann R et al. Technical developments in intraoperative monitoring for the preservation of cranial motor nerves and hearing in skull base surgery. Neurological Research. 1999; 21: 11–24. 39. Schlake HP, Milewski C, Goldbrunner RH, Kindgen A, Riemann R, Helms J et al. Combined intra-operative monitoring of hearing by means of auditory brainstem responses (ABR) and transtympanic electrocochleography (EcochG) during surgery of intra and extrameatal acoustic neurinomas. Acta Neurochirurgica. 2001; 143: 985–95. 40. Legatt AD. Mechanisms of intraoperative brainstem auditory evoked potential changes. Journal of Clinical Neurophysiology. 2002; 19: 396–408. 41. Moller AR. Intraoperative neurophysiologic monitoring. In: Brackmann DE, Shelton C, Arriaga MA (eds). Otologic surgery. Philadelphia: WB Saunders, 2001: 645–61. 42. Nguyen BH, Javel E, Levine SC. Physiologic identification of eighth nerve subdivisions: Direct recordings with



754 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



 



bipolar and monopolar electrodes. American Journal of Otology. 1999; 20: 522–34. 43. Mann WJ, Maurer J, Marangos N. Neural conservation in skull base surgery. Otolaryngologic Clinics of North America. 2002; 35: 411–24. 44. Radtke RA, Erwin CW, Wilkins RH. Intraoperative brainstem auditory evoked potentials: Significant decrease in postoperative morbidity. Neurology. 1989; 39: 187–91. 45. Harper CM, Harner SG, Slavit DH, Litchy WJ, Daube JR, Beatty CW et al. Effect of BAEP monitoring on hearing preservation during acoustic neuroma resection. Neurology. 1992; 42: 1551–3.







46. Fischer G, Fischer C, Remond J. Hearing preservation in acoustic neurinoma surgery. Journal of Neurosurgery. 1992; 76: 910–7. 47. Colletti V, Fiorino FG, Mocella S, Policante Z. EcochG, CNAP and ABR monitoring during vestibular schwannoma surgery. Audiology. 1998; 37: 27–37. 48. Jackson LE, Roberson Jr. JB. Acoustic neuroma surgery: Use of cochlear nerve action potential monitoring for hearing preservation. American Journal of Otology. 2000; 21: 249–59. 49. Colletti V, Fiorino FG. Advances in monitoring of seventh and eighth cranial nerve function during posterior fossa surgery. American Journal of Otology. 1998; 19: 503–12.



60 Optical coherence tomography MARIAH HAHN AND BRETT E BOUMA



Introduction Optical coherence tomography: System operation OCT imaging of the larynx: Preliminary results Polarization-sensitive optical coherence tomography Summary



755 756 757 757 759



Key points Deficiencies in current knowledge and areas for future research Acknowledgements References



759 760 760 760



SEARCH STRATEGY AND EVIDENCE-BASE The data in this chapter are supported by a PubMed search using the key words optical coherence tomography, needle biopsy, surgical guidance, nasal, vocal cord and/or otology. Levels of evidence are not really applicable to this area, but the human studies that underlie the discussion of optical coherence tomography as applied to otorhinolaryngology are observational or at best non-randomized.



INTRODUCTION Medical imaging technology has advanced rapidly over the past twenty years, providing physicians with essential information on the macroscopic anatomy of patients. Techniques capable of imaging subepithelial structure in situ, such as conventional x-ray radiography, magnetic resonance imaging (MRI), computed tomography (CT) and ultrasonography have allowed the noninvasive investigation of relatively large-scale structures in the human body, with resolutions ranging from 100 mm to 1 mm. Resolution on this scale, however, is insufficient to detect the subtle changes in tissue microstructure characteristic of many ear, nose and throat (ENT) pathologies. In the field of laryngology, light endoscopy currently forms the cornerstone of clinical imaging and biopsy guidance, yet conventional light endoscopic techniques are unable to reveal information concerning subepithelial tissue. Since many laryngeal pathologies originate near the boundary between the epithelium and the underlying mucosa or within the mucosa itself, the inability to image subepithelial tissue represents a serious limitation of



conventional light endoscopy. Even as a method for guiding biopsy, conventional light endoscopy gives only a relatively coarse indication of prospective biopsy locations. Frequently, a large number of biopsies are required to achieve high diagnostic accuracy, a situation that is often undesirable or unfeasible. This is particularly true for the vocal folds, which have a specialized and delicate microstructure that is highly intolerant of trauma. These limitations restrict the effectiveness of light endoscopy for the diagnosis, monitoring and treatment of many laryngeal pathologies. Clinical laryngology would thus greatly benefit from a noninvasive imaging technology capable of resolving subepithelial tissue microstructure in the range of conventional biopsy. Optical coherence tomography (OCT) is a relatively new optical imaging modality that allows high-resolution, crosssectional imaging of tissue microstructure. OCT can image with an axial resolution of 1–15 mm and has an imaging depth of 2–3 mm in nontransparent tissue. OCT was first applied in 1991 to imaging optically transparent structures, such as the anterior eye and retina.1, 2 Subsequent technological advances have enabled high-resolution



756 ] PART 11 RECENT ADVANCES IN TECHNOLOGY imaging of nontransparent tissue in the cardiovascular system3 and the gastrointestinal,4 urinary5 and female reproductive tracts.6 In addition, the application of OCT to surgical guidance7 and carcinoma detection6, 8 has been explored. The aforementioned imaging studies have demonstrated that OCT is particularly informative in tissues in which nonkeratinized epithelium is separated from underlying stroma by a smooth basement membrane zone, suggesting that OCT may have a strong clinical relevance in laryngology. This indication is further strengthened by several features of the OCT system itself.  OCT imaging can be performed in situ and nondestructively, enabling the imaging of tissue for which biopsy should be avoided or is impossible.  OCT images are high resolution, 10–100 times that of conventional MRI or ultrasound.  Imaging can be performed in real time, without the need to process a specimen, as in conventional biopsy, and without the need for a transducing medium, as in ultrasound imaging.  OCT is fibreoptically based and can thus be interfaced to a wide range of instruments including catheters, endoscopes, laparoscopes and surgical probes.  OCT systems can be engineered to be compact, portable and low cost, depending on the desired system specifications. Although this chapter will focus primarily on the potential applications of OCT to laryngeal tissue, OCT has also been applied to imaging of the middle ear9 and of rat cochlea10 ex vivo with promising results. In addition, OCT has strong potential for improved visualization of nasal tissues.11 Before presenting and discussing OCT images of larynges ex vivo, it is useful to briefly describe the basis by which OCT systems form subepithelial images.



to conventional OCT, polarization-sensitive OCT (PSOCT), exploits birefringence as an additional contrast mechanism for imaging tissue. PS-OCT will be discussed in further detail under Polarization-sensitive optical coherence tomography. The analogy with ultrasound is a useful starting point for understanding the basics of OCT. In ultrasound, a high frequency acoustic pulse travels into the tissue and is reflected or backscattered from internal structures having different acoustic properties. The magnitude and the delay time of the echoes are electronically detected, and the structural properties of the internal tissues are determined from the measured signals. In OCT, imaging is performed by measuring the echo delay time and magnitude of light backreflected or backscattered from internal structures with distinct optical properties. Unlike in ultrasound, though, the speed of light is very high, rendering electronic measurement of the echo delay time of the reflected light impossible. OCT systems circumvent this limitation by using low coherence interferometry, also known as white light interferometry, to characterize optical echoes. The most common OCT echo detection scheme is based on a Michelson interferometer set up with a scanning reference delay arm, shown in Figure 60.1. Within the interferometer, the beam leaving the optical light source is split into two parts, termed the reference and sample beams, at the beam splitter. The reference beam then travels to a mirror, located at a known distance from the detector and subsequently returns to the beam splitter. The sample beam travels to the tissue sample and is reflected back toward the detector by scattering sites within the tissue. Light reflecting from deeper tissue layers has traversed a greater optical pathlength (optical distance) and therefore arrives at the detector at a later time. In addition, the various backreflected parts of the sample beam will have different amplitudes based on the Reference mirror



OPTICAL COHERENCE TOMOGRAPHY: SYSTEM OPERATION Optical coherence tomography performs high-resolution tomographic imaging by measuring light backscattered or backreflected from internal tissue structures. OCT imaging is analogous to ultrasound B-mode imaging, but is based on the detection of infrared light waves, instead of sound. Like acoustic waves, light is characterized by its propagation direction. Light is distinct, however, in that it has an additional vector characteristic known as polarization. The polarization direction is orthogonal to the propagation direction and can be influenced by the medium in which the light propagates. This is known as ‘birefringence’. Polarization measurements can be used to provide additional insights into the microscopic structure and integrity of tissues. An adjunct



Light source Beam splitter Sample



Detector



Figure 60.1 Schematic of a optical coherence tomography system using a Michelson interferometer with scanning reference delay arm.



Chapter 60 Optical coherence tomography



] 757



differing strengths of the scattering sites. The reflected portions of the sample beam then return to the beam splitter where they interact with the reflected reference beam and are directed toward the detector. When two waves from the same source recombine in this manner, an interference pattern will result if the optical pathlength travelled by the two waves is identical. The output detector measures the intensity of this interference pattern and thus only measures information contained in the portion of the sample beam that has travelled the same optical distance as the light in the reference arm. Information about the remaining tissue layers can be extracted by moving the mirror in the reference arm, which changes the optical pathlength travelled by the reference beam. Two- or three-dimensional images are produced by scanning the beam across the sample and recording the optical backscattering versus depth at different transverse positions. The resulting data is a two- or three-dimensional representation of the optical backscattering of the sample on a micron scale.



OCT IMAGING OF THE LARYNX: PRELIMINARY RESULTS To assess the potential value of OCT imaging technology in improving visualization and pathology, diagnosis and treatment of a given organ system, it is essential to first image normal tissue from that particular organ and then reference the resulting OCT images to parallel histology. Initial studies of the laryngeal imaging using OCT12, 13 have focused on the following structures due to clinical relevance:  the laryngeal surface of the epiglottis;  the false folds;  the inferior surface of the true folds. The images presented are based on a similar focus. It should be mentioned, however, that since OCT can be readily adapted to fibreoptic-based probes, other areas of traditionally poor visibility in light endoscopy, such as the anterior commissure, the subglottis and the laryngeal ventricle, could also be readily imaged using this technology. In Figure 60.2, an OCT image of the laryngeal surface of an ovine (sheep) epiglottis and the associated histological section are shown. Several different layers of the epiglottic tissue are visible, with epithelium, lamina propria and cartilage being clearly demarcated. Glandular structures and vessels are also often visible. Figure 60.3 presents an OCT image taken along the inferior portion of a porcine true fold, as evidenced by the transition from free-edge nonglandular mucosa to subglottal glandular mucosa observed in the OCT image and corresponding histological section. This image illustrates the ability of OCT to allow visualization of the epithelium, basement membrane and lamina propria of



Figure 60.2 (a) OCT image of an ovine epiglottis; (b) corresponding haematoxylin and eosin histology section. Scale bars represent 150 mm.



the true fold, as well as of certain structural features, such as glands and vessels. Note in particular that local epithelial thickness and transparency of the true fold can be readily deduced from these images, implying that OCT may prove useful for the diagnosis of hyperplasia, early-stage keratosis and papillomas, and other pathologies resulting in epithelial abnormalities not normally visible by conventional light endoscopy until an advanced stage. Since many laryngeal pathologies originate at the border between the mucosa and the epithelium, the ability to observe the integrity of the basement membrane via OCT imaging may have important implications for the use of this technology in aiding the diagnosis and monitoring of such laryngeal disorders, possibly including early stage carcinoma.



POLARIZATION-SENSITIVE OPTICAL COHERENCE TOMOGRAPHY A monochromatic light wave carries three potential sources of information contained in the amplitude, phase and polarization state of its electrical field vector. Conventional OCT, by detecting only the intensity of the interference pattern resulting from interaction of the sample and reference optical beams, fails to quantify the effects of tissue on the polarization state of the sample light beam. An adjunct to conventional OCT, known as polarization-sensitive OCT, therefore has focussed on



758 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



Figure 60.3 (a) OCT image of the inferior surface of a porcine true fold; (b) corresponding haematoxylin and eosin histology section; (c) phase-contrast viewing of the stained histology section. Scale bars represent 150 mm.



exploiting the information contained in the polarization state of the light backreflected or backscattered from the sample relative to that of the reference beam to obtain information about the tissue microstructure. In order to appreciate the power of PS-OCT and its potential applications, it is helpful to describe briefly the main mechanisms by which tissue alters the polarization state of light. Two mechanisms dominate the changes in the polarization state of light propagating through biological tissue: scattering and birefringence. Scattering changes the polarization of light mainly in a random manner, so it is generally not useful in mapping tissue structure. Organized linear structures, on the other hand, such as collagen fibre bundles with a clear orientation, can exhibit birefringence, resulting in a predictable change in polarization state. Many biological tissues exhibit birefringence, such as tendons, muscle, nerve,



bone, cartilage and teeth.14 Since changes in birefringence may, for instance, indicate changes in functionality, structure, or viability of tissues, PS-OCT images may contain information which can be used to evaluate tissue microstructure and health beyond the level possible with the use of OCT alone.15 The potential implications for clinical laryngology of the sensitivity of PS-OCT to organized linear structures, such as oriented collagen fibre bundles, are numerous. To illustrate, the human vocal fold lamina propria is generally subdivided into the superficial, intermediate and deep layers, with each layer having a characteristic composition and functionality. The relative thickness and integrity of these various layers appear to have a profound impact on voice. This idea is supported by the fact that the average adult will experience marked thinning of the superficial lamina propria relative to the intermediate and deep layers with increasing age,16 with a concomitant reduction in voice quality. Histological studies of normal adult human vocal fold tissue have repeatedly noted increased density and longitudinal orientation of collagen fibres as one progresses from more superficial to deeper lamina propria layers.16 PS-OCT, due to its sensitivity to oriented collagen fibre bundles, may thus allow indirect detection of these lamina propria layers in situ and hence monitoring and possibly improving treatment of voice ageing. In the case of vocal fold scarring, the biomechanical integrity of the lamina propria is disrupted, resulting in reduced vocal fold pliability and thus decreased vocal range and quality. Scar tissue has collagen fibre orientation and density that is generally distinct from that of normal tissue and therefore PS-OCT may also prove useful in visualizing, localizing and treating vocal fold scars within the lamina propria. These capabilities could have a profound impact in the area of voice management. Figure 60.4 compares a PS-OCT image of the inferior surface of a human true fold with the corresponding OCT image and histological section. Note the banding pattern apparent in the PS-OCT image of the vocal fold. A transition from white to black banding indicates a change in the polarization state of light. In particular, observe the marked increase in the thickness of the initial subepithelial white band in the leftmost portion of the PS-OCT image (corresponding to the true fold free edge) relative to that in the more subglottic tissue. The spatial location of this thickened initial band corresponds closely with the increased prominence of the superficial lamina propria in this region. Since collagen in normal superficial lamina propria is sparse and relatively randomly oriented, a net change in the polarization state of light will be slow to occur in this region. Hence, the thickness of the superficial lamina propria can potentially be inferred from the thickness of the initial subepithelial PS-OCT band. The true fold superficial lamina propria features critically in vocal health and disease. Although the presented results are not definitive, the potential of PS-OCT to detect



Chapter 60 Optical coherence tomography



] 759



by this imaging adjunct, PS-OCT could prove very useful to the study of lamina propria health and in the development of targeted scar treatment strategies.



SUMMARY OCT and OCT adjuncts have enormous potential for improving the diagnosis, monitoring and both surgical and nonsurgical management of diseased ENT tissues. Conventional OCT provides objective information concerning the structure of tissue elements with distinct optical backscattering and backreflection properties. With regard to the larynx, conventional OCT studies of healthy laryngeal tissues ex vivo have demonstrated that OCT can detect spatial changes in the thickness and transparency of the epithelium, the content of the connective tissues in terms of glands and vessels as well as pronounced transitions in connective tissue type and the architecture of the basement membrane.17 PS-OCT holds promise for permitting indirect visualization of the layers of the true fold lamina propria, as well as of vocal fold scar beds. In addition, a number of other OCT adjuncts, such as spectroscopic OCT18 and phase dispersion tomography19 exist or are currently in development, each with distinct imaging advantages and capabilities over conventional OCT imaging. For ENT tissue visualization, the OCT imaging modality represents a rich and vibrant area of future research and discovery.



KEY POINTS



Figure 60.4 (a) Haematoxylin and eosin histology section of the inferior surface of a human true fold; (b) corresponding OCT image; (c) corresponding PS-OCT image. Scale bars represent 150 mm.



changes in the superficial lamina propria thickness and collagen alignment is an exciting area for future study. Again, it must be stressed that further studies must be carried out to determine what the observed vocal fold polarization patterns truly indicate. In the case that these studies show, that the relative thickness and/or health of the lamina propria layers can indeed be indirectly assessed



 The diagnosis and treatment of many ENT pathologies would benefit from the ability to image subepithelially and with high resolution in situ.  Optical coherence tomography enables subepithelial imaging of tissue with a resolution ranging from 1 to 15 mm and is fibreoptically based, allowing OCT systems to be interfaced with a wide range of instruments, including catheters and endoscopes.  With respect to laryngeal tissues, conventional OCT appears to be able to detect spatial changes in the thickness and transparency of the epithelium; the content of the connective tissue, including the presence of glands and vessels; and the integrity of the basement membrane.  Polarization-sensitive OCT, with its sensitivity to organized linear structures, such as oriented collagen fibre bundles, may prove useful for vocal fold visualization and health assessment.



760 ] PART 11 RECENT ADVANCES IN TECHNOLOGY  OCT and OCT adjuncts hold substantial promise for improved diagnosis and treatment of many ENT disorders, a potential which remains largely untapped and constitutes an exciting avenue for future research, development and discovery.



Deficiencies in current knowledge and areas for future research Although OCT and OCT adjuncts hold substantial promise for improved visualization of ENT-related tissues, OCT/ OCT-adjunct feasibility studies for most of these tissues are either extremely limited or entirely lacking. For those ENT tissues that have been imaged with OCT, the primary focus of the imaging studies has been the necessary but insufficient first step of assessing the degree to which normal tissue microstructure can be accurately identified from OCT images. Before the OCT imaging modality can be regarded as a diagnostic tool in ENT, rigorous correlation of OCT/OCT-adjunct images of diseased tissue with parallel histology must be carried out to demonstrate that a given disease state can be accurately and reliably identified from OCT image structures. In the case of the larynx, although conventional OCT images of pathological laryngeal structures in vivo have been recorded,12 the corresponding histology necessary to guide OCT image interpretation has most often been lacking. Thus, whether OCT/OCT-adjuncts can reproducibly and accurately identify laryngeal pathology or other ENT pathologies remains to be determined and constitutes an exciting and vital area of future work. In addition to pathology identification, OCT or certain OCT adjuncts may prove useful in guiding laser and nonlaser-based microsurgery of ENT tissues. In laser microsurgery, PS-OCT may be particularly applicable, since laser damage of collagenous tissue has been shown to alter the tissue’s effect on the polarization state of incoming light.20 This is yet another exciting avenue for future research. In brief, the following are a few of the many vital areas for current and future investigation: rigorous correlation of histology with OCT images of pathologic laryngeal specimens; further investigation of the potential use of PS-OCT in laryngology and correlation of PS-OCT images to laryngeal microstructure; investigation of other OCT adjuncts, such as spectroscopic OCT, to laryngeal imaging; further exploration of the possible applications of OCT/OCT adjuncts to ear-related imaging and a feasibility study of OCT applied to nasal tissue; exploration of the use of OCT or PS-OCT systems to ENT surgical guidance.



ACKNOWLEDGEMENTS The authors thank Dr Steven Zeitels of the Massachusetts Eye and Ear Infirmary for fruitful discussions. We also thank Joe Seidel, PhD, for cutting and staining the histological sections appearing in this chapter, and Diane Jones of the Massachusetts Eye and Ear Infirmary for helping to procure human laryngeal autopsy specimens. This work was funded in part by the Advisory Board Foundation, Eugene B Casey Foundation and the National Science Foundation Graduate Fellowships.



REFERENCES



$ $ $ $ $











1. Hee MR, Izatt JA, Swanson EA, Huang D, Schuman JS, Lin CP et al. Optical coherence tomography of the human retina. Archives of Opthalmology – Chicago. 1995; 113: 325–32. 2. Swanson EA, Izatt JA, Hee MR, Huang D, Lin CP, Schuman JS et al. In vivo retinal imaging by optical coherence tomography. Optics Letters. 1993; 18: 1864–6. 3. Jang IK, Bouma BE, Kang DH, Park SJ, Park SW, Seung HK et al. Visualization of coronary atherosclerotic plaques in patients using optical coherence tomography: comparison with intravascular ultrasound. Journal of the American College of Cardiology. 2002; 39: 604–9. 4. Bouma BE, Tearney GJ, Compton CC, Nishioka NS. Endoscopic optical coherence tomography of the gastrointestinal tract. Gastrointestinal Endoscopy. 1999; 49: 390 Part 2. 5. Tearney GJ, Brezinski ME, Southern JF, Bouma BE, Boppart SA, Fujimoto JG. Optical biopsy in human urologic tissue using optical coherence tomography. Journal of Urology. 1997; 157: 1915–9. 6. Pitris C, Goodman A, Boppart SA, Libus JJ, Fujimoto JG, Brezinski ME. High-resolution imaging of gynecologic neoplasms using optical coherence tomography. Obstetrics and Gynecology. 1999; 93: 135–9. 7. Boppart SA, Bouma BE, Pitris C, Southern JF, Brezinski ME, Fujimoto JG. Intraoperative assessment of microsurgery with three-dimensional optical coherence tomography. Radiology. 1998; 208: 81–6. 8. Jesser CA, Boppart SA, Pitris C, Stamper DL, Nielsen GP, Brezinski ME et al. High resolution imaging of transitional cell carcinoma with optical coherence tomography: Feasibility for the evaluation of bladder pathology. British Journal of Radiology. 1999; 72: 1170–6. 9. Pitris C, Saunders KT, Fujimoto JG, Brezinski ME. Highresolution imaging of the middle ear with optical coherence tomography – A feasibility study. Archives of Otolaryngology – Head and Neck. 2001; 127: 637–42. Recently published study of conventional OCT as applied to ENT-related tissues. 10. Wong BJF, de Boer JF, Park BH, Chen Z, Nelson JS. Optical coherence tomography of the rat cochlea. Journal of



Chapter 60 Optical coherence tomography



11.



 12.



 13.



14.



Biomedical Optics. 2000; 5: 367–70. Recently published study of conventional OCT as applied to ENT-related tissues. Mahmood U, Ridgway J, Jackson R, Guo SG, Su JP, Armstrong W et al. In vivo optical coherence tomography of the nasal mucosa. American Journal of Rhinology. 2006; 20: 155–9. Gladkova ND, Shakhov AV, Feldschtein F. Capabilities of optical coherence tomography in laryngology. In: Bouma B, Tearney G (eds). Handbook of optical coherence tomography, Basel: Marcel Dekkar, 2002: 705–24. Recently published study of conventional OCT as applied to ENT-related tissues. Pitris C, Brezinski ME, Bouma BE, Tearney GJ, Southern JF, Fujimoto JG. High resolution imaging of the upper respiratory tract with optical coherence tomography – A feasibility study. American Journal of Respiratory and Critical Care Medicine. 1998; 157: 1640–4. Recently published study of conventional OCT as applied to ENTrelated tissues. de Boer JF, Srinivas SM, Nelson JS, Milner TE, Ducros MG. Polarization Sensitive OCT. In: Bouma B, Tearney G (eds). Handbook of optical coherence tomography. Basel: Marcel Dekkar, 2002: 237–74.



 15.



16.



17.



18.



19.



20.



] 761



Drexler W, Stamper D, Jesser C, Li XD, Pitris C, Saunders K et al. Correlation of collagen organization with polarization sensitive imaging of in vitro cartilage: Implications for osteoarthritis. Journal of Rheumatology. 2001; 28: 1311–8. Potential uses of PS-OCT are discussed in the context of a clinical example. Hammond TH, Gray SD, Butler JE. Age- and gender-related collagen distribution in human vocal folds. Annals of Otology Rhinology and Laryngology. 2000; 109: 913–20 Part 1. Burns JA, Zeitels SM, Anderson RR et al. Imaging the mucosa of the human vocal fold with optical coherence tomography. Annals of Otology Rhinology and Laryngology. 2005; 114: 671–6. Morgner U, Drexler W, Kartner FX, Li XD, Pitris C, Ippen EP et al. Spectroscopic optical coherence tomography. Optics Letters. 2000; 25: 111–3. Yang CH, Wax A, Dasari RR, Feld MS. Phase-dispersion optical tomography. Optics Letters. 2001; 26: 686–8. de Boer JF, Milner TE, vanGemert MJC, Nelson JS. Twodimensional birefringence imaging in biological tissue by polarization-sensitive optical coherence tomography. Optics Letters. 1997; 22: 934–6.



61 Contact endoscopy MARIO ANDREA AND OSCAR DIAS



Introduction Technical details Contact endoscopy of the larynx Contact endoscopy of the nasal cavity Contact endoscopy of the nasopharynx Oral cavity and oropharynx



762 763 763 764 766 767



Summary Key points Deficiencies in current knowledge and areas for future research References



767 767 767 767



SEARCH STRATEGY AND EVIDENCE-BASE The data in this chapter are supported by a Medline search using the key words contact endoscopy, in vivo diagnosis focusing on the clinical impact of contact endoscopy in the assessment and understanding of diseases of the upper aero digestive tract. The evidence level of contact endoscopic assessment is correlated with its learning curve involving knowledge of cytological, pathological and micro vascular patterns. It is difficult to classify the contact endoscopy findings by levels of evidence because we have direct access to the cell images in its more pure form. In some aspects the level of evidence is 1. However, in general, we would classify it as level 3 and 4 evidence.



INTRODUCTION Contact endoscopy was first described by Desormeaux in 1865,1, 2 who managed to obtain a direct view of the bladder mucosa. Interest in this technique then waned for almost a century, until technological advances had been made and Jaupitre3 began to promote it once more. More recently, Hamou4, 5 developed and described his technique of microcolpohysteroscopy, in which contact endoscopy played a pivotal role. Its use in research to study the microvasculature of the larynx6, 7, 8, 9, 10 prompted further clinical applications within the upper airway. Contact endoscopy has now been employed to assess vocal cord pathology, the nasal mucosa, nasopharynx, oral cavity, oropharynx and the trachea.11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22 Contact endoscopy allows in vivo and in situ observation of the mucosal blood vessels and superficial cells of the epithelium which have been previously stained with methylene blue. The experienced otolaryngologist is able



to perform a pathological evaluation during planned endoscopy and in many cases, an immediate diagnosis can be made. The information acquired does not render conventional biopsy and histological evaluation obsolete, rather it gives additional information to complement it. With contact endoscopy it is possible to detect microvascular changes and/or alterations in surface cellular structure that are suggestive of subclinical stages of disease. It is worth noting that because cells migrate towards the surface, most pathological processes can be seen by examining the superficial layers23, 24, 25 and it is precisely this that contact endoscopy addresses. The surgeon is able to assess the microscopic structure of the entire mucosa, allowing a more complete interpretation of the disease process. In other words, it offers the clinician a global perspective of the disease process with the facility of cellular and vascular mapping at any number of sites in the region. Furthermore, contact endoscopy has the advantage that it can be undertaken



Chapter 61 Contact endoscopy



with the patient awake in the outpatient setting or while anaesthetized in the theatre. Contact endoscopy of the upper aerodigestive tract is still in its infancy. More and more centres are using this technique and its principles are being applied in other specialties as well. No doubt further technological advances will be forthcoming and the process will be refined. It is possible that in the future, contact endoscopy will occupy a central role in diagnosis, therapeutic planning and the follow-up of patients with a number of upper aerodigestive tract diseases. This chapter describes the current state of the art of contact endoscopy.



TECHNICAL DETAILS Contact endoscopy can be undertaken in either the conscious or anaesthetized patient. Topical anaesthesia is all that is required. Several endoscopes have been developed and are commercially available but, for most purposes in the upper airway, only two endoscopes are required (7215 AA and 7215 BA, Karl Storz, Tuttlingen, Germany). With these endoscopes the clinician has direct access to the subsurface microvascular plexus and to the surface epithelium which can be viewed at a magnification of 60 and 150. The mucosal surface to be examined is carefully cleaned by gentle suction or with a swab that has been moistened with saline. The surface is then stained with 1 percent methylene blue applied on a fragment of Spongostans. The tip of the endoscope is placed gently against the surface of the mucosa allowing examination. Staining lasts for approximately four to five minutes at most sites before gradually disappearing. In the nasal cavity, the stain disappears more quickly as it is moved away towards the nasopharynx along the mucociliary pathway.



CONTACT ENDOSCOPY OF THE LARYNX Contact endoscopy of the larynx and hypopharynx is performed under general anesthesia with conventional endotracheal intubation and synchronous microlaryngoscopy.



Normal appearance Squamous cells at the vocal cord edge and in the hypopharynx have a polyhedric shape, being contiguous with each other. Their nuclei are round, darkly stained and the cytoplasm has a light blue tone. The nuclear: cytoplasmic ratio is regular and the overall morphologic pattern is homogeneous. Ciliated epithelium is present in most of the larynx. The nuclei of the cells are round, but the limits of the cytoplasm are difficult to define. Bundles



] 763



of cilia appear as filamentous structures that can be displaced by the tip of the contact endoscope. The transition from squamous to ciliated epithelium can be observed. Islands of squamous cells in the middle of ciliated epithelium may be seen and are not abnormal. The duct orifices of glands can also be identified. What is and what is not normal varies between patients and is very much influenced by such factors as age and diet and habits such as voice use and smoking.



Abnormal appearance Metaplastic substitution of ciliated epithelium by squamous epithelium is most commonly seen in heavy smokers. It also occurs in patients with gastroesophageal reflux where the ciliated epithelium of the posterior commissure is replaced by squamous epithelium. It would seem that this interferes with normal mucus clearance. When patients with chronic laryngitis are studied by this method, the epithelial pattern is found to be homogeneous, but epithelial cells have larger nuclei than normal and an increased nuclear:cytoplasmic ratio. As cell turnover is accelerated by inflammation, it is possible to visualize immature cells at the surface that are similar to those usually present in the intermediate layers of the normal epithelium. There are more blood vessels within the mucosa, but the vascular network keeps its normal pattern. In some cases of chronic laryngitis, filaments of fungal hyphae are identified and, if seen, should prompt treatment with antifungal agents. Keratosis is detected very easily. Distinct and different stages of keratinization may be seen in the same patient. In the initial stages of keratinization, isolated cells without nuclei are observed. More advanced stages show groups of cells without nuclei, but with distinct cells still identifiable. With further progression of the disease it is not possible to distinguish the borders of individual cells and only large areas of an amorphous or laminar structure are visible. In frank leukoplakia, degrees of cellular abnormality are identified that include heterogeneity of cell populations with nuclei of different colour, size and shape. The variation seen is in agreement with the histopathological concept of leukoplakia, in which different pathological alterations such as hyperkeratosis and dysplasia can develop simultaneously.24 Contact endoscopy may not detect dysplasia when alterations are confined solely to the deep epithelial layers. However, in most cases, dysplasia is associated with involvement of the superficial layers as well and these are accessible to the endoscope. The most significant dysplastic changes23, 24, 25 are variations in nuclear size, shape and colour, together with changes of the nuclear: cytoplasmic ratio (dyskaryosis and anysokaryosis). The more pronounced the changes identified by contact endoscopy, the more severe the dysplasia. Vascular



764 ] PART 11 RECENT ADVANCES IN TECHNOLOGY changes also develop with dysplasia, where the normal vascular pattern is substituted by vessels of varying size. Unfortunately, when intense keratosis accompanies dysplasia it is not possible to observe the vessels beneath the surface. In carcinoma, cell morphology is very variable with extreme heterogeneity of nuclear size, shape and staining characteristics. The nuclear:cytoplasmic ratio also changes from cell to cell. Nuclear inclusion bodies, prominent nucleoli and mitoses may be seen. The normal microvascular pattern and architecture is disturbed. Atypical vessels of different sizes and shapes are seen, as are thromboses, ectasias and rupture. In this situation, contact endoscopy enables the assessment of transitional zones between normal and abnormal mucosa and, as a result, a better evaluation of early stage disease. Surgeons also find it useful to guide biopsies and the collection of cytological samples so that safe margins may be identified. Interestingly, adjacent areas with normal squamous and ciliated cells often exhibit vascular changes (vascular loops) that signify an imminent change in status. Contact endoscopy has also had an impact on the assessment and management of laryngeal papillomata. Typical vascular loops in the core of the papillomata can be seen as long as the associated keratosis is not too severe. In certain phases of the disease it is possible to identify koilocytes (ballooned cells) and inflammatory infiltrate. New lesions and incipient spread can also be detected. It would be hoped that eventually this information would translate into more accurate excision of papillomata and a reduced risk of residual or recurrent disease. In summary, contact endoscopy helps the surgeon understand the pathophysiology of several common laryngeal conditions, such as metaplasia, caused by smoking, gastroesophageal reflux or vocal abuse, premalignant lesions, the early stages of laryngeal carcinoma, human papilloma virus (HPV) lesions and the association of cancer and HPV infection. In the future, tumour staging may well expand to encompass information on the cellular changes detected by contact endoscopy (Figure 61.1).



CONTACT ENDOSCOPY OF THE NASAL CAVITY The anterior part of the nasal cavity may be examined easily with the patient awake using topical anaesthesia.



Normal appearance As in the larynx, it is possible to evaluate the distribution and morphological appearance of the squamous and ciliated epithelia, glandular ostia, mucus and submucosal vascular networks. Squamous epithelium is usually present at the anterior tip and inferior border of the



inferior turbinate, the septum and nasal vestibule. This epithelium consists of polyhedric cells with dark blue nuclei and a light blue cytoplasm. Ciliated epithelium is present in most of the nasal cavity. The transition between ciliated and squamous epithelium is usually well defined, although in some cases isolated patches of squamous epithelium are seen in the middle of the ciliated epithelium. The dynamic variation in volume of the turbinates caused by the nasal cycle explains the presence of small folds and papillas that tend to move in front of the endoscope. The duct orifices of glands are most prominent at the anterior end of the turbinates. Characteristically, they appear as concentric accumulations of cells with the innermost cell layers on a deeper plane. Dynamic changes in the glandular ostia are sometimes observed when manipulating the contact endoscope. The orifices of the glands become more noticeable when mucus is expelled. Mucus itself is seen as a moving layer that stains light blue and emanates from distinct papillas. It follows the contours of the papillas and carries away cell debris which either detaches spontaneously or as a result of the movement of the endoscope. Characteristic features of the microvascular network have also been described. The papillas observed in the turbinates usually contain a small blood vessel. In the anterior part of the septum there are vessels of a very distinct calibre that are linked by anastomoses.



Abnormal appearance Contact endoscopy in the nasal cavity is probably of most use in the study and management of chronic rhinosinusitis, allergic rhinitis, nasal polyposis and mucocilliary diseases. In these conditions, it can be employed to assess the response to different types of treatment.26 In patients with chronic rhinitis, squamous epithelium covers most of the inferior and middle turbinates and anterior septum. The nuclei of these cells look larger than normal. Areas of keratosis predominate in regions exposed to turbulent air flow, such as the anterior end of the inferior surface of the turbinates. Usually in these patients there is overproduction of mucus and the methylene blue stain clears very quickly. The density of mucous glands is increased and consequently more ostia are seen than in normal patients. Inflammatory cell infiltrates can be identified and differentiation of this infiltrate may help define the nature of the inflammatory process. Mucus production is also increased in patients with allergic rhinitis. The papillae of glands are larger than normal and a U-shaped vascular axis is seen in the papillae. Ciliated epithelium is usually preserved in allergic rhinitis and there are fewer inflammatory changes than in the chronic nonallergic rhinitis group. In chronic nonallergic rhinitis, vascular congestion is diffuse and



Chapter 61 Contact endoscopy



] 765



Figure 61.1 Contact endoscopic assessment of a laryngeal tumour. (a) Laryngeal tumour (T3 N2 Mx); (b–h) Contact endoscopy (60). (b) Right arytenoid region: normal squamous epithelium. Note the regular appearance of the cells and their nuclei. (c) Left arytenoid region: tumour present. The nuclei of the epithelium are irregular in size and shape. Numerous mitoses are visible. (d) Right vocal cord: malignant changes are easily visible, as well as variations in nuclear size and shape. (e) Left vocal cord: tumour pattern. (f) Right false cord: transition from normal epithelium to tumour pattern.



independent of the papillae. Very enlarged vessels are sometimes identified. Examination of nasal polyps has shown that the anterior face is covered by squamous epithelium while the



rest of the polyp is covered with ciliated cells. Metaplastic changes may be present and are related to the duration of exposure to turbulent air flow. Nasal polyposis is usually accompanied by inflammatory cell infiltrates.



766 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



Figure 61.1 Contact endoscopic assessment of a laryngeal tumour. (Continued) (g) Right false cord: islands of atypical cells. (h) Medial aspect of the right arytenoid: tumour pattern associated with presence of koylocytes. (i) Contact endoscopy (150), koylocytes (ballooned cells), medial aspect of the right arytenoid. (j) Contact endoscopy (60), vascular loops typical of HPV, right arytenoid region.



Contact endoscopy can, and perhaps should, be used during functional endoscopic sinus surgery to assess the state of the mucosa. The type and degree of abnormalities can help the surgeon decide which areas of mucosa to remove and which to retain.



CONTACT ENDOSCOPY OF THE NASOPHARYNX Contact endoscopy of the nasopharynx can be performed through the nose under local anaesthesia. Its main use is in the long-term follow-up of patients who have been treated for nasopharyngeal carcinoma.



Normal appearance Although the majority of the nasopharynx is lined with ciliated epithelium, it is usual to see areas of squamous epithelium, typically in the central and inferior part of the



posterior wall. Orifices of glandular ducts are present throughout the nose.



Abnormal appearance An irregular vascular pattern is seen in areas with nasopharyngeal carcinoma. Atypical vessels, ectasias, thrombosis, blood cell aggregates and increased vascular fragility are common. The tissue is extremely fragile and will bleed easily if probed too firmly with the endoscope. As with all malignancy, the stained epithelium of the nasopharynx shows anysokaryosis, heterochromasia and hyperchromasia. There are often clusters of malignant cells with hyperchromatic nuclei. Most nuclei are irregular in shape, others fusiform in shape (fibroblastlike) and some are large and oval. Occasionally, it is possible to observe the nucleolus, nuclear inclusions and even mitoses. There is almost always a marked lymphocytic infiltrate. The transitional zone between the tumour



Chapter 61 Contact endoscopy



and normal tissues is very important and it is usual to see inflammatory infiltrates and/or keratosis. Small islands of malignant cells are sometimes found in the middle of the apparently normal ciliated epithelium in areas immediately adjacent to the tumour. It is worth emphasizing that a number of centres regularly use contact endoscopy for both the diagnosis and follow-up of their patients with nasopharyngeal tumours.27, 28 In particular, this technique is used in areas where nasopharyngeal carcinoma is endemic as an officebased procedure to confirm the diagnosis, identify submucosal disease when malignancy is not detected endoscopically and to identify persistent and recurrent nasopharyngeal carcinoma. Contact endoscopy is also used to select areas for biopsy.



ORAL CAVITY AND OROPHARYNX Contact endoscopy of the oral cavity and oropharynx is performed either with or without topical local anaesthesia.



Normal appearance It is important to remember that the morphology of the oral mucosa varies from site to site. In some parts it is keratinized, the so-called masticatory mucosa, while in others it is not, as in the lining of the mucosa or specialized mucosa. These different patterns must be recognized if pathological changes are to be understood and interpreted. The transition from the keratinized epithelium of the lip to the nonkeratinized epithelium of the vestibule is seen easily. The greater part of the mucosa of the lip, the alveoli, cheek, floor of the mouth, ventral surface of the tongue and soft palate is also nonkeratinized squamous epithelium. Some keratinization is to be expected in the cheek opposite the occlusal plane. The masticatory mucosa covering the hard palate and gingiva is a keratinized epithelium. On the dorsum of the tongue, filiform and fungiform papillae can be seen. Contact endoscopy in the oral cavity and oropharynx has great potential in areas such as the diagnosis of early cancer, the study of tumour margins, the assessment of the response to radiotherapy and chemotherapy and to identify subclinical stages of disease.



SUMMARY Contact endoscopy allows in vivo and in situ assessment of the mucosa and underlying microvascular network. With experience, the operator can distinguish between normal and abnormal tissues. It represents a new phase in the development of endoscopy. This technique can



] 767



improve the accuracy of diagnosis and give useful information about the extent of disease. It helps identify earlier subclinical stages of disease than was previously possible. Many centres use this technique and global experience is increasing.27, 28, 29, 30, 31, 32 The accuracy and clinical applicability of contact endoscopes will continue to be enhanced by improvements in optical systems, new cell dyes, markers, fluorescent products,22 light sources, image processing and better recording techniques.



KEY POINTS  Contact endoscopy is a simple, noninvasive technique that offers real-time information about disease processes.  Contact endoscopy allows the mapping of an entire mucosal surface in the conscious or anaesthetized patient.  Contact endoscopy may eventually be used to monitor changes that develop in the premalignant stages of the disease.



Deficiencies in current knowledge and areas for future research



$ $ $



Contact endoscopy has great potential for the early diagnosis of nasopharyngeal cancer. However, more experience needs to be gained to define its role as a screening procedure. More knowledge is needed on the pathogenesis of squamous metaplasia and atypical hyperplasia. Articulation of contact endoscopy with other technologies such as fluorescence, confocal and narrow band imaging will further enhance the accuracy of in vivo and in situ diagnosis.



REFERENCES 1. Reuter M, Reuter H, Engel R. History of endoscopy. Publication of the Max Nitze Museum, Stuttgart and International Nitze Leiter Research Society of Endoscopy, Vienna, 1999. 2. Desormeaux AJ. De l’endoscope et de ses applications au diagnostique et au traitement des affe´ctions de l’ure´thre et de la vessie. Paris, 1865 (from History of endoscopy. Publication of the Max Nitze Museum, Stuttgart and International Nitze Leiter Research Society of Endoscopy, Vienna, 1999).



768 ] PART 11 RECENT ADVANCES IN TECHNOLOGY



 







3. Jaupitre M. La cystocine´matographie en couleur. Communications au Congre´s Franc- ais d’Urologie. October 1955, 128 (from History of endoscopy. Publication of the Max Nitze Museum, Stuttgart and International Nitze Leiter Research Society of Endoscopy, Vienna, 1999). 4. Hamou JE. Microendoscopy and contact endoscopy. Brevet Franc- ais 79,04168, Paris 1979. International patent PCT/ FR 80/0024, Paris, 1980. US patent 4,385,810, Washington DC, 1983. 5. Hamou JE. Hysteroscopy and microcolpohysteroscopy, text and atlas. Appleton & Lange, 1991. 6. Andrea M. ‘Vascularizac- a˜o arterial da laringe, distribuic- a˜o macro e microvascular’. PhD dissertation, University of Lisbon, Portugal, 1975. 7. Andrea M, Guerrier Y. L’Epiglotte et ses amarrages. Cahiers d’ORL et de Chirurgie Cervico-Facial. 1979; 14: 793–803. 8. Andrea M, Guerrier Y. Microvascularization de la muqueuse larynge´e et trache´ale. Introduction a` la physiopathologie des le´sions ste´nosantes. Annales d’Otolaryngologie et de Chirurgie Cervico Faciale. 1980; 97: 409–21. 9. Andrea M, Guerrier Y. The anterior commissure of the larynx. Clinical Otolaryngology. 1981; 6: 259–64. 10. Andrea M. Vasculature of the anterior commissure. Annals of Otology, Rhinology, and Laryngology. 1981; 90: 18–20. 11. Andrea M, Dias O, Pac- o J, Santos A, Fernandes A. Anatomical borders in staging – Rigid endoscopy associated to microlaryngeal surgery. In: Smee R, Bridger G. (eds). Laryngeal cancer. Proceedings of the II World Congress of Laryngeal Cancer, Sidney, Australia. Elsevier Science BV, 1994: 233–5. 12. Andrea M, Dias O, Santos A. Contact endoscopy during microlaryngeal surgery. A new technique for endoscopic examination of the larynx. Annals of Otology, Rhinology, and Laryngology. 1995; 104: 333–9. 13. Andrea M, Dias O, Pac- o J. Endoscopic anatomy of the larynx. In: Fabian R, Gluckman J (eds). Current review of otolaryngology and head and neck surgery. London: Churchill Livingstone, 1995: 271–5. 14. Andrea M, Dias O, Santos A. Contact endoscopy of the vocal cord. Normal and pathological patterns. Acta Otolaryngologica. 1995; 115: 314–6. 15. Andrea M, Dias O. Rigid and contact endoscopy associated to microlaryngeal surgery. In: Fried M (ed.). The larynx: A multidisciplinary approach, 2nd edn. Mosby Year Book, 1996: 75–9. 16. Andrea M, Dias O. Atlas of rigid and contact endoscopy in microlaryngeal surgery. Philadelphia: Lippincott-Raven Publishers, 1995. 17. Andrea M, Dias O. Rigid and Contact endoscopy during microsurgery. In: Yanagisawa E (ed.). Color atlas of diagnostic endoscopy in otorhinolaryngology. Tokyo: Igaku Shoin, 1996: 168–73.



















18. Andrea M, Dias O. Endoscopic assessment of early vocal cord cancer. Paper presented at the Proceedings of IV International Conference of Head and Neck Tumors, Toronto, 1996. 19. Andrea M, Dias O. Newer techniques of laryngeal assessment. In: Cummings CW, Frederikson JM, Krause CJ, Harker LA, Schuller DE, Richardson MA (eds). Otolaryngology head and neck surgery. St Louis: Mosby Year Book, 1998: 1967–78. 20. Andrea M, Dias O. Rigid and contact endoscopy of the larynx. In: Ferlito A (ed.). Diseases of the Larynx. London: Arnold, 2000: 1001–111. 21. Andrea M, Dias O. La endoscopia rı´gida y de contacto en la evaluacio´n de las lesiones premalignas de la laringe. In: Suarez Nieto C (ed.). Tratado de otolaringologia y cirurgia de cabeza y cuello. Madrid: Proyectos Me´dicos, 1999: 3006–9. 22. Arens C, Malzahn K, Dias O, Andrea M, Glanz H. Endoskopische bildegebende verfahren in der diagnostik des kehlkopfkarzinoms und seiner vorstufen. Laryngorhinootologie. 1999; 78: 685–91. 23. Crissman J. Pathology of the upper aerodigestive tract mucosa. In: Paparella M, Schumrick D, Gluckman J, Meyerhoff W (eds). Otolaryngology, head and neck, 3rd edn. Vol. III, Philadelphia: WB Saunders, 1991. 24. Ferlito A. Neoplasms of the larynx. New York: Churchill Livingstone, 1993. 25. Kleinsasser O. Tumors of the larynx and hypopharynx. George Thieme Verlag, 1988. 26. Andrea M, Dias O, Macor C, Santos A, Varandas J. Contact endoscopy of the nasal mucosa. Acta Otolaryngologica. 1997; 117: 307–11. 27. Xiaoming H, Haiqiang M, Manquan D, Jianyong S, Yong S, Kela L et al. Examination of nasopharyngeal epithelium with contact endoscopy. Acta Otolaryngologica. 2001; 121: 98–102. 28. Pak MW, To KF, Leung SF, van Hasselt CA. In vivo diagnosis of nasopharyngeal carcinoma using contact rhinoscopy. Laryngoscope. 2001; 11: 1453–8. 29. Pau HW, Dommerich S, Just T, Beust M. Cholesteatoma recurrences caused by intraoperative cell seeding? Contact endoscopic and cytologic studies. Laryngorhinootologie. 2001; 80: 499–502. 30. Richtsmeier WJ, Huang P, Scher RL. In situ identification of normal visceral tissues using contact telescopic microscopy. Laryngoscope. 1999; 109: 216–20. 31. Wardrop PJ, Sim S, Mclaren K. Contact endoscopy of the larynx: a quantitative study. Journal of Laryngology and Otology. 2000; 114: 437–40. 32. Carriero E, Galli J, Fadda G, Di Girolamo S, Ottaviani F, Paludetti G. Preliminary experiences with contact endoscopy of the larynx. European Archives of Oto-RhinoLaryngology. 2000; 257: 68–71.



PART



12



PAEDIATRIC OTORHINOLARYNGOLOGY EDITED BY RAY CLARKE



62 Introduction Ray Clarke



771



63 The paediatric consultation Ray Clarke and Ken Pearman



776



64 ENT input for children with special needs Francis Lannigan



783



65 Head and neck embryology T Clive Lee



792



66 Molecular otology, development of the auditory system and recent advances in genetic manipulation Henry Pau



811



67 Hearing loss in preschool children: screening and surveillance Kai Uus and John Bamford



821



68 Hearing tests in children Glynnis Parker



834



69 Investigation and management of the deaf child Sujata De, Sue Archbold and Ray Clarke



844



70 Paediatric cochlear implantation Joseph G Toner



860



71 Congenital middle ear abnormalities in children Jonathan P Harcourt



869



72 Otitis media with effusion George Browning



877



73 Acute otitis media in children Peter Rea and John Graham



912



74 Chronic otitis media in childhood John Hamilton



928



75 Management of congenital deformities of the external and middle ear David Gault and Mike Rothera



965



76 Disorders of speech and language in paediatric otolaryngology Ray Clarke and Siobhan MCMahon



990



77 Cleft lip and palate Chris Penfold



996



78 Craniofacial anomalies: genetics and management Dean Kissun, David Richardson, Elizabeth Sweeney and Paul May



1019



79 Vertigo in children Gavin AJ Morrison



1040



80 Facial paralysis in childhood SS Musheer Hussain



1052



81 Epistaxis in children Ray Clarke



1063



82 Nasal obstruction in children Michelle Wyatt



1070



83 Paediatric rhinosinusitis Glenis Scadding and Helen Caulfield



1079



84 The adenoid and adenoidectomy Peter J Robb



1094



85 Obstructive sleep apnoea in childhood Helen M Caulfield



1102



86 Stridor David Albert



1114



87 Acute laryngeal infections Susanna Leightony



1127



88 Congenital disorders of the larynx, trachea and bronchi Martin Bailey



1135



89 Laryngeal stenosis Michael J Rutter and Robin T Cotton



1150



90 Paediatric voice disorders Ben Hartley



1167



91 Juvenile-onset recurrent respiratory papillomatosis Michael Kuo and William J Primrose



1174



92 Foreign bodies in the ear and the aerodigestive tract in children A Simon Carney, Nimesh Patel and Ray Clarke



1184



93 Tracheostomy and home care Michael Saunders



1194



94 Cervicofacial infections in children Ben Hartley



1210



95 Diseases of the tonsil William S McKerrow



1219



96 Tonsillectomy William S McKerrow and Ray Clarke



1229



97 Salivary gland disorders in childhood Peter D Bull



1242



98 Tumours of the head and neck in childhood Fiona B MacGregor



1251



99 Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma Peter D Bull



1264



100 Gastro-oesophageal reflux and aspiration Haytham Kubba



1272



101 Diseases of the oesophagus, swallowing disorders and caustic ingestion Lewis Spitz



1282



102 Imaging in paediatric ENT Neville Wright



1295



103 Medical negligence in paediatric otolaryngology Maurice Hawthorne



1305



62 Introduction RAY CLARKE



Introduction History Societies The scope of paediatric practice Paediatric otolaryngologists Developing ENT services for children The general/specialist otolaryngologist



771 771 773 773 774 774 774



Training the otolaryngologists of the future Key points Deficiencies in current knowledge and areas for future research Acknowledgements References



775 775 775 775 775



SEARCH STRATEGY The author has a personal bibliography, which formed the basis for this chapter. The opinions expressed are his own.



INTRODUCTION



HISTORY



Caring for children has been an integral part of otolaryngological practice since the specialty began. Textbooks of paediatric otolaryngology have a shorter history. Although Douglas Guthrie published a small booklet Diseases of the ear, nose and throat in childhood in 1921,1 Wilson’s 1955 Diseases of the ear, nose and throat in children was the first substantial English-language text devoted to paediatric otorhinolaryngological disorders.2 This important little book provides early descriptions of aminoglycoside-induced deafness, the technique of tracheobronchoscopy in children and obstructive sleep apnoea, then known as ‘aprosexia’. It is a measure of the maturity of paediatric otolaryngology as a discrete subspecialty, and a recognition that the pathophysiology, management and expectations from treatment of diseases in children are often very different from what obtains in adults2 that this is now the third edition of Scott-Brown’s Otorhinolaryngology, Head and neck surgery which includes a separate section covering paediatric aspects of our specialty.



Mastoid surgery for the treatment of otological sepsis was popularized by Sir William Wilde (1815–1878) who described what we now know as otitis media with effusion – ‘strumous otitis’.3 Wilde recognized the association between ‘strumous otitis’ and Eustachian tubal dysfunction, described tympanocentesis as a treatment and popularized the use of a myringotomy knife which is not dissimilar to that used today.4 The Victorian and Edwardian laryngologists, notably Morrel McKenzie and Sir Felix Semon in London, had large paediatric practices and dealt with often-fatal upper respiratory tract pathologies, such as diphtheria. McKenzie provides us with the first description of laryngeal papillomatosis. He noted the post-mortem findings in a child who died of airway obstruction in what he poignantly described as a ‘home for the friendless’.5 In Europe, Gustav Killian pioneered suspension laryngoscopy and tracheobronchoscopy in Freiburg, Bresgau at the beginning of the twentieth century. The principles were soon extended to children. Chevalier



772 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Figure 62.1 Chevalier Jackson teaching in Paris. Reproduced by kind permission of the John Q Adams Center for the History of Otolaryngology—Head and Neck Surgery, American Academy of Otolaryngology—Head and Neck Surgery Foundation, r 2007. All rights reserved.



Jackson (Figure 62.1) in Philadelphia established a reputation throughout the United States for the skill with which he could extract bronchial foreign bodies. A brilliant teacher, he illustrated his work in his own hand and was probably the single most important figure to popularize airway endoscopy in children on both sides of the Atlantic. The pioneering work of British physicist Harold Hopkins in the design of modern ‘rod-lens’ telescopes (Figure 62.2) moved paediatric airway surgery to new levels. Modern-day paediatric otolaryngologists are acutely aware of the debt they owe to paediatricians and anaesthesiologists. Sophisticated diagnostic and therapeutic procedures in children are only possible because of the huge advances made in improving the survival and care of small babies. It is difficult for those of us trained in recent times to appreciate just how harrowing airway disorders in children could become before the equipment and skills which we now take for granted were developed. Joseph O’Dwyer (New York) is credited with the first successful emergency endotracheal intubation in a child. In 1886 he presented 50 cases with a 70 percent mortality. The technique remained controversial but was enthusiastically taken up by a Chicago physician, Dr Frank Waxham, who used it to provide an alternative airway for children with diphtheria.6 Long-term nasal endotracheal intubation as an alternative to tracheotomy was popularized only from the 1960s onwards. In 1955, Wilson wrote of emergency tracheotomy in children, ‘these are desperate cases at best, and it may be a comfort to remember that the worst thing which can happen is that the patient will die. This is unfortunately a likely event in any case.’2 The late Sylvan Stool, a pioneer American paediatric otolaryngologist and historian recounts his earliest experiences with the removal of bronchial foreign bodies: ‘The telescopes and cameras of today make endoscopy easy. Anybody can remove a foreign body and look like a technical master. You should have been around when we were changing that little bulb on the end of the Jackson bronchoscope three or four times during a case.’6



Figure 62.2 A modern ventilating bronchoscope incorporating a ‘Hopkins rod’ telescope.



Airway endoscopy in children is now a highly skilled – and safe – undertaking. This is due to parallel advances in equipment and the training of personnel – otolaryngologists, anaesthesiologists and support staff. The equipment required for this work is now considerable (see Chapter 86, Stridor). The first paediatric ENT ward was opened in Poland in 1895. This was at the Children’s Hospital in Warsaw under the supervision of Dr Jan Gabriel Danielewicz. An independent children’s ENT department was established in Budapest immediately after the Second World War. In Italy, Dr Carlo Mancini headed a department in Brezia in 1950. Children’s ENT services were quickly established in Czechoslovakia, Slovakia, and Bucharest. In Britain ENT departments were founded in the early part of the twentieth century in the children’s hospitals at Great Ormond Street London, Alder Hey Liverpool and in Bristol. Advances in diagnosis and rehabilitation have greatly improved outcomes and expectations for the deaf child. Although audiological medicine has developed as a separate specialty, and a multiplicity of professionals



Chapter 62 Introduction



] 773



Figure 62.3 Surviving document from the founding of the British Association for Paediatric Otolaryngology (BAPO). Reproduced with kind pemission of Mr John Graham FRCS, London, UK.



work in tandem to care for children with hearing loss, this remains a core part of the workload of otolaryngologists who look after children. Hence, much of the foregoing section is devoted to this topic.



SOCIETIES As new specialties emerge, the societies and associations which promote their aims and allow colleagues to share views, advances and collective wisdom develop. Loose gatherings of otolaryngologists with an interest in children were formed in Eastern Europe from the beginning of the twentieth century. In the United States the Society for Ear Nose and Throat Advances in Children (SENTAC) met in 1977. The American Society for Pediatric Otolaryngology (ASPO) (www.aspo.us) was founded in 1985. The European Working Group in Pediatric ENT was founded in 1977 and later became the European Society for Pediatric Otolaryngology (ESPO) – a forum for discussion and advances in paediatric ORL through international meetings and via its journal the International Journal of Pediatric Otolaryngology. In Britain it was not until 1991 that the British Association for Paediatric Otolaryngology (BAPO) (www.bapo.org/uk) was formed. The idea for a British society came about at a meeting of colleagues attending the international congress at Ghent in 1990. Surviving mementos include a signed menu (Figures 62.3 and 62.4) from the Chez Armand restaurant (John Graham FRCS, personal communication).



THE SCOPE OF PAEDIATRIC PRACTICE Almost all otolaryngologists look after children. For most this represents a vital and rewarding part of their clinical workload. The spectrum of diseases ranges from common problems such as adenotonsillar hypertrophy,



Figure 62.4 The BAPO logo.



rhinosinusitis and otitis media to more esoteric and challenging presentations such as subglottic stenosis, tumours of the salivary glands and congenital deformities of the ear. A general otolaryngologist or an otolaryngologist with a broad-based practice but an interest in children can satisfactorily look after most children. Those with complex or rare conditions, the very young and those with multiple medical pathologies requiring the attention of paediatricians, neonatologists and intensivists are best managed in a designated children’s centre in close liaison with other professionals. Hence the skills of ‘paediatric otolaryngologists’ are increasingly required at these centres.



774 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



PAEDIATRIC OTOLARYNGOLOGISTS Paediatric otolaryngologists look after ‘special problems, in special children in a special institution’.7 Needless to say, they will have had training that reflects this remit.



Special problems The ‘special problems’ that paediatric otolaryngologists look after include congenital and acquired airway disease and the otorhinolaryngology (ORL) pathologies which occur in children who require management on a special care baby unit (SCBU) or in the paediatric intensive care unit (PICU). Much of this work is diagnostic and supportive. It lacks the glamour and appeal of laryngotracheal reconstructions, cricotracheal resections, etc., which are procedures now rarely done outside of highly specialized units. Day-to-day paediatric ORL involves liaison with paediatricians, paediatric anaesthesiologists, intensivists, nursing and allied staff and, most of all, often distraught parents and carers. The work is by its nature unpredictable and often occurs ‘out of hours’. The required skills are not just technical but relate to the judgement and experience needed to make decisions on small and often very ill babies and to contribute to a multidisciplinary team.



Special children All children are special but some of the children with whom a paediatric otolaryngologist deals are very special indeed because of their complex medical needs. These include syndromic children, children with genetic diseases, children with often severe locomotor and neurological dysfunction, and children with communication difficulties and learning disabilities, which range from mild to profound. All have parents/carers who are intimately and uniquely involved in the child’s health care decisions and who simply cannot be appropriately looked after in a general ENT clinic setting. These children get the range of ORL problems with which we are all familiar, for example adenotonsillar disease, rhinitis, otitis media, often against a background of severe multisystem disease. They need a completely different management approach to that needed for otherwise healthy children. They are best looked after by clinicians with skills that can only be developed by working with such children and parents and who have an appreciation of the natural history of diseases in such settings. If they require surgery it is often wholly inappropriate that this should take place outside of a specialist children’s centre where facilities for anaesthesia, recovery and postoperative intensive care are available as required. Special needs children also develop unique problems such as sialorrhoea, obstructive sleep apnoea



and aspiration which are best managed by experienced personnel in a designated setting.



Special hospitals Children’s hospitals or ‘centres’ are ‘special’ places not because of their geographic location but because of the availability of skilled personnel in a unique multidisciplinary setting. Support services such as anaesthesia, speech and language therapy, paediatric imaging, children’s specialist nurses, feeding clinics, regional facilities such as craniofacial surgery and neonatal surgery are simply not found outside these centres. ‘Centres’ include not only the specialist children’s hospitals but also paediatric facilities in the few large teaching hospitals where supraregional services for children are concentrated. Perhaps the defining features of such a centre are the availability of a dedicated paediatric surgical team and the presence of a designated PICU. A paediatric otolaryngologist will have the major proportion of his/ her clinical commitments at such a centre.



DEVELOPING ENT SERVICES FOR CHILDREN It is essential that otolaryngologists are to the fore in local discussions on service planning that have an impact on children – and that is most discussions in the sphere of health care. Up to 30–50 percent of our workload involves the care of children.8 Otolaryngologists have, and must maintain, an important role as strong advocates for children.9, 10 Otherwise healthy children with common pathologies where anaesthetic risks are minimal are best looked after near their homes in settings that are convenient for parents and carers. Provided support services are adequate there is no merit in relocating all this work to already overstretched children’s centres. Children with airway obstruction who can be cared for locally – for example those with acute infections, croup, the older child with sleep apnoea – are appropriately managed by a designated otolaryngologist with a special paediatric interest. Increasingly, one or more members of a team of otolaryngologists will assume a lead role in looking after such children and may also lead the development and management of audiology services for children, given the shortage of audiology professionals.



THE GENERAL/SPECIALIST OTOLARYNGOLOGIST What of the general ENT surgeon or the otolaryngologist with a special interest in, say head and neck disease, skull base surgery or rhinology? Can he/she continue to include



Chapter 62 Introduction



children in his/her clinical practice? Emphatically yes. It is essential that the skills of these surgeons are made available as needed for often rare and complex problems in children. Special local arrangements may also be needed for cochlear implantation where otologists with a largely adult practice may be best placed to establish a service with a mixed adult and children base. Rhinologists with an adult base may be best to deal on an ad hoc basis with some of the complex nasal problems that occur in children. Head and neck problems in children are very different to the range of disease seen in adults. There is much to be said for developing the expertise of a very small number of otolaryngologists who take a particular interest in these challenging cases rather than regarding them as within the ambit of all adult head and neck oncologists.



TRAINING THE OTOLARYNGOLOGISTS OF THE FUTURE The fortunes of paediatric otolaryngology are very much in the ascendancy. The specialty is particularly popular with trainees, assuring us a bright future. Training programmes in ORL must include sufficient exposure to the diagnosis and management of ENT disease in children to enable otolaryngologists to continue with a broadbased practice that includes the management of children. The specialist paediatric otolaryngologists of the future will need more advanced training. Fellowship programmes are well established in the United States and represent a model that could usefully be adopted elsewhere. Otolaryngologists – paediatric or not – need to act as feisty advocates for children. It is easy for paediatric aspects of services to be marginalized and for children to be disadvantaged, particularly when ENT services or training programmes are reconfigured. Continued vigilance is essential.



Deficiencies in current knowledge and areas for future research



$ $



 Paediatric otolaryngology is as old as the specialty of otolaryngology itself.  Looking after children is an integral part of the work of an ENT specialist.  The pathophysiology and natural history of disease may be very different in children.  Some children will require the specialist knowledge and skill of a paediatric otolaryngologist.  Training in the generic skills required to care for children is essential to ORL practice.



There is an increasing trend to centralize children’s surgery. Otolaryngologists must ensure they are to the fore in local service planning. There is a need for careful workforce planning and the establishment of fellowship training programmes to maintain the advances we have made.



ACKNOWLEDGEMENTS I am grateful to Professor George Browning, Editor, Clinical Otolaryngology and Allied Sciences and to Blackwell Publications for permission to use material previously published as an editorial in the journal Clinical Otolaryngology and Allied Sciences April 2005.



REFERENCES







 KEY POINTS



] 775



1. Guthrie D. Diseases of the ear nose and throat in childhood. London: AC Black, 1921. 2. Wilson TG. Diseases of the ear nose and throat in children. London: William Heinemann, 1955. 3. Wilde WR. Practical observations on aural surgery and the nature and treatment of diseases of the ear. London: John Churchill, 1853. 4. Clarke RW. Irish literary otolaryngologists. In: Pirsig W, Willemot J, Weir N (eds). Ear, nose and throat mirrored in medicine and arts. Ostend: Wayenborgh Publishing, 2005: 221–36. 5. Mackenzie M. Diseases of the pharynx, larynx and trachea. A manual of the diseases of the throat and nose. New York: William Wood and Company, 1880. 6. Allen GC, Stool SE. History of paediatric airway management. Otolaryngologic Clinics of North America. 2000; 33: 1–14. A readable and entertaining account of the developments which have made paediatric airway endoscopy the sophisticated procedure it is today. 7. Bluestone CD. Paediatric otolaryngology: past, present, and future. Archives of Otolaryngology – Head and Neck Surgery. 1995; 121: 505–8. 8. Osman EZ, Aneeshkumar MK, Clarke RW. Paediatric otolaryngology services in the UK: a postal questionnaire survey of ENT consultants. Journal of Laryngology and Otology. 2005; 119: 259–63. 9. Clarke RW, Osman E. British paediatric otolaryngology – coming of age. Clinical Otolaryngology and Allied Sciences. 2005; 30: 94–7. 10. Bluestone CD. Humans are born too soon: impact on pediatric otolaryngology. International Journal of Pediatric Otorhinolaryngology. 2005; 69: 1–8.



63 The paediatric consultation RAY CLARKE AND KEN PEARMAN



Introduction Requirements for the clinic Preparation for the consultation The consultation Consent in children



776 777 777 777 780



Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



781 781 781 782



SEARCH STRATEGY This chapter was complied largely from the authors’ personal experience as clinicians and teachers, supplemented by their personal bibliographies. The websites of the Departments of Health (www.doh.gov.uk) and Education and Skills (www.dfes.gov.uk) were consulted for the sections on consent, for information on the National Service Framework for Children and for ‘Every Child Matters’.



INTRODUCTION Major recent developments in UK government policies that relate to children have been informed by the findings of the Kennedy1 and Laming2 inquiries. Their recommendations, together with those of subsequent government publications (the National service framework for children and Every child matters) provide a comprehensive blueprint for best practice in children’s services and are universally applicable. While many of their recommendations have been routinely practised for a long time, they are a useful reference source and what follows is in keeping with their spirit.3, 4 Children and parents often remember vividly their first encounter with a ‘specialist’. For many, it will be the child’s first contact with doctors/clinics and hospitals. It is worthwhile investing time and effort into making the experience as pleasant and productive as possible. The principles that make for a satisfactory and productive visit to the otorhinolaryngology (ORL) clinic apply to both adults and children, but some features of the paediatric consultation are unique.



The decision to seek advice will have been made not by the patient but by the child’s parent(s) or carer. With infants and young children the history and a discussion concerning diagnosis and management is by proxy, i.e. involving the parent/carer (usually the mother) rather than the child. Often the mother will be anxious, perplexed and ‘worried if she is doing the right thing’. While the primary purpose of the visit is for the doctor to make a diagnosis and advise on an investigative/treatment plan, a good consultation is far more than that. It represents an opportunity to establish a rapport with the parent and child that in some cases may last for much of the patient’s childhood. It is the forum for explaining, in simple and understandable terms, what the doctor’s opinion is on the child’s condition and for discussion of treatment options. It provides the chance to familiarize the mother and child with the hospital/clinic, to introduce them to other members of staff who may be involved in the child’s care and to reassure them that the child will be well looked after should they need to come into hospital. Most hospitals demand that a trained children’s nurse be available for all paediatric consultations. Specialist



Chapter 63 The paediatric consultation



nurses who work in paediatric ORL are invaluable. Staff numbers need to be sufficient not only to support the working of the clinic, but also to ensure the safe supervision of patients and their siblings while parents are preoccupied. Thus paediatric clinics need more nurse input than general ENT clinics.



REQUIREMENTS FOR THE CLINIC The waiting area A paradox of planning hospitals and clinics for children is that despite children’s small size their requirements for space are much greater than those of adults. In an ideal world, children would only be seen in dedicated clinics designed with their unique requirements in mind but this is not always possible. Whatever the setting, seating has to be comfortable and suitable for all ages. Wheelchair access is essential as are facilities for breastfeeding and for changing of babies’ clothes. Children become bored and fractious if they wait too long. As well as organizing clinic appointments to minimize waiting, a well-equipped waiting area with toys, paper, coloured pencils and computer games will help to keep children occupied and reduce parental stress. Play therapists greatly enhance the quality of a child’s hospital visit. Some progressive hospitals employ clowns and entertainers. Adolescents may feel uncomfortable surrounded by hordes of small children and need to have their particular needs catered for as well. Some hospitals have separate clinics for adolescents scheduled for after school times and they usually appreciate a separate ward or section of a ward if they require inpatient care.



‘Special’ clinics There is much to be said for running separate clinics for some categories of patients. These may include ‘special needs’ clinics and clinics where multidisciplinary input is required, for example audiology, cleft palate, etc. Thought needs to be given to the organization of time and space in such clinics because a balance needs to be struck between making the most efficient use of time by all concerned and the need not to overtire or overwhelm the child by seeing too many adults in one room at one time.



The clinic rooms An examining room should be able to accommodate not just the patient, doctor and nurse but two parents, one or more siblings, sometimes in ‘Moses baskets’ or pushchairs, and often a grandparent. In the case of special needs children, there may be a need to accommodate a wheelchair, oxygen cylinders and the various bits and



] 777



pieces the mother needs for tracheotomy care and gastrostomy management. Ideally, the examining room should have a small play area as well where the child and siblings can occupy themselves while the mother gives the history and the doctor can quietly observe the child. Instruments other than those in frequent use should be discreetly put away. Small children will be frightened at the sight of an array of picks and hooks. Facilities for hand washing are essential. An operating microscope is essential, either in the room or nearby. Endoscopy – both flexible and rigid – is now so frequently performed in an outpatient setting in children that it can be regarded as a mandatory requirement in a paediatric clinic. A suitable range of scopes with facilities for safe storage and ideally a monitor and image capture system should be available, not just for specialized airway or voice clinics but as a routine requirement for any paediatric consultation. The physical environment must be safe for the child, with no spirit lamps, sharp instruments or corners.



PREPARATION FOR THE CONSULTATION A hospital visit is a routine event in the life of the doctor, but a major event for the mother and child. Arrangements may have been made for sibling childcare, the mother will have had to discuss the visit with the child’s teacher and often one or both parents will have taken a day off work. Parents or children must never feel rushed. If you need to hurry them along, the clinic has not been properly planned. It makes for a smoother consultation if the doctor has taken the trouble to read the case notes and learn the child’s name before the consultation starts. If the child has a syndrome or an unusual medical condition read up about it in advance. Make sure you have checked any investigations the child may have had, including hearing tests. As far as possible, children and parents appreciate continuity and like to be able to see the same doctor on successive visits. It is not the authors’ place to advise on dress code, suffice to say that your best tailored suit and crisp, cotton dress shirt will neither impress the average eight year old or continue to look crisp at the end of a busy morning!



THE CONSULTATION Welcome the child and carer(s), make eye contact and start by introducing yourself. Introductions should include others in the room and include permission for medical/nursing students to be present and for them to examine the child. Multidisciplinary clinics can be particularly intimidating as by their very nature several specialists are present. It is good practice for all to wear a name badge and many hospitals will demand this. This does not excuse the need to introduce yourself and other members of the team. Learn and use the child’s name.



778 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Establish who is with the child and if not the parents note names, for example of foster carers or grandparents. A well-conducted consultation where the mother/carer feels she has been listened to and had the child’s symptoms properly and simply explained is an important part of the management of many childhood ailments. Many doctors regard themselves as excellent communicators because of their capacity to talk well. Fewer listen well! Listen, watch the child, look at the mother’s facial expressions, pick up as many nuances as you can to help you in your decision making and to improve your rapport with mother and child. Resist the temptation to conduct the entire consultation without involving the child directly in the discussion. Listen and respond to both the child and the family and encourage the child to give her own views. Good consultation skills can be taught, learned and improved upon. Parents may be angry, upset, uncooperative, seeming not to listen, evasive or challenging in a variety of other ways, but it goes without saying that as the child’s doctor you must never become shorttempered no matter how difficult the exchange or how great you feel the provocation is. If interpreters are needed, it is best to make arrangements in advance. Pay attention to the seating plan, especially with signing interpreters. Professional interpreters are preferable to relatives and friends, and using elder siblings to interpret is undesirable. Speak slowly and clearly to the patient’s mother at normal volume and use plain English without jargon or colloquial terms.5



The history Doctors are taught to take a history in a structured didactic way. This is not always suited to paediatric work. It is best to start with an open question, such as ‘what are your worries about Jennifer?’ and let the mother explain in her own words, without interruption until she feels she has got her message across. More direct questions can then be put. Always ask about the impact of symptoms on the child herself and on family life. If parents volunteer to show you the child’s ‘health book’ or growth charts, make sure you look at these even if you feel they are not directly relevant. Parents understandably feel any record of their child’s health and progress must be important.



The examination Most children are happy to be examined (Figure 63.1). An astute clinician will have commenced the process as soon as the child enters the room. Stridor, mouth breathing, syndromic features and the child’s general alertness may give clues. Smaller children are best examined sitting on the mother’s knee. Tell the child in an age-appropriate way what you are going to do at all stages of the examination – do not assume they have been listening and expect it. Children do not need to be



Figure 63.1 Examining a young child.



restrained for clinical examination and if the child is not cooperative it is best not to press or coerce them. When potentially uncomfortable procedures such as the removal of wax or foreign bodies are to be performed, it is important to be truthful and not to tell the child it will not hurt if it may do. If you do hurt the child explain and, where appropriate, apologise.



Otoscopy Most children will tolerate the gentle introduction of an electric auriscope. Some units use video-otoscopy which has the advantage of enabling the mother and child to see an image of the eardrum on a monitor and also permits pictorial record keeping.



The nose Children do not like the Thudicum’s speculum. A good view of the nasal cavities can be obtained by elevating the tip of the nose or by using an otoscope with a large speculum (Figures 63.2 and 63.3). Rigid endoscopy may be tolerated by the very young, where a fine-calibre telescope will be needed, and in older children who will cooperate if the procedure is explained to them. Local anaesthetic sprays often make matters worse and if a child finds the procedure uncomfortable it is best to just abandon it. The nasal airway can be assessed by gently occluding one nostril at a time and asking the child to breathe gently. The spatula test in which a cold tongue depressor is placed under the nostrils and the condensation pattern inspected to determine whether there is occlusion of the nasal airway is especially helpful in very young children (Figure 63.4).



The pharynx and larynx Try to obtain a view without using a tongue depressor. Many young children will tolerate mirror examination of



Chapter 63 The paediatric consultation



] 779



Figure 63.2 Anterior rhinoscopy.



Figure 63.4 The spatula test.



Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma and Chapter 97, Salivary gland disorders in childhood. Figure 63.3



Using an aural speculum for nasal examination.



Investigations the post-nasal space, but flexible endoscopy is increasingly used and often well tolerated. The adenoids, hypopharynx and larynx can be seen in this way. The function of the velopharyngeal sphincter can also be assessed and recorded. Some children will tolerate mirror examination but often the view is unsatisfactory and, if the child will permit examination via the mouth a 701 rod lens telescope is very useful as its slim shaft fits easily between large tonsils.



The neck It is important to inspect and palpate the neck, but remember that enlarged lymph nodes in children are so common as to be normal. Other swellings in the neck include cystic hygroma, thryoglossal cysts and salivary gland swellings. These are dealt with in Chapter 99,



Other than audiometry and tympanometry, few investigations are needed for the common presentations at a children’s ENT clinic. Hearing tests are considered in Chapter 68, Hearing tests in children.



Management plan The parents have come to see you to hear your opinion on their child’s condition and to discuss a management plan with you. This part of the consultation is vital and must never be rushed. Explanations should be straightforward and easy to understand, involving the child where appropriate. Models, diagrams, printed and audiovisual material can aid this process greatly. For many interventions, such as tonsillectomy, adenoidectomy and insertion of tympanostomy tubes, there will be more than one management option and it is important that these are



780 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY discussed in an open and honest way. This sometimes means doctors have to admit doubts and uncertainties which are best explained without embarrassment so that a way forward can be agreed by consensus. The treatment will be a matter for negotiation between the otolaryngologist and the mother and/or child and questions should be encouraged. Some parents see the ORL consultation as a means to confirm a treatment option they have already decided on with their family doctor, for example tonsillectomy for recurrent sore throats. Others are reluctant to consider any surgery, but all will greatly appreciate an informed discussion and an honest presentation of the evidence for current practice focused on the specific needs of their child. In general, parents appreciate a discussion with the senior clinician, but this needs to be balanced with the need for residents and juniors to see patients and improve their diagnostic and consultation skills.



Functional problems Problems for which no organic cause can be found are encountered in paediatric ENT practice, particularly in adolescents. They often take the form of unexplained coughs, pains or abnormal hearing tests. It is important to investigate for underlying organic pathology. Where none is found, the patient should not be led to feel that the clinician believes that the problem is entirely imaginary because this breaks down confidence. An approach in which the problem is treated as a disorder of body function rather than mental function is more likely to be helpful.



Breaking bad news If bad news has to be given, such as a diagnosis of deafness, the need for long-term tracheotomy or suspicion of malignancy, the help of a senior clinician should be sought immediately. Parents and children remember such episodes with chilling clarity – often for a lifetime. Great harm can be done by insensitive or even wellmeaning but inexperienced handling of such situations. Many hospitals have policies on the breaking of bad news. In the clinical situation, it is not always possible to follow these blueprints to the letter, but every consideration should be shown to the family. The setting in which the news is to be broken should be considered beforehand and appropriate supporting staff briefed. The consultation must not be rushed. Parents will often be shellshocked by unexpected bad news and may not remember detailed information. Further discussion should be offered if needed and it is helpful to make a clear management plan beforehand, so that there is no uncertainty about the next steps to be taken or the timing of the next appointment.6



Prescribing for children In general, otolaryngologists do not prescribe a large range of drugs in the outpatient setting. Where drugs are prescribed, it is essential to be aware of the correct dosage and presentation for children and that the child’s mother understands how to give the medication. Evidence-based prescribing is not always possible for children, and where unlicensed and off-label medicines are used, local safety standards should be adhered to.



Recording and communication of the outcome of the consultation Hand-written notes should be legible, signed, dated and clearly identify the writer. Subject to the need for confidentiality, there is often a need to inform more people of the outcome of the consultation in paediatric work than there is with adults (e.g. educators, community paediatricians, geneticists, etc.). Parents should be made aware of this and correspondence should be copied to all appropriate parties including the parents/carers themselves.



Child protection issues All professionals involved in paediatric work need to be aware of the potential for children to be subject to abuse or neglect. Of those children subject to physical abuse some 75 percent suffer injuries to the head and neck.7, 8 Examples of presentations that may arouse suspicion include auricular haematomas, traumatic perforations of the eardrum, nasal injuries, pharyngeal lacerations, dental trauma and tears of the lingual frenulum. Children are naturally active and some of these injuries occur in the rough and tumble of normal childhood but if there is a pattern of injury or if the presentation seems inappropriate clinicians have not only an ethical but a legal duty to raise concern. Needless to say, such concerns must be handled with great sensitivity, but prompt action should be taken and comprehensive notes, detailing the injuries and the reasons for suspicion, should be made. All hospitals which look after children should have an agreed mechanism for dealing with such concerns, usually involving the help of a senior and experienced paediatrician.2 Clinicians who have little direct experience of child protection will need the advice and support of a more experienced colleague.



CONSENT IN CHILDREN Competence It is axiomatic that any interventions proposed in children require consent. While the law has primacy this will vary according to the administration or health service



Chapter 63 The paediatric consultation



that one is working in and it is imperative that doctors who care for children are thoroughly conversant with the law relating to consent and parental responsibility. Whatever the legal requirements it is important to respect the legitimate concerns of parents, be they mothers, fathers, single, married or divorced, to be cognisant of cultural and individual family dynamics and to proceed by consensus wherever possible. What is unique about consent in paediatric practice is that young children will not have the understanding required to give informed consent, although it is good practice at all times to involve the child as much as possible in decision making. The following is based on English law and on custom and practice in the UK. Once children have reached the age of 16 years they are deemed to be competent and can therefore give consent themselves. Nevertheless, it is wise to encourage teenagers to involve their families in decision making. ‘Competence’ implies the capacity to comprehend and retain information material to a decision, and the ability to weigh this information in decision making. Some 16- and 17-yearolds may not be competent by reason of unconsciousness, pain or severe learning disability. In such cases, a person with parental responsibility may take a decision for the child, but over the age of 18 nobody can give consent for another person, and if such a person is not competent clinicians may have to take decisions without consent that are in the best interests of the individual. Some children under the age of 16 will be able to understand what an intervention involves. The courts have determined that such children are competent if they have ‘sufficient understanding and intelligence to enable him or her to understand fully what is proposed’.9 This is colloquially known as ‘Gillick competence’. The determination as to whether the child has Gillick competence rests with the clinician.10



Parental responsibility If a child is not competent, only a person exercising parental responsibility can give valid consent for that child. This is usually one or both of the child’s parents. In England and Wales ‘parental responsibility’ is as defined by the Children’s Act 1989 and amended by the Adoption and Children Act 2002. The child’s natural mother has parental responsibility as has the father if he is married to the mother or was married to the mother at the time of the child’s birth. Unmarried fathers may acquire parental responsibility by order of a court or by a ‘parental responsibility agreement’. For births registered after December 2003, a father named on the birth certificate has parental responsibility irrespective of his marital status. Grandparents and others who look after children cannot give consent unless special legal arrangements have been made. Consent is only legally required from one parent, but where applicable it is good practice to



] 781



ensure that both parents and the child are involved in the decision-making process. Consent in children is an issue that causes great sensitivities. The reader is referred to the Department of Health guidelines (www.doh.gov.uk/consent) and trainee surgeons in particular are advised to seek the advice of senior colleagues in the event of any uncertainties.



KEY POINTS  A clinic visit is a routine event for the doctor. It may be a major episode in the life of the parent and child.  The organization of children’s clinics require more space, more preparation and more attention to detail than is the case for adult clinics.  Everyone can learn better consultation skills.  Consent is a sensitive issue in children.  Clinicians need to be familiar with the law in the jurisdiction in which they work and with best practice.



Best clinical practice [ Prepare for the consultation before the child comes into the examining room.



[ Always introduce yourself to the parent and child. [ Explain the current knowledge base for every proposed intervention and listen to the views of the parent/child. [ A suitable range of flexible and rigid endoscopes, with facilities for safe storage and ideally a monitor and image capture system should be available, not just for specialized airway or voice clinics, but as a routine requirement for any paediatric consultation. [ It is important to know the procedures to be followed in the event of concerns about child protection, and the law on consent for the jurisdiction in which you work.



Deficiencies in current knowledge and areas for future research



$



Parents are becoming more and more involved in decision making for their children in health care. The old patriarchal model of medical practice is fast disappearing. Doctors will need to discuss the natural history of diseases, treatment options and the uncertainties around evidence for existing practices in much greater detail than before.



782 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



REFERENCES







1. Kennedy I. Learning from Bristol: the report of the public inquiry into children’s heart surgery at the Bristol Royal Infirmary 1984–1995. Command Paper: CM 5207, 2001. Available from: www.bristol-inquiry.org.uk. A far-reaching set of recommendations to guide clinicians and health care managers regarding the provision of clinical services for children. 2. Laming H. The Victoria Climbie´ Inquiry; Report of an Inquiry by Lord Laming. Command Paper CM5730, 2003. Available from: www.victoria-climbie-inquiry.org.uk. 3. British Red Cross, with the advice and funding from the Department of Health. Emergency multilingual phrasebook. Rochester, UK: British Red Cross, 2004. 4. National Service Framework for Children, Young People and Maternity Services. Available from: www.doh.gov.uk.











5. Department for Education and Skills. Every child matters. London: The Stationery Office, 2003. Available from: www.dfes.gov.uk/everychildmatters. 6. Faulkner A. When the news is bad. A guide for health professionals. Cheltenham: Nelson Thornes, 1998. Excellent handbook with guidance applicable to both adults and children. 7. Crouse CD, Faust A. Child abuse and the otolaryngologist: Part 1. Otolaryngology Head and Neck Surgery. 2003; 128: 305–11. 8. Crouse CD, Faust A. Child abuse and the otolaryngologist: Part 2. Otolaryngology Head and Neck Surgery. 2003; 128: 311–8. 9. Gillick v. Norfolk and Wisbech AHA [1985] 3 WLR 830, 3 All ER 402. 10. DOH website on consent. Available from: www.doh.gov.uk. Required reading for all UK practitioners.



64 ENT input for children with special needs FRANCIS LANNIGAN



Introduction Clinical environment Communication and rapport General considerations The ear and hearing The upper aerodigestive tract



783 783 784 785 785 786



Drooling Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



788 789 789 790 790



SEARCH STRATEGY The opinions in this chapter are supported by a Medline search using the key words cerebral palsy, mental retardation, handicap, special needs, Down syndrome, hearing loss, hearing aids, obstructive sleep apnoea/disorder, adenoidectomy/ tonsillectomy, tracheostomy, drooling, sialorrhoea.



INTRODUCTION



Terminology



Developments in neonatology and improved paediatric intensive care management are such that more infants with extreme prematurity or multiple organ pathologies now survive.1 Similarly, modern treatments of many genetic and congenital diseases have improved outcomes. A proportion of such infants and children will have multiple medical conditions. These include neuromuscular dysfunction, sensory impairment including deafness, communication disorders and developmental delay. Epilepsy and learning disability are also more common in this group of children.1 The prospect of managing a child with special needs may appear daunting, and is often associated with disappointing outcomes. However, in the appropriate clinical setting and with the intent of attaining maximal objective evidence to support a proposed intervention, the disappointment will be minimized. Down syndrome (trisomy 21) raises many of the issues common to managing ENT problems in children with special needs (Figure 64.1 and Table 64.1).



The Oxford English dictionary defines a ‘handicap’ as a condition that markedly restricts a person’s ability to function physically, mentally or socially. A ‘disability’ is a physical or mental condition that limits a person’s movements, senses or activities. Some parents and children in the UK in particular dislike the terms handicap and disability as they feel these emphasize negative aspects of the child’s needs and potential. The term ‘special needs’ is now more commonly accepted but much of the literature still relates to children with disabilities or multiple handicaps and these terms are widely accepted internationally.



CLINICAL ENVIRONMENT An increasing proportion of clinical time is spent by the paediatric otolaryngologist–head and neck surgeon as part of a multidisciplinary team in the provision of



784 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY treatment to children with special needs, usually in a tertiary setting. It is neither possible nor desirable to adequately assess a child with multiple pathologies in a general paediatric ENT clinic. These children often have clinical records that extend to several volumes, which require time for meaningful review. It is, therefore, desirable to set aside a specific clinic for evaluating children with special needs. Needless to say this clinic should be conducted by the consultant or senior attending physician. Even with the best clinic administration, some children will ‘slip through the net’ into a general clinic. It is the author’s



practice to explain such an error to the parents/carer and, if appropriate, reschedule the planned consultation. It goes without saying that wheelchair access, facilities for oxygen delivery, and seats for parents/carers and siblings are mandatory. Clinical examination may be especially challenging; a skilled paediatric nurse is invaluable.



COMMUNICATION AND RAPPORT The parents/carers of these children are often highly discerning and probe carefully the merits and risks of any proposed medical or surgical intervention. On occasion, more time is required to explain why an accepted surgical intervention may not be appropriate for a child than would be spent obtaining informed consent for that same procedure. The carers and children have multiple interactions with health care staff and will often develop a warm rapport with clinicians. Many will be familiar with the details and expectations of often-rare syndromes and will be affiliated with parent/carer groups that provide information and enable parents to discuss issues common to particular conditions with each other. In Britain, a joint project between parents and carers and the Royal College of Paediatrics and Child Health – ‘Contact a Family’ – maintains a website (www.cafamily.org.uk) which provides information and support for families of children with a variety of specific conditions.



Planned surgery



Figure 64.1



Table 64.1



It should be remembered that the child with multiple pathologies will usually represent a greater anaesthetic and surgical risk with increased postoperative morbidity. They require the skills of a paediatric anaesthesiologist and may be unsuitable for same-day hospital care – overnight stay or even longer may be required predominantly due to delay in re-establishment of adequate oral intake and increased respiratory complications of



Child with Down syndrome.



ENT problems in Down syndrome.



General Frequent upper respiratory tract infections (URTI) Thyroid disease Sialorrhoea Atlantoaxial instability (manipulate the head with extreme care under anaesthesia)



The airway



Ear disorders



Airway obstruction Narrow subglottis and trachea – may need smaller endotracheal tube during anaesthesia Midfacial hypoplasia – increased incidence of obstructive sleep apnoea syndrome (OSAS) Hypotonic pharyngeal airway Macroglossia



Deafness Frequent acute otitis media and otitis media with effusion Small pinna and external meatus Premature ossification of the cochlea Facial nerve dehiscence common Ossicular fixation Mondini malformation more common



Chapter 64 ENT input for children with special needs



anaesthesia.2 Some will need paediatric intensive care facilities postoperatively and this will need to be planned in advance. Often they require multiple surgical input, e.g. dental treatment, ear surgery and attention to a gastrostomy, and it may be appropriate to coordinate any planned ENT intervention with other treatments/ disciplines. What are routine investigations in the otherwise healthy child may require general anaesthesia in the child with multiple pathologies.



GENERAL CONSIDERATIONS Common childhood ENT disorders may significantly interfere with the normal development of any child and their impact may be significant. In the child with special needs these common conditions may have a disproportionate impact on development and family life. Conversely, apparently minor clinical gains may have dramatic impacts for the child and their parents/carers, e.g. reduced nocturnal seizures following treatment of obstructive sleep disorder or improved hearing following grommet insertion. In the child with special needs the maxim remains: maximize input in order to facilitate development to maximum potential. [**/*]



is often uncertain. There are some crude indicators: the necessity for nasogastric tube feeding is a significant adverse prognostic indicator, interestingly unaffected by conversion to gastrostomy feeding. Paradoxically, the addition of a tracheostomy (despite its inherent mortality and morbidity) in the presence of nasogastric/gastrostomy feeding improves outcome, presumably due to improved access for bronchial toilet.3 These indicators obviously only relate to the most severe end of the spectrum; however, such information may facilitate difficult decisions for parents/carers. Otherwise there is little evidence base to support interventions, other than that available for such interventions in otherwise healthy children. The prognosis of many childhood medical conditions, for example cystic fibrosis, metabolic diseases, renal failure and congenital cardiac lesions, has been greatly improved by active intervention and the involvement of multidisciplinary teams.1 [****/**] It has been demonstrated in other areas that timely intervention may in the long-term reduce expenditure (cf. Chapter 70, Paediatric cochlear implantation). Whilst outcome measures are difficult to assess, there is now evidence that multidisciplinary intervention in children with Down syndrome improves intermediate/long-term hearing outcomes.4 [**/*]



Ethical issues



THE EAR AND HEARING



It is necessary to address some difficult ethical issues when treating children with special needs. These children will naturally make greater demands on the health-care system, irrespective of the quality of care. ‘There is no point in treating children like this.’ ‘The resources spent treating children like this would be better spent elsewhere.’ These are common comments that are often heard in relation to the management of such children. It goes without saying that such opinion must never be communicated in any way to the patient or the parent/ carer. A full discussion of the ethical and health economic issues is beyond the scope of this chapter; nevertheless, some practical guidelines are required. It is probably better if health-care professionals who hold negative opinions regarding intervention, irrespective of their role, excuse themselves from the management of these children; to do otherwise is inviting problems.



Assessment



Prognosis The available evidence related to the care of these children is predominantly in the form of retrospective case series. [**/*] Other than for some specific conditions, prognosis



] 785



Evaluation of the auditory pathway in the child with multiple pathologies may prove challenging, but is essential in overall management. Many referrals of these children will be from developmental paediatricians who wish to exclude significant hearing loss from a child’s spectrum of disability (and also as part of a screening protocol for ‘at risk’ infants). Where access for audiological assessment is provided without direct assessment by a paediatric otolaryngologist–head and neck surgeon, systems must be in place to initiate appropriate clinical assessment and habilitation/rehabilitation of an identified hearing loss. Conventional age-appropriate audiological evaluation is performed with the understanding that results may not be accurate and that general anaesthesia with objective audiological evaluation is more likely to be required. Under general anaesthesia, any external or middle ear component of the problem should be minimized simultaneously to provide maximal accuracy of obtained thresholds (e.g. ear toilet, myringotomy plus/ minus grommet insertion). Little controversy exists when it comes to obtaining accurate hearing thresholds, and apportioning conductive/sensorineural components to hearing loss. The same is not true when it comes to management.



786 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Management Concern has been raised with respect to the use of ventilation tubes in children with special needs, in particular Down syndrome. Some reports have documented a higher rate of morbidity following ventilation tubes in these children, notably otorrhoea and residual perforation.5 Overall, however, the majority of children in these reports still benefit from intervention. [**] The narrow external meatus and the often difficult anatomy of the ear canal associated with Down syndrome make any transcanal middle ear surgery technically difficult (if not impossible); the narrow, and often hairy, meatus also contributes to failure of clearance of skin debris and wax. The proposed alternative is to use amplification for the management of hearing loss; this may prove to be a very acceptable alternative for some children. Often children with Down syndrome develop sensorineural loss in the teenage years and if amplification is anticipated this may be an argument for considering it earlier. Amplification is not trouble-free. The presence of a hearing aid mould in a congenitally narrow and hairy external meatus may predispose to wax impaction and recurrent otitis externa – ear toilet will often require general anaesthesia in such children. Amplification may not, therefore, have the desired effect of avoiding otorrhoea and/or intervention under general anaesthesia. A rarely mentioned issue with respect to amplification is the small, but potentially disastrous, risk of ingestion of hearing aid batteries by the wearer or a sibling; this should be communicated to parents/carers.6 The judgement is individual to each child and will involve full discussion of the advantages and disadvantages with the parent/carer and child if he/she is old enough to participate in the decision. [*] If amplification is the chosen modality for treatment, then regular observation is required for the early detection of the development of retraction changes/ atelectasis, which should prompt reconsideration of surgical intervention.5



Cholesteatoma Childhood cholesteatoma is a miserable condition for the patient, the parent/carer and all but the most enthusiastic otologist; this is even worse in the child with special needs. Not least of the expected problems encountered when cholesteatoma occurs is the unusual temporal bone anatomy associated with some syndromes, for example Down syndrome.7 Computed tomography (CT) scanning is essential to anticipate anatomical anomalies and facilitate planning surgical management – a conventional post-aural approach is ill-advised if there is a sclerotic mastoid, with middle fossa dura at the level of the second genu of the facial nerve. Peroperative facial nerve monitoring is especially valuable in this group. [*] In



view of the difficulties associated with cholesteatoma, it is the author’s practice to treat retraction pockets/atelectasis aggressively in an incremental manner: from ventilation tubes and management of nasal conditions, to excision and reinforcement grafting. [*]



Otalgia Recurrent otalgia, particularly at night, is a common feature of otitis media. This symptom may result in sleep deprivation for the child and family. It may also cause daytime behaviour problems, e.g. head-banging. Prolonged conservative management will not alleviate this situation which will often have been present for some time before otherwise asymptomatic middle ear disease is identified. Surgical intervention in the form of tympanostomy tube insertion may have dramatic benefit in such a circumstance. [**/*]



THE UPPER AERODIGESTIVE TRACT Issues relating to feeding problems, drooling, stridor, nocturnal airway obstruction, dysphonia, aspiration and long-term airway access, may all present either as a single entity or in any combination.



Feeding problems Most tertiary paediatric institutions have ‘feeding teams’. These teams comprise paediatric gastroenterologists, paediatric neurologists, speech and swallowing therapists, dieticians and psychologists. They will manage issues related to feeding from the time of discharge from the neonatal unit through to adolescence. The paediatric otolaryngologist–head and neck surgeon will therefore usually only become involved with purely feeding issues when a surgical problem has already been identified or when feeding problems are associated with aspiration requiring tracheostomy for bronchial toilet. Great care must be taken when explaining to a parent/carer the postoperative management required when tracheostomy is performed for this reason: continued aspiration of oropharyngeal secretions is likely to occur and increase demands for tracheostomy care. Some parents/carers may, on reflection, decline the option of tracheostomy. In this instance, an upper endoscopy should be recommended in order to exclude any other structural or functional abnormality (other than neuromuscular incoordination) contributing to aspiration. If not already performed, a thorough upper endoscopy is mandatory prior to tracheostomy for the same reasons; particular care must be given to the exclusion of an occult laryngeal cleft by the use of an appropriate laryngeal probe for palpation of the posterior glottis and cricoid cartilage (Figure 64.2).



Chapter 64 ENT input for children with special needs



Figure 64.2 Blunt laryngeal probe and measuring device (Lowinger laryngeal probe; Xomed-Medtronic) – useful in older infants and children; not readily manipulated in smaller laryngoscopes where a fine blunt suction probe will suffice.



Often by the time tracheostomy is performed, some degree of reduced respiratory reserve is present; however, once past the anaesthetic and perioperative period, there is often apparent improvement in respiratory function due to reduced dead space and reduced airway resistance. Whilst this apparent improvement remains it may provide considerable respite for the parent/carer.3



Laryngotracheal disease Dysphonia and stridor will often be wrongly attributed to neuromuscular problems intrinsic to the child’s overall condition. It is important that the paediatric otolaryngologist–head and neck surgeon should not perpetuate such assumptions; thorough upper endoscopy is required as would be the case for otherwise healthy children. It must be emphasized, however, that any concomitant neuromuscular incoordination will adversely impact upon therapeutic interventions – it is always difficult to be precise about the degree of this adverse impact, and not just because of the varying degree of neuromuscular incoordination. For example, correction of a subglottic stenosis may have a disappointing outcome due to arytenoid prolapse secondary to neuromuscular incoordination; changing the dynamics of the upper airway may have a very unpredictable outcome. Nowhere is this principle better illustrated than in the management of snoring and sleep-related airway obstruction.



Obstructive sleep apnoea The simplistic notion that snoring equates to adenotonsillar hypertrophy and that each child may be safely treated with adenotonsillectomy will result in disappointment more commonly in the child with multiple pathologies than in the general paediatric population.8 Adenotonsillectomy alone fails to reduce the Respiratory Disturbance Index below 5 in approximately 15 percent of children undergoing surgery for a presumed obstructive sleep disorder;9 and it is into this group that the child with multiple pathologies is more likely to fall.8 Airway noise is



] 787



common in these children and is often a respiratory monitor for parents/carers! It is always worth directly enquiring about features of nocturnal airway obstruction, but the usual clinical features (which are notoriously unreliable in the healthy child9) may be even more difficult to interpret in the child with a complex medical history. Initial examination should include consideration of dysmorphic features and neuromuscular coordination in addition to the more usual clinical examination. Simple measures to eradicate any bacterial rhinitis or to institute treatment for atopy should be used as appropriate. The myriad of potential causes of sleep disruption, some of which may coexist with an element of an obstructive sleep disorder, makes clinical judgement suspect. Formal polysomnography may prove extremely helpful in the selection of children who are likely to benefit from intervention; it may also identify other causes for sleep disruption.10 It is common practice for a parent/carer to board with their child for this investigation; the validity of the investigation may be assessed by asking the parent/ carer if the night of the sleep study was a typical night for their child. If polysomnography demonstrates significant obstruction, then the next step is to determine the level of that obstruction. Once again it is important to emphasize that whilst adenotonsillar hypertrophy may be significant, and indeed adenotonsillectomy is likely to be the intervention of first choice, other causes of airway obstruction are more likely to coexist.11 Sleep endoscopy has been suggested to facilitate localization of the level of obstruction, which is more likely to be multiple in this type of patient. This procedure requires experience to be of value, but may demonstrate unexpected levels of airway compromise; and may be of particular importance following failed earlier intervention.12 Syndromic children are more likely to require tracheostomy in the management of an obstructive sleep disorder, for example, Crouzon syndrome (Figures 64.3 and 64.4). However, it may be possible to avoid tracheostomy by management of other levels of obstruction. Modified conservative uvulopalatopharngyoplasty (UPPP),8, 13 tongue base reduction, hyoid suspension10 and laser aryepiglottoplasty14 may have a role depending on the findings at sleep endoscopy. [**/*] Using a stepwise incremental targeted procedure has been reported to avoid tracheostomy in up to 80 percent of children with cerebral palsy.10 Great caution must be taken with any procedure that has the potential to impact upon velopharyngeal function; there is a much higher risk of postoperative velopharyngeal insufficiency in this group of children.15 Partial (or superior) adenoidectomy using an endoscopic-powered debrider may minimize this risk.16, 17 [**/*]



The multidisciplinary sleep clinic It is the author’s practice to coordinate each stepped intervention with the paediatric sleep physician; partial



788 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY surgical successes may permit the use of positive airway support devices that had previously been unsuccessful. It is the author’s experience that an increasing number of such children require multimodality therapy to adequately control nocturnal airway obstruction without tracheostomy. Many special needs children have sleep disorders not due to airway obstruction. These may have a profound impact on quality of life for families. Optimum treatment of these disorders involves an interested paediatrician and a multidisciplinary team.10



DROOLING



Figure 64.3 Obstructive sleep disorder in a child with Crouzon syndrome treated with tracheostomy.



Drooling, or sialorrhoea, is physiological in the early part of life. It will occur in relation to teething, oropharyngeal ulceration and upper respiratory tract infections/upper aerodigestive tract infections. Physiological drooling usually settles around the age of two years when the primary dentition has erupted; however, it may persist on a minor basis in some children up to the age of four. This physiological occurrence is minor by comparison to the sialorrhoea that may occur with cerebral palsy and other forms of severe neurological impairment. Approximately 10 percent of children with cerebral palsy will have problematic drooling.18 In order to determine this it is necessary to question the parent/carer regarding their management of the drooling – how many times are changes of clothes required per day; are ‘quilted’ bibs required for management? Some carers will complain that the school-going child soils books and papers and in older children there is the added difficulty of peer acceptance. If the condition has little impact on clothing management, then the drooling is minimal and simple reassurance will usually suffice. The resourcefulness, and tolerance, of parents/carers in dealing with this problem never ceases to amaze the author! Excessive drooling will macerate skin and therefore give rise to dermatological problems around the mouth and along the line that gravity carries it on an individual child. It can be a miserable problem for both parents and the child. General examination should include consideration of posture as well as a full otolaryngology evaluation. Particular consideration should be given to evaluation of nasal obstruction, and the presence of oropharyngeal ulceration (including gingivitis), which may be overlooked or considered noncontributory in the presence of severe neurological impairment.



Treatment



Figure 64.4 Successful decannulation of child in Figure 64.3 following midfacial advancement.



Nonsurgical treatments are often disappointing, and have usually failed by the time a child arrives at the paediatric otolaryngologist–head and neck surgeon. Anticholinergic agents to reduce salivary flow are often poorly tolerated



Chapter 64 ENT input for children with special needs



due to nausea, and ocular side effects. Speech and swallowing therapy should also have been tried for a period of six months; occasionally simple posture changes (sometimes using appliances) will improve oral continence.19 Surgical treatments utilize the concepts of: the division of parasympathetic neurological supply to major salivary glands; mechanical obstruction of the salivary ducts; and rediversion of salivary flow.



Neurectomy Division of the anterior branches of the tympanic plexus on the promontory via a relatively simple tympanotomy will reduce, but not eliminate, secretions from the parotid gland. It should be remembered that these branches (usually two) may be covered by a thin layer of bone (more commonly the anterior branch). This necessitates formal identification and division of two nerves in the mucosa overlying the promontory; failing this, a microdrill is used to find the branch(es) which are always superficial in the bone.20 It is also possible to reduce the parasympathetic drive to the submandibular salivary gland with simultaneous division of the chorda tympani nerve; however, this necessitates removal of taste from the anterior two-thirds of the tongue. In a child who already has severe limitation of sensory input, it may be difficult to justify the deliberate removal of one route of input. It can be difficult to obtain informed consent, particularly given the usually disappointing results produced by this intervention.21 In its favour, this technique can be performed bilaterally, with little in the way of postoperative morbidity. The author no longer recommends or performs denervation procedures unless it is the preferred option of the parent/carer. [**/*]



Botulinum toxin Botox has been used by some authors. Direct injection into the parotid and submandibular glands may reduce salivary flow.22 Evidence is currently sparse and controlled trials are needed before this can be recommended for routine clinical use.



] 789



initially swell and subsequently undergo atrophy.23 It is interesting to note that a radionucleotide study of submandibular gland function following duct transposition reported 50 percent of the glands are atrophied and nonfunctioning following surgery.24 In the past 20 years or so, transposition of the submandibular ducts to the tonsillar fossae has become the most popular and reported procedure for drooling. From its initial description, the procedure has been refined to include intraoral excision of the sublingual salivary glands. The sublingual glands drain via the distal portion of the submandibular salivary duct; therefore leaving the sublingual glands in situ may predispose to postoperative ranula formation. Tonsillectomy (if not already performed) is carried out as part of the procedure to facilitate duct transposition. There is often significant postoperative swelling and significant postoperative morbidity may occur. Parental satisfaction is very high with this procedure.19 Long-term follow-up suggests that improvement in drooling is maintained.25 There are reports of increased dental caries following submandibular duct transposition; this is presumed to be due to loss of the salivary ‘puddle’.26, 27 It should be mentioned to parents/carers that there will be the need for diligent dental hygiene and increased risk of dental caries following transposition of the submandibular ducts. [**/*]



KEY POINTS  Children with special needs require extra time for clinical evaluation.  The maxim remains: maximize sensory input in order to maximize achievement of potential.  Apparent small clinical gains may have dramatic benefits for the child and family.  Where possible, maximal information should be obtained to support planned intervention.  Conventional intervention is less likely to produce the anticipated outcome in a child with special needs.



Mechanical diversion or obstruction of salivary flow Overall, mechanical diversion or obstruction of salivary flow from the major glands is reported to produce the best outcomes. Ligation of the submandibular and parotid ducts is reported to be an effective treatment, with surprisingly few sequelae. The procedure is performed intraorally, with postoperative antibiotic prophylaxis. The glands



Best clinical practice [ Rehabilitation of hearing presents special challenges in this group of children and management must be tailored for a particular child. [ Airway obstruction is more likely to be complex and multilevel in syndromic children.



790 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY [ Submandibular duct transposition with excision of the







sublingual glands is the current procedure of choice for excessive drooling. [ Children with Down syndrome present challenges to otolaryngologists and anaesthesiologists and special care is needed when considering surgery in this group.



 Deficiencies in current knowledge and areas for future research



$ $ $



There is an increasing trend to centralize the care of special needs children around a tertiary centre where support services including paediatric anaesthesia are available. Developmental paediatrics is a growing specialty; better classification of developmental disorders and attendant improvements in predicting outcome are already underway. Multidisciplinary involvement – essential to the optimum management of these children – will improve outcome and hopefully the evidence base for treatment.







REFERENCES











1. Colvin M, Maguire W, Fowlie PW. Neurodevelopmental outcomes after pre-term birth. British Medical Journal. 2004; 329: 1390–3. Excellent review of expected outcomes and improved survival of babies from SCBUs. 2. Goldstein NA, Armfield DR, Kingsley LA, Borland LM, Allen GC, Post JC. Postoperative complications after tonsillectomy and adenoidectomy in children with Down syndrome. Archives of Otolaryngology – Head and Neck Surgery. 1998; 124: 171–6. 3. Strauss D, Kastner T, Ashwel S, White J. Tube feeding mortality in children with severe disabilities and mental retardation. Pediatrics. 1997; 99: 358–62. 4. Shott SR, Joseph A, Heithaus D. Hearing loss in children with Down syndrome. International Journal of Pediatric Otorhinolaryngology. 2001; 61: 199–205. 5. Iino Y, Imamura Y, Harigai S, Tanaka Y. Efficacy of tympanostomy tube insertion for otitis media with effusion in children with Down syndrome. International Journal of Pediatric Otorhinolaryngology. 1999; 49: 143–9. 6. Litovitz TL. Battery ingestions: product accessibility and clinical course. Pediatrics. 1985; 75: 469–76. 7. Kanamori G, Witter M, Brown J, Williams-Smith L. Otolaryngologic manifestations of Down syndrome. Otolaryngology Clinics of North America. 2000; 33: 1285–92. An excellent overview of this condition.







8. Seid AB, Martin PJ, Pransky SM, Kearns DB. Surgical therapy of obstructive sleep apnea in children with serve mental insufficiency. Laryngoscope. 1990; 100: 507–10. A good series. 9. Suen JS, Arnold JE, Brooks LJ. Adenotonsillectomy for treatment of obstructive sleep apnea in children. Archives of Otolaryngology – Head and Neck Surgery. 1995; 121: 525–30. 10. Cohen SR, Lefaivre JF, Burstein FD, Simms C, Kattos AV, Scott PH et al. Surgical treatment of obstructive sleep apnea in neurologically compromised patients. Plastic and Reconstructive Surgery. 1997; 99: 638–46. Comprehensive multi-procedure series, with good work-up. 11. Magardino TM, Tom LW. Surgical management of obstructive sleep apnea in children with cerebral palsy. Laryngoscope. 1999; 109: 1611–5. 12. Myatt HM, Beckenham EJ. The use of diagnostic sleep endoscopy in the management of children with complex upper airway obstruction. Clinical Otolaryngology. 2000; 25: 200–8. A good series of an emerging technique. 13. Kosko JR, Derkay CS. Uvulopalatopharyngoplasty: treatment of obstructive sleep apnea in neurologically impaired pediatric patients. International Journal of Pediatric Otorhinolaryngology. 1995; 32: 241–6. 14. Hui Y, Gaffney R, Crysdale WS. Laser aryepiglottoplasty for the treatment of neurasthenic laryngomalacia in cerebral palsy. Annals of Otology, Rhinology and Laryngology. 1995; 104: 432–6. 15. Kavanagh KT, Kahane JC, Kordan B. Risks and benefits of adenotonsillectomy for children with Down syndrome. American Journal of Mental Deficiency. 1986; 91: 22–9. 16. Murray N, Fitzpatrick P, Guarisco JL. Powered partial adenoidectomy. Archives of Otolaryngology – Head and Neck Surgery. 2002; 128: 792–6. 17. Finkelstein Y, Wexler DB, Nachmani A, Ophir D. Endoscopic partial adenoidectomy for children with submucous cleft palate. Cleft Palate and Craniofacial Journal. 2002; 39: 479–86. 18. O’Dwyer TP, Conlon BJ. The surgical management of drooling – a 15 year follow-up. Clinical Otolaryngology. 1997; 22: 284–7. 19. Crysdale WS, Raveh E, McCann C, Roske L, Kotler A. Management of drooling in individuals with neurodisability: a surgical experience. Developmental Medicine and Child Neurology. 2001; 43: 379–83. A large series and clinical experience. 20. Hollinshead WH. The ear. In: Anatomy for surgeons: the head and neck, 3rd edn. Vol 1. London: Harper and Row, 1982: 174–5. 21. Parisier SC, Blitzer A, Binder WJ, Friedman WF, Marovitz WF. Tympanic neurectomy and chorda tympanectomy for the control of drooling. Archives of Otolaryngology. 1978; 104: 273–7.



Chapter 64 ENT input for children with special needs 22. Cordivari C, Misra VP, Catania S, Lee AJ. New therapeutic indications for botulinus toxins. Movement Disorders. 2004; 19: S157–61. 23. Klem C, Mair EA. Four-duct ligation; a simple and effective treatment for chronic aspiration from sialorrhea. Archives of Otolaryngology – Head and Neck Surgery. 1999; 125: 796–800. 24. Hotaling AJ, Madgy DN, Kuhns LR, Filipek L, Belenky WM. Postoperative technetium scanning in patients with submandibular duct diversion. Archives of Otolaryngology – Head and Neck Surgery. 1992; 118: 1331–3.



] 791



25. Burton MJ, Leighton SE, Lund WS. Long-term results of submandibular duct transposition for drooling. Journal of Laryngology and Otology. 1991; 105: 101–3. 26. Hallett KB, Lucas JO, Johnston T, Reddihough DS, Hall RK. Dental health of children with cerebral palsy following sialodochoplasty. Special Care in Dentistry. 1995; 15: 234–8. 27. Arnrup K, Crossner CG. Caries prevalence after submandibular duct retroposition in drooling children with neurological disorders. Pediatric Dentistry. 1990; 12: 98–101.



65 Head and neck embryology T CLIVE LEE



General embryology Skull Pharyngeal arches Pharyngeal pouches Pharyngeal clefts Face Palate Mouth Nose and paranasal sinuses



792 794 795 796 797 797 798 799 803



Larynx Thyroid gland Ear Key points Deficiencies in current knowledge and areas for further research Acknowledgements References



803 805 805 809 809 809 809



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words head, skull, ear, nose, palate, tongue, tooth, salivary gland, neck, thyroid, pharynx, oesophagus and larynx and focussing on level 1 evidence on embryology, growth and development. Morphological details are described and their relation to embryological stages, size and age are summarized in Table 65.1. References to the underlying genetic and molecular mechanisms governing embryological development are cited.



GENERAL EMBRYOLOGY Development begins with fertilization when the sperm and oocyte, each haploid with 23 chromosomes, unite to form the diploid zygote containing 46 chromosomes. Chromosomal abnormalities and gene mutations cause major craniofacial defects. Three, rather than two, copies of chromosomes cause trisomy syndromes – trisomy 13, 18 and 21 (Down syndrome). Partial chromosome deletions result in cri-du-chat syndrome (chromosome 5), Angelman syndrome (chromosome 15), Miller–Dieker syndrome (chromosome 17) and Shprintzen syndrome (chromosome 22). Breakage of chromosomes (fragile X syndrome) and gene mutations (MSX2) also cause craniofacial abnormalities.2



The zygote divides, producing a morula which cavitates to form a blastocyst that implants in the endometrium during the second week. The inner cell mass of the blastocyst gives rise to the embryo. Initially the inner cell mass forms a bilaminar germ disc which then, by gastrulation, leads to a trilaminar disc comprised of ecto-, meso- and endoderm layers in week 3 (Figure 65.1). Weeks 3–8 comprise the embryonic period during which the main organ systems are established and are vulnerable to teratogens such as drugs (thalidomide, alcohol), infectious agents (rubella, HIV) and physical agents (x-rays) which can cause major birth defects. Ectoderm gives rise to tissues and organs which maintain contact with the outside world – the nervous system, skin, the sensory epithelium of the ear, nose and



Chapter 65 Head and neck embryology Table 65.1 Carnegie embryonic stages in the development of the human head and neck, based on ultrasonic studies.1



] 793



Buccopharyngeal membrane Amniotic cavity



Stage



Age (days)



Size (mm)



Ectoderm



1 5 8b



1 7 23



0.1–0.15 0.1–0.2 1.01–1.5



9 10



25 28



1.5–2.5 2–3.5



12 13 14



30 32 33



3–5 4–6 5–7



15



36



7–9



16



38



8–11



17



41



11–14



18



44



13–17



19



20 22 23



46



49 53 56



Major features



16–18



18–22 23–28 27–31



Fertilization Implantation Neural folds and groove appear Otic disc, somites appear Neural fold fusion; two pharyngeal arches; otic pit 3–4 pharyngeal arches Otic vesicle Endolymphatic appendage, cochlear duct begins Nasal pit; auricular hillocks beginning Utricosaccular diverticulum, nasal sacs face ventrally Tubotympanic recess, six auricular hillocks, nasofrontal groove 1–3 semicircular ducts, stapes and stapedius, tip of nose; ossification begins oronasal membrane Cartilaginous otic capsule, malleus and incus Tensor tympani External ear developing Cochlear duct with 2.5 turns, head more rounded



Notochord



Mesoderm Endoderm Yolk sac



Cloacal membrane



Figure 65.1 Highly diagrammatic representation of a longitudinal section of the developing embryonic disc. The structure never looks like this at any one time since development is progressing at different rates. Nevertheless, mesoderm lies between ectoderm and endoderm, except at the buccopharyngeal and cloacal membranes where the two layers are in contact. The notochord, a derivative of the ectodermal layer, lies within the mass of mesoderm. A connecting stalk attaches the developing disc with its amniotic cavity and yolk sac to the uterine wall, and these structures lie within the extra-embryonic coelom at this stage of development.



Amniotic cavity Neural groove



(a)



Notochord plate



Extra-embryonic mesoderm Intra-embryonic mesoderm



Yolk sac Deepening neural groove Neural crest



(b)



Developing notochord Neural crest



Neural tube



Embryonic period: weeks 3–8 (days 8–56); foetal period: weeks 9–38.



Somite



eye, and tooth enamel. Neurulation is the process by which ectoderm forms a neural plate that folds to form the neural tube, giving rise to the brain and spinal cord, and the neural crest (Figure 65.2). Neural crest changes into mesenchyme and contributes to the connective tissue and bones of the face and skull. Lateral to the neural tube, paraxial mesoderm forms pairs of somites each of which give rise to its own sclerotome (bone and cartilage), myotome (muscle) and dermatome (dermis) component. The occipital bone and cervical vertebrae are derived from sclerotomes, and the related myo- and dermatomes are segmentally innervated. Endoderm provides the epithelial lining of the gastrointestinal and respiratory tracts, including the tympanic



(c)



Notochord



Figure 65.2 Transverse section of the developing embryonic disc at various stages of development. In the early presomite stage (a) a midline neural groove is present above the notochordal plate, which is fused with the endodermal layer. The neural groove deepens (b) and specialized neural crest cells develop on the lips of the groove. The notochord separates from the endodermal layer. By the somite stage (c) the neural tube has formed and the neural crest cells are about to migrate to form cell clusters which, at the head end of the embryo, become the cranial sensory nerve ganglia. Cells from the neural crest also form the posterior root ganglion of the spinal cord and the cells of the sympathetic ganglia.



794 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY cavity and auditory tube, and the parenchyma of the thyroid and parathyroid glands. The foetal period runs from the ninth week of gestation until birth at 38 weeks of gestation or 40 weeks after the onset of the last menstruation and is characterized by growth and maturation of tissues and organs. Abnormalities may result from disruption of blood supply or mechanical deformation. Major congenital malformations, due to genetic, environmental or, most commonly, unknown causes, occur in approximately 4 percent of live births, but minor abnormalities are found in a further 15 percent.3 A syndrome is a group of anomalies occurring together with a specific common cause, while in an association this cause has yet to be determined. Due to the common germ layer origin of a variety of organs, or the contemporaneous development of different systems, the presence of minor abnormalities may indicate more serious underlying defects. This is particularly true of external ear abnormalities which feature in most common chromosomal syndromes.2



SKULL Skull development is influenced by both genetic factors and the mechanical forces generated by brain growth, development of the pharynx and muscle activity. The skull is divided into the neurocranium, which protects the brain, and the viscerocranium, derived from pharyngeal arches, which forms the face and jaws. Both neurocranium and viscerocranium ossify partly by intramembranous and partly by endochondral ossification.1 In intramembranous ossification, mesenchyme differentiates directly into bone. This mesenchyme is derived from neural crest cells in the roof and sides of the skull and from paraxial mesoderm in the occiput and posterior otic capsule. Primary ossification centres appear as spicules in the centre of these flat bones and radiate to the periphery. At birth, membranous sutures separate these bony plates, enabling moulding to occur during birth. Osteoblasts lay down new bone on outer surfaces and osteoclasts resorb bone from inner surfaces to permit the vault to grow and accommodate the expanding brain. Where sutures meet at the angles of the parietal bone are six membranous intersections, termed fontanelles. The largest is the anterior fontanelle at bregma, between the parietal and frontal bones, which is a useful indicator of raised intracranial pressure or dehydration and provides access to the underlying superior sagittal venous sinus. It is obliterated by two years, while the posterior fontanelle at lambda, between the parietal and occipital bones, closes at about six months after birth.4 Premature closure of sutures results in craniosynostosis.5 Early closure of the coronal suture causes acrocephaly, or tower skull; early closure of the sagittal suture causes a long narrow skull, scaphocephaly (Figure 65.3) and unilateral premature closure of the coronal and



Figure 65.3 Scaphocephaly. In this child the sagittal suture is completely fused, while the remaining sutures of the cranial vault are normal.



lambdoid sutures results in asymmetry – plagiocephaly (see Chapter 78, Craniofacial anomalies: genetics and management). In endochondral ossification, mesenchyme forms a cartilaginous model which is then replaced by bone. Anterior to the pituitary gland in the midline, neural crest cells give rise to the ethmoid, inferior nasal concha and body of the sphenoid. Posterior to the pituitary, paraxial mesoderm forms the clivus and base of the occipital bone.



Chapter 65 Head and neck embryology



Laterally, the ala orbitalis and ala temporalis form the lesser and greater wings of the sphenoid, while the otic capsule develops around the membranous labyrinth and gives rise to the petrous and mastoid parts of the temporal bone.2 The sphenooccipital junction remains cartilaginous until puberty, while the cartilage of the foramen lacerum does not undergo ossification. Incomplete ossification may lead to herniation of the meninges, or meningocoele, but larger defects can include neural tissue (meningoencephalocoele) (see Chapter 78, Craniofacial anomalies: genetics and management) or even brain and part of a ventricle (meningohydroencephalocoele).2 The viscerocranium, forming the face and jaws, is derived from the first two pharyngeal arches.



PHARYNGEAL ARCHES Arches of mesenchyme derived from paraxial and lateral plate mesoderm and neural crest cells appear in the fourth and fifth weeks of development. They are covered externally by ectoderm, which forms clefts between successive arches, and internally by endoderm which forms pouches between arches. Morphologically, they resemble gills in fish, but as true gills (branchia) are never formed, the term pharyngeal arches is used in humans. Pharyngeal arches play a role in the formation of the face, ear and neck (Figure 65.4) and their derivatives are summarized in Table 65.2.5



8 7



9 10



] 795



First pharyngeal arch The first arch has a dorsal maxillary process and a ventral mandibular process. The maxillary process gives rise to the premaxilla, maxilla, zygomatic bone, the zygomatic process and squamous part of the temporal bone by intramembranous ossification. The mandibular process contains Meckel’s cartilage which persists as the malleus, anterior ligament of malleus, incus and sphenomandibular ligament. The majority of the mandible is formed by membranous ossification of mesenchyme around Meckel’s cartilage. This entraps the inferior alveolar branch of the mandibular nerve which enters the bone through the mandibular foramen. Its terminal branch, the mental nerve, exits via the mental foramen. The right and left mandibular processes meet in the midline at the symphysis menti which achieves bony union six months after birth.4 The trigeminal (V) nerve supplies sensation to the first arch connective tissues via its ophthalmic, maxillary and mandibular branches. Only the mandibular division has a motor root and supplies eight first arch muscles: four muscles of mastication (temporalis, masseter, medial and lateral pterygoid), two tensors (tympani and palati), mylohyoid and the anterior belly of digastric.



Second pharyngeal arch The second arch gives rise to the stapes, styloid process of the temporal bone, stylohyoid ligament, lesser cornu and upper body of the hyoid bone. The facial (VII) nerve supplies sensation to second arch connective tissue in the external auditory canal, but is mainly motor supplying: stapedius, stylohyoid, posterior belly of digastric, occipitofrontalis and the muscles of facial expression.



11 12 1



13



2



14



3



15



4 6



16



Third pharyngeal arch The third arch gives rise to the greater cornu and lower body of the hyoid bone. The glossopharyngeal (IX) nerve supplies sensation to third arch connective tissues in the posterior third of the tongue and is motor to glossopharyngeus.



5



Figure 65.4 A diagram to show the derivatives of the pharyngeal arch cartilages. 1, First arch cartilage; 2, second arch cartilage; 3, third arch cartilage; 4, fourth arch cartilage; 5, tracheal rings; 6, sixth arch cartilage; 7, Meckel’s cartilage; 8, malleus; 9, incus; 10, stapes; 11, styloid process; 12, stylohyoid ligament; 13, lesser horn and upper body of hyoid bone; 14, greater horn and lower body of hyoid bone; 15, thyroid cartilage; 16, cricoid cartilage. Modified from Ref. 2, with permission.



Fourth and sixth pharyngeal arches The cartilagenous components of the fourth and sixth arches fuse to form the laryngeal cartilages: thyroid, cricoid, arytenoid, corniculate and cuneiform. The superior laryngeal branch of the vagus (X) supplies sensation to fourth arch connective tissue from the valleculae and epiglottis to the true vocal cords and is motor to levator palati, pharyngeal constrictors (partially) and



796 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Table 65.2 Arch



Derivatives of the pharyngeal arches and pouches.1, 2, 5 Skeleton



Nerve



Muscles



Artery



Pouch



First



Maxilla, Meckel’s cartilage, malleus, anterior ligament of malleus, incus, sphenomandibular ligament, portion of mandible



V. Trigeminal mandibular division



Maxillary



Eustacian tube; middle ear cavity; Tubotympanic recess (first cleft forms external auditory meatus)



Second



Staples; styloid process; stylohyoid ligament; lesser horn and upper portion of body of hyoid bone



VII. Facial



External carotid



Bed of palatine tonsil



Third



Greater horn and lower portion of body of hyoid bone Laryngeal cartilages (thyroid, cricoid, arytenoid, corniculate, cuneiform)



IX. Glossopharyngeal



Mastication (temporalis, masseter; medial and lateral pterygoids); mylohyoid; anterior belly of digastric, tensor palati, tensor tympani Facial expression (buccinator; auricularis; frontalis; platysma; orbicularis oris; orbicularis oculi); posterior belly of diagastric; stylohyoid; stapedius Stylopharyngeus



Common carotid and first part of internal carotid Aortic arch (fourth left)



Thymus; inferior parathyroid



First part subclavian (fourth right) Ductus arteriosus (sixth left) Proximal right pulmonary (sixth right)



Ultimobranchial body (fifth)



Fourth–sixth



V. Vagus. Superior laryngeal branch (nerve to fourth arch) Recurrent laryngeal branch (nerve to sixth arch)



cricothyroid. The recurrent laryngeal branch of the vagus supplies sensation to sixth arch derivatives, notably the infraglottic larynx, and is motor to the other muscles of the larynx.



PHARYNGEAL POUCHES On each side, between the six arches, lie five pharyngeal pouches lined by endoderm (Figure 65.5). Their derivatives are summarized in Table 65.2. The first pouch extends laterally to form the Eustachian tube, the middle ear cavity and the tubotympanic recess, which extends as far as the tympanic membrane. The second pouch forms the bed of the palatine tonsil. The third pouch forms ventral and dorsal wings. The epithelium of the ventral wing differentiates into



Cricothyroid; levator palati; constrictors of pharynx (fourth) Intrinsic muscles of larynx (sixth)



Superior parathyroid (fourth)



the thymus, while that of the dorsal wing forms the inferior parathyroid gland. The thymus separates from the pharyngeal wall and descends inferomedially to unite with contralateral thymic tissue behind the sternum. The inferior parathyroid descends to lie posterior to the inferior pole of the ipsilateral thyroid lobe. Thymic tissue may be found embedded in the thyroid, while the inferior parathyroids may lie at sites from the bifurcation of the common carotid artery to the anterior mediastinum behind the sternum. The dorsal wing of the fourth pouch differentiates into parathyroid tissue which descends to lie posterior to the superior pole of the ipsilateral thyroid lobe. The fifth pouch (or ventral wing of the fourth) forms the ultimobranchial body, which is incorporated into the thyroid gland and gives rise to parafollicular calcitoninsecreting cells.



Chapter 65 Head and neck embryology



] 797



6



7



1B



1B 1C



1A



11



9



2B



2B 12



2C 2A



3B



3B



13



3C 3A 4A 8



14



10



4B 4C



4B



5C



15 16



Pharyngeal diverticulum The structural weakness between thyropharyngeus and cricopharyngeus, Killian’s dehiscence, may be the site of an acquired pulsion diverticulum which may be described erroneously as a ‘pharyngeal pouch’.6 As neural crest cells contribute to both pharyngeal and heart formation, craniofacial and cardiac abnormalites are common to a number of syndromes (see Chapter 78, Craniofacial anomalies: genetics and management). First arch structures are affected in Treacher Collins and Goldenhar syndromes (see Chapter 80, Facial paralysis in childhood), and the Pierre Robin sequence. The DiGeorge sequence affects the third and fourth pouches, as well as the first arch and external ears.1, 2



PHARYNGEAL CLEFTS On each side, between the first five arches, lie four pharyngeal clefts lined by ectoderm. Normally, only the first cleft persists and its dorsal end gives rise to the external auditory meatus, separated from the first pouch by the tympanic membrane. The ventral end is normally obliterated, but may persist as a sinus, cyst or fistula. Such a fistula extends from below the auricle, through the parotid gland and opens into the external auditory meatus. It has a variable relationship with the facial (VII) nerve.1 The second arch overgrows the second, third and fourth clefts, forming the cervical sinus which then resorbs. However, if this sinus persists, it gives rise to cervical cysts along the anterior border of sternocleidomastoid (Figure 65.6).7 If the cysts communicate with the



Figure 65.5 A diagram to show the development and derivatives of the pharyngeal pouches and clefts. Note the way in which the second, third and fourth clefts are buried and, if not obliterated, form the cervical sinus. 1A, first cleft; 1B, first arch; 1C, first pouch; 2A, second cleft; 2B, second arch; 2C, second pouch; 3A, third cleft; 3B, third arch; 3C, third pouch; 4A, fourth cleft; 4B, fourth arch; 4C, fourth pouch; 5C, fifth pouch; 6, maxillary process; 7, mandibular process; 8, epicardial ridge; 9, external auditory meatus; 10, cervical sinus; 11, pharyngotympanic tube; 12, palatine tonsil; 13, inferior parathyroid gland; 14, thymus; 15, superior parathyroid gland; 16, ultimobranchial body. Modified from Ref. 2, with permission.



skin, they form external branchial fistulae. The cervical sinus may also communicate with the second pouch in the bed of the palatine tonsil to form an internal branchial fistula. Such fistulae pass over the glossopharyngeal (IX) and hypoglossal (XII) nerves and run between the external and internal carotid arteries (Figure 65.7) (see Chapter 99, Branchial arch fistulae, thryoglossal duct anomalies and lymphangioma).1, 8 The third cleft may give rise to a fistula extending from the anterior border of sternocleidomastoid, passing superficial to the hypoglossal (XII) nerve, deep to the carotid arteries and piercing the thyrohyoid membrane to open into the piriform fossa.8 Very rarely a fistula from the fourth cleft may extend from the anterior border of sternocleidomastoid, pass around the arch of the aorta on the left side, or right subclavian artery, ascend in the neck running superficial to the hypoglossal (XII) nerve, deep to the carotid arteries and open into the lower pharynx (see Chapter 99, Branchial arch fistulae, thryoglossal duct anomalies and lymphangioma).1



FACE The face develops from five growth centres or prominences – the frontonasal in the midline and the right and left maxillary and mandibular prominences. The frontonasal prominence forms the forehead and, inferiorly, lies above the primitive mouth or stomodeum with a nasal placode on either side. The nasal placodes invaginate to form nasal pits and their margins form lateral and medial nasal processes. The first arch gives rise to the maxillary prominences lateral to the stomodeum and the mandibular prominences inferiorly.9



798 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



5 I



I 1



8



2



2



II



II 9



6



3



3 7



4



(a)



(b)



(c)



Figure 65.6 A diagram to show a lateral cervical (branchial) cyst opening onto the surface as a fistula (a). (b) Location of cysts and fistulae anterior to the sternocleidomastoid muscle. (c) Internal opening of a fistula by the palatine tonsil. 1, external auditory meatus; 2, palatine tonsil; 3, lateral cervical (branchial) cyst; 4, external branchial fistula; 5, region of preauricular fistulae; 6, region of lateral cervical cysts and fistulae; 7, sternocleidomastoid muscle; 8, tubotympanic recess; 9, internal branchial fistula. Modified from Ref. 2, with permission.



The maxillary prominences grow medially, pushing the medial nasal processes towards the midline and fusing with them to form the upper lip (Figure 65.8). The philtrum is derived from the medial nasal process with the maxillary prominence providing the lateral lip and cheek. An alternative theory is that the maxillary prominences fuse in the midline superficial to the frontonasal prominence, thus accounting for the maxillary nerve supplying sensation to all of the upper lip, including the philtrum.1 The nasolacrimal groove lies between the maxillary prominence and lateral nasal process. This groove separates, forming a duct, and allows the lateral nasal process and maxillary prominence to fuse. The lower lip and jaw are formed by the fusion of the right and left mandibular prominences. The right and left halves meet anteriorly at the symphysis menti, but fusion, starting at the outer and inferior surfaces, is usually complete by six months after birth.4 The cheeks develop from the maxillary and mandibular processes which accounts for their sensory innervation. The musculature, however, is derived from mesenchyme of the second pharyngeal arch and so receives its motor supply from the facial (VII) nerve.



form the triangular primary palate which carries the upper medial and lateral incisors and extends posteriorly to the incisive foramen – the premaxilla of the adult. The secondary palate comprises the remainder of the adult hard and soft palates and is formed in three stages where the palatal shelves are, respectively, vertical, horizontal or fused.10 In stage I, vertical palatal shelves arise from the maxillary prominences and lie lateral to the tongue. The two shelves expand by cell division, hydration and electrostatic forces in the orofacial cavity which is growing in height but not in width.11 Descent of the tongue as the first arch elongates, combined with shelf expansion, results in the elevation of the shelves to the horizontal position – stage II (Figures 65.9). This movement is rapid and occurs in a matter of minutes or hours.12 In stage III, the palatal shelves fuse separating the nasal and oral cavities.12 The epithelium on the superior, nasal surface is pseudostratified, ciliated columnar, while on the inferior, oral surface are stratified, squamous, nonkeratinizing cells.10 Bone that forms in the hard palate belongs to the maxillae and palatine bones. The soft palate and uvula probably develop by epithelial displacement due to mesenchymal activity, rather than by midline fusion.1



Facial clefts PALATE The palate separates the oral and nasal cavities and is formed in two phases, primary and secondary, under separate genetic and epigenetic controls (see Chapter 77, Cleft lip and palate).1 The medial nasal processes fuse to



Cleft palate may result from disturbances at any stage of palatal development.5, 11 Failure of fusion of the maxillary prominence and medial nasal process gives rise to clefts anterior and posterior to the incisive foramen. Anterior clefts may involve the lip alone or include the palate, may



Chapter 65 Head and neck embryology



] 799



Forehead Lateral nasal fold



Frontonasal process



Medial nasal fold



Eye Primative posterior naris



Maxillary process



Rathke’s pouch



Figure 65.8 The roof of the stomodeum of a 12 mm human embryo illustrating the development of the primitive palate and posterior nares by approximation of the maxillary processes to the lateral and medial nasal folds. The previous site of the attachment of the buccopharyngeal membrane is represented by a dotted line and part of the left maxillary process has been removed. After Hamilton WJ and Mossman HW (eds). Human embryology. Cambridge: Heffer, 1972.



(a) Internal carotid artery External carotid artery



Hypoglossal nerve



Internal jugular vein



Vagus nerve (b)



Figure 65.7 Operation for second branchial cleft fistula (a) showing two horizontal incisions, the lower one to include the fistulous opening; (b) the fistula is seen crossing the hypoglossal nerve to pass between the internal and external carotid arteries.



extend into the nose and may be bilateral.13 Posterior clefts lie in the midline and may involve the uvula alone, or separate both palatine shelves and extend into the nose. Anterior and posterior clefts may be combined. Cleft types are summarized in Figure 65.10.14 A variety of mechanisms have been implicated in clefting, with failure of shelf elevation occurring in the majority of cases.11 Failure of fusion of the maxillary prominence and lateral nasal process exposes the nasolacrimal duct to form an oblique facial cleft. However, they may be due to irregular tearing accompanying rupture of the amnion.1 Craniofacial clefts are summarized in Figure 65.11.15 Rarely, the medial nasal processes may fail to fuse causing a median cleft lip. The entrapment of epithelium along lines of fusion may result in nasolabial, globulomaxillary, median alveolar and median palatal cysts which usually do not present until adulthood (Figure 65.12).



MOUTH The primitive mouth or stomodeum is bounded by the forebrain superiorly and the pericardial cavity inferiorly. The buccopharyngeal membrane separating it from the foregut breaks down and so connects the amniotic cavity and the primitive gut (Figure 65.13). The cranial part of the foregut, often termed the pharyngeal gut, thus extends from the buccopharyngeal membrane to the tracheobronchial diverticulum.2 Laterally, pharyngeal mesenchyme segments to form the pharyngeal arches.



Salivary glands Three pairs of salivary glands arise from the oropharyngeal epithelium and follow similar patterns of development to produce branching ducts and secretory acini.16 The submandibular gland originates as an epithelial cord in the groove between tongue and mandible which branches into the surrounding mesenchyme. Serous acini appear early, followed by mucous acini postnatally.1, 17 The sublingual and parotid glands develop in a similar fashion, with the site of opening of the parotid duct dependent upon skeletal development. There is no evidence to suggest that the gland is composed of two lobes, rather the parotid ductules grow around the facial (VII) nerve, making it intimately entwined with the gland.18



Teeth The epithelium of the oral cavity forms the dental lamina on the upper and lower jaws. On each jaw, ten diverticula



800 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Septal cartilage Developing inferior turbinate Vomeronasal cartilage Tongue Palatal process of maxilla Tooth bud Meckel's cartilage



(a)



Septal cartilage Developing turbinate Vomeronasal organ Fused palatal and septal processes



Tongue



Meckel's cartilage (b)



of the lamina push into the jaw mesenchyme giving rise to a permanent tooth bud and the dental bud from which a deciduous tooth develops in a complex interaction of genetic and epigenetic factors.19 The deep surface of each bud invaginates to form a dental cap lying on the dental papilla, a neural crest derivative. The cap comprises outer and inner epithelial layers separated by the stellate reticulum. The central part of the dental papilla forms the pulp of the tooth, while cells deep to the inner dental epithelium differentiate into odontoblasts which produce dentine and push themselves inwards into the papilla. The outer dental epithelial cells differentiate into ameloblasts which deposit enamel prisms over the dentine, beginning



Figure 65.9 Section through the developing palate of (a) a 20-mm and (b) a 48-mm human foetus. Reprinted from Hamilton WJ and Mossman HW (eds). Human embryology. Cambridge: Heffer, 1972.



at the apex and spreading towards the neck. As the enamel thickens, the ameloblasts are pushed outwards into the stellate reticulum (Figure 65.14). The dental epithelial cells form a root sheath which expands into the jaw mesenchyme. Odontoblasts lay down dentine, continuous with that in the crown, which narrows the pulp chamber to a root canal. In the jaw mesenchyme, cells in contact with dentine differentiate into cementoblasts and secrete cementum. Outside the cementum, the periodontal ligament forms to connect the tooth to the jaw bone. None of the deciduous or primary teeth are impeded by covering bone and lengthening of the root causes the crowns to erupt in sequence from six



Chapter 65 Head and neck embryology



(a)



(b)



(d)



(e)



(g)



(h)



] 801



(c)



(f)



to 24 months after birth.20 The permanent tooth buds remain dormant until the age of six years when they grow, pushing the deciduous teeth up and out.



Tongue Two lateral lingual swellings from the first pharyngeal arch fuse with the midline tuberculum impar to form the anterior two-thirds of the tongue. In the posterior third, the third branchial arch overgrows the second. The Vshaped sulcus terminalis separates these regions with its apex at the foramen caecum from which the thyroid gland originates. General sensation to the anterior two-thirds of the tongue is from the lingual branch of the mandibular (V) nerve and taste from the chorda tympani branch of the facial (VII) nerve, while both taste and sensation to the posterior third are from the glossopharyngeal (IX) nerve. However, the glossopharyngeal nerve supplies taste and sensation to the vallate papillae anterior to the sulcus (Figure 65.15). Behind the posterior third lie the valleculae and epiglottis, fourth arch derivatives innervated by the superior laryngeal nerve from the vagus (X). The musculature is derived from four occipital somites innervated by the hypoglossal (XII) nerve.



Figure 65.10 Kernahan and Stark classification of cleft lip and palate. (a) Group A, incomplete unilateral cleft of primary palate; (b) group B, complete unilateral cleft of primary palate; (c) group C, complete bilateral cleft of primary palate; (d) group D, incomplete midline cleft of secondary palate; (e) group E, complete midline cleft of primary and secondary palates; (f) group F, complete unilateral cleft of primary and secondary palates; (g) group G, complete bilateral clefts of primary and secondary palates; (h) group H, cleft of primary palate and incomplete cleft of secondary palate.



Oropharynx and oesophagus The formation of an oesophagotracheal septum divides the foregut into a respiratory diverticulum anteriorly and the oesophagus posteriorly (Figure 65.16). The oesophagus descends with the heart and septum transversum and so is lengthened. It is lined by endoderm which changes from stratified columnar in the embryo, to cuboidal and then to stratified squamous by birth, although ectopic gastric muscosa may also be present. The surrounding mesoderm gives rise to a muscular coat that is striated in its upper two thirds, innervated by the nucleus ambiguus of the vagus (X), and smooth in its lower two thirds and innervated by the dorsal nucleus of the vagus (X), via the oesophageal plexus.6 Peristalsis is immature at birth along the oesophagus and at the lower oesophageal sphincter, permitting regurgitation of food. Deviation of the oesophagotracheal septum may give rise to oesophageal atresia and/or tracheooesophageal fistula. Atresia prevents passage of amniotic fluid into the gut resulting in hydramnios, while five types of fistula may result.2, 21, 22 The most common comprises a blind upper oesophagus with the lower oesophagus joining the trachea near its bifurcation. In the second type, there is a



802 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Stellate reticulum Ameloblasts Enamel



13 14 1 12 10 2 11



9



Dentine Odontoblasts 8



12



3



Pulp Stellate reticulum



7 6



4



5 7?



Figure 65.14 Developing human tooth. 30



(a)



14 13 12 01 2 3 11



10 9 8



3 4 5



6 7 7b



1 2 3



9 10



4 0 1



(b)



5



23



6



Figure 65.11 Tessier classification of craniofacial clefts. Localization of (a) soft-tissue clefts; (b) skeletal (bony) clefts. Dotted lines represent either uncertain sites or uncertain clefts.



7



8



11 12 13 14 15



16 Median alveolar cyst



Globulomaxillary cyst



Median palatal cyst



Figure 65.15 The base of tongue and valleculae from above. 1, tongue; 2, vallate papillae; 3, sulcus terminalis; 4, foramen caecum; 5, base of tongue (pharyngeal part); 6, median glossoepiglottic fold; 7, lateral glossoepiglottic fold; 8, pyriform fossa; 9, folia linguae; 10, palatoglossal fold; 11, palatine tonsil; 12, palatopharyngeal fold; 13, vallecula; 14, epiglottis; 15, aryepiglottic fold; 16, posterior pharyngeal wall.



Nasolabial cyst



Figure 65.12 Development of cysts of the maxilla.



Forebrain Rathke’s pouch Buccopharyngeal membrane Foregut



Stomodeum



Figure 65.13 Sagittal section of human embryo showing early development of oral cavity.



blind upper and lower oesophagus, with no connection (fistula) to the trachea. The third type consists of a proximal tracheooesophageal fistula with a blind lower oesophagus. In the H-type abnormality, there is a proximal and a distal tracheooesophageal fistula. In the fifth type, there is a tracheooesophageal fistula without oesophageal atresia (Figure 65.17). Vertebral, anal, tracheooesophageal and radial (VATER) abnormalities are frequently found in the same individual – a VATER association (see Chapter 101, Diseases of oesophagus, swallowing disorders and caustic ingestion).1



Chapter 65 Head and neck embryology



3



2 1



4



5



(a)



(b)



6



Figure 65.16 The development of the tracheobronchial (respiratory) diverticulum and oesophagus. (a) Laryngotracheal groove appearing in the ventral aspect of the foregut. (b) The lips of the groove closing in to form the oesophagotracheal septum separating the respiratory tract from the alimentary canal. 1, foregut; 2, laryngotracheal groove; 3, oesophagotracheal septum; 4, trachea; 5, lung bud; 6, stomach.



NOSE AND PARANASAL SINUSES The development of the medial and lateral nasal prominences causes the deepening of the nasal pits which open into the stomodeum between medial and lateral nasal processes (Figure 65.8). The nasal pits deepen to form nasal sacs which open at the anterior nares and will form the nasal cavities (see Figure 104.1 in Chapter 104, Anatomy of the nose and paranasal sinuses). Posteriorly, the medial and lateral nasal processes fuse to form a nasal fin of epithelium which then regresses to leave the oronasal membrane caudally.23 The development of the primary palate (premaxilla) and secondary palate, from the maxillary shelves, separates the oral and nasal cavities. Rupture of the oronasal membrane enables the nasal sac to communicate with the oropharynyngeal cavity via the primary choanae. The definitive posterior choanae will lie further posteriorly due to palatal growth. Unilateral or bilateral choanal atresia may be due to persistence of the oronasal membrane or due to persistence of epithelial cells which normally plug the nasal cavity during the embryonic period (see (see Chapter 82, Nasal obstruction in children).23, 24 A midline ridge develops from the posterior edge of the frontonasal prominence in the roof of the nasal cavity and extends posteriorly to the adenohypophyseal (Rathke’s) pouch, which ascends from the roof of the oral cavity to form part of the pituitary gland (see Figure 104.4 in Chapter 104 Anatomy of the nose and paranasal sinuses). Anteriorly, the septum is grooved forming the vomeronasal organ, which becomes a blind, tubular pouch 2–6 mm in length, involved in the sensation of pheromones.1 The primitive septum is initially made



] 803



entirely of cartilage. The superior part ossifies to form the perpendicular plate of the ethmoid. Posteroinferiorly, the vomer ossifies from two centres on either side of the cartilage leaving two bony alae. The remaining septal cartilage is quadrilateral and lies anteroinferiorly, surmounted by the two nasal bones. Conchal cushions appear on the lateral walls of the nasal cavity and fuse to form the conchae or turbinate bones. The superior and middle conchae originate with the ethmoid bone, while a supreme concha is present in up to one-third of cases. The inferior concha may be regarded as a separate bone, but functionally is a detached piece of the ethmoid.4 A depression or meatus develops beneath each concha and diverticula from these form the paranasal sinuses. The maxillary sinus arises from the middle meatus at the beginning of the foetal period, but most expansion develops after birth with development of the middle third of the face and eruption of the secondary dentition. A few ethmoid cells are present at birth and the frontal sinus, an enlarged anterior ethmoid air cell, invades the frontal bone after birth, while the sphenoid develops during childhood (Figure 65.18). The nasolacrimal duct appears along, but independently of, the border of the maxillary and frontonasal prominences to drain into the inferior meatus of the nose. The frontal prominence gives rise to the bridge and septum of the nose, the fused medial nasal processes form the bridge and tip, while the alae are derived from the lateral nasal processes. Agenesis, clefting, duplication of the nose and supernumerary nostrils are anomalies of embryological development.25



LARYNX The respiratory system develops as a ventral outgrowth of the foregut and grows caudally, on a lengthening stalk.26 The cephalic end of the stalk develops into the glottis and infraglottis, and the rest becomes the trachea. Above the respiratory diverticulum, pharyngeal mesoderm compresses the foregut lumen ventrodorsally as far cranially as the fourth pharyngeal pouches, forming an epithelial lamina with a narrow pharyngoglottic duct along its dorsal border. This mesoderm also raises an epiglottic and two arytenoid swellings in the pharyngeal floor at the level of the fourth pouches. The triangular laryngeal caecum is bounded by these swellings and grows caudally between the lamina and epiglottis. The epithelial lamina separates cephalocaudally, connecting the laryngeal caecum with the pharyngoglottic duct to form the laryngeal vestibule; when the separation is complete, the vestibule is continuous with the infraglottic cavity.27 However, incomplete separation gives rise to congenital laryngeal webs (see Chapter 88, Congenital disorders of the larynx, trachea and bronchi).28 Meanwhile, bilateral pouches arising from the caudal end of the caecum form the



804 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



(a)



(c)



(b)



(d)



Figure 65.17 Types of oesophageal atresia. (a) Oesophageal atresia with distal trachaeooesophageal fistula, 87 percent; (b) oesophageal atresia without distal trachaeooesophageal fistula, 6–7 percent; (c) oesophageal atresia with proximal trachaeooesophageal fistula, 2 percent; (d) oesophageal atresia with proximal and distal trachaeooesophageal fistula, o1 percent (continued over).



Chapter 65 Head and neck embryology



] 805



supplied by the external branch of the superior laryngeal nerve. Pharyngeal mesoderm surrounding the laryngeal cavity gives rise to the laryngeal cartilages and intrinsic musculature. The laryngeal cartilages are derived from the fourth and sixth pharyngeal arches. The cricoid chondrifies bilaterally from a single centre in the ventral arch of a precartilaginous template which encircles the infraglottic cavity, and on meeting forms the dorsal lamina. Each arytenoid chondrifies from a single centre and each lamina of the thyroid cartilage chondrifies from two.27, 28 At birth, the epiglottis is close to the soft palate and apposition of the two structures can close off the oral cavity, so protecting the airway during suckling when breathing and drinking occur simultaneously.1



THYROID GLAND



(e)



Figure 65.17 Types of oesophageal atresia (continued). (e) trachaeooesophageal fistula without oesophageal atresia, 3–4 percent.



Figure 65.18 Coronal section through the head of a 150-mm human foetus, showing early cavitation of maxillary sinus and ethmoids (arrowed).



ventricles, the lower lips of which become the vocal folds. The vestibule and ventricles receive sensory innervation from the internal branch of the superior laryngeal nerve of the fourth arch, while sensation to the infraglottic larynx is from the recurrent laryngeal nerve of the sixth arch. Motor supply is from the recurrent laryngeal nerve, with the exception of cricothyroid muscle which is



A diverticulum develops behind the tuberculum impar of the tongue and descends anterior to the pharynx and hyoid bone. It then ascends deep to the hyoid and descends again anterior to the thyroid and cricoid cartilages to the level of the upper tracheal rings and expands to form lateral lobes connected by a central isthmus. Epithelial cells develop into thyroid follicles, with hormone synthesis beginning early in the foetal period.1 The ultimobranchial body from the fourth or fifth pharyngeal pouch provides parafollicular cells which secrete calcitonin. The descent of the thyroid is marked by the thyroglossal duct which connects the gland to the foramen caecum. It is divided into suprahyoid and infrahyoid portions by the growth of the hyoid bone from the second and third pharyngeal arches. The duct usually atrophies but may persist as a fibrous cord or as a tube.8 Thyroglossal cysts may develop along its course, usually median in position (Figures 65.19 and 65.20), which move on swallowing and protrusion of the tongue.29 A small or absent thyroid results in congenital cretinism (see Chapter 99, Branchial arch fistulae, thryoglossal duct anomalies and lymphangioma). Failure of normal descent may result in a lingual thyroid, causing dysphagia or it may be sublingual, prelaryngeal or suprasternal. Between the isthmus and hyoid bone a pyramidal lobe or a fibromuscular band, the levator glandulae thyroidea, may be found.6 The isthmus, or one lobe, may be absent.29 More than half of abnormal neck masses are of thyroid origin.1 Accessory thyroid tissue may be found in the neck, thymus, mediastinum, heart, trachea, oesophagus or liver, due to attachment to an adjacent organ and also, as part of a teratoid tumour, in the ovary.1



EAR The internal, middle and external ear each have a different origin.



806 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



1 2 3



4



5



5 5



Figure 65.20 Midline thyroglossal cyst.



6



(a)



7-somite, 22 days Neural groove Otic pit



Otic placode



(a)



Notochord



(b)



1 2 3 Otocyst (c)



4



(d) 30 somite, 30 days



5



6



(b)



Figure 65.19 Possible sites of thyroglossal cysts. 1, Base of tongue; 2, intralingual; 3, suprahyoid; 4, infrahyoid; 5, prethyroid; 6, pretracheal.



Internal ear Ectoderm over the hindbrain thickens to form otic discs (placodes) which invaginate to form pits and then separate from the surface to form otocysts (Figure 65.21).



Figure 65.21 Diagram to represent the development of the otocyst from the otic placode, which in turn is derived from the ectoderm cranial to the first occipital somite. During the seven days of its development, the neural groove has also become converted into the future brain stem.



Each otocyst subsequently divides into a ventral part, which gives rise to the saccule and cochlear duct, and a dorsal part which forms the utricle, semicircular canals and endolymphatic duct (Figure 65.22). These epithelial derivatives are collectively termed the membranous labyrinth. Neural crest cells migrate away from the developing otocyst to form the statoacoustic ganglion from which the vestibulocochlear nerve is derived.2 The cochlear duct develops as an outpouching of the saccule and penetrates the surrounding mesenchyme in a spiral fashion forming 2.5 turns. The ductus reuniens connects the saccule and cochlea. This mesenchyme



Chapter 65 Head and neck embryology



Saccus endolymphaticus



] 807



Superior vertical canal Ductus endolymhaticus



Crus commune Horizontal canal



Utricle Otic vesicle



Saccule



Ampulla of posterior vertical canal



Cochlea 5 weeks 6 weeks 10 weeks



Figure 65.22 Development of the membranous labyrinth from the otocyst. This rapid and complex change occurs over a six-week period.



differentiates into cartilage which then vacuolizes to form two perilymphatic spaces, the scala vestibuli and scala tympani, around the cochlear duct or scala media. The vestibular membrane separates the duct from the scala vestibuli, while the basilar membrane separates the duct from the scala tympani (Figure 65.23). An inner sulcus divides the epithelial cells of the cochlear duct into an inner ridge, the future spiral limbus, and an outer ridge. The outer ridge forms one row of inner and three to four rows of outer cells, separated by the tunnel of Corti. On either side of the tunnel, pillar cells give mechanical rigidity, while the sensory cells develop a cluster of



ScV SV RM ScM TZ TM Spiral lamina



IS OC



SP OS BM



ScT



Figure 65.23 Diagram illustrating the structures and relationship of the cochlear duct. BM, basilar membrane; IS, inner sulcus; OC, organ of Corti; OS, outer sulcus; RM, Reissner’s membrane; ScM, scala media; ScT, scala tympani; ScV, scala vestibuli; SP, spiral prominence; SV, stria vascularis; TM, tectorial membrane; TZ, transitional zone. Yellow areas show bone.



stereocilia and are termed hair cells. These are covered by the tectorial membrane which extends from the spiral limbus. The hair cells and tectorial membrane comprise the organ of Corti which transmits sensory impulses to the spiral ganglion and then, via the cochlear fibres of the vestibulocochlear (VIII) nerve, to the brain. Differentiation proceeds in the basal to apical direction as the duct elongates. The dorsal part of the otocyst flattens and outpouchings appear to form semicircular canals. The walls between adjacent canals resorb to give rise to three canals – superior, posterior and lateral – emerging from the utricle. One end of each canal dilates to form a crus ampullare in which a fibrogelatinous cupola develops. The crus nonampullare of the lateral canal persists, while those of the superior and posterior canals fuse to form the crus commune nonampullare. In each ampulla, sensory cells form a crest or crista ampullaris. Similar sensory areas, the maculae, develop in areas where vestibular fibres of the vestibulocochlear (VIII) nerve enter the walls of the utricle and saccule. Two cell types differentiate; sensory cells with a single kinocilium and many stereocilia and supporting cells which produce a gelatinous matrix and are responsible for the formation of otoconia.30 Congenital malformations of the inner ear ranging from complete cochlear aplasia to incomplete partition of the cochlea, first described by Mondini in 1791, are reviewed by Graham et al.31



Middle ear As the first pharyngeal pouch grows laterally to meet the floor of the first pharyngeal cleft, the ossicles and



808 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY tympanic branch of the facial (VII) nerve develop in mesenchyme between them. The proximal part of the pouch forms the narrow auditory or Eustachian tube, while its distal part widens to form the tubotympanic recess which gives rise to the primitive tympanic cavity. Cleft ectodem, mesenchyme and pouch endoderm form the three layers of the future tympanic membrane, with the handle of the malleus embedded in it, traversed by the chorda tympani. The malleus and incus are derived from the first pharyngeal arch cartilage and begin to ossify at 16 weeks, reaching adult size and form by 25 weeks. The stapes is derived from the second arch (Reichert’s) cartilage and begins to ossify at 18 weeks. The base of the stapes appears in the lateral wall of the otic capsule, surrounding the otocyst, where it occupies the oval window, and hence may be involved in otosclerosis. Rarely the stapedial artery, passing between the crura, may persist. The ossicular ligaments develop within endodermal reflections connecting the ossicles to the walls of the tympanic cavity. The tensor tympani, attached to the malleus, and stapedius, attached to the stapes, are supplied by the mandibular and facial nerves, respectively. The middle ear reaches adult size before birth.1 Epithelium of the tubotympanic recess invades the mastoid process to form the antrum and, postnatally, the mastoid air cells. Type 1 malformations of the middle ear are due to abnormal development of the ring surrounding the tympanic membrane. In type 2 malformations, the space between the mandibular fossa and the front of the mastoid process is absent, resulting in an abnormal course of the facial (VII) nerve which may be associated with cochlear malformations.1, 31 Evolutionary and mechanical influences on middle ear development are discussed by Prendergast.32



External ear The external auditory meatus develops from the dorsal portion of the first pharyngeal cleft and canalizes from medial to lateral. Epithelial cells proliferate to form a meatal plug which may persist after birth causing conductive deafness. Auricular hillocks, three from the first and three from the second pharyngeal arch, surround the cleft below the level of the mandible. As they ascend, some mandibular hillocks contribute to the tragus, while the remainder form the bulk of the auricle (Figure 65.24), but abnormalities such as auricular appendages and fistulae are common (Figure 65.25). Congenital malformations of the ear are often features of common chromosomal syndromes, such as Trisomy 18, Trisomy 21, Fragile X and Treacher Collins syndrome.2



Helix Antihelix Tragus Antitragus



Figure 65.24 Early stages in the development of the auricle. Six cartilaginous hillocks are visible around the first pharyngeal groove. During subsequent development these become the cartilage of the auricle, although the bulk of the auricle appears to be derived from the cartilage of the second arch.



Temporal bone The temporal bone is composed of four parts: petromastoid, styloid, squamous and tympanic.4 The petromastoid originates from the capsule surrounding the otocyst and progresses from mesenchyme to cartilage to bone. It gives rise to the petrous which encloses the bony labyrinth, the tegmen tympani which roofs the tympanic cavity, and forms the mastoid process at one to two years after birth. The styloid process is derived from the second pharyngeal arch cartilage. The squamous part ossifies intramembranously and includes the mandibular fossa and zygomatic process. The tympanic part may be derived from the first pharyngeal arch and develops as an incomplete bony ring which unites with the squamous part before birth at the squamotympanic fissure.1 The tegmen separates them in part and first arch derivatives pass through the petrotympanic fissure to the spine of the sphenoid and then, as the sphenomandibular ligament, to the lingula of the mandible.



Figure 65.25 Accessory auricle in an infant.



Chapter 65 Head and neck embryology



KEY POINTS  Major organ systems are established during the embryonic period, weeks 3–8.  Pharyngeal arch, pouch and cleft development explain motor and sensory innervations and common abnormalities.  Palatogenesis explains most craniofacial clefting.  Recent advances in laryngeal embryology correlate well with clinical data.  Thyroglossal duct cysts are the most common congenital cervical abnormality.  Ear abnormalities are commonly associated with chromosomal syndromes.



Deficiencies in current knowledge and areas for further research Morphological descriptions of embryological development are well established.2, 9, 15, 17, 18, 27, 28, 29, 30 The underlying genetic and molecular mechanisms are now beginning to be understood.5, 10, 12, 19 Detection of molecular markers may lead to earlier and more effective intervention while increased understanding of mesenchymal stem cell and matrix interactions offer the prospect of tissue engineered replacement surgery within the next five years. By clearly understanding the mechanisms by which organs develop, the effects of teratogens may be predicted and birth defects avoided.



ACKNOWLEDGEMENTS I would like to thank my colleagues, Professor EJ Guiney, Professor MJ Ryan and Professor BE McCartan, for providing illustrations and for their helpful comments on the text.



REFERENCES 1. O’Rahilly R, Muller F. Human embryology and teratology, 3rd edn. New York: Wiley-Liss, 2001. 2. Sadler TW. Langman’s medical embryology, 8th edn. Philadelphia: Lippincott Williams & Wilkins, 2000. 3. Hobbs CA, Cleves MA, Simmons CJ. Genetic epidemiology and congenital malformations: from the chromosome to the crib. Archives of Pediatrics and Adolescent Medicine. 2002; 156: 315–20.











] 809



4. Scheuer L, Black S. Developmental juvenile osteology. San Diego: Academic Press, 2000. 5. Wilkie AO, Morriss-Kay GM. Genetics of craniofacial development and malformation. Nature Reviews Genetics. 2001; 2: 458–68. This combines embryology, evolution and mouse genetics to describe head development and explain skull vault and pharyngeal arch malformations and clefting. 6. Standring S (ed.) Gray’s anatomy, 39th edn. Edinburgh: Churchill Livingstone, 2005. 7. Davenport M. ABC of general surgery in children. Lumps and swellings of the head and neck. British Medical Journal. 1996; 312: 368–71. 8. Ellis PDM. Branchial cleft anomalies, thyroglossal cysts and fistulae. In: Kerr AG (ed.). Scott-Brown’s otolaryngology, 6th edn. Oxford: Butterworth-Heinemann, 1997: 6.30.1–6.30.13. 9. Hinrichsen K. The early development of morphology and patterns of the face in the human embryo. Advances in Anatomy, Embryology and Cell Biology. 1985; 98: 1–79. 10. Kerrigan JJ, Mansell JP, Sengupta A, Brown N, Sandy JR. Palatogenesis and potential mechanisms for clefting. Journal of the Royal College of Surgeons of Edinburgh. 2000; 45: 351–8. A comprehensive review of the molecular and morphological mechanisms involved in palate formation and the vulnerable points which may lead to clefting. 11. Ferguson MWJ. Palate development: Mechanisms and malformations. Irish Journal of Medical Science. 1987; 156: 309–15. 12. Sandy JR. Molecular, clinical and political approaches to the problem of cleft lip palate. The Surgeon, Journal of the Royal Colleges of Surgeons of Edinburgh and Ireland. 2003; 1: 9–16. 13. Summerfield Kung T. The anatomy of hare-lip in man. Journal of Anatomy. 1954; 88: 1–12. 14. Kernahan DA, Stark RB. A new classification for cleft lip and palate. Plastic and Reconstructive Surgery. 1958; 22: 435–41. 15. Tessier P. Anatomical classification of facial, cranio-facial and latero-facial clefts. Journal of Maxillofacial Surgery. 1976; 4: 69–92. 16. Denny PC, Ball WD, Redman RS. Salivary glands: A paradigm for diversity of gland development. Critical Reviews in Oral Biology and Medicine. 1997; 8: 51–75. 17. Guizetti B, Radlanski RJ. Development of the submandibular gland and its closer neighboring structures in human embryos and fetuses of 19–67 mm CRL. Anatomischer Anzeiger. 1996; 178: 509–14. 18. Guizetti B, Radlanski RJ. Development of the parotid gland and its closer neighboring structures in human embryos and fetuses of 19–67 mm CRL. Anatomischer Anzeiger. 1996; 178: 503–8. 19. Tucker AS, Sharpe PT. Molecular genetics of tooth morphogenesis and patterning: The right shape in the right place. Journal of Dental Research. 1999; 78: 826–34.



810 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY 20. Ranly DM. Early orofacial development. Journal of Clinical Pediatric Dentistry. 1998; 22: 267–75. 21. Mansfield LE. Embryonic origins of the relation of gastroesophageal reflux disease and airway disease. American Journal of Medicine. 2001; 111: 3S–7S. 22. Spitz L. Diseases of the oesophagus. In: Kerr AG (ed.). Scott-Brown’s otolaryngology, 6th edn. Oxford: Butterworth-Heinemann, 1997: 6.29.1–6.29.5. 23. Lund VJ. Anatomy of the nose and paranasal sinuses. In: Kerr AG (ed.). Scott-Brown’s otolaryngology, 6th edn. Oxford: Butterworth-Heinemann, 1997: 1.5.1–1.5.7. 24. Keller JL, Kacker A. Choanal atresia, CHARGE association and congenital nasal stenosis. Otolaryngologic Clinics of North America. 2000; 33: 1343–51. 25. Williams A, Pizzuto M, Brodsky L, Perry R. Supernumerary nostril: A rare congenital deformity. International Journal of Pediatric Otorhinolaryngology. 1998; 44: 161–7. 26. O’Rahilly R, Muller F. Chevalier Jackson lecture. Respiratory and alimentary relations in staged human embryos. New embryological data and congenital anomalies. Annals of Otology, Rhinology and Laryngology. 1984; 93: 421–9. 27. Zaw-Tun HA, Burdi AR. Reexamination of the origin and early development of the human larynx. Acta Anatomica. 1985; 122: 163–84.



 28.



 29.



30.



 31.



32.



Milczuk HA, Smith JD, Everts EC. Congenital laryngeal web: Surgical management and clinical embryology. International Journal of Pediatric Otorhinolaryngology. 2000; 52: 1–9. This illustrates and summarizes recent advances in laryngeal embryology which explain web formation and applies this knowledge to the surgical management of five cases. Organ GM, Organ Jr. CH. Thyroid gland and surgery of the thyroglossal duct: Exercise in applied embryology. World Journal of Surgery. 2000; 24: 886–90. A review of thyroid and thyroglossal duct embryology, explaining the rationale behind their surgical management in general and the Sistrunk operation in particular. Wright A. Anatomy and ultrastructure of the human ear. In: Kerr AG (ed.). Scott-Brown’s otolaryngology, 6th edn. Oxford: Butterworth-Heinemann, 1997: 1.1.1–1.1.11. Graham JM, Phelps PD, Michaels L. Congenital malformations of the ear and cochlear implantation in children: Review and temporal bone report of common cavity. Journal of Laryngology and Otology. 2000; 114: 1–14. A review of the development of the inner, middle and outer ears and the practical aspects of congenital malformations in cochlear implant surgery. Prendergast PJ. Mechanics applied to skeletal ontogeny and phylogeny. Meccanica. 2002; 37: 317–34.



66 Molecular otology, development of the auditory system and recent advances in genetic manipulation HENRY PAU



Introduction The development of the ear Recent advances in genetic manipulation in the treatment of congenital deafness



811 813 817



Key points Deficiencies in current knowledge and areas for future research References



818 819 819



SEARCH STRATEGY The following databases were consulted: OMIM, PubMed, Web of Science, MedlinePlus, Hereditary Hearing Loss homepage http://webhost.ua.ac.be/hhh/ using the key words congenital deafness, gene therapy and auditory development.



INTRODUCTION Table 66.1 provides a glossary of terms used in genetics.



Types of deafness Deafness is a common childhood problem. Serious hearing impairment is found in one in 800 newborns.1 Abnormal genetic make up accounts for approximately 50 percent of permanent childhood deafness.2 In addition, several genes have been found that, when mutated, either cause or predispose to progressive hearing loss,3 suggesting that this type of deafness may also have a significant genetic basis. Genetic deafness, therefore, poses a problem both at birth and later in life. It is therefore important to identify the mutated genes involved and to understand the ear pathology caused by these defective genes. Deafness can be broadly categorized into syndromic (deafness associated with other symptoms) or nonsyndromic (deafness associated with no other symptoms). Many more loci have been found for syndromic deafness



than nonsyndromic deafness, despite the fact that these make up the minority of all inherited deafness cases (about 30 percent of deafness is thought to be syndromic). This is likely due to the fact that syndromic cases of deafness are more accurately diagnosed due to the additional features of the syndromes. Nonsyndromic types of deafness are classified on the basis of the mode of inheritance as follows: DFNA (autosomal dominant), DFNB (autosomal recessive) and DFN (X-linked forms). At present, 39 DFNB, 51 DFNA, eight DFN, one modifier DFNM14 and two mitochondrial loci (regions on the mitochondrial chromosome) have been identified,5 and 36 causative genes have been reported (Table 66.2).5, 6 Clinically, deafness is generally categorized on the basis of which auditory structures are affected. Conductive deafness refers to defects found in the outer or middle ear, whereas sensorineural hearing loss (SNHL) refers to disruptions in the sound transmission from the inner ear to the cortex of the brain. Most SNHL is due to abnormalities at the level of the inner ear. Inner ear pathology is further categorized into three groups: morphogenetic, cochleosaccular and neuroepithelial.7



812 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Table 66.1



Glossary of terms used in genetics.



Term



Name



Explanation



Autosomal dominant gene



DFNA



Autosomal recessive gene



DFNB



A single, abnormal gene on one of the autosomal chromosomes (one of the first 22 ‘non-sex’ chromosomes) from either parent can cause certain diseases. One of the parents will usually have the disease (since it is dominant) in this mode of inheritance. Only one parent must have an abnormal gene in order for the child to inherit the disease. An abnormal gene on one of the autosomal chromosomes (one of the first 22 ‘non-sex’ chromosomes) from each parent is required to cause the disease. People with only one abnormal gene in the gene pair are called carriers, but since the gene is recessive they do not exhibit the disease. The normal gene of the pair can supply the function of the gene so that the abnormal gene is described as acting in a recessive manner. Both parents must be carriers in order for a child to have symptoms of the disease; a child who inherits the gene from one parent will be a carrier. The genetic composition of an organism Technique employed to identify gene by using primers (manufactured genetic markers) Transfer of certain genes using vectors, such as viruses Similar technique to genetic mapping in gene identification by using primers A region on a chromosome Genes that do not cause disease but have the ability to influence the expression of other genes resulting in abnormalities Abnormalities caused by alteration of genome The physical appearance of an organism A carrier employed to enter and introduce genetic materials into a cell, e.g. a virus The insertion or deletion of a gene at the embryonic stage of development resulting in different phenotypes Genes that can only be inherited from the maternal side, e.g. mitochondrial genes



Genotype Genetic ‘mapping’ Gene transfer Linkage analysis Locus Modifier genes



DFNM1



Mutation Phenotype Transfer ‘vector’ Transgenesis X-linked



DFN



Morphogenetic defects, characterized by abnormal development of the ear leading to a malformation, only occur in 15–20 percent of profoundly deaf humans.8 This is because important morphogenetic genes are often involved in diverse development processes and mutations in these genes are likely to lead to antenatal death. Cochleosaccular refers to abnormalities in the stria vascularis, which regulates the ionic balance of the endolymph. The third and probably the most common type of inner ear pathology, neuroepithelial, is caused by defects in the organ of Corti, in which the hair cells responsible for the first step of the pathway of transducing sound information to neural code are situated.8



Using mouse mutants in understanding human genetic deafness Identifying the gene that causes a particular type of deafness involves, first of all, localizing the mutation to a specific region of a chromosome (genetic mapping). In human populations, this process is confused by the fact that many types of deafness, particularly nonsyndromic forms, are difficult to diagnose as there are few defining features. To minimize grouping cases of deafness together in which the causative gene may be different, genetic



mapping must be done either in large consanguineous families or in populations that have been isolated from immigration.9 To a certain degree, this approach has been successful in that researchers have been able to localize over 100 nonsyndromic forms of deafness.5 However, the step between localizing the gene to a region of the chromosome and actually identifying the mutated gene is a large and difficult one. The mouse appears to be the best model for understanding genetic hearing and balance defects in humans for several reasons. Firstly, because large numbers of mice which all carry the same mutation can be produced, the region of the chromosome in which the mutation is located can be more specifically pinpointed than in humans, making the mouse a powerful tool for gene identification. Due to genomic conservation between mice and humans, the mutated genes can be positioned and identified in the mouse and the homologous human genes can be identified and sequenced for possible mutations. Secondly, the mouse cochlea is structurally very similar to humans in its organization and specialization. Both species have similar types of inner ear pathology, and in some cases, the orthologous (same gene, but different species) gene is involved in deafness in the two species. Thirdly, there have been recent advances in the manipulation of the mouse genome, such that it is the most appropriate vertebrate in which to perform



Chapter 66 Molecular otology, development of the auditory system and recent advances in genetic manipulation Table 66.2



] 813



Human syndromes with known causative genes.



Syndromes



Causative genes



Clinical features



Alport syndrome



COL4A5, COL4A3, COL4A4



Branchio-oto-renal syndrome



EYA1



Norrie syndrome



Norrin



Pendred syndrome Stickler syndrome



SLC26A4 COL2A1, COL11A2, COL11A1



Treacher Collins syndrome



TCOF1



Usher syndrome



MYO7A, USH1C, CDH23, PCDH15, USH2A, USH3 FGFR1



Nephritis, hearing loss, lenticonus and other eye disorders, immunologic abnormality of skin, disorders of platelets, abnormalities of white blood cells, or smooth muscle tumours Auricular pits/fistulae, blocked/absent nasolacrimal duct, branchial cleft/sinus/cysts, cleft palate, cochlear/saccular abnormalities, deafness, double ureters, facial weakness, hydronephrosis, microphthalmia, atretic auditory canal, renal agenesis/dysplasia or ectopic kidneys Pseudotumour of the retina, retinal hyperplasia, hypoplasia and necrosis of the inner layer of the retina, cataracts, phthisis bulbi, progressive sensorineural hearing loss and mental disturbance Goitre, hypothyroidism, deafness Progressive myopia, vitreoretinal degeneration, premature joint degeneration with abnormal epiphyseal development, midface hypoplasia, irregularities of the vertebral bodies, cleft palate deformity and variable sensorineural hearing loss Coloboma of the lower eyelid (the upper eyelid is involved in Goldenhar syndrome), micrognathia, microtia, hypoplasia of the zygomatic arches, macrostomia and inferior displacement of the lateral canthi Hearing impairment and retinitis pigmentosa – there are at least three clinical subtypes



Pfeiffer syndrome, Kallmann syndrome, craniosynostosis Waardenburg syndrome



Jervell and Lange–Nielson Maternally inherited diabetes and deafness (MIDD) Mitochondrial encephalopathy, lactic acidosis and stroke-like episodes (MELAS) Myoclonic epilepsy and ragged red fibres (MERRF) Kearns–Sayre syndrome (KSS)



Progressive myoclonic epilepsy, ataxia and deafness



PAX3, MITF, SLUG, EDNRB, EDN3, SOX10 KVLQT1, KCNE1 (IsK) tRNALeu, tRNALys, tRNAGlu tRNALeu



tRNALys, tRNALeu Several tRNA genes



tRNASer



Craniosynostosis, ocular proptosis, mixed deafness, mid-face dysgenesis Dystopia canthorum (lateral displacement of the inner canthus of each eye), pigmentary abnormalities of hair, iris and skin (often white forelock and heterochromia iridis), sensorineural deafness Long QT syndrome and deafness Deafness, diabetes, seizures, migraines, short stature, mental retardation and stroke-like episodes Deafness, mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes. Deafness, myoclonus, ataxia, progressive external ophthalmoplegia and seizures Weakness of facial, pharyngeal, trunk and extremity muscles, deafness, small stature, electroencephalographic changes, markedly increased cerebrospinal fluid proteinophthalmoplegia, pigmentary degeneration of the retina and cardiomyopathy Deafness, ataxia and seizures



Modified from Ref. 5.



transgenesis, a feature which can be useful in both the identification of a mutated gene as well as in the production of potential mouse models. Finally, the mouse is useful in embryological and developmental studies, and invasive physiological tests can also be performed in anaesthetized mice, which provide important clues to the basis of the dysfunction and could not be carried out in humans.



THE DEVELOPMENT OF THE EAR The branchial apparatus In the fourth and fifth week of development in the human embryo, five ridges develop in the side walls of the primitive pharynx (Figure 66.1). These ridges are known as branchial arches. Initially they consist of a mesodermal



814 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



The middle ear 1



1



2



2 3



3



4



4 6



Branchial pouches Branchial clefts



Figure 66.1 Coronal section through the developing pharynx. The branchial arches are numbered 1, 2, 3, 4, 6. Redrawn from Ref. 10, with permission.



core surrounded by epithelium of endodermal and epidermal origin. These arches are separated by deep grooves, known as branchial clefts. At the same time as the development of the arches and clefts, a number of outpocketings known as the branchial pouches appear on the lateral wall of the pharynx. These pouches gradually penetrate the surrounding mesenchyme, but do not establish an open communication with the exterior. The mesodermal core of the branchial arches is pleuripotential in that it can give rise to any number of structures, for example cartilage, muscles or arteries. The development of the branchial apparatus is covered in Chapter 65, Head and neck embryology. Those features pertinent to the development of the ear are reviewed here.



The outer ear The pinna of the outer ear is formed by a gradual alteration and partial fusion of the first and second branchial arches, which include arch ectoderm and mesoderm (Figure 66.1). The groove between the two arches deepens and gives rise to the external acoustic canal (EAC). The development of the outer ear is hence controlled, in part, by genes that determine the identity of the first and second arches. This regulation is mediated through the segmentation of the developing hindbrain into rhombomeres. The hindbrain is the source of the neural crest cells that populate the branchial arches. Mutations in the genes that control the development of the fourth segment of the hindbrain, rhombomere 4, can cause abnormalities of the outer ear.11 These genes include Hoxa1,12 Hoxb1,13 goosecoid (Gsc)14, 15 and genes for the secreted peptide, endothelin 1 and its related molecules.16, 17, 18, 19



The embryological origins of the middle ear are diverse. Middle-ear morphogenesis can be altered under various conditions that cause congenital abnormalities in head development. Genes that are expressed in the first and second branchial arches and in the hindbrain segments that give rise to these arches, appear to have a significant role in patterning the middle ear. These include Gsc,14, 15 Dlx2,20 retinoic acid receptors (Rars),21 Prx1,22 Otx2,23 Hoxa1,12 Hoxb113 and endothelin-related molecules.16 The tympanic membrane is formed by the apposition of two epithelia provided by the EAC and the middle ear cavity, with a fibrous layer between them. Embryologically, the epithelia of the EAC and middle ear cavity derive from the ectoderm of the first branchial cleft and the endoderm of the corresponding branchial pouch respectively (Figure 66.1). The fibrous layer originates from the branchial arch mesenchyme.24 A functional tympanic membrane is formed via successive reciprocal interactions between the branchial arch mesenchyme and the overlying epithelia, and one of the main steps is the insertion of the handle of the malleus into the tympanic membrane. A central role of this process is played by the tympanic ring, which is a C-shaped membranous bone that develops from midbrain and rhombomere 1 neural crest and provides physical support to the tympanic membrane.25 Ossicles are developed from the mesenchyme of the proximal area of the branchial arches. The malleus, originating from the midbrain and rhombomeres 1 and 2 crest cells, and the incus, originating from rhombomeres 1 and 2 crest cells, both derive from the first branchial arch. The second branchial arch element of the middle ear, the stapes, derives from rhombomere 4.25 The ossicles are derivatives of the neural crest. Proper development of the branchial arches, and thus the ossicles, depends on information provided by the neural crest, the mesoderm, and the overlying epithelia. The vertebrate ear develops from a complex convergence of tissues from all three germ layers (endoderm, mesoderm and ectoderm), as well as the neural crest. A major contribution to the outer ear is made from the first and second branchial arches, which include the arch ectoderm and mesoderm. The middle ear is formed from cells arising from the neural crest and paraxial mesodermal cells that initially migrated to the branchial arches and then coalesced to form the middle ear components. It also contains contributions from the endoderm, as it is lined by mucosal tissue that expands into the middle ear from the oral cavity. Embryologically, this is derived from the first branchial pouch. The epithelial tissues and neurons of the inner ear are formed of placodal ectoderm, which invaginates to form the otocyst, a cartilaginous capsule that eventually ossifies and surrounds the epithelial core of the inner ear. Most of the otic capsule is derived from paraxial mesoderm, and forms through epithelial–mesenchymal interactions with



Chapter 66 Molecular otology, development of the auditory system and recent advances in genetic manipulation



] 815



absence of most of the malleus, provided that the EAC is present. On the medial side of the ossicular chain, the formations of the stapes and the oval window appear to be controlled by the Hoxa2 gene;27 the insertion of the stapes into the oval window seems to be regulated by the Pou3f4 gene.28



Midbrain Hindbrain R1 R2 R3 R4 R5



The inner ear



R6



B1



R7 B2 B3/B4



Otocyst Neural crest origin R3



Posterior midbrain



R4



Posterior midbrain + R1



R5



R1



R6



R1 + R2



R7



Posterior midbrain + R1 + R2



R6 + R7



Figure 66.2 Segmental emigration of cranial neural crest. B1–4, branchial arches 1 to 4; R1–7, rhombomeres 1 to 7. Modified from: http://ucl.ac.uk/wibr/2/research/funcdev/gko.htm



the membranous labyrinth that surrounds and protects (Figure 66.2). Several molecules including endothelin 1, Fgf8 and retinoic acid (RA) have been shown to play specific signalling roles in the development of the branchial arches and middle ear structures.26 The timing of the migration of neural crest cells that will form each of the different middle ear components follows a typical sequence. The first neural crest cells to migrate are those contributing to the head of the malleus, followed by those forming the head of incus. The next to migrate are rhombomere 4 crest cells contributing to the stapedial footplate. The crest cells that make the tympanic ring and the handle of malleus start their migrations shortly after. This is then followed by the migration of crest cells that form the stapedial arch and those contributing to the neck of the malleus. The last cells to migrate are those that form the long process of the incus.25 The process involves more than just the formation of individual components. These have to be connected in a normal and orderly fashion, so that the middle ear can function properly. The insertion of the handle of the malleus between the two epithelial layers of the tympanic membrane relies on the proper formation of the EAC. The handle of the malleus can be formed even in the



During embryonic development, the inner ear arises from a simple epithelium adjacent to the hindbrain, the otic placode, that is formed through inductive interactions with surrounding tissues. The otic placode, which is a thickened area of ectoderm next to the hindbrain rhombomeres 5 and 6 in 8–10 somite embryos, invaginates to form the otic cup, an ellipsoid-shaped structure lined by a pseudostratified epithelium. The cochlear and vestibular neurons are formed by delamination from the otic cup. Initially, they form a single structure, the cochleovestibular ganglion, which divides into the cochlear and vestibular ganglia as development proceeds. The otic cup then ‘pinches off ’ to form a hollow ball of cells called the otic vesicle (otocyst) (Figure 66.2). The inner ear then enlarges, while undergoing complex changes, in order to assume its final morphology. Hindbrain segmentation occurs at the same time that the otic placode arises and the hindbrain could be crucial for otic placode specification. By the time the otocyst that is adjacent to rhombomere 5 has formed, hindbrain segmentation is complete. At this time, neural crest cells migrate out of the hindbrain and stream around the adjacent otocyst as they descend ventrally to form the branchial arches, together with their surrounding mesodermal cells, providing information transfer from neural ectoderm to the otic field.11 The otic vesicle (otocyst) enters an early proliferation phase followed by a differentiation phase, during which morphogenesis and patterned development of specific cell types occurs. Basic processes in the development of the otic vesicle have been shown to include the initiation and completion of each of the following overlapping steps.29  Placodal competence is when the prospective placodal ectodermal stripe is able to develop into any placode derivatives, if placed in the adequate position.  Specification of the otic field then enables the placodally competent ectoderm to be restricted to the prospective ear region.  Determination occurs when the otic field progressively acquires its identity and becomes committed to its otic fate.  Regional and cell-fate specification of the otic vesicle may be related to the earliest stage of regional gene expression and to the first events of cell differentiation, respectively.



816 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Structurally in mice, the cochlear duct develops from a tubular out-pocketing of the ventral portion of the otocyst on embryonic day 11 (E11) and starts to coil. By embryonic day 17 (E17), it attains its mature shape of one and three-quarters turns. The organ of Corti develops from the thickened epithelial cells of the cochlear duct. This thickened area is formed of tall columnar cell mounds: the lesser and the greater epithelial ridges. It is believed that inner hair cells (IHC) derive from the greater epithelial ridge and the outer hair cells (OHC) derive from the lesser ridge. Pax2 has been shown to be responsible for the outgrowth of the cochlea. The main events of semicircular canal formation in a developing mouse occur between 11.5 and 13 days embryonically (E11.5–E13). From the medial aspect of the otocyst, an uncoiled tube grows out and projects dorsally. The first two semicircular canals to form, the superior and the posterior, develop from a shared pocketshaped rudiment. This out-pocketing begins to develop from the dorsal part of the otocyst by E11.5 and shows the first signs of conversion into a pair of semicircular canals. The canals are formed by fusion of the lateral and medial epithelial surfaces of the out-pocketing and the subsequent disappearance of the fused region of epithelia. The superior canal forms first, followed closely by the posterior; the horizontal canal forms last from a separate rudiment. By E13, all the major changes that



the inner ear undergoes are complete and the inner ear labyrinth resembles a miniature replica of its adult form.30 The utricle and its macula are derived from the upper middle third of the medial and lateral walls of the otocyst. The saccule and its macula are derived from the lower middle third of the medial wall of the otocyst (Figure 66.3).



Origin of the sensory organs The inner ear houses the receptor cells for two distinct sensory pathways: the auditory and vestibular systems. Patterning of the inner ear into prospective auditory and vestibular sensory areas is associated with restricted gene expression domains during the stages of development from otic placode to otocyst. Ventromedial areas, where the putative auditory epithelium forms, express Pax2, Dlx4, Fgf2, Fgf3, BMP4, notch and Ncam genes. Dorsolateral areas which form vestibular epithelium express Hmx2, Hmx3, Dlx3, MshC, MshD, Sox9, p75, Lmx1, Gbx3, Sek1, BMP7 and Igf1 genes.31 Once the sensory hair cells have been specified, their continued differentiation requires transcription factor Pou4f3. In the absence of this gene, the hair cell phenotype aborts at an early stage, eventually leading to the death of hair cells and the supporting cells.32



M P



A



*



L



E



S



CC



H



U SA



C (a)



(b)



(c)



Figure 66.3 Schematic transected diagrams of inner ear morphogenesis during the period of semicircular canal development (left inner ear). (a) The walls of the canal plate have met in the region of the superior semicircular canal (arrow). (b) The fused layer of canal plate epithelium in the region of the superior canal has now disappeared (arrow), during which time the posterior canal walls have come together and are now apposed (asterisk). (c) All fused regions of superior and posterior canal plate have now disappeared leaving two canal tubes that are united centrally by the crus commune. A, anterior; C, cochlea; CC, crus commune; E, endolymphatic duct; H, horizontal; L, lateral; M, medial; P, posterior; SA, saccule; S, superior; U, utricle. Redrawn from Ref. 30, r 1993, with permission from Elsevier.



D



Chapter 66 Molecular otology, development of the auditory system and recent advances in genetic manipulation



Modifier genes affecting the development of the auditory system More often than not, a given genetic disorder may manifest in a variety of ways. Variable phenotypes can result from genetic modifier loci. Modifier genes affect the phenotypic outcome of a given genotype by interacting in the same or parallel biological pathway as a disease gene. The effect can be enhancing, leading to a more severe mutant phenotype, or suppressive, reducing the mutant phenotype even to the extent of completely restoring the normal condition. Modifier genes can also alter the clinical presentations of a given disease, resulting in different combinations of traits. Moreover, multiple modifier genes may act in combination to create a cumulative effect on the expression of a phenotype. Human modifiers of hearing include DFNM1 locus which suppresses hearing loss among individuals with DFNB26 mutation4 and a locus on chromosome 8, which modifies maternally inherited aminoglycosideinduced deafness associated with a mutation (A1555G) in the mitochondrial 12S ribosome RNA.33



RECENT ADVANCES IN GENETIC MANIPULATION IN THE TREATMENT OF CONGENITAL DEAFNESS The purpose of any medical research is to apply the findings to clinical use. It may one day be possible to reverse certain types of defect that affect the early development of the auditory and vestibular systems. A prerequisite for treatment will almost certainly be early and accurate diagnosis. Screening programmes can then be established using molecular techniques. Groups with strong family histories of vestibular disorders and/or deafness with any related syndromes and premature newborns should be targeted for screening in the first instance. Molecular techniques can scan for the genetic defect at several levels. They may be used in linkage analysis, where the location of the gene on a particular chromosome is identified; or the techniques may identify the actual gene, the mutation within the gene and the gene product.34 Some childhood hearing loss may be progressive during the first few months or years of life. Early diagnosis of hearing impairment in children would be essential if medical or surgical interventions were to stand a good chance of preserving some hearing ability. Intervention would depend on us having a much greater understanding of the likelihood of developing a serious hearing impairment than we presently have, so that a realistic prognosis could be provided to aid a family’s decision.35



Gene transfer The prospect of therapeutic insertion of genetic material into the inner ear promises to be an attractive technology



] 817



for clinical applications. It offers the hope of preventing, arresting, reversing or curing vestibular or hearing disorders caused by hereditary diseases or environmental insults. There are currently very few treatment options for vestibular disorders and sensorineural hearing loss and, therefore, it is important to investigate and develop new technologies for inner ear disease. In utero gene transfer may indeed be an option, but this technology will have many technical and ethical issues that remain to be overcome. A number of different gene transfer vectors have been studied in vivo for their efficacy, utility and safety in intracochlear gene transfer. Vectors successfully studied include cationic liposomes, adeno-associated virus, adenovirus, lentivirus, herpes simplex virus and vaccinia virus, with variable intensities of transgene expression in different parts of the organ of Corti.36 For experimental purposes as well as future clinical application, it is important to develop a vector delivery system that preserves hearing and vestibular functions, as well as inner ear architecture. Systemic administration of the vectors has the disadvantage of systemic toxicity including viral hepatitis and encephalitis.37 Direct instillation, on the other hand, allows high vector concentration to be delivered to the target organ, using a small amount of vector solution, which increases the likelihood of gene delivery to the targeted tissue, while minimizing leakage of the vector to the surrounding organs. Several techniques of intracochlear vector delivery have been developed and their efficacy and safety have been compared and contrasted.36 They include miniosmotic pump infusion or microinjection into the scala tympani via the round window, infusion or microinjection into the scala tympani through a cochleostomy, injection into the endolymphatic sac accessed from the posterior fossa and gel foam application to the round window membrane. The first three methods have the advantage of a controlled steady intracochlear perfusion, but have the disadvantage of potentially physically damaging the architecture of the cochlea. Gelfoam soaked with the vector solution applied to the round window is noninvasive and nondestructive, and does not cause disruption to the anatomy of the inner ear.36 However, the exact mechanism of vector transfer through the round window membrane is poorly understood and hence there is no control over the rate of delivery. It has also been shown that transgene expression in the contralateral, noninjected cochlea occurs, suggesting that systemic dissemination could potentially be associated with direct delivery methods. Three potential routes of spread from the cochlea have been suggested and they include viral spread via the cochlear aqueduct which connects the scala tympani of the basal turn of the cochlea to the cerebrospinal fluid (CSF) space, transmission of vectors through the temporal bone marrow spaces and haematogenous dissemination.38 To date, there has been no report of gene transfer being used to successfully treat deafness in an animal model.



818 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Stem cell and gene therapy In addition to single gene transfer, there has been great interest in the potential of the twin technologies of stem cell and gene therapy for medicine. Hair cells in the cochlea are very vulnerable to disruption of their homeostasis and tend to die if they cannot function normally.39 Once dead, hair cells are not naturally replaced. One approach to treating deafness could be to trigger regeneration of cochlear hair cells. However, there is no point in stimulating the regeneration of hair cells that are unable to function because of a genetic defect. Accurate diagnosis combined with dual treatment of stem cell insertion and gene transfer would be vital for this approach to succeed. Over recent times, the potential of somatic stem cells for therapeutic applications has become almost infinite, limited only by the ingenuity of investigators in the manipulation of their genomes and culture conditions.40 Cochlear gene transfer can offer several potential applications including the study of specific expression and function of certain genes and the delivery of therapeutic agents. However, demonstration of transgene expression in tissues outside the treated cochlea raises concerns about the safety of its application regarding systemic dissemination of the vectors used.36 Better techniques to improve and enhance specific vector delivery into the cochlea need to be developed and tested.



Timing of intervention In humans, the otic placode starts to develop when the embryo has reached the seventh somite stage (about 22 days) and by 25 weeks, the development of the organ of Corti and the vestibular system are complete and resemble the adult form. Prenatal tests for congenital abnormalities including Down syndrome, cystic fibrosis and spina bifida are performed by the end of the first trimester and it would be too late if interventions were to be implemented to the developing ear at this stage. The rudimentary pinna has formed by 60 days and in the fourth month convolutions have attained their adult form, although further enlargement continues during the remaining months of gestation and also in the post-natal period.41 Timing of genetic intervention would therefore be important. Direct gene transfer or stem cell insertion into the inner ear at the embryonic stage would involve some form of intrauterine surgery and surgical intervention on the human foetus has been performed for more than two decades in the United States.42 In the past, only foetuses with life-threatening defects, for example, giant foetal cervical teratoma43 have been considered as candidates for prenatal correction. However, more foetal surgical procedures are now being performed for nonlethal conditions, such as congenital diaphagmatic hernia



using fetoscopic repair techniques.42 Surgical intervention involving inner ear gene transfer on human embryos seems a good idea, but is still rather far-fetched with the technology available to us at present. It may be more feasible to treat post-natal progressive deafness and/or vestibular disorders by using this approach, but the exact pattern of expression and functions of the genes involved must be understood prior to such intervention. In addition, the attitudes of the patients and their families and their views on such intervention must be respected. The deaf community in the United Kingdom is very advanced, with its own culture and there has been anecdotal evidence that deaf couples do not necessarily want their children to be able to hear if it means that they will be excluded from the deaf community to which their parents belong (see Chapter 69, Investigation and management of the deaf child).



Drug therapy An alternative approach is based on drugs, which, by definition, are small molecules with ready access to the target cells. We may be able to exploit alternative pathways to carry out the task that is affected by a mutation. For example, connexin 26 mutation is thought to the most common cause of nonsyndromic deafness in the Caucasian population. Another connexin may be capable of substituting for connexin 26 in forming gap junctions, but its gene may not normally be expressed in the cochlea. A drug might therefore be developed to activate expression of the alternative connexin gene in the cells needing to form gap junctions. Clues about which alternative pathways might be worth exploiting will come from investigations of interacting genes.35



KEY POINTS  Congenital deafness is a common childhood problem.  Genetic abnormalities cause not only congenital deafness; several genes have been found that, when mutated, either cause or predispose to progressive hearing loss.  Mice appear to be the best model for the study of genetic deafness.  Understanding the molecular biology of the auditory system may be the key to the future development of genetic counselling and treatment for congenital deafness.  Surgical intervention involving inner ear gene transfer on human embryos seems a good idea, but is still rather far-fetched with the technology currently available.



Chapter 66 Molecular otology, development of the auditory system and recent advances in genetic manipulation



Deficiencies in current knowledge and areas for future research Genetic testing for deafness is now a reality and has changed the paradigm for evaluating deaf patients. It will be used by surgeons for diagnostic purposes and as a basis for treatment and management options. Mutation screening is currently available for only a limited number of genes, such as connexin 26. In these cases, diagnosis, carrier detection and reproductive risk counselling can be provided. In the future, there will be an expansion of the role of genetic testing and counselling will not be limited to reproductive planning. Treatment and management decisions will be made based on specific genetic diagnoses. New discoveries and technologies will expand and increase the complexity of genetic testing options. Otolaryngologists will therefore have to familiarize themselves with current discoveries and accepted protocols for genetic testing.



REFERENCES











1. Fortnum HM, Summerfield AQ, Marshall DH, Davis AC, Bamford JM. Prevalence of permanent childhood hearing impairment in the United Kingdom and implications for universal neonatal hearing screening: Questionnaire based ascertainment study. British Medical Journal. 2001; 323: 536–40. A clinically relevant article explaining the incidence of congenital deafness and the need for universal neonatal hearing screening 2. Marazita ML, Ploughman LM, Rawlings B, Remington E, Arnos KS, Nance WE. Genetic epidemiological studies of early-onset deafness in the U.S. school-age population. American Journal of Medical Genetics. 1993; 46: 486–91. 3. Steel KP. Progress in progressive hearing loss. Science. 1998; 279: 1870–1. 4. Riazuddin S, Castelein CM, Ahmed ZM, Lalwani AK, Mastroianni MA, Naz S et al. Dominant modifier DFNM1 suppresses recessive deafness DFNB26. Nature Genetics. 2000; 26: 431–4. 5. Van Camp G, Smith R. Hereditary Hearing Loss homepage. 2005. http://webhost.ua.ac.be/hhh/ The most up-to-date source of information in the genes identified in congenital deafness. 6. Verpy E, Masmoudi S, Zwaenepoel I, Leibovici M, Hutchin TP, Del Castillo I et al. Mutations in a new gene encoding a protein of the hair bundle cause non-syndromic deafness at the DFNB16 locus. Nature Genetics. 2001; 29: 345–9. 7. Steel KP, Bock GR. Hereditary inner-ear abnormalities in animals. Relationships with human abnormalities. Archives of Otolaryngology. 1983; 109: 22–9. 8. Steel KP. Inherited hearing defects in mice. Annual Review of Genetics. 1995; 29: 675–701.



] 819



9. Kiernan A, Steel K. Mouse homologues for human deafness. In: Kitamura K, Steel K (eds). Genetics in otorhinolaryngology. Advances in otorhinolaryngology. Basel: Karger, 2000; 56: 233–43. 10. Shanahan D. Cyberanatomy of the branchial arches. Newcastle-upon-Tyne: Newcastle University. Last updated 19th June 2007; cited May 2007. Available from: http:// anatome.ncl.ac.uk/tutorials/clinical/arch/text/page1.html 11. Fekete DM, Wu DK. Revisiting cell fate specification in the inner ear. Current Opinion in Neurobiology. 2002; 12: 35–42. 12. Lufkin T, Dierich A, LeMeur M, Mark M, Chambon P. Disruption of the Hox-1.6 homeobox gene results in defects in a region corresponding to its rostral domain of expression. Cell. 1991; 66: 1105–19. 13. Gavalas A, Studer M, Lumsden A, Rijli FM, Krumlauf R, Chambon P. Hoxa1 and Hoxb1 synergize in patterning the hindbrain, cranial nerves and second pharyngeal arch. Development. 1998; 125: 1123–36. 14. Rivera-Perez JA, Wakamiya M, Behringer RR. Goosecoid cell acts autonomously in mesenchyme-derived tissues during craniofacial development. Development. 1999; 126: 3811–21. 15. Yamada G, Mansouri A, Torres M, Stuart ET, Blum M, Schultz M et al. Targeted mutation of the murine goosecoid gene results in craniofacial defects and neonatal death. Development. 1995; 121: 2917–22. 16. Kurihara Y, Kurihara H, Suzuki H, Kodama T, Maemura K, Nagai R et al. Elevated blood pressure and craniofacial abnormalities in mice deficient in endothelin-1. Nature. 1994; 368: 703–10. 17. Yanagisawa H, Yanagisawa M, Kapur RP, Richardson JA, Williams SC, Clouthier DE et al. Dual genetic pathways of endothelin-mediated intercellular signaling revealed by targeted disruption of endothelin converting enzyme-1 gene. Development. 1998; 125: 825–36. 18. Clouthier DE, Williams SC, Yanagisawa H, Wieduwilt M, Richardson JA, Yanagisawa M. Signaling pathways crucial for craniofacial development revealed by endothelin-A receptor-deficient mice. Developmental Biology. 2000; 217: 10–24. 19. Thomas T, Kurihara H, Yamagishi H, Kurihara Y, Yazaki Y, Olson EN et al. A signaling cascade involving endothelin-1, dHAND and msx1 regulates development of neural-crestderived branchial arch mesenchyme. Development. 1998; 125: 3005–14. 20. Qiu M, Bulfone A, Martinez S, Meneses JJ, Shimamura K, Pedersen RA et al. Null mutation of Dlx-2 results in abnormal morphogenesis of proximal first and second branchial arch derivatives and abnormal differentiation in the forebrain. Genes and Development. 1995; 9: 2523–38. 21. Lohnes D, Mark M, Mendelsohn C, Dolle P, Dierich A, Gorry P et al. Function of the retinoic acid receptors (RARs) during development (I). Craniofacial and skeletal abnormalities in RAR double mutants. Development. 1994; 120: 2723–48.



820 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY 22. Martin JF, Bradley A, Olson EN. The paired-like homeo box gene MHox is required for early events of skeletogenesis in multiple lineages. Genes and Development. 1995; 9: 1237–49. 23. Matsuo I, Kuratani S, Kimura C, Takeda N, Aizawa S. Mouse Otx2 functions in the formation and patterning of rostral head. Genes and Development. 1995; 9: 2646–58. 24. Mallo M. Formation of the middle ear: Recent progress on the developmental and molecular mechanisms. Developmental Biology. 2001; 231: 410–9. 25. Mallo M. Embryological and genetic aspects of middle ear development. International Journal of Developmental Biology. 1998; 42: 11–22. 26. Francis-West P, Ladher R, Barlow A, Graveson A. Signalling interactions during facial development. Mechanisms of Development. 1998; 75: 3–28. 27. Rijli FM, Mark M, Lakkaraju S, Dierich A, Dolle P, Chambon P. A homeotic transformation is generated in the rostral branchial region of the head by disruption of Hoxa-2, which acts as a selector gene. Cell. 1993; 75: 1333–49. 28. de Kok YJ, van der Maarel SM, Bitner-Glindzicz M, Huber I, Monaco AP, Malcolm S et al. Association between X-linked mixed deafness and mutations in the POU domain gene POU3F4. Science. 1995; 267: 685–8. 29. Torres M, Giraldez F. The development of the vertebrate inner ear. Mechanisms of Development. 1998; 71: 5–21. 30. Martin P, Swanson GJ. Descriptive and experimental analysis of the epithelial remodellings that control semicircular canal formation in the developing mouse inner ear. Developmental Biology. 1993; 159: 549–58. 31. Represa J, Frenz DA, Van De Water TR. Genetic patterning of embryonic inner ear development. Acta Otolaryngologica. 2000; 120: 5–10. 32. Xiang M, Gan L, Li D, Zhou L, Chen ZY, Wagner D et al. Role of the Brn-3 family of POU-domain genes in the development of the auditory/vestibular, somatosensory, and visual systems. Cold Spring Harbor Symposia on Quantitative Biology. 1997; 62: 325–36.















33. Bykhovskaya Y, Estivill X, Taylor K, Hang T, Hamon M, Casano RA et al. Candidate locus for a nuclear modifier gene for maternally inherited deafness. American Journal of Human Genetics. 2000; 66: 1905–10. 34. Smith SD, Kimberling WJ, Schaefer GB, Horton MB, Tinley ST. Medical genetic evaluation for the etiology of hearing loss in children. Journal of Communication Disorders. 1998; 31: 371–88; quiz 388–9. 35. Steel KP. Science, medicine, and the future: New interventions in hearing impairment. British Medical Journal. 2000; 320: 622–5. A beautifully written account of the science behind the various modes of possible gene therapy. 36. Lalwani AK, Jero J, Mhatre AN. Current issues in cochlear gene transfer. Audiology and Neurootology. 2002; 7: 146–51. A good summary of experiments that have been carried out in cochlear gene transfer in the world literature. 37. Lalwani AK, Walsh BJ, Carvalho GJ, Muzyczka N, Mhatre AN. Expression of adeno-associated virus integrated transgene within the mammalian vestibular organs. American Journal of Otology. 1998; 19: 390–5. 38. Kho ST, Pettis RM, Mhatre AN, Lalwani AK. Safety of adeno-associated virus as cochlear gene transfer vector: Analysis of distant spread beyond injected cochleae. Molecular Therapy. 2000; 2: 368–73. 39. Steel KP. Perspectives: Biomedicine. The benefits of recycling. Science. 1999; 285: 1363–4. 40. Lemoine NR. The power to deliver: stem cells in gene therapy. Gene Therapy. 2002; 9: 603–5. An interesting and easy to follow article on current issues in stem cell/ gene therapy in general. 41. Wright A. Anatomy and ultrastructure of the human ear. In: Kerr A. (ed.). Scott-Brown’s otolaryngology, Vol. 1. 1997: pp.1/1/6–8. 42. Farmer D. Fetal surgery. British Medical Journal. 2003; 326: 461–2. 43. Hirose S, Sydorak RM, Tsao K, Cauldwell CB, Newman KD, Mychaliska GB et al. Spectrum of intrapartum management strategies for giant fetal cervical teratoma. Journal of Pediatric Surgery. 2003; 38: 446–50.



67 Hearing loss in preschool children: screening and surveillance KAI UUS AND JOHN BAMFORD



Introduction Principles and definitions Prevalence and risk factors The rationale for screening Newborn hearing screening The eight-month IDT screen The school-entry screen



821 821 823 825 826 828 829



Surveillance Surveillance for otitis media with effusion Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



829 830 830 831 831 831



SEARCH STRATEGY The data in this chapter are supported by a Medline and CINAHL search for relevant articles published in English, using key words infant or newborn and hearing screening and hearing loss.



INTRODUCTION



PRINCIPLES AND DEFINITIONS



Any discussion of screening must be firmly grounded in an understanding of the principles of screening as a public health exercise, an understanding of some of the key evidence regarding the epidemiology of the condition of interest, in this case childhood hearing loss, and the evidence for or against different approaches to screening and surveillance for that condition. Thus, this chapter starts with an outline of the key definitions and principles of screening, before proceeding to review evidence on the prevalence of different types of childhood hearing loss. The case for newborn screening is summarized and evidence for current performance of newborn screens reviewed. Other approaches to screening and surveillance in the preschool years are also reviewed (the eight-month screen, the school entry screen) and the position with regard to temporary childhood hearing loss discussed.



Surveillance is a process of ongoing or regular observation of the health of individuals or populations. Traditionally health service professionals have considered surveillance essentially as secondary prevention through early detection. However, the concept that child health depends upon continuous vigilance and supervision by health professionals is fairly narrow and supervisory; child health surveillance is now regarded as just one component of child health promotion programmes which aim to promote partnership between parents, children and health professionals, in which parents are empowered to make use of services and expertise according to their needs. Screening is a public health activity in which members of a defined population, who do not show any obvious symptoms, are offered a test in order to identify those individuals who are more likely to be helped than harmed



822 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY by further assessment or treatment to diminish the risk of a disease (or impairment) or its complications (or consequences). Accurate early diagnosis of serious conditions offers the opportunity to initiate treatment before the disease progresses (or to intervene to mitigate the consequences of the impairment). When screening is done well, it can be very beneficial. Even though screening can potentially help to save lives or improve the quality of life through early diagnosis of serious conditions, it is not an infallible process. Screening can potentially reduce the risk of developing a disease (or impairment) or its complications (or consequences). However, in all screening programmes there are inevitably individuals who are referred as having the condition but do not (false positives), and who are not referred but do have the condition (false negatives). From an ethical viewpoint, screening differs from responsive clinical practice, since screening is targeting individuals who appear healthy and is offering to help individuals make an informed choice about their health or the health of their children. It is particularly important for individuals who consent to be screened or have their child screened to be aware of the risks involved and have reasonable expectations of a screening programme. In the UK, the National Screening Committee ‘assesses proposed new screening programmes against a set of internationally recognised criteria covering the condition, the test, the treatment options and the effectiveness and acceptability of the screening programme. Assessing programmes in this way is intended to ensure that they are more beneficial than harmful, and can be implemented at reasonable cost. When research has shown screening to be effective in reducing mortality and morbidity from a particular condition, the National Screening Committee may decide to pilot the proposed screening programme. This provides valuable information on the effectiveness, feasibility and public acceptability of screening when performed in an ordinary health service setting rather than a specialist research site. Pilot studies enable informed decisions to be made about policy priorities; ensuring resources are targeted on services that are proven to work’.1 The principles that a screening programme should satisfy have been identified in seminal work by Wilson and Junger.2 These were extended by Haggard and Hughes3 in their review of screening for hearing loss in children.  The condition (hearing impairment) should be an important health problem.  There should be an accepted treatment, i.e. an acceptable means of habilitation for those identified by the screen.  Facilities for assessment, diagnosis and treatment should be available.  The hearing impairment should be recognizable at an early stage.



 There should be a suitable test for use as the screen.  The test should be acceptable to the parents and to the child.  The natural history of the condition should be known and understood.  There should be an agreed policy on who to treat.  The cost of case finding (including all consequential costs of the screening programme) should not be disproportionate to overall healthcare costs of care for the hearing-impaired child.  Case finding should be seen as a continuing process.  The incidental harm should be small compared to the overall benefits.  There should be guidelines on how to explain results to parents with appropriate support.  All hearing screening arrangements should be reviewed in the light of changes in demography, epidemiology and other factors.  Cost and effectiveness of hearing screening should be examined on a case-type basis to maximize the effectiveness and benefit for each type before considering overall costs, effectiveness and benefits. There are a number of key definitions relevant to screening programmes which should be used in the quality assurance and audit of any screening programme: Coverage is the proportion of the target population who undergo the screen. Screen positive result is a screening result that is greater than or equal to a specified cut-off level. In the case of permanent childhood hearing impairment, there is debate about the appropriate cut-off level; at present in the UK the cut-off level is the estimated hearing threshold (averaged across 0.5, 1, 2 and 4 kHz) of 40 dB HL. This level is based upon evidence for the benefits of early intervention in moderate or greater hearing loss. Further research on the effects of and early intervention for mild hearing loss in the 20–40 dB HL range is ongoing. A truepositive is an individual with a screen-positive result who has the condition (impairment). A false-positive is an individual with a screen-positive result who does not have the target condition (impairment). Screen negative result is a screening result that is less than the specified cut-off level. A true-negative is an individual with a screen-negative result who does not have the condition (impairment). A false-negative is an individual with a screen-negative result who does in fact have the condition (impairment). Effective screening programmes attempt to reduce false-negatives to the irreducible minimum, while keeping false-positives within manageable service levels. The aim of a screen is to refer on a manageable proportion of the population for further (diagnostic) tests, that proportion being likely to contain as many of the true cases in the population as possible. Sensitivity is the rate of true-positives or the proportion of individuals with the target condition in



Chapter 67 Hearing loss in preschool children: screening and surveillance



the population who are correctly identified by the screen. The term can be applied to a screening test, if (as is the case with newborn hearing screening) the screen consists of more than one test, or to the screen as a whole, or indeed to the screening programme. Test sensitivity is the proportion of individuals who were given the test, have the condition (impairment) and were detected by the test. Screen sensitivity is the proportion of individuals who completed the screen who have the condition (impairment) and who were detected by the screen (which may consist of more than one test with conditional pass/refer rules). Programme sensitivity is the proportion of the subjects in the whole population with the specified condition (impairment) who are detected by the screening programme. It is a product of the screen sensitivity and the coverage achieved by the screen programme. Specificity is the rate of true-negatives or the proportion of individuals free of the target condition in the population who are correctly identified as such by the screen. As with sensitivity, the term specificity may be applied at the level of test, screen or programme. The latter is a product of specificity and coverage. Positive predictive value is the proportion of individuals with a positive test result who have the target condition. Yield is used to indicate the number of cases identified by a screen (e.g. in the first 12 months of the first phase of the Newborn Hearing Screening Programme (NHSP) in England, 40 true cases of bilateral permanent hearing loss were identified via the screen). The yield is sometimes expressed as the number of cases identified via the screen per 1000 individuals screened, thus allowing comparison with published prevalence figures, and acting as a surrogate for sensitivity (since sensitivity can only be established in retrospect, once all false-negative and missed cases have been found). The yield of a screen is affected by coverage and sensitivity, and affects the cost per case identified. Finally, the incremental yield is the number of true cases referred by a screen when any true cases that would have been or were identified by preceding screening, surveillance programmes or responsive services are excluded. Incidence is the number of new instances of the condition (impairment) occurring during a certain period (e.g. a year) in a specified population. Thus, in an average-sized health community with 5000 births per year, an incidence of between five and ten cases of congenital permanent moderate or greater bilateral hearing loss might be expected. Prevalence is the total number of individuals who have a given disease or condition (impairment) at a given point in time per population figure (e.g. per 1000 live births). Davis et al.4 have suggested a figure of 1.12 per 1000 for congenital permanent bilateral hearing impairment of moderate or greater degree (see under Prevalence and risk factors below for further discussion of prevalence figures).



] 823



The extended Wilson and Junger screening criteria identify costs and effectiveness as important issues for screening programmes (points 9 and 14 above). Cost-effectiveness analysis is an economic analysis used to compare effectiveness and cost of health interventions in which either (1) effects of the interventions are known to be equal and so the option to be recommended is that which is least (or less) costly (sometimes known as ‘costminimization analysis’) or (2) effects and costs differ across interventions; hence, the option to be recommended is that with the lowest (or lower) ratio of cost per unit of health gain, as implementation of this option will lead to the most (or more) effective use of a fixed budget.



PREVALENCE AND RISK FACTORS Congenital permanent bilateral hearing loss The quality of published studies on the prevalence of congenital permanent hearing loss is variable, with the poorer studies marred by lack of clarity on case definition, uncertain methodology and doubtful case ascertainment. Of the better studies, those reported by Davis et al.4 and Fortnum et al.5, 6 are probably the most extensive and reliable. Fortnum et al.5 report the results of a retrospective ascertainment of all cases of permanent bilateral hearing loss of moderate or greater degree in children born in the UK’s Trent health region between 1985 and 1993. Considerable effort was put into the ascertainment process such that the authors believe that over 90 percent of all known cases were found. Based on the birth cohorts for 1985–1990, the prevalence of moderate-to-profound congenital permanent bilateral hearing loss was 1.12 (95 percent CI 1.01–1.23) per 1000 live births (0.64 per 1000 for moderate hearing loss; 0.23 per 1000 for severe hearing loss; 0.24 per 1000 for profound hearing loss). An even more extensive study is reported by Fortnum et al.6 In this, the authors carried out a full national retrospective ascertainment of all cases of bilateral permanent hearing loss of moderate or greater degree in the UK, born between 1980 and 1997. Data were collected in years 1998–1999 within a strict ethical framework from both health (audiology) and education (support services for hearing impaired children). Once duplicate data had been excluded, there were 17,160 cases, giving an overall prevalence of 1.33 per 1000 births (95 percent CI 1.22–1.45). However, prevalence varied as a function of age, being 0.91 per 1000 (95 percent CI 0.85–0.98) for the three-year-old cohort and rising to 1.65 per 1000 (95 percent CI 1.62–1.68) for the nine-year-old cohort, where it levelled off. The authors argue that the rise in prevalence with age is not because of any change in prevalence with year of birth, nor can it be ascribed only to later acquired



824 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY losses, and they conclude that there are more late-onset and progressive permanent childhood hearing losses than previously suspected. Furthermore, statistical adjustment of the data indicated the possibility of under-ascertainment, such that the final prevalence figure at age nine years might be as high as 2.0 per thousand. While these data support the use of newborn hearing screening for the identification of the 1.12 (or thereabouts) congenital cases per 1000 births, other processes will be needed for the post-natal onset cases. [**]



Risk factors Risk factors for permanent congenital hearing loss are well established. The risk factors for permanent hearing loss in neonates as suggested by the US Joint Committee on Infant Hearing position statement 2000 are as follows.7  An illness or condition requiring admission of 48 hours or greater to a NICU.  Stigmata or other findings associated with a syndrome known to include a sensorineural and/or conductive hearing loss.  Family history of permanent childhood sensorineural hearing loss.  Craniofacial anomalies, including those with morphological abnormalities of the pinna and ear canal.  In utero infection such as cytomegalovirus, herpes, toxoplasmosis or rubella. The three major summary risk factors are: 1. history of treatment in an neonatal intensive care unit (NICU) or special care baby unit (SCBU) for more than 48 hours; 2. family history of early childhood deafness; 3. craniofacial anomaly (e.g. cleft palate) associated with hearing impairment.4, 8 About 60 percent of congenital bilateral permanent hearing loss of moderate degree or greater is associated with one or more of these three risk factors, in the proportions 29.3 percent NICU, 26.7 percent family history and 3.9 percent craniofacial anomaly. [**] The high proportion of cases with risk factors led, in the early 1990s, to the widespread introduction of ‘at risk’ newborn screening in which attempts were made to screen all those babies (perhaps some 10 percent of the birth cohort) with risk factors, in order to identify early this 60 percent of the target true cases of congenital permanent loss. However, in practice, due to the difficulty experienced by maternity services in reliably identifying a family history of permanent childhood hearing loss, the proportion of the target population identified by at risk screening was rarely above 40 percent.4 [**]



Acquired and late-onset permanent bilateral hearing loss Acquired hearing loss is a hearing loss acquired postnatally which, on the basis of case history, was not considered to be present and detectable using appropriate tests at or very soon after birth. Post-meningitic hearing loss and hearing loss due to trauma are the most common causes of acquired hearing loss in children. The study by Fortnum et al.6 has indicated that progressive or late-onset hearing losses may be more prevalent than suspected, having an effect through the preschool and early school years (see Prevalence and risk factors). Any newborn hearing screening programme will of course not identify these cases and this therefore points to the importance of addressing other means of finding them.



Unilateral permanent hearing loss There are few reliable studies on the prevalence of unilateral permanent hearing loss in preschool children. However, there is some reason to suspect that prevalence may be relatively high. In a retrospective study of children up to the age of ten years, Vartiainen and Karjalainen9 found a prevalence of 1.7 per 1000 live births for permanent unilateral sensorineural or mixed hearing loss (425 dB HL at 0.5–4 kHz, including mixed hearing impairments with bone conduction thresholds of Z25 dB HL). [**] The yield for permanent unilateral moderate or greater hearing loss in the Newborn Hearing Screening Programme in England is 0.64 (95 percent CI 0.37–0.91) per 1000 screened. This is new information and has raised a number of issues about the appropriate early management of unilateral permanent hearing loss that have not been faced before.10 [**]



Auditory neuropathy/dys-synchrony Auditory neuropathy/auditory dys-synchrony (AN/AD) is caused by damage to inner hair cells, the synaptic juncture between the inner hair cells, auditory neurons in the spiral ganglion, the VIIIth nerve fibres, or any combination. This is observed on clinical audiological tests as normal otoacoustic emissions (OAEs) in the presence of an absent or severely abnormal auditory brainstem response (ABR). The prevalence of AN/AD is not known. It has been reported that up to 10 percent of all children with confirmed permanent hearing loss have auditory neuropathy.11, 12 Infants and children diagnosed with AN/AD typically have a remarkable medical history. It is a condition found predominantly in the NICU population: three in 1000 high-risk babies present with AN/AD13 and it is considered relatively rare in the well-baby population.12 [**]



Chapter 67 Hearing loss in preschool children: screening and surveillance



Temporary childhood hearing loss Temporary childhood hearing loss due to otitis media with effusion (OME) is extremely common, with a point prevalence of some 20 percent and a period prevalence in the under five-year-olds of some 80 percent.3, 14 [**] The major risk factors are season, passive smoking, bottlefeeding, upper respiratory tract infections, admission to NICU as a newborn, day care and siblings having had OME.3 [**] The large scale TARGET (Trials of Alternative Regimes in Glue Ear Treatment) studies currently being carried out in the UK are likely to provide much-needed evidence in a number of domains; careful recruitment of over 300 cases aged 3.5 to 7 years has provided more robust data than have been available previously. Refinement of risk factor analyses has indicated season, passive smoking, siblings having had OME and snoring and mouth-breathing as the best indicators for persistence.15 The TARGET studies indicate material effects of persistent OME on hearing difficulties, speech/language delay, disturbed sleep patterns, behaviour and (consequent on these) parental and child quality of life. [****]



THE RATIONALE FOR SCREENING In this section, we will consider screening only with respect to permanent childhood hearing loss. The critical review (1997) of newborn hearing screening carried out as part of the UK’s Health Technology Assessment programme4 identified eight broad reasons for the introduction of newborn hearing screening for all babies. 1. Outcomes for at least some children with congenital permanent hearing loss are the cause of considerable concern. 2. Median identification age in the UK for moderate or greater bilateral congenital permanent hearing loss based on current screening tests was around 22 months. 3. There is evidence that intervention in the first six months of life for children with moderate or greater permanent bilateral hearing loss can improve at least some outcomes. 4. More precise and detailed neural connections depend upon appropriate early stimulation; myelination of auditory pathways by six months of age is delayed by almost any chronic insult. 5. Early identification allows early and more appropriate management decisions, made not from a starting point of developmental deficit. 6. Costs of newborn screening look to be broadly acceptable. 7. Evidence from parents of deaf children strongly suggest that they would have welcomed very early identification.



] 825



8. Evidence from parents suggest that they would welcome a newborn hearing screen.



Early deprivation and early plasticity Studies with animals have shown that deafferentation of the auditory system during early sensitive periods of development results in marked anatomic and physiological changes that occur extensively throughout the central auditory system.16, 17, 18 [****] The first study to demonstrate that, within the sensitive period, changes wrought by loss of afferent input could be reversed was carried out by Pasic and Rubel.19 They showed that in neonatal gerbils, changes in cochlear nucleus cell size produced by auditory nerve blocking by tetrodotoxin were completely reversible after seven days. [****] There appears to be considerable plasticity in the whole pathway during early development, but in the adult subject plasticity at lower neural levels (e.g. brainstem, midbrain) is considerably reduced and possibly lost.20 This indicates that the early post-natal period is crucial for the establishment of auditory pathways that can accurately represent complex sounds at the cortical level. There appears to be a critical period of plasticity at lower pathways of the auditory system. [****]



Outcomes There is good evidence that outcomes in a number of domains are affected by permanent hearing loss. In most cases this is a severity-dependent effect, with greater effects for more severe hearing losses; there are, however, very great individual differences, with some children performing at levels appropriate to their age and others performing very poorly. These differences are not always a function of degree of hearing loss, but will be a result of a range of intrinsic and extrinsic factors. One such factor is the age at which the hearing loss was identified, as well as habilitative support provided. The evidence of compromised outcomes associated with congenital permanent hearing loss comes from studies looking at communication skills,21 literacy,22 behaviour,23 educational achievement,24 mental health,25 family dynamics26 and quality of life.23 [**]



Late identification Since the late 1950s, the UK has had a universal eightmonth screen for childhood hearing loss based on the infant distraction test (IDT) (see Chapter 68, Hearing tests in children) and usually carried out by health visitors in community clinic settings. With the introduction of newborn hearing screening this is to be phased out; there is good evidence that it has poor sensitivity,



826 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY poor specificity and is not cost effective in terms of cost per case found.5, 27, 28 The retrospective study of permanent childhood hearing impairment greater than or equal to 40 dB HL in children born between 1985 and 1993 and resident in Trent health region, showed that the median ages at referral, confirmation of the impairment, prescription of hearing aids and fitting of hearing aids were, respectively, 10.4, 18.1, 24.4 and 26.3 months.5 [**] While much of the delay was clearly due to delays in the assessment process after referral, nonetheless the median age of screen referral for children with congenital hearing loss was little short of one year.



Costs



Benefits of early identification and intervention



Parental wishes



Yoshinaga-Itano et al.29 compared the receptive and expressive language abilities of 72 children with hearing loss who had been identified by six months of age with 78 children whose hearing losses were identified after six months and showed that early-identified children demonstrated significantly better language scores than the later-identified children. For children with normal cognitive abilities, this language advantage was found across all test ages, communication modes, degrees of hearing loss and socioeconomic backgrounds. It was also independent of gender, ethnicity and the presence or absence of additional disabilities. [***] This finding is consistent with a number of similar studies of variable quality, all reporting significantly better language scores for children whose hearing losses were identified earlier.30, 31, 32, 33, 34, 35, 36, 37 [***/**] Even though evidence from randomized controlled trials is not available to address the question of whether earlier rather than later intervention is better for children with hearing loss, the consistency of findings from a number of quasi-experimental studies provides consistent and convincing evidence about the benefits of earlier intervention.



Assessment and management Some aspects of audiological assessment are easier in the first few months of life: for example, unsedated electrophysiological tests such as the auditory brainstem response, and habilitative procedures such as real ear measurements for hearing aid fittings, or ear impressions for earmoulds. Perhaps more importantly, the earlier that parents and services know about the child’s hearing loss, its degree and the child’s progress, the earlier can decisions be made about management (e.g. appropriate hearing aids, first language, extent and type of family support).



While further cost-effectiveness studies are required, it is clear that the costs of newborn screening are broadly acceptable.4 Stevens et al.28 compared the costs of newborn hearing screening of all babies in three areas already implementing such a service, with costs of the IDT screen in seven areas, and found that the mean service costs for the universal newborn hearing screen and the eight-month IDT screen were £13,881 and £24,519, respectively, for a standardized district of 1000 live births; the cost per case found was much lower for newborn screening due to the higher sensitivity of the screen. [**]



Finally, there is the issue of parental rights to significant information about their child. This has been given expression in a number of surveys, the largest of which is that by Watkin et al.38 He investigated the attitudes of parents of deaf children to newborn hearing screening. If such a procedure had been available to them when their child was born, 89 percent said they would have wanted it. It is also known that early identification tends to avoid the parental anxiety and anger that may be associated with delayed detection.38, 39 [**]



NEWBORN HEARING SCREENING Current screening tests available The tests used for newborn hearing screening are currently one or more of:    



automated otoacoustic emissions (AOAE); transient evoked otoacoustic emissions (TEOAE); distortion product otoacoustic emissions (DPOAE); automated auditory brainstem response (AABR).



TEOAE reflect the activity in the outer hair cells of the cochlea, while AABR waveforms are affected by cochlear and neural lesions. These methods are described in detail in Chapter 68, Hearing tests in children and Chapter 69, Investigation and management of the deaf child. It is important to note that a newborn hearing screening programme protocol that regards a clear OAE result as a screen pass will miss neonates with auditory neuropathy, since the latter will manifest with OAE present, but ABR absent. Because auditory neuropathy is predominantly found in the NICU/SCBU population (and according to Sininger13 as many as three per 1000 neonates enrolled in NICU/SCBU may have auditory neuropathy), AABR is the method of choice in the NICU/ SCBU population. [**] In the UK Newborn Hearing Screening Programme (www.hearing.screening.nhs.uk),



Chapter 67 Hearing loss in preschool children: screening and surveillance



both AABR and TEOAE is required for NICU babies, while for well babies the two tests are done in series, with only those without clear bilateral responses on OAE going on to AABR. This ordering is driven by the time required and the costs associated with AABR as opposed to OEA.



Cut-off level and case definition While many North American programmes40, 41 target all babies with permanent hearing loss, including mild and unilateral hearing losses, the Newborn Hearing Screening Programme in England aims to identify all children with a permanent hearing loss of 40 dB HL or greater averaged over the frequencies 0.5, 1.0, 2.0 and 4.0 kHz for the better hearing ear.4 As a byproduct, the screen will also identify a number of babies who have unilateral and in some cases mild hearing loss, but the consequences of delay in identification are not well established in these infants, and most are not candidates for hearing aids or other therapies associated with early identification.42 [**] However, this is lack of evidence for effects rather than evidence of no effects, and research is underway to identify best practice for the early management of congenital permanent mild and unilateral hearing loss.



Models for universal newborn hearing screening The most commonly used model to date has been a hospital-based screen employing a team of dedicated screeners performing the screening tests on neonates on the maternity unit before discharge. This is backed up by a recall clinic for the babies missed by the initial screen.27, 42 However, in some European countries, a communitybased screening model integrated into an existing framework of preventive health care has been used with considerable success.43, 44 Owen et al.45 reported that it was feasible for health visitors to perform neonatal hearing screening in the community and achieve very high coverage and low false-positive rates. The assumption is that community-based screening also results in higher parental satisfaction, but this is yet to be established. [**] Some areas in England are using a community-, rather than hospital-based model and costeffectiveness comparisons are being undertaken.



‘At-risk’ screening Risk factors for permanent congenital hearing loss are well established. Davis and Wood8 showed that babies admitted to a neonatal intensive care unit for more than 48 hours were 10.2 (95 percent CI 4.4–23.7) times more likely to have a permanent hearing loss (greater than 50 dB HL in this study) than those who did not undergo



] 827



intensive care. Fortnum et al.5 showed that prevalence was increased 14-fold for children with a family history of early permanent childhood deafness. Babies with craniofacial anomalies associated with hearing impairment (e.g. cleft palate) are also at high risk for hearing loss. The large-scale Wessex study27 reported that, of 25,000 newborns screened, 8.1 percent fulfilled high-risk criteria for permanent childhood hearing loss. Thus, targeted newborn hearing screening looks like a desirable option, as only a small proportion of babies would need to be screened. The yield is potentially high: just 86–208 high-risk babies need to be screened to find one baby with hearing loss as opposed to a yield of one per 2041–2794 low-risk babies screened.27, 46, 47 Thus, it has been argued that targeted newborn hearing screening is a relatively inexpensive way of improving the age of identification for a significant proportion of children with congenital hearing loss.28 However, there are a number of disadvantages to targeted newborn hearing screening. Stevens et al.28 noted that some ‘at risk’ infants, in particular those with a family history of hearing loss, are not being referred for screening. More importantly, some 40 or 50 percent of babies born with permanent hearing loss demonstrate no risk factors. Numerous studies agree that around half of all affected infants have no risk factors at birth and thus would be missed by a targeted hearing screening.5, 12, 41, 48



Performance of universal newborn hearing screening Davis et al.4 reported in their critical review of the evidence that high coverage of over 90 percent was possible with hospital-based newborn screening programmes. The current newborn hearing screening programme in England has completed the screening in 96 percent of all target babies,10 [**] which is well within the set quality standards. Davis et al.4 also concluded that screen sensitivity was estimated to range from 80 to nearly 100 percent and that screen specificity was above 90 percent. Specificity has been shown to gradually but significantly improve as advances in technology, use of two-stage screening (i.e. two tests) and training methodology evolve.41, 49 [****/**] Davis et al.50 concluded that all methods of neonatal hearing screening appear to have high test and screen sensitivity, which can be extended to programme sensitivity in both universal and targeted neonatal hearing screening programmes given sufficient resources. The authors also concluded that TEOAE screens and TEOAE/ AABR combined screens in particular have high screen specificity for full-term babies shown in several different implementations. [**] Screen specificity is influenced by whether the protocol requires a clear response in both ears or in either ear. For example in the Wessex study, to facilitate lower failure rates, more rapid screening and



828 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY greater coverage, a unilateral clear TEOAE response was accepted as a pass in infants younger than 48 hours.27 [****] Programme sensitivity has been shown to be around 80 percent.4 [***] The yield has been shown to be very high. To find one child with bilateral moderate-toprofound hearing loss 755–1422 newborns have to be screened.27, 42, 47 Yield is even higher, of course, if unilateral and mild cases are included.41, 51 [**] The median age of identification for those screened neonatally is of the order of two months.10, 42, 46, 52 [**] It is important to note that these data are derived from a rather few, though large, published studies of universal newborn screening.27, 42, 51 Only the Wessex study27 is a controlled trial (but not a randomized controlled trial). The extent to which these encouraging results can be translated into a national screen is an issue being addressed by the national evaluation of the first phase of the Newborn Hearing Screening Programme (NHSP) in England.



Cost-effectiveness Stevens et al.28 found that in comparison to the IDT screen, universal newborn hearing screening has acceptable costs. They calculated the mean service price for universal hearing screening in three services and, for the IDT test screen, at 1994 prices, these were £13,881 and £24,519, respectively, for a standardized district of 1000 live births. [**] The cost of universal newborn hearing screening compares favourably with screening for other congenital conditions, e.g. hypothyroidism, phenylketonuria or haemoglobinopathy.53 It has been argued that universal newborn hearing screening will have paid for itself within ten years in terms of the long-term savings made in special education and social interventions.53 However, longitudinal data do not yet exist. [**]



for other professionals, plain English information leaflets, written parental consent to the screen and so on. The extent to which these have minimized levels of anxiety in the mothers of babies passing the screen at different stages, and those whose babies are referred by the screen, is being investigated. The effects of early identification on the family are being investigated in true cases by in-depth interviews with parents.



Quality assurance There is an ethical duty within a screening programme (particularly a national programme) to ensure that good quality services are in place prior to the start of the screening programme and that they continue to meet national guidelines. It is most important to promote an evaluative culture to enable continued service improvement in both technical and family-friendly service delivery. Quality standards for early identification and management of congenital permanent hearing loss have been developed in the USA7 and the UK.58, 59 Consensus statements on newborn hearing screening have been developed in the USA7 and in Europe.60 Some of the quality standards adopted by the current Newborn Hearing Screening Programme in England are shown in Table 67.1.



THE EIGHT-MONTH IDT SCREEN The eight-month IDT screen has been in place as an almost universal screen across the UK since the late 1950s. It is being phased out as the Newborn Hearing Screening Programme is implemented. While protocols and test Table 67.1 Quality standards for the NHS Newborn Hearing Screening Programme in England. Quality standards



‘Family-friendliness’ Since screening takes place over the very first days or even hours after the baby’s birth, one of the main concerns associated with newborn hearing screening is the risk of causing unnecessary anxiety for the family. Several studies on parental opinions show that in general parents were very positive about newborn hearing screening and that the risk of disturbing the parent–child relationship by early screening seemed small and could be further minimized by improved information and rapid and effective follow-up.10, 38, 54, 55, 56, 57 [**] The current implementation of the NHSP in England has developed a number of processes for ensuring high levels of information for mothers/parents, including a national screenertraining programme, training and information sessions



Initial programme coverage >90% of all live births by discharge Screen coverage including call back >95% of all births by six months Coverage of NICU/SCBU 99% of live discharge Referral rate for non-NICU/SCBU o3% Referral rate for NICU/SCBU o5% All screen referrals followed up with audiological assessment by three months of agea True cases confirmed by six months of age for all cases Hearing aid fitted for all true cases within four weeks of confirmationb Education services informed of all true cases within three working days a b



Unless deliberately delayed for diagnostic reasons or follow-up refused. Unless deliberately delayed for management reasons.



Chapter 67 Hearing loss in preschool children: screening and surveillance



details vary from area to area, the most typical format involves a two-person distraction test, using frequencyspecific stimuli (low, mid and high frequencies) presented at quiet levels (e.g. 35 dBA) to the side and slightly behind (at 451) the eight-month-old infant who is seated on the parent’s knees. A full localization response to all stimuli on both sides represents an acceptable pass. Failures are usually retested once at a later date, except in the case of obvious concern, and two test failures constitutes a screen referral. This may be made to a secondary or tertiary audiology department, to a general practitioner for onward referral, or to an ear, nose and throat (ENT) department. The screen cannot be used in the first six months of life and therefore it cannot support the start of intervention before six months of age. Evidence on the performance of this screen is reviewed by Davis et al.4 The median age of identification through the IDT screen varies from 12 to 20 months; coverage falls in the range of 80–95 percent, although there may be some urban areas where coverage is as low as 60 percent. Sensitivity estimates of the IDT screen vary from 18 to 88 percent according to the service area. Screen sensitivity is greatly influenced by the degree of hearing loss, with higher sensitivity for more severe losses; even so, there have been too many cases of profound hearing losses passing the screen. Low coverage and low sensitivity result in unacceptably low incremental yield, in the range of 25–40 percent. The failure rate is high at 5–10 percent, largely because of poor practice and because of a high number of referrals due to transient OME. The referral rate has considerable resource implications for services. [**] Because of its low yield, the IDT screen has not proved to be cost-effective. Davis et al.4 estimate the cost per child detected with permanent bilateral hearing loss to be between £81,700 and £102,100 ($166,366 and $207,906), as opposed to £9900–£19,700 ($20,159 and $40,115) per child detected through a hospital-based newborn hearing screening programme. [****] There have been consumer concerns about the quality of the IDT screen. Studies have shown that the subjectivity of the test has caused parents not to take the test seriously, which has contributed to low coverage and low interest in attending follow-up appointments.61, 62 [**] Because of widespread dissatisfaction with the IDT screen, there have been some attempts to substitute ‘vigilance’ programmes based on questionnaires to parents. Sutton and Scanlon.63 showed that a vigilance programme was likely to perform as well as the IDT screen (but no better) for severe and profound hearing losses, but less well than the IDT screen in referring moderate hearing losses. [**]



THE SCHOOL-ENTRY SCREEN The school-entry hearing screen is currently employed universally across the UK. It consists of a modification of



] 829



standard pure-tone audiometry, with test signals presented at specific frequencies and at fixed levels (the ‘sweep’ test). The cut-off level is set at 20 or 25 dB HL. One or two tests may be performed, usually during the first year of school entry. The screen is usually carried out by a school nurse, but may be done by a school doctor or audiology technician. Testing is generally done within the school and ambient noise may limit specificity. There is currently very little monitoring or evaluation of the school entry screen. Guidelines for training and testing for non-audiology professionals have been issued.64 Apart from referring children with progressive and acquired permanent hearing loss that have not been picked up by parental observation and responsive services, the school-entry screen also has the potential to identify children with mild and unilateral hearing loss, as well as high-frequency hearing loss and other hearing losses with unusual configurations. It may also have a role in identifying children with marked hearing loss due to OME, and is thought to be a reasonably cost-effective way of raising awareness of the importance of good hearing at a crucial stage of the child’s education and development. However, these assertions remain to be tested by good quality research.



SURVEILLANCE As we saw earlier, child health surveillance is regarded as one component of child health promotion programmes which encourage partnership between parents, children and health professionals and in which parents are empowered to make use of services and expertise according to their needs. Surveillance for childhood hearing loss is of considerable importance, since (1) newborn hearing screens will not find all those with preschool permanent hearing loss, because of late onset and progressive hearing loss; (2) any screening programme will miss an irreducible minimum of true cases; and (3) surveillance is a more justifiable approach to identification of children with persistent OME than a screening programme. At present, there are only rather general guidelines for ongoing surveillance for childhood hearing loss.65 A UK working group is currently developing more detailed recommendations. These require a hearing assessment by an appropriately trained audiology team at around eight months of age for all babies who did not give clear bilateral responses in the newborn screen, in particular those found to have permanent unilateral or permanent bilateral mild loss. Naturally all babies who missed the newborn screen or audiological follow-up for whatever reason (e.g. declined the newborn hearing screen, moved out of the area, etc.) need a follow-up by an audiologist at eight months of age. They also require that all children who are at risk for late onset or progressive deafness be offered regular hearing assessments by audiologists at



830 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY appropriate ages, irrespective of the newborn screen outcome. These include infants with family history of childhood hearing loss;66 a history of severe hyperbilirubinaemia;67 evidence or suspicion of congenital infection;68, 69 presence of neurodegenerative or neurodevelopmental disorders; all infants with confirmed or suspected meningitis;70 those at high risk of persistent middle ear problems as well as sensorineural loss (e.g. children with cleft palate or Down syndrome);71 and other craniofacial anomalies, including chromosomal or syndromic conditions, including branchial arch and cervical spine anomalies.72 Surveillance policies should encourage services to be highly responsive to parental or professional concern about a child’s hearing or their development of auditory, vocal or communicative behaviour; such infants should be offered a full hearing assessment at any age by an appropriately trained team. The importance of ongoing surveillance by health visitors, family doctors, community health teams, as well as parents (reinforced by appropriate check lists such as those given to parents shortly after the birth of their child), should not be underestimated, since it will help to deal with the increasing prevalence with age found by Fortnum et al.6 in their extensive prevalence study.



SURVEILLANCE FOR OTITIS MEDIA WITH EFFUSION Screening for either OME (the pathology) or for the consequent hearing loss (the impairment) fails to meet the criteria for screening programmes outlined under Principles and definitions above. On the other hand, we know that there are material effects upon speech/ language, behaviour, physical well-being and quality of life for both parent and child in persistent cases. Some sort of surveillance system needs therefore to be in place for identifying and intervening with those 5 or 10 percent of cases that have OME with a persistence or severity likely to interfere with development. The TARGET group has shown that a screening questionnaire to parents on OME history and hearing loss gives only slightly lower sensitivity but similar specificity to that from schoolentry screening. For OME, such questionnaire approaches to surveillance gain over screening, since the fluctuating nature of the condition tends to make a one-off screen misprecise, while parental reports tend usefully to average the observed effects over time.15 [****] ‘Watchful waiting’ for three months has been shown to be an effective strategy for confirmed cases picked up by responsive services. Some 50 percent of such cases resolve after three months and do not justify further management unless the condition recurs.14, 15 The only exception to the advisability of having a watchful-waiting period arises in children with a pure-tone average in the better ear Z30 dB HL at their initial visit, who have a prior audiogram to



confirm OME, and who are sent to ENT between August and December. In these children the probability of persistence is greater than 80 percent. [****] If good audiometry is not available in primary care (because of lack of equipment or expertise), tympanometry has been confirmed as sensitive and fairly specific in identifying children with material hearing loss associated with OME.15 [****] On the other hand, it has been shown that little equipment is available or used to support OME assessment in UK primary care. However, it has also been shown that the use by family doctors of a simple checklist and a training video significantly improves the positive predictive value of their referrals.15 [****] The aim for services must be to identify those children who have persistent OME to the extent of being at risk for hearing and speech/language deficits, behaviour problems, quality of life, and, in a few cases, educational deficits.



KEY POINTS  At least one in 1000 children is born with permanent bilateral hearing loss Z40 dB HL.  If permanent bilateral congenital hearing loss is identified before six months of age and habilitation started soon thereafter the adverse effects are lessened.  By the age of nine years, the prevalence of permanent bilateral hearing loss Z40 dB HL is at least 1.65 per 1000 children.  Newborn hearing screening of all babies is the most cost-effective method of delivering early identification of congenital permanent hearing loss of at least moderate or greater degree.  More evidence is required on the outcomes for children with permanent mild or unilateral hearing loss, and on alternative approaches to management.  Screening at eight months using the infant distraction test is not cost-effective.  School-entry screening may be cost-effective, but there is little evidence to judge as yet.  Since prevalence continues to rise in the first few years of life, other methods based on surveillance and responsive services need to be in place.  OME is a very common condition which may lead to hearing loss, delays in speech/ language, behaviour problems and reduced quality of life in persistent cases.  Identification of persistent OME cases depends upon good surveillance systems in primary care and responsive services.



Chapter 67 Hearing loss in preschool children: screening and surveillance



Best clinical practice



$



[ Newborn hearing screening of all babies is the most cost-effective way to identify congenital hearing loss.



[ The eight-month infant distraction test screen is not cost-effective and cannot be justified.



[ A surveillance programme needs to be in place to [ [



[ [



help identify late onset and acquired hearing losses, backed up by fast responsive services. Every effort should be made to minimize family anxiety by good-quality information and rapid and effective follow-up. Good quality services for true-positive cases based upon informed choices for parents are central to the introduction of a screening programme at any age, but particularly for newborn screening. Management of true-positive cases will involve close cooperation between health, education and social services. Parents and professionals in primary care need to be alert to the presence of persistent OME in some children; identification of these children is improved by the use of checklists and a training video by family doctors, and by a standardized ‘screening’ questionnaire to parents at varying times.



$



$



$



$ $



Even though there is evidence to show that parental anxiety associated with newborn hearing screening is fairly minimal, ways of reducing it further need to be investigated. In addition, more understanding of the experiences of the families whose babies are identified with hearing loss at a very early age is needed, in order to ensure that the service received at this early and sensitive stage is ‘family-friendly’. There are a number of different ways to run a successful newborn hearing screening programme, in terms of equipment, tests and protocols. More work on the modelling of cost-effectiveness under different conditions (e.g. urban versus rural, size of the birthing hospital, the pattern of discharge, etc.) would be helpful. Auditory neuropathy is a relatively new condition to audiology services. More research is needed concerning the prevalence, aetiology, outcomes, diagnostic and habilitative options for children with this condition. Currently, children with mild and unilateral hearing loss will be identified as a by-product of newborn hearing screening for permanent bilateral hearing losses of Z40 dB HL. More research is needed about



the prevalence, effects, outcomes and habilitative options for these children. There is evidence to show that progressive and acquired permanent childhood hearing loss is more prevalent than previously thought. Confirmatory data are needed and evidence on the risk factors for progressive and acquired permanent childhood hearing loss is needed in order to improve surveillance or targeted assessment programmes. The school entry screen requires a more extensive evidence base.



REFERENCES







 Deficiencies in current knowledge and areas for future research



] 831







1. National Screening Committee. Webpage: http:// www.nsc.nhs.uk. 2. Wilson JMG, Junger G. Principles and practice of screening for disease. Geneva: World Health Organisation, 1968. 3. Haggard MP, Hughes E. Screening children’s hearing. London: HMSO, 1991. 4. Davis A, Bamford J, Wilson I, Ramkalawan T, Forsaw M, Wright S. A critical review of the role of neonatal hearing screening in the detection of congenital hearing impairment. Hearing Technology Assessment. 1997; 1: 1–105. 5. Fortnum H, Davis A. Epidemiology of permanent hearing impairment in Trent region, 1985–1993. British Journal of Audiology. 1997; 31: 409–46. 6. Fortnum HM, Summerfield AQ, Marshall DH, Davis AC, Bamford JM. Prevalence of permanent childhood hearing impairment in the United Kingdom and implications for universal neonatal hearing screening: questionnaire based ascertainment study. British Medical Journal. 2001; 323: 536–9. 7. Finitzo T, Sininger Y, Brookhouser P, Epstein S, Erenberg A, Roizen N et al. Year 2000 position statement: Principles and guidelines for early hearing detection and intervention programs. Pediatrics. 2000; 106: 798–817. 8. Davis A, Wood S. The epidemiology of childhood hearing impairment: Factors relevant to planning services. British Journal of Audiology. 1992; 26: 77–90. 9. Vartiainen E, Karjalainen S. Prevalence and etiology of unilateral sensorineural hearing impairment in a Finnish childhood population. International Journal of Pediatric Otorhinolaryngology. 1998; 43: 253–9. 10. Bamford J, Ankjell H, Crockett R, Marteau T, McCracken W, Parker D, et al. Evaluation of the newborn hearing screening programme in England: Studies, results and recommendations. Report to Department of Health and National Screening Committee, May 2005, 242. 11. Rance G, Beer DE, Cone-Wesson B, Shepherd RK, Dowell RC, King AM et al. Clinical findings for a group of infants and young children with auditory neuropathy. Ear and Hearing. 1999; 20: 238–52.



832 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY 12. Mehl AL, Thomson V. The Colorado Newborn Hearing Screening Project, 1992–1999: On the threshold of effective population-based universal newborn hearing screening. Pediatrics. 2002; 109: e7. 13. Sininger Y. Auditory neuropathy in infants and children: Implications for early hearing detection and intervention programs. Paper presented at the 2nd International Conference on Newborn Hearing Screening Diagnosis and Intervention, Como, Italy. 2002: 25. 14. Zeilhuis GA, Heuvelmans EW, Rach GH. Environmental risk factors for otitis media with effusion in preschool children. Scandinavian Journal of Primary Health Care. 1989; 7: 33–8. 15. Haggard MP, Gannon MM, Birkin JA, Bennett KE, Nicholls EE, Browning GG et al. Risk factors for persistence of bilateral otitis media with effusion. Risk factors for persistence of bilateral otitis media with effusion. Clinical Otolaryngology. 2001; 26: 147–56. 16. Trune DR. Influence of neonatal cochlear removal on the development of mouse cochlear nucleus. 1. Number, size, and density of its neurons. Journal of Comparative Neurology. 1982; 209: 409–24. 17. Harrison RV, Stanton SG, Ibrahim D, Nagasawa A, Mount RJ. Neonatal cochlear hearing loss results in developmental abnormalities of the central auditory pathways. Acta Otolaryngologica. 1993; 113: 296–302. 18. Tierney TS, Russell FA, Moore DR. Susceptibility of developing cochlear nucleus neurons to deafferentationinduced death abruptly ends just before the onset of hearing. Journal of Comparative Neurology. 1997; 378: 295–306. 19. Pasic TR, Rubel EW. Cochlear nucleus cell-size is regulated by auditory-nerve electrical activity. Otolaryngology – Head and Neck Surgery. 1991; 104: 6–13. 20. Harrison RV. Representing the acoustic world within the brain: Normal and abnormal development of frequency maps in the auditory system. Paper presented at the 2nd International Conference: A Sound Foundation Through Early Amplification. 2001: 199–212. 21. Gregory S. Deaf children and their families. Cambridge: Cambridge University Press, 1995. 22. Conrad R. The deaf school child. London: Harper & Row, 1979. 23. Hind S, Davis A. Outcomes for children with permanent hearing impairment. Proceedings of an International Conference: A Sound Foundation Through Early Amplification. 2000: 199–212. 24. Powers S. Deaf pupils’ achievement in ordinary subjects. Journal of British Association of Teachers of the Deaf. 1996; 20: 111–23. 25. Laurenzi C, Montiero B. Mental health and deafness – the forgotten specialism? ENT News. 1997; 6: 22–4. 26. Gregory S, Bishop J, Sheldon L. Deaf young people and their families. Cambridge: Cambridge University Press, 1995. 27. Kennedy CR, Kimm L, Dees DC, Campbell MJ, Thornton ARD, Bamber J et al. Controlled trial of universal neonatal



28.



29.



30.



31.



32. 33.



34.



35.



36.



37.



38.



39.



40.



41. 42.



43.



44.



screening for early identification of permanent childhood hearing impairment. Lancet. 1998; 352: 1957–64. Stevens JC, Hall DMB, Davis A, Davies CM, Dixon S. The costs of early hearing screening in England and Wales. Archives of Disease in Childhood. 1998; 78: 14–9. Yoshinaga-Itano C, Sedey AL, Coulter DK, Mehl AL. Language of early- and later-identified children with hearing loss. Pediatrics. 1998; 102: 1161–71. Markides A. Age at fitting of hearing aids and speech intelligibility. British Journal of Audiology. 1986; 20: 165–8. Ramkalawan TW, Davis A. The effects of hearing loss and age of intervention on some language metrics in young hearing-impaired children. British Journal of Audiology. 1992; 26: 97–107. Kuhl PK. Learning and representation in speech and language. Current Opinion in Neurobiology. 1994; 4: 812–22. Kuhl PK, Williams KA, Lacerda F, Stevens KN, Lindblom B. Linguistic experience alters phonetic perception in infants by six months of age. Science. 1992; 255: 606–8. Downs MP. Universal new-born hearing screening – the Colorado story. International Journal of Pediatric Otorhinolaryngology. 1995; 32: 257–9. Robinshaw HM. Early intervention for hearing impairment: Differences in the timing of communicative and linguistic development. British Journal of Audiology. 1995; 29: 315–34. Apuzzo ML, Yoshinaga-Itano C. Early identification of infants with significant hearing loss and the Minnesota Child Development Inventory. Seminars in Hearing. 1995; 16: 124–39. Mayne AM, Yoshinaga-Itano C, Sedey AL, Carey A. Expressive vocabulary development of infants and toddlers who are deaf or hard of hearing. Volta Review. 2000; 100: 1–28. Watkin PM, Beckman A, Baldwin M. The views of parents of hearing-impaired children on the need for neonatal hearing screening. British Journal of Audiology. 1995; 29: 259–62. Magnuson M, Hergils L. The parents’ view on hearing screening in newborns – Feelings, thoughts and opinions on otoacoustic emissions screening. Scandinavian Audiology. 1999; 28: 47–56. White KR, Maxon AB. Universal screening for infant hearing impairment: Simple, beneficial and presently justified. International Journal of Pediatric Otorhinolaryngology. 1995; 32: 201–11. Mehl AL, Thomson V. Newborn hearing screening: the great omission. Pediatrics. 1998; 101: E4. Watkin P. Neonatal otoacoustic emission screening and the identification of deafness. Archives of Disease in Childhood. 1996; 74: 16–25. Van Zanten B. Community based otoacoustic emission screening in the Netherlands. Baby BSA, British Society of Audiology Annual Conference, Winchester, 1996. Stappaerts L, van Kerschaven E. Achieving efficient and effective early hearing detection in Flanders. 2nd



Chapter 67 Hearing loss in preschool children: screening and surveillance



45.



46.



47.



48.



49.



50.



51.



52.



53. 54.



55.



56.



57.



International Conference on Newborn Hearing Screening Diagnosis and Intervention, Como, Italy. 2002: 59–60. Owen M, Webb K, Evans K. Community based universal neonatal hearing screening by health visitors using otoacoustic emissions. Archives of Disease in Childhood. 2001; 84: 157–62. Dalzell L, Orlando M, MacDonald M, Berg A, Bradley M, Cacace A et al. The New York State universal newborn hearing screening demonstration project: Ages of hearing loss identification, hearing aid fitting, and enrollment in early intervention. Ear and Hearing. 2000; 21: 118–30. Prieve B, Dalzell L, Berg A, Bradley M, Cacace A, Campbell D et al. The New York State universal newborn hearing screening demonstration project: Outpatient outcome measures. Ear and Hearing. 2000; 21: 104–17. Berg AL, Spivak LG. Universal newborn hearing screening: Should we leap before we look? Pediatrics. 1999; 104: 351–2. Maxon AB, White KR, Behrens TR, Vohr B. Referral rates and cost efficiency in a universal newborn hearing screening program using transient evoked otoacoustic emissions. Journal of American Acadademy of Audiology. 1995; 6: 271–7. Davis A, Bamford J, Stevens J. Performance of neonatal and infant hearing screens: Sensitivity and specificity. British Journal of Audiology. 2001; 35: 3–15. Vohr BR, Carty LM, Moore PE, Letourneau K. The Rhode Island hearing assessment program: Experience with statewide hearing screening (1993–1996). Journal of Pediatrics. 1998; 133: 353–7. McClelland RJ, Watson DR, Lawless V, Houston HG, Adams D. Reliability and effectiveness of screening for hearingloss in high-risk neonates. British Medical Journal. 1992; 304: 806–9. Mehl AL. Universal newborn hearing screening: Should we leap before we look? Pediatrics. 1999; 104: 352–4. Watkin PM, Baldwin M, Dixon R, Beckman A. Maternal anxiety and attitudes to universal neonatal hearing screening. British Journal of Audiology. 1998; 32: 27–37. Hergils L, Hergils A. Universal neonatal hearing screening – Parental attitudes and concern. British Journal of Audiology. 2000; 34: 321–7. Uilenburg NN, Kauffman de Boer MA, de Ridder-Sluiter JG, van der Ploeg CPB, Lanting CI, Verkerk PH. Parents opinions on neonatal hearing screening and diagnostic research after positive screening. Paper presented at the 2nd International Conference on Newborn Hearing Screening Diagnosis and Intervention, Como, Italy. 2002: 45. Van der Ploeg CPB, Lanting CI, de Ridder-Sluiter JG, Kauffman de Boer MA, Verkerk PH. Parental concern after positive results of neonatal hearing screening. Paper



58. 59. 60.



61.



62.



63.



64.



65. 66. 67.



68.



69.



70.



71.



72.



] 833



presented at the 2nd International Conference on Newborn Hearing Screening Diagnosis and Intervention, Como, Italy. 2002: 45–6. National Deaf Children’s Society (NDCS). Quality standards in paediatric audiology, I. London: NDCS, 1994. National Deaf Children’s Society (NDCS). Quality standards in paediatric audiology, IV. London: NDCS, 2000. Grandori F. European consensus statement on neonatal hearing screening. Journal of Laryngology and Otology. 1998; 112: 1219. Johnson A, Sherratt F, Holmes S. Parents’ attitudes to developmental screening in the first year of life. Child: Care, Health and Development. 1991; 17: 197–211. Robertson C, Aldridge S, Jarman F, Saunders K, Poulakis Z, Oberklaid F. Late diagnosis of congenital sensorineural hearing impairment – Why are detection methods failing. Archives of Disease in Childhood. 1995; 72: 11–5. Sutton GJ, Scanlon PE. Health visitor screening versus vigilance: Outcomes of programmes for detecting permanent childhood hearing loss in west Berkshire. British Journal of Audiology. 1999; 33: 145–56. Smith PA, Evans PIP. Hearing assessment in general practice, schools and health clinics: Guidelines for professionals who are not qualified audiologists. British Journal of Audiology. 2000; 34: 57–61. Hall D, Elliman D. Health for all children. Oxford: Oxford University Press, 2003. Lucas D. Progressive loss. In: Newton V (ed.). Paediatric audiological medicine. London: Whurr, 2002: 269–93. Berlin CI. Managing patients with auditory neuropathy/ auditory dys-synchrony. Available at: http:// .medschool.lsumc.edu/otor/dys.html 1999. Hickson LMB, Alcock D. Progressive hearing-loss in children with congenital cytomegalovirus. Journal of Paediatrics and Child Health. 1991; 27: 105–7. Wild NJ, Sheppard AS, Smithells RW, Holzel AH, Jones G. Onset and severity of hearing-loss due to congenitalrubella infection. Archives of Disease in Childhood. 1989; 64: 1280–3. Fortnum H, Davis A. Hearing impairment in children after bacterial-meningitis – Incidence and resource implications. British Journal of Audiology. 1993; 27: 43–52. Zlotogora J, Sagi M, Schuper A, Leiba H, Merin S. Variability of Stickler syndrome. American Journal of Medical Genetics. 1992; 42: 337–9. Cremers CWRJ. Craniofacial syndromes and hearing impairment with special reference to syndromes with external ear anomalies. In: Newton V (ed.). Paediatric audiological medicine. London: Whurr, 2002: 224–31.



68 Hearing tests in children GLYNNIS PARKER



Introduction Electrophysiological testing Behavioural observation audiometry The distraction test Visual reinforcement audiometry Performance testing Pure tone audiometry



834 834 835 835 838 839 840



Auditory speech discrimination tests Hearing assessment in children with special needs Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



841 841 842 842 842 842



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words audiometry/child and hearing test/child.



INTRODUCTION



ELECTROPHYSIOLOGICAL TESTING



It is essential that suspected hearing loss in young children is promptly investigated. Accurate assessment of hearing is fundamental to diagnosis, investigation and rehabilitation. The techniques applicable to adults and older children are often inappropriate for the young child and for the older child or adult with special needs, particularly learning disability. Skilled testing by trained personnel in a suitable test environment is essential. Electrophysiological events in response to a sound stimulus may be recorded (electrophysiological testing). This is considered in detail in Chapter 67, Hearing loss in pre-school children: screening and surveillance and Chapter 233, Evoked physiological measurement of auditory sensitivity. The focus of this chapter is on those tests that require skilled observation and recording of the child’s response to one or more sound stimuli. Testing is increasingly required to monitor speech perception in children with hearing aids and cochlear implants. These ‘functional hearing tests’ are also considered.



Key developmental age: 0–6 months, up to adult if appropriate Electrophysiological techniques including otoacoustic emission (OAE), auditory brainstem response (ABR) and cortical evoked response audiometry (CERA) are discussed in Chapter 67, Hearing loss in pre-school children: screening and surveillance and Chapter 233, Evoked physiological measurement of auditory sensitivity. These methods are mainly employed in newborn screening (see Chapter 67, Hearing loss in pre-school children: screening and surveillance) and diagnostic testing of infants in the first six months.1, 2 They can also prove of value in children of any age, when behavioural testing has failed to produce reliable results, in particular, those with severe learning or communication difficulties.3 They may require general anaesthesia. Electrophysiological techniques are also used to confirm hearing thresholds, for example in children with profound hearing loss, prior to fitting of high power hearing aids or cochlear



Chapter 68 Hearing tests in children



implantation or where nonorganic factors are suspected. Whilst electrophysiological testing has the advantage of being objective in terms of the child’s response, behavioural or speech discrimination testing remain the only functional measures for assessing the complete auditory system.



BEHAVIOURAL OBSERVATION AUDIOMETRY Key developmental age: 0–6 months In behavioural observation audiometry (BOA), changes in activity are observed in response to a sound stimulus. Response behaviours in infants up to four months might include eye widening, eye blink (auropalpebral reflex), arousal from sleep, startle or shudder of the body or definite movement of the arms, legs or body. From four to seven months, lateral inclination of the head towards the sound or a listening attitude or stilling may be observed. The test is usually performed with the child cradled in the parent’s lap. The child’s attention may be lightly engaged in front by a distractor. The sound stimulus is presented for o2 seconds, in a horizontal plane, 15 cm from the child’s ear, out of peripheral vision. The distractor observes evidence of a response. A range of sound stimuli may be sometimes employed, including narrow band and warble tones or less frequency-specific examples, such as a small bell, squawky squeeze toy or banging wooden bricks.4 [***] Reliability as a diagnostic test is obviously a concern. A wide variability in judgement of response between testers due to misinterpretation of random movements and a tendency to underestimate hearing thresholds has been demonstrated.5 Attempts have been made to reduce observer bias by use of video recording of the procedure and scoring the playback without knowledge of the sound or no sound trials.6 [***] The use of BOA in infants under six months has been largely superseded by the availability of electrophysiological techniques. It does, however, have a continuing role in the assessment of children with severe learning difficulties, who have not reached an appropriate developmental stage for distraction or visual reinforcement audiometry. It may also be of value in cases of auditory neuropathy when ABR is a poor indicator of functional hearing levels.7



THE DISTRACTION TEST Key developmental age: 6–18 months The test is based on the principle that the normal response observed when sound is presented to a baby is a head turn to locate the source of sound. The test is



] 835



suitable for babies from 6 to 18 months, corresponding to the stage when the child can sit erect unsupported and perform head turns in a horizontal plane. An appropriate allowance should be made for prematurity. Habituation is increasingly likely to occur after 12 months, although the technique may prove to be useful in older children with learning or communication difficulties where other methods have been unsuccessful. The credibility of the infant distraction test as a tool for screening for hearing impairment has become questionable. In many areas, it has been discontinued in favour of electrophysiological testing in the newborn period1 (see Chapter 67, Hearing loss in pre-school children: screening and surveillance). Nevertheless, it remains a valuable diagnostic technique when used by trained, experienced testers in appropriate settings. The infant distraction test was first described by Ewing and Ewing in 1944,8 but was subsequently modified by McCormick9 who placed particular emphasis on the use of frequency-specific, calibrated sound stimuli. The test is described with reference to the guidelines outlined by McCormick.10 [***/**/*]



Test method The test should be performed in a tidy, uncluttered, suitably sized room (recommended minimum 416 m2) with ambient noise levels o30 dB(A). Two testers are required, one to present the sound stimuli out of vision and the other to control the baby’s attention in the forward direction (the distractor). The latter should be responsible for directing the test. The arrangement is shown in Figure 68.1. The child sits on the parent’s knee, facing forward and erect, lightly supported around the waist. The distractor directs the attention of the child to a simple activity usually performed on a low table. Suitable examples include spinning a brightly coloured object, using finger puppets or gently pushing a miniature car. In the classic distraction test described by McCormick, the item is covered by the hands which maintain a fine attention control by moving the fingers slightly. The sound stimulus is presented by the second tester, half a second after the item is covered. The distractor observes the child’s response. The second tester should be positioned strictly out of the child’s visual field, which can be assumed to extend at least 901 to either side of the child’s midline. The sound stimulus should be presented in the horizontal plane to the ears at an angle, set back 451 between 1 m and 15 cm from the child’s ear. The shorter distance increases the head shadow effect, thereby increasing the sound stimulus on one side relative to the other. Evidence of asymmetric hearing levels may be indicated by a difference in response thresholds or difficulty in localizing the sound source. Particular care must be taken when presenting the sound close to the ear to avoid visual or tactile clues. The sound



836 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Sound level meter on stand Tester (standing) Right sound source



Left sound source



Parent (seated)



45°



Child



45°



Low table



Distractor (seated)



Figure 68.1



Test arrangement for the distraction test.



stimulus should continue up to 10 seconds if there is not an immediate response. During this interval, the distractor should maintain the fine attention control at the front until a response is observed. The normal response expected is a full head turn in the direction of the sound. This may be rewarded by a smile or vocal praise or gentle tickle from the second tester. Some warble tone generators incorporate a flashing light, which can be activated in response to the turn, providing an incentive to the child and thereby acting as a ‘response reinforcer’. The child’s attention should then be brought back to the front by the distractor. [**]



The stimulus A wide range of sound stimuli can be used to elicit a response, including voice, musical toys, ‘everyday sounds’, narrow band noise and warble tones (see Figure 68.2). Pure tones should be avoided due to the potential creation of standing waves, resulting in unpredictable sound levels. Responses to any sound stimulus may be valid, provided the intensity and the frequency spectrum, as delivered at the level of the ear, can be established. A sound level meter should be employed to check the intensity of the stimulus by accurately reproducing the



sound and the distance from the ear. Although it has been demonstrated that babies are more likely to respond to wide band sounds, this will inevitably provide less information regarding frequency-specific hearing as required for diagnosis and possible amplification prescription. Sound generators delivering calibrated narrow band and frequency-modulated warble tones have been demonstrated to be effective at eliciting a response and are therefore generally preferred.11 Initially, a sound stimulus is presented which is anticipated to be likely to be suprathreshold, for example at 70 dB(A) in a child with probable normal hearing. A sound is then presented at the anticipated minimal response threshold, for example, 30 dB(A). The intensity for sound field testing is generally measured on the dB(A) scale, but this can be converted to dBSPL (sound pressure level) if required, particularly for hearing aid prescription fitting.12 There is no prescribed order for frequency and intensity of sound presentations. The number of reliable responses elicited by distraction testing may be limited, particularly in children over 12 months who may habituate to the test. It is therefore important to prioritize and focus on those thresholds which would be most valuable for each clinical situation, but would normally aim to include high-, medium- and low-frequency information. This, however,



Chapter 68 Hearing tests in children



] 837



Soft ‘mm’ 500 Hz warble 500 Hz NBN 1 kHz warble 2 kHz warble 4 kHz warble 4 kHz NBN Soft ‘ss’ Soft Manchester rattle 125



250



500



1000



2000



4000



8000



Frequency in Hz



may need to be balanced against the benefit of varying the order and side of presentation simply to sustain the child’s interest. [Grade D]



No sound control trials No sound trials are essential for the validation of response thresholds in a distraction test. Some children exhibit checking or searching behaviour or simply turn towards their parents for reassurance during the test. This can clearly be misleading if it coincides with a sound presentation. It is therefore appropriate to introduce control trials in which the test is performed in the usual manner by both testers, but no sound is generated. If the child appears to make a response, then it must be assessed how this compares to the responses observed to sound trials. Usually a difference is noted which can be used to confirm a true response. However, if the checking responses are indistinguishable from the sound responses, the results must be interpreted with appropriate caution. Table 68.1



Figure 68.2 The distraction test: examples of frequency spectra for commonly used sound sources. With thanks to JC Stevens, Department of Medical Physics, Royal Hallamshire Hospital, Sheffield, UK.



Tactics such as increasing the interest of the distracting activity and/or maintaining that activity whilst the sound/ no sound is presented may avoid ‘clueing in’ the child. This may be particularly helpful in children over one year. Evidence of visual, tactile or even olfactory clues must also be considered and addressed. [*]



Recording the results The outcome of each valid presentation and no sound trial should be recorded in chart form, specifying the sound source, frequency range if known and intensity. A tick or a cross may be used to indicate whether a response was observed. The response threshold is regarded as the quietest level at which two out of three clear responses were recorded for each particular sound stimulus. Other observations such as the ability to localize the sound source correctly or any concerns about the test reliability should also be recorded. Table 68.1 lists the most commonly encountered pitfalls in distraction testing and avoidance measures. [**]



Pitfalls encountered in distraction testing.



Pitfalls Visual, tactile, olfactory clues from parent, distractor or second tester ‘Checking’ responses by child Loss of interest in test, suprathreshold responses



Inaccurate estimation of frequency and intensity of stimulus Extraneous noise, leading to false responses or only suprathreshold response



Avoidance measures Critical observation of all parties No sound trials Critical assessment of responses No sound trials, if necessary take a break, or sustain distraction activity Vary the sound stimulus Exchange roles of distractor or second tester Use broader band or more interesting sounds Use calibrated sound sources Measure accurately with sound level meter, trying to reproduce conditions Use quiet facilities, ideally a sound-treated room with ambient noise o30 dB(A) Observe other auditory clues, e.g. clicking switch on warbler No sound trials



838 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



VISUAL REINFORCEMENT AUDIOMETRY



to provide an observation area with a one-way window, with access to a communication system between the rooms, such as a radio link. If however, this facility is not available, then it is still possible to perform the test within one room, provided the tester operating the audiometer minimizes the distraction to the child and does not provide any additional clues. A suitable arrangement is shown in Figure 68.3. The child is seated on the parent’s lap or on a low chair in front of a low table, with the parent seated slightly behind. A second tester may sit on the other side of the table or adjacent to the child to provide low level play activity. For sound field testing, the speakers may be placed at 45, 60 or 901 from the child and at the same height as the child’s head at a distance of at least 1 m from the ear. Frequency-specific calibration of the sound signal from a test point equivalent to the child’s head position is essential. Alternatively, a calibrated signal may be presented via insert earphones using foam tips or the child’s own ear moulds, if available. This has the advantage of potentially providing more precise ear-specific response thresholds and even offers an opportunity to introduce masking by appropriate presentation of narrow band noise in the contralateral ear. This is particularly valuable in amplification selection and setting.23 A standard bone conductor may also be used in conjunction with VRA to differentiate between a conductive and sensorineural hearing loss. [**]



Key developmental age: 6–36 months Visual reinforcement audiometry (VRA) incorporates the principle that young children can be trained by operant conditioning to produce a localizing turn to a visual stimulus in response to a sound stimulus. A technique of conditioned orientation reflex (COR) audiometry was initially described by Suzuki and Ogiba13 and was further developed by Liden and Kankkunen14 who introduced the term ‘visual reinforcement audiometry’. Further developments and modifications to the technique have been described.15, 16 VRA is now established as a standard technique within the test battery for hearing assessment in preschool children.17, 18, 19 A comparison of VRA with the infant distraction test in children aged 12–25 months has demonstrated that VRA generates more auditoryevoked head turns than an unrewarded sound stimulus and delays habituation so that the test was more likely to be completed.20 The method is described with reference to Widen et al.21 and Shaw.22 [***/**/*]



The test arrangement The test room should be sound-treated to ensure low levels of ambient noise and should ideally be partitioned



Parent (seated)



Right speaker



Left speaker



Child



Visual reinforcer



Visual reinforcer



Low table



Observation window



Test area Distractor (seated)



Audiometer Observation area



Tester



Figure 68.3



The test arrangement for visual reinforcement audiometry.



Reinforcer control



Chapter 68 Hearing tests in children



Thompson and Folsom24 demonstrated that in VRA, unlike distraction testing, the complexity of the signal, including the bandwidth, had no effect on response rate once the child had been correctly conditioned. Frequency-modulated warble tones are generally employed for sound field testing, but pure tones may be used in conjunction with insert earphones or a bone conductor. The frequency sequence should be prioritized according to the clinical situation, but would most often include 0.5, 1.0, 2.0 and 4.0 kHz and in some cases 0.25 and 3 kHz. The duration of the signal is usually two to three seconds. Visual reinforcers are generally placed adjacent to or above each speaker. The reinforcer acts as a reward and therefore increasing the attractiveness or appeal to the child and introducing variety is likely to sustain interest and reduce habituation. Commercially available reinforcers, include toys with eyes that light up and puppets in darkened glass cabinets which illuminate and dance. Monitors showing appropriate video displays can also be effective. The reinforcer is generally remotely activated by the tester as required. [**]



Test method Using the arrangement as described above, the first task must be to establish the conditioned response. Having drawn the child’s attention to the front by gentle play activity on the table, a suprathreshold auditory signal should be presented concurrently with the visual reward, for example, for a child with probable near-normal hearing, one might start with a 1 kHz warble tone at 70 dBHL. The child may turn to locate the sound (orientation response) and thereby view the activated reinforcer, but if necessary, the child’s attention should be directed towards it. This sequence is repeated several times. The sound stimulus should then be presented alone and the reinforcer only activated after the child has produced an appropriate turning response. Praising the child and making it a game will further help to reinforce the response. If this can be reliably repeated two or three times, then operant conditioning has been established. Children under one year or those with learning difficulties may take longer to successfully condition. If it appears that the initial auditory stimulus may be inadequate, this can be cautiously increased or a different frequency or vibrotactile stimulus used. Once the child has demonstrated successful conditioning, the aim is to establish minimal response levels (MRL) for each frequency, as prioritized by the clinical situation. One reliably obtained MRL is more valuable than a number of ‘maybes’. A flexible descending/ascending technique, similar to that used in play audiometry may be applied. Responses should be charted as described for the distraction test, using the two out of three valid



] 839



response rule to determine the MRL. No sound control trials should be included to verify a true response and similar measures taken in the event of frequent ‘checking behaviour’. MRL less than 20 dBHL can routinely be recorded for frequencies between 0.5 and 4 kHz in normal hearing infants using insert earphone VRA.25 Primus26 found test–retest reliability to be good with 50 percent repeatability within 10 dB. Widen21 demonstrated that MRL for at least four frequencies were successfully recorded in more than 90 percent of over 3000 infants aged 8–12 months, tested using VRA. [***]



PERFORMANCE TESTING Key developmental age: 2–5 years The performance test was first described by Ewing and Ewing8 as a transitional technique suitable for children from 2.5 years and in some cases younger, until cooperation with pure tone audiometry can be achieved. It can be used to assess unaided or aided sound field thresholds. It follows the simple principle that the child is conditioned to wait for a sound and then to respond with a play activity. The test method is described with reference to McCormick.10 [**]



Test method The child should be seated on a low chair adjacent to the parent in an uncluttered room with low levels of ambient noise. A toy is placed on a low table in front of the child (Figure 68.4). Toys which involve a simple repetitive activity are most suitable, such as the classic ‘men in a boat’, balls on sticks, knocking down skittles or pegs in a board. The conditioning sequence starts with the tester engaging the child’s attention by holding the response item, e.g. the wooden man, poised waiting in front of the child. After a few seconds, a suprathreshold sound stimulus is presented and the tester responds by an appropriate activity, e.g. placing the man in the boat. This sequence is repeated several times often supported by gestures such as a stop sign using the palm of the hand and a cupped hand to the ear to indicate listening. This has the advantage of avoiding dependence on spoken language. The child is then offered the response item and guided to wait and perform the task as shown. Vocal praise should be used to reinforce a correct response. The number of repetitions required to successfully condition the child will depend on their age, developmental status, willingness to co-operate and in particular their ability to inhibit the response until the signal is detected. It may be necessary to increase the intensity of the signal or initiate conditioning using a vibrotactile stimulus.



840 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Figure 68.4 The performance test.



Once the child has been successfully conditioned to perform the test without guidance, then a flexible descending/ascending technique as described for pure tone audiometry (see Chapter 232, Psychoacoustic audiometry) can be applied to determine the minimal response threshold. The interval between presentations must be randomly varied to avoid a predictable rhythm. No sound trials should be included. It may be necessary to reinstruct the child by guiding a few responses when the stimulus is changed or if the child appears to have lost concentration. It may also be necessary to introduce several changes of activity to sustain interest. This provides the tester with an opportunity to be adaptable and inventive and is part of the joy of audiology! Examples might include knocking the man off the table, flicking a small ball into a ‘goal’, or even poking the tongue out at the tester. The aim is to be fun enough to motivate the child, whilst being accurate enough to provide valid responses.



The stimulus In the original description of the performance test, the signal used was a voiced ‘go’ (low frequency), or a sibilant ‘s’ (high frequency).8 When using voiced signals it is essential that the face is totally outside the child’s visual field and the mouth is shielded to prevent any awareness of air movement. The tester should therefore adopt a position similar to that described for distraction testing. Frequency-modulated warble tones with stimulus duration of one to three seconds are now the preferred option, providing low–mid and high frequency-specific signals by means of portable sound field noise generators. These should ideally also be used out of vision with care to avoid clues such as a clicking switch. Alternatively, loudspeakers can be utilized in a similar arrangement to that described for VRA. Whatever the signal, it is essential that the test conditions are reproduced accurately to



calibrate the intensity delivered to the child’s ear. The sequence of presentation would normally aim to include stimuli at 0.5, 1, 2 and 4 kHz and possibly 0.25, 3, 6 and 8 kHz.



PURE TONE AUDIOMETRY Key developmental age: 3 years onwards The techniques employed in performance testing naturally lead into pure tone audiometry, as the child matures. This progression should be adapted to the individual child’s ability and co-operation and to the clinical priority. It may be possible, for example, to introduce a bone conductor before a child will tolerate headphones or alternatively, a child who is accustomized to wearing hearing aids may comply with insert earphone audiometry via their ear moulds, as a progression from VRA. Insert earphones have the advantage of increasing transcranial attenuation reducing ‘cross over’ and of reducing the effect of ear canal collapse which may occur with supra-aural headphones. Again, it is important to sustain interest and concentration by the appropriate use of toys and praise. The recommended procedures for pure tone audiometry based on the Hugson and Westlake descending/ ascending technique using 10/5 dB steps should be flexibility adapted (see Chapter 232, Psychoacoustic audiometry). There is no standard for the interpretation of results in children and the accepted limits of normality range from 15 to 30 dBHL in clinical and screening practice. Nielsen and Olsen27 reported that it was possible to obtain at least six ear-specific air conduction thresholds from virtually 75 percent of three year olds. The introduction of standard masking techniques, where appropriate, will depend on the child’s cognitive developmental level and co-operation, but may not be reliable until the age of seven years. [**]



Chapter 68 Hearing tests in children



AUDITORY SPEECH DISCRIMINATION TESTS The ability to discriminate speech signals may be a valuable measure of functional hearing in children with normal to moderate degrees of impairment and is also increasingly used in the evaluation of more severely affected children supplied with hearing aids or cochlear implants. The test batteries developed for use in these latter situations are discussed elsewhere (see Chapter 232, Psychoacoustic audiometry). The tests described below represent a selection from a wide range available and are for use in more general clinical situations.



The co-operative test LINGUISTIC DEVELOPMENTAL LEVEL: 18–30 MONTHS



This is another test originally described by Ewing and Ewing8 and requires the child to discriminate three different simple instructions, e.g. having been handed a small toy, asked to ‘give it to Mummy’ or ‘give it to teddy’ or ‘give it to baby’. Starting at a suprathreshold level, the voice is then dropped and visual clues removed by covering the mouth. A child with normal hearing may discriminate the instruction at 35–40 dB(A). [Grade D]



Toy discrimination tests LINGUISTIC DEVELOPMENT LEVEL: 30 MONTHS ONWARDS



Several generations of toy or picture tests have been progressively developed and refined including the Stycar test,28 the Manchester picture test29 and the McCormick toy test30 which is widely used in the UK. This employs seven pairs of similar sounding nouns such as /cup/ and /duck/ or /tree/ and /key/, each represented by a small easily recognizable toy, placed on a table in front of the child in a quiet room. Using live voice, the child is asked to ‘show me the y spoon’, etc. Having established an understanding, the voice level is dropped and visual clues removed by covering the mouth. The speech discrimination score is taken as the quietest level at which the child correctly identifies 80 percent of the toys including the paired consonants. For a child with normal hearing this would be expected to occur at r40 dB(A). A digitized recorded version of the test ‘the Parrot’ utilizes a small loudspeaker to deliver the instruction at variable intensities.31 [***]



Consonant discrimination tests LINGUISTIC AGE: OVER 6 YEARS



A number of test packages have been designed to specifically examine auditory discrimination of



] 841



consonants and intonation patterns, such as the imitative test of speech pattern contrast perception (IMSPAC)32 and the four alternative auditory feature (FAAF) test.33 There is increasing use of interactive computer-based programmes, but these have so far mainly been employed in the evaluation of hearing aid benefit and for research purposes. [***] Older children from around the age of eight years may be compliant with formal speech audiometry, in quiet or noise, as described in Chapter 232, Psychoacoustic audiometry.



HEARING ASSESSMENT IN CHILDREN WITH SPECIAL NEEDS Approximately 30 percent of hearing-impaired children have an additional disability.34 Hearing assessment in children with special needs may be challenging and techniques must be appropriately adapted. For children with learning difficulties, the test most suited to their developmental level, rather than chronological age should be selected (www.nhsp.info).35 Clear simple demonstrations of conditioning tasks may require multiple repetitions to achieve understanding. In children with motor delay or physical disabilities, such as cerebral palsy, the required response must be tailored to the child’s capability. In distraction testing, for example, a partial head turn, eye glide or stilling may be judged as an acceptable response provided it is reproducible and validated by no sound trials. In VRA, it may be necessary to bring the visual reward closer and to accept an eye glance response. For performance testing and pure tone audiometry, the conditioning task may need to be modified, such as knocking over a skittle on a stand with a head movement, hitting a drum or even a reproducible postural change. Visually impaired children will turn to locate a sound from around eight months, but are more likely to respond to familiar sounds or if the response is reinforced by a tactile reward, such as touching the sound generator. The use of brightly illuminated reinforcers in a darkened room may facilitate VRA in the partially sighted. Assessment of hearing in a child with autistic spectrum disorder may present the greatest challenge, due to the characteristic self-directed behaviour and lack of shared attention, making interaction with the tester difficult. The repetitive nature of performance testing may particularly appeal to some, but in others distraction testing may be effective, particularly using electronically generated sounds. The child may not comply with sitting on a chair at the table and it may be necessary to take the test to the child, for example sitting on the floor in the corner of the room. Many autistic children are hypersensitive to noise and care must be taken to avoid distress. [**/*]



842 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY of accuracy.20, 26 Comparison of behavioural test response thresholds with subsequent pure tone audiometry thresholds cannot take account of fluctuating hearing levels during the intervening months or years. There are therefore inherent challenges in conducting large-scale controlled trials. In the UK, emphasis has more recently been placed on establishing guidelines for good practice, promoting training and availability of equipment with the aim of improving equity of service (www.nhsp.info; www.mchas.man.ac.uk).



KEY POINTS  Accurate testing of hearing is essential in the diagnosis and management of childhood hearing loss.  Testing and accurate assessment of the test results requires considerable skill and expertise.  A ‘menu’ of tests is available depending on the child’s age, the availability of expertise and the purpose of the test.  More than one test method may be needed.  Testing children with special needs is particularly challenging.



REFERENCES



 Best clinical practice [ Be age-appropriate: use the test technique that is [ [ [ [



most suited to the child’s age, developmental status and willingness or ability to co-operate. Be specific: know what you are testing. Calibrate the frequency and intensity of the sound stimulus. Be valid: reproducibility and use of ‘no sound trials’ are essential. Avoid other clues. Children can make good cheats. Be adaptable: nonstandard techniques may be best in nonstandard children. Be fun: children easily get bored and frightened. Their co-operation is required. Use toys and smiles – make it a game.



Deficiencies in current knowledge and areas for future research The only level 1 evidence and grade A recommendations cited in this chapter relate to the use of electrophysiological testing (ABR, OAE) for screening and diagnostic assessment in infants under six months of age.1 Other test methods are essentially descriptive, having been progressively modified, based on expert experience.4, 10, 22 Whilst specific elements of the test protocols have been subjected to small-scale controlled studies, for example the use of different noise stimuli in sound field testing,11 the only large-scale studies available examine the feasibility of obtaining response thresholds using a test protocol, rather than its sensitivity or specificity.22 There is clearly a difficulty in performing blind controlled trials of obviously differing techniques. There are also no subject-specific definitive response thresholds available for comparison. Reliance is placed on intra- and intertest/retest reproducibility, as an indication







1. Davis A, Bamford J, Wilson I, Ramkalawan T, Forshaw M, Wright S. A critical review of the role of neonatal hearing screening in the detection of congenital hearing impairment. Health Technology Assessment. 1997; 1: 10. 2. Sininger YS. Audiological assessment in infants. Current Opinion in Otolaryngology and Head and Neck Surgery. 2003; 11: 387–2. 3. Mason S, McCormick B, Wood S. The auditory brainstem response (ABR). Archives of Diseases of Childhood. 1988; 63: 465–7. 4. Northern JL, Downs MP. Clinical audiologic testing in children. In: Northern JL, Downs MP (eds). Hearing in children, 4th edn. Baltimore, MD: Williams and Wilkins, 1991: 131–77. 5. Thompson G, Weber B. Responses of infants and young children to behavioural audiometry (BOA). Journal of Speech and Hearing disorders. 1974; 39: 140–7. 6. Gans DP. Improving behaviour observation audiometry testing and scoring procedures. Ear and Hearing. 1987; 8: 92–100. 7. Madden C, Rutter M, Hilbert L, Greinwald Jr. JH, Choo DI. Clinical and audiological features in auditory neuropathy. Archives of Otolaryngology – Head and Neck Surgery. 2002; 128: 1026–30. 8. Ewing IR, Ewing AWG. The ascertainment of deafness in infancy and early childhood. Journal of Laryngology and Otology. 1947; 59: 309–38. 9. McCormick B. Hearing screening by health visitors: a critical appraisal of the distraction test. Health Visitor. 1983; 59: 143–4. 10. McCormick B. Behavioural tests for infants in the first years of life. In: McCormick B (ed.). Paediatric Audiology 0–5 years, 3rd edn. London: Whurr, 2004: 67–107. 11. Thompson G, Foulsom RC. Reinforced and nonreinforced head turn responses of infants as a function of stimulus bandwidth. Ear and Hearing. 1985; 6: 125–9. 12. Lutman ME, McCormick B. Converting free-field A weighted sound levels to hearing levels. Journal of the British Association of Teachers of the Deaf. 1987; 11: 127. 13. Suzuki T, Ogiba Y. Conditioned orientation reflex audiometry. Archives of Otolaryngology. 1961; 74: 84–90.



Chapter 68 Hearing tests in children







14. Liden G, Kankkunen A. Visual reinforcement audiometry. Acta Otolaryngologica. 1969; 67: 281–92. 15. Moore JM, Wilson WR, Thompson G. Visual reinforcement of head turn responses in infants under 12 months of age. Journal of Speech and Hearing Disorders. 1977; 42: 328–34. 16. Thompson G, Thompson M, McCall A. Strategies for increasing response behaviour of one and two year old children during visual reinforcement audiometry (VRA). Ear and Hearing. 1992; 13: 236–40. 17. National Deaf Children’s Society. Quality standards in paediatric audiology, 4. Guidelines for the early identification and the audiological management of children with hearing loss. London: National Deaf Children’s Society, 2000. 18. American Speech–Language Hearing Association. Guidelines for the audiological assessment of children from birth through 36 months of age. ASHA. 1991; 33: 37–43. 19. Gravel JS, Traquina DN. Experience with audiologic assessment of infants and toddlers. International Journal of Paediatric Otorhinolaryngology. 1992; 23: 59–71. 20. Gliddon ML, Martin AM, Green R. A comparison of some clinical features of visual reinforcement audiometry and the distraction test. British Journal of Audiology. 1999; 33: 355–65. 21. Widen JE, Folsom RC, Cone-Wesson B, Carty L, Dunnell JJ, Koebsell K et al. Identification of neonatal hearing impairment hearing status at eight to twelve months corrected age using a visual reinforcement audiometry protocol. Ear and Hearing. 2000; 21: 471–87. 22. Shaw P. Visual reinforcement audiometry. In: McCormick B (ed.). Paediatric audiology 0–5 years, 3rd edn. London: Whurr, 2004: 108–50. 23. Seewald RC, Moodie KS, Sinclair ST, Scollie SD. Predictive validity of a procedure for pediatric hearing instrument fitting. American Journal of Audiology. 1999; 8: 143–52.











] 843



24. Thompson G, Folsom RC. Reinforced and nonreinforced head-turn responses of infants as a function of stimulus bandwidth. Ear and Hearing. 1985; 6: 125–9. 25. Parry G, Hacking C, Bamford J, Day J. Minimal response levels for visual reinforcement audiometry in infants. International Journal of Audiology. 2003; 42: 413–7. 26. Primus MA. Repeated infant thresholds in operant and nonoperant audiometric procedures. Ear and Hearing. 1991; 12: 119–22. 27. Neilsen SE, Olsen SO. Validation of play-conditioned audiometry in a clinical setting. Scandinavian Audiology. 1997; 26: 187–91. 28. Sheridan MD. Simple clinical hearing tests for very young or mentally retarded children. British Medical Journal. 1958; 2: 999. 29. Hickson FS. The Manchester picture test (1984): a summary. Journal of the British Association of Teachers of the Deaf. 1987; 11: 161–6. 30. McCormick B. The toy discrimination test: an aid for screening the hearing of children above the age of two years. Public Health. 1977; 91: 67–73. 31. Shaw P. The Parrot speech discrimination test. British Society of Audiology News. 1997; 20: 44–5. 32. Kosky C, Boothroyd A. Validation of an on-line implementation of the imitative test of speech pattern contrast perception (IMSPAC). Journal of American Academy of Audiology. 2003; 14: 72–83. 33. Foster JR. The four alternative auditory feature test (FAAF) – linguistic and psychometric properties of the material with normative data in noise. British Journal of Audiology. 1987; 21: 165–74. 34. Fortnum H, Davis A. Epidemiology of permanent hearing impairment in Trent region 1985–1993. British Journal of Audiology. 1997; 31: 409–14. 35. Coninx F, Lancioni GE. Hearing assessment and aural rehabilitation of multiply handicapped deaf children. Scandinavian Audiology. 1995; 24: 41.



69 Investigation and management of the deaf child SUJATA DE, SUE ARCHBOLD AND RAY CLARKE



Introduction Epidemiology Aetiology The newly diagnosed deaf child Support and education



844 844 845 849 853



Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



856 856 857 857



SEARCH STRATEGY The references for this chapter were compiled from a Medline search using the terms deaf, hearing loss or hearing impairment and child. The World Health Organization (WHO) website (www.who.int), the NHS Newborn Hearing Screening Programme website (www.nhsp.info), the deafness and hereditary hearing loss overview website (www.geneclinics.org/profiles/deafness-overview/details) and the NHS Modernising Childrens’ Hearing Aid Service website (www.psych-sci.manchester.ac.uk/mchas) were also consulted.



INTRODUCTION This chapter focusses on the general principles of management of a moderately to profoundly deaf child. Sensorineural deafness may occur before (prelingual deafness) or after the child has acquired the ability to speak (postlingual deafness). This distinction has important implications for the child’s education and management. Earlier detection with universal neonatal screening is now practised in most developed countries. Advances in detection, genetics, imaging and treatment – including amplification and cochlear implantation – for these children have meant that new guidelines and ways of working are needed for healthcare professionals. There has also been more understanding of the deaf culture, the deaf community and their concerns with these advances. The implication of all levels and types of deafness for education has increasingly been recognized, together with the implications of managing the modern technologies in educational settings. The education of deaf children has



also been challenged with a worldwide move towards the integration of deaf children into mainstream education wherever possible.



EPIDEMIOLOGY Approximately one child in 1000 (some 900 per year in the UK) is born with a bilateral permanent childhood hearing impairment (PCHI). This is defined as confirmed permanent bilateral hearing impairment exceeding 40 dBHL (hearing level) (average of pure tone thresholds at 0.5, 1, 2 and 4 kHz in the better hearing ear). About 60 percent of these children have a moderate (41–60 dBHL) hearing loss, while the remainder have a severe (61–80 dBHL) or profound (481 dbHL) loss (www.who.int/pbd/ deafness/hearing_impairment_grades/en/index). Hearing loss in a child may be present at birth (congenital) or may develop after birth (acquired). The prevalence of PCHI increases with age, suggesting that a further one in 1000 children develop acquired or progressive hearing



Chapter 69 Investigation and management of the deaf child



impairment.1 The terms ‘deafness’ and ‘deaf ’ in this chapter are used to cover both moderate, severe and profound hearing loss. The WHO estimated in 2005 that there were 278 million people worldwide with bilateral moderate to profound hearing loss, of whom 62 million had deafness that began in childhood. Two-thirds of people with moderate to severe hearing impairment live in developing countries. In the developing world, the greatest proportion of childhood hearing loss is caused by infection. This includes congenital conditions, such as rubella and cytomegalovirus, and acquired childhood infections such as mumps, measles, meningitis and chronic otitis media. In the developed world, about half of children with PCHI have a genetic cause for their deafness.



AETIOLOGY Most children with permanent childhood hearing impairment have sensorineural hearing loss. Causes of conductive hearing loss include chronic otitis media and some rare congenital conditions, such as bilateral aural atresia. The causes of PCHI are shown in Table 69.1.



Table 69.1



] 845



Causes of permanent childhood hearing impairment.



Causes of hearing impairment Congenital disorders Genetic Syndromic



Nonsyndromic



Congenital rubella syndrome Cytomegalovirus Congenital syphilis Environmental causes Perinatal Hypoxia causes Hyperbilirubinemia Low birth weight Acquired Infections disorders



Autosomal recessive Autosomal dominant X-linked Mitochondrial Autosomal recessive Autosomal dominant X-linked Mitochondrial



Nongenetic



Chronic otitis media Meningitis Mumps Measles AIDS



Ototoxic drugs Trauma Neoplastic disease



Congenital causes Idiopathic



GENETIC



Genetic disease results from aberrations in the coding function or processing of human DNA. In the western world, congenital sensorineural hearing loss is accounted for by genetic factors in approximately half of cases. About 77 percent of these cases of hereditary hearing loss are autosomal recessive, 22 percent are autosomal dominant and 1 percent are X-linked.2 In addition, a small fraction (less than 1 percent) represents those families with mitochondrial inheritance in which the trait is passed through the maternal lineage. In general, autosomal recessive forms of hearing loss are prelingual. Autosomal dominant forms tend to present later and result in progressive, postlingual hearing loss.3 Approximately 30 percent of patients will have a pattern of additional medical anomalies such as to constitute a syndrome – syndromic hearing loss. Nonsyndromic hearing loss constitutes most cases of congenital sensorineural loss. A number of gene mutations thought to be responsible for nonsyndromic hearing loss have recently been identified. The disorder DFNB1, caused by mutations in the GJB2 gene (which encodes the protein connexin 26) and the GJB6 gene (which encodes the protein connexin 30), accounts for 50 percent of autosomal recessive nonsyndromic hearing loss. The carrier rate in the general population for a recessive deafness-causing GJB2 mutation is about one in 33.4



Syndromic sensorineural hearing loss is found in conditions such as Pendred’s syndrome, Branchio-otorenal syndrome, Usher syndrome and Wardenburg syndrome, amongst others. Conductive hearing loss is commonly encountered in Down syndrome and Treacher Collins syndrome.



NONGENETIC



Perinatal factors In the developed world, approximately half of cases of congenital hearing loss are due to environmental factors. In previous decades, infectious agents such as rubella and cytomegalovirus were common environmental causes of hearing loss. While the frequency of such infections has dropped with the introduction of vaccines, a concomitant improvement in the survival rates of preterm babies has increased the proportion of babies with hearing loss related to perinatal events, including stays in the neonatal intensive care unit (NICU). Thus, the overall incidence of nongenetic deafness has remained approximately unchanged. Preterm and low-birth-weight infants are particularly susceptible to factors, such as hypoxia and



846 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY hyperbilirubinaemia. Davis and Wood5 showed that babies admitted to a neonatal intensive care unit for more than 48 hours were 10.2 (95 percent, confidence interval 4.4 – 23.7) times more likely to have a permanent hearing loss (greater than 50 dBHL in this study) than those who did not undergo intensive care. Hypoxia is associated with apnoea, difficult delivery, low Apgar scores and use of ventilation. It is known to be associated with neurodevelopmental deficits, but the exact mechanism of hypoxia-related hearing loss is unclear. Hyperbilirubinemia is an independent risk factor for sensorineural hearing loss in infants. High levels of unconjugated bilirubin cross the immature blood–brain barrier and deposit in the grey matter causing neurotoxicity. This is thought to lead to sensorineural hearing loss, which is usually permanent, although some cases may be reversed once bilirubin levels return to normal.6 These infants are more often than not nursed on a neonatal intensive care unit, where they may receive ototoxic drugs, such as aminoglycoside antibiotics (e.g. gentamicin, tobramycin and amikacin) and diuretics (e.g. furosemide). There is some evidence of a genetic susceptibility to the damaging effects of these drugs explaining the variability in hearing loss that affects infants treated in neonatal intensive care units.7 Maternal infections Infection is still responsible for a significant proportion of acquired hearing loss, especially in developing countries, although vaccination has greatly reduced the incidence. Congenital cytomegalovirus (CMV) is the most common cause of nonhereditary sensorineural hearing loss in the developed world.8, 9 Between 22 and 65 percent of symptomatic and 6 and 23 percent of asymptomatic children will have hearing loss following congenital CMV infection.10 It has been estimated that 12 percent of congenital sensorineural hearing loss is due to CMV infection. The hearing loss can be of delayed onset with threshold fluctuations and/or progressive loss.3 Congenital rubella syndrome (CRS) is probably the most important cause of nongenetic congenitally acquired hearing loss in countries with no rubella vaccination programme. CRS occurs when there is maternal infection with the rubella virus in the first trimester of pregnancy. It leads to a number of abnormalities in the child including deafness, ocular defects (cataracts, glaucoma), cardiovascular anomalies (patent ductus arteriosus, pulmonary artery stenosis and ventricular septal defects), central nervous system problems (microcephaly, global retardation) and characteristic skin changes (Figure 69.1).11 The WHO has recommended vaccination against rubella (www.who.int/immunization/topics/rubella/en/index1). Two approaches have been suggested: (1) prevention of CRS only, through immunization of adolescent girls and/or women of childbearing age; or (b)



Figure 69.1 Congenital rubella with characteristic skin rash. The child was profoundly deaf. Courtesy of J Verbov, Emeritus Professor of Paediatric Dermatology, Royal Liverpool Children’s Hospital, Alder Hey.



elimination of rubella, as well as CRS through universal vaccination of infants and young children (with/without mass campaigns), surveillance and assuring immunity in women of childbearing age.12 This is the norm in most developed countries where the incidence of CRS deafness is now very rare.13 Syphilis is caused by the bacterium Treponema pallidum. It is a common infection in most of the developing world and has recently re-emerged in parts of the developed world. The main route of transmission is sexual contact, but syphilis can also be transmitted from an infected mother to her baby (congenital syphilis). Congenital infection can occur at any stage during pregnancy, but the highest likelihood of damage to the foetus is when infection occurs and is untreated during the first or second trimesters. During primary syphilis, the rate of vertical transmission in untreated women is 70–100 percent; this drops to 10–40 percent in the latent stage of the disease. Thus the poorest prognosis is for an infant infected during the first or second trimester by a mother in the primary or secondary stages of disease.



Chapter 69 Investigation and management of the deaf child



] 847



Hearing loss is a late feature of congenital syphilis and often appears as one of a group of signs known as ‘Hutchinson’s triad’. These signs include inflammation of the cornea giving it an opaque appearance, which leads to loss of vision, peg-shaped upper incisors (Hutchinson’s teeth) and eighth cranial nerve deafness. Hearing loss is the least common component of Hutchinson’s triad and occurs in around 3 percent of children with late congenital syphilis. It typically appears when the child is eight to ten years of age, although occasionally it may be delayed until adulthood. Onset is sudden and damage to the cranial nerve is thought to result from a persistent and ongoing inflammatory response to the infection. Loss of hearing may be unilateral or bilateral and initially involves higher frequencies, with normal conversational tones affected later.14 Congenital syphilis can be treated with high-dose penicillin, but is best prevented by identification and eradication of the infection in the mother. Prenatal screening for syphilis is commonplace in developed countries and is the main priority for prevention of the disease in developing countries.



Acquired causes Causes of permanent acquired hearing loss include bacterial meningitis, chronic otitis media, mumps, measles, trauma and ototoxic drugs.



MENINGITIS



The most common cause of acquired PCHI is childhood meningitis. The risk of developing significant sensorineural impairment after bacterial meningitis has been estimated at approximately 10 percent.15, 16, 17, 18 Early assessment of hearing after meningitis is recommended so that appropriate rehabilitation can be initiated. If cochlear implantation is necessary, this should be done with minimal delay as ossification of the cochlear duct can make implantation difficult or even impossible (see Chapter 70, Paediatric cochlear implantation).



Figure 69.2 Child with maculopapular rash characteristic of measles. Courtesy of J Verbov, Emeritus Professor of Paediatric Dermatology, Royal Liverpool Children’s Hospital, Alder Hey.



Measles has been reported as a major aetiological factor for severe to profound bilateral hearing loss in deaf children. As it was known that mucous membranes all over the body were affected, the observed hearing loss was previously thought to be conductive and attributable to suppurative otitis media, chronic perforation and mastoiditis. There are reports of measles virus being found within the cochlea, thus providing the needed evidence in favour of the sensorineural component. This explains the often-reported severe and sometimes progressive/delayed sensorineural hearing loss among affected children. The WHO and United Nations Children’s Fund (UNICEF) have emphasized primary prevention of measles in its global campaign against the prevailing childhood illnesses in the developing world for several decades. Immunization for measles can be administered as a single vaccine or as the triple MMR (measles, mumps and rubella) vaccine (www.who.int/vaccines-documents/ DocsPDF01/www573).



MEASLES



Measles is a highly infectious viral illness, which presents acutely with high fever, running nose, characteristic Koplik’s spots on the buccal mucosa and a distinctive generalized maculopapular rash (Figure 69.2). It occurs worldwide but its incidence has reduced significantly in developed countries since the introduction of an effective vaccine in 1968. It is predominantly a disease of infants and young children and occurs mostly after the age of six months. Globally, measles is the leading cause of vaccinepreventable child deaths.



MUMPS



Mumps, caused by infection with the mumps virus, is a nonsuppurative enlargement of the salivary glands, particularly the parotids. The infection may be subclinical in up to one-third of the cases and in these the first presentation may be the appearance of complications. Sensorineural hearing loss as a result of mumps is mostly unilateral, although bilateral loss has been described. The incidence of mumps-related sensorineural hearing loss has been documented as 5/100,000.19



848 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Mode of presentation It has been shown that early diagnosis of PCHI has an important bearing on outcome. A landmark study has demonstrated that children whose hearing losses were identified by six months of age demonstrated significantly better receptive and expressive language skills than did children whose hearing losses were identified after the age of six months.20 [***]



SCREENING



Since the recent introduction of universal neonatal screening in most developed countries, PCHI is identified mainly through this modality. However, some children with PCHI are still missed until a later age. In countries where neonatal screening has not been implemented, diagnosis is still dependent upon parental suspicion, distraction testing and preschool screening. In addition, children with progressive hearing loss may be identified by surveillance and parental/teacher suspicion. Children who develop PCHI as a result of childhood illness are usually identified by a high degree of suspicion by parents and carers and by surveillance. Prior to the introduction of universal neonatal screening in the UK in 2001/2002, almost a quarter of children born deaf were not identified until they were over 3.5 years of age.21 As a result of universal neonatal screening in the UK, the median age of detection of PCHI is now ten weeks.22 Currently in the UK, all children are screened with automated otoacoustic emissions (AOAE) as close to the time of birth as possible by a trained operator. The result of the test is either a ‘pass’ or ‘refer’. In addition, all infants in neonatal intensive care units have automated auditory brainstem response (AABR) testing. This is because there is a proportion of children with auditory neuropathy (estimated to be up to three in 1000 in the neonatal intensive care unit) whose AOAE test would give a ‘pass’ result.23 A child who does not pass the OAE screening test in both ears would go on to have AABR testing. If this too is not passed then the child is referred for further testing. More information on screening can be found in Chapter 67, Hearing loss in preschool children: screening and surveillance.



course, unilateral hearing loss can also be acquired and this occurs as a result of trauma, infection, ototoxicity, metabolic disorders and surgery. Unilateral hearing loss results in an inability to hear from the direction of the deaf side, difficulty in hearing speech in noise and poor localization of sound. A review carried out in 2004 concluded that school children with unilateral hearing loss appear to have increased rates of grade failures, need additional educational assistance and may show behavioural issues in the classroom. Speech and language delays may occur in some children, but it is unclear if children ‘catch up’ as they grow older.24 It would appear that the impact of unilateral hearing loss on the language development and educational performance varies considerably. Current practice is that unilateral sensorineural hearing loss is investigated to identify progressive pathology or conditions that might affect the other ear, e.g. dilated vestibular aqueducts. At the time of writing, children with single-sided deafness are monitored in the UK, but not treated unless the child complains of a problem with localization or perception of speech in noise. DELAYED PRESENTATION



The infant distraction test (IDT), which was the screening test used in the UK until recently, had been in place since the 1960s. It involved a health visitor check at the age of nine months within the community. A behavioural test was used which had neither high sensitivity nor specificity. As a result, a large number of children with normal hearing were referred for further investigations. More worryingly, a large number of children with PCHI were not being identified. There are no data available yet on how many children are detected late since the introduction of universal neonatal screening in the UK. In countries where universal neonatal screening has not yet been established, identification of children with PCHI is still dependent on parental or teacher suspicion, locally arranged behavioural screening programmes and preschool screening. In a questionnaire-based study in Nigeria, only 12 percent of parents of a child with hearing loss suspected hearing difficulty by the age of six months. Parental suspicion occurred mostly at 12–24 months, compared with 8–14 months in developed countries. The most common mode of detection was a child’s failure to respond to sound (49 percent).25



SINGLE-SIDED DEAFNESS



Screening has also resulted in early detection of unilateral sensorineural hearing loss. In the past, unilateral hearing loss was considered to be of little consequence because speech and language presumably developed appropriately with one normal-hearing ear. The yield for permanent unilateral moderate or greater hearing loss in the Newborn Hearing Screening Programme in England is 0.64 (95 percent, CI 0.37–0.91) per 1000 screened.22 Of



PROGRESSIVE HEARING LOSS



Some children have normal hearing or only a mild loss at birth, but develop progressive hearing loss with time. A study performed in the UK just before the introduction of universal neonatal screening showed the prevalence of confirmed PCHI to be 1.07 per 1000 (95 percent confidence interval 1.03–1.12) for three-year-olds. Amongst children aged 9–16 years though, the prevalence



Chapter 69 Investigation and management of the deaf child



increased to 2.05 per 1000 (2.02–2.08). This increase in prevalence must be due to either acquired or progressive hearing loss. It is likely that a number of children have progressive hearing loss and would not necessarily be identified by neonatal or early screening.1 Progressive hearing loss may be due to genetic or environmental causes. In some syndromes such as Pendred’s syndrome, the hearing loss is typically progressive. Nonsyndromic hearing loss related to a variety of known and unknown mutations can also be fluctuant or progressive.26 Some infectious causes such as congenital CMV and congenital syphilis can cause late-onset or progressive hearing loss. This group poses a challenge, as neonatal screening will not identify them. Identification of these cases therefore is dependent upon a high degree of suspicion in children with known associated syndromes, surveillance and prompt assessment in response to parent or teacher concern. HEARING LOSS DUE TO CHILDHOOD ILLNESS



Acquired PCHI can be caused by meningitis, viral infections such as mumps and measles, chronic otitis media, trauma or ototoxicity (streptomycin, cisplatin). Early detection of hearing loss in these groups is based upon surveillance of at-risk groups (e.g. children undergoing chemotherapy) and parent/teacher/medical staff suspicion.



THE NEWLY DIAGNOSED DEAF CHILD Parents and healthcare professionals are faced with a series of dilemmas and challenges when it is established that a child is deaf. Parents remember with chilling clarity the day they were told their baby was deaf, and cumbersome inexpert handling of this scenario can have profound adverse effects. An experienced empathic team that works in a coordinated way with the family and can present a knowledgeable and nonpartisan approach to the various options and resources available makes for the best outcome for parents and child. Now the diagnosis is likely to take place very early in the child’s life before parents have adjusted to being parents and it is even more important that time is made available for them and that they are supported in the early days.



Reactions to the diagnosis For the majority of parents, having a child with hearing loss is completely unexpected. Parents, when confronted with a diagnosis of hearing loss in their child, go through a grieving process akin to a death experience; they have lost the expectation of the ‘perfect’ baby. For the parents of a child with a disability, this time frame is less rigid and



] 849



the grief can be nonfinite and last the lifespan of the child, emerging at critical points. In a survey of parental reactions to the news that their baby had failed a hearing screening test, the feelings elicited were of fear, shock, confusion, depression, frustration, anger, loneliness, sadness, blame and even agression.27 Large surveys have shown that the vast majority of parents would like to know whether their child has permanent hearing impairment as close to birth as possible.28 This is now an expectation, but it places great responsibility on the professionals. Parents will have questions regarding the development of the child, especially speech and future education needs. Many of these questions cannot be answered straight away. The parents are then bombarded with information regarding intervention and are required to make decisions often without understanding or taking in the information they have been given. The presence of a child with hearing loss is often linked with psychosocial stress in the parents and other family members.29 Once a hearing loss has been confirmed, these emotions become even more heightened. The degree of hearing loss does not seem to influence the parents’ ability to cope. In fact, stress levels were found to be higher in the parents of children with less severe hearing loss.30 This might be because the child may at least sometimes respond to sound, which can delay the parents’ acceptance of the hearing loss and confuse the diagnosis. The initial grief may be followed by feelings of anger, despair and helplessness as the diagnosis becomes apparent. Sadness, sorrow, even depression may occur. Parental guilt is also a feature of the reaction to the diagnosis. The expectation of having to make decisions that will hugely impact the future of their child and the amount of information from health professionals that they have to take in within a short span of time leads to confusion and sometimes insecurity. Stress and quality of life remain factors even after the parents have come to terms with the diagnosis. The impact on the parents of earlier diagnosis of hearing loss as a result of neonatal screening is a cause for concern. It has been suggested that such an early diagnosis of hearing loss may interfere with parental bonding with a child that is usually otherwise perfectly healthy, and which we know is vital to support the early development of communication skills, the foundations of later language development, whether spoken or signed. This concern is reflected by the experience of some parents. The risk of disturbing the parent–child relationship by early screening could be minimized by improved information and rapid and effective follow up. Young and Tattersall31 used a narrative approach to parents as evaluators of newborn hearing screening. They reaffirmed the excellence of screening practice but also raised new questions about how to pre-empt the experiences of the minority for whom screening will not be a satisfactory process.



850 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Investigations In the UK, national guidelines for investigating the aetiology of permanent childhood hearing impairment were developed by the British Association of Audiological Physicians in conjunction with the British Association of Community Doctors in Audiology (BACDA) (www.baap. org.uk/Guidelines/BAAP_guidelines_Jan_2001). It is recommended that core investigations are offered to the parents of all children with newly diagnosed bilateral sensorineural hearing loss and thresholds over 70 dB in the better ear averaged across 500, 1000, 2000 and 4000 Hz, while ‘additional’ investigations are dictated by individual circumstances and the findings of core investigations (Tables 69.2 and 69.3). Parents may not wish to pursue investigations, the results of which in many cases have no immediate impact on the management of the child, but it is essential that healthcare professionals are aware of the baseline investigations and can discuss the rationale behind various tests with parents so they can make an informed choice. There is some evidence, however, that these guidelines are not being fully implemented. This is mainly due to issues of funding and parental choice. In an audit of children diagnosed with severe or profound PCHI between 2002 and 2004, 41 percent were not offered imaging of the inner ears. A total of 47.1 percent accepted and 52.9 percent declined electrocardiograph (ECG) evaluation. A total of 70.6 percent accepted and 29.4 percent declined connexin mutations testing.32 The aims of aetiological investigations are as follows:  to try to answer parents who ask ‘why is my child deaf?’;  to identify and treat medical conditions, e.g. 8th nerve aplasia, congenital infection, Jervell and Lange–Nielsen syndrome, Alport’s syndrome, neurofibromatosis type 2, Usher syndrome and vestibular hypofunction;  to assist the family in making decisions about the most appropriate communication mode, educational placement and counselling on cochlear implantation, e.g. in 8th nerve aplasia, Usher syndrome;  to inform genetic counselling;  to inform epidemiological research. The recommendation is that patients should undergo core investigations with additional investigations as dictated by the clinical scenario. [**] CLINICAL EXAMINATION



All children with hearing loss of unknown cause should be evaluated for features associated with syndromic deafness. Important features include branchial cleft pits, cysts or fistulae, preauricular pits, telecanthus, heterochromia iridis, white forelock, pigmentary anomalies, high myopia, pigmentary retinopathy, goitre and craniofacial anomalies.



Table 69.2 Core investigations for all cases of bilateral severe to profound sensorineural hearing loss. Investigation Paediatric history:



Family history:



Clinical examination:



Audiology:



Imaging:



Detailed history of pregnancy delivery and postnatal period. Developmental milestones including speech and language and motor milestones, pre- and postnatal noise exposure, history of ototoxic medications, head injuries, ear disease, meningitis, viral illness and immunization status Deafness or risk factors associated with hearing loss in first- and second-degree relatives Inspection and physical measurement of craniofacial region, assessment of the neck, skin and nails, limbs, chest and abdomen. Developmental examination Age-appropriate assessment including tympanometry. Audiometry on first-degree relatives Magnetic resonance imaging of inner ears and/or computed tomography of petrous temporal bones.



Electrocardiograph Urine for microscopic haematuria Connexin 26 and 30 mutations testing with access to clinical genetics service for counselling Ophthalmic assessment Referral to clinical geneticist Vestibular investigations



AUDIOLOGY



Skilled audiological testing must be performed to assess the severity of the hearing loss and to determine whether it is conductive, sensorineural or mixed. This is discussed in detail in Chapter 68, Hearing tests in children.



IMAGING



Imaging can detect treatable pathologies such as perilymph fistula, inner ear malformations with an associated



Chapter 69 Investigation and management of the deaf child Table 69.3



] 851



Additional investigations.a



Investigations Haematological and biochemical tests Serological test for congenital rubella and cytomegalovirus infections Thyroid tests Immunology tests Metabolic screen blood and urinalysis Renal ultrasound Chromosomal analysis a



Only used when medical indications are present.



risk of meningitis and the enlarged vestibular aqueduct. Both computed tomography (CT) and magnetic resonance imaging (MRI) can be useful. Imaging modalities and techniques are discussed in Chapter 102, Imaging in paediatric ENT. Radiological abnormalities either on CT or MRI are found in about one-third of children with PCHI (Figure 69.3). On CT scans, a large vestibular aqueduct is the most common isolated finding, while cochlear dysplasia is the most common abnormality in scans showing multiple abnormalities. Other abnormal findings include lateral semicircular canal dysplasia, otic capsular lucency, small internal auditory canals and hypoplastic cochlea. MRI scans are more likely to detect abnormalities primarily involving the central nervous system.33 There is a slightly higher yield of radiological abnormalities with increasing severity of hearing loss and a significantly higher imaging yield with unilateral sensorineural hearing loss than with bilateral.34 At least 40 percent of children with large vestibular aqueduct (LVA) will develop profound sensorineural hearing loss. The identification of LVA should raise the suspicion of Pendred syndrome and thyroid abnormality. Patients with LVA are at risk for progressive deafness after minor head trauma. Identifying this anomaly influences management as it prompts a search for treatable thyroid function anomalies and parents can be counselled with respect to the dangers of incidental head trauma.35



SEROLOGY



Universal neonatal screening will detect less than half of all sensorineural hearing loss caused by congenital CMV infection36 as the hearing loss in a large proportion of children with congenital CMV is progressive. Laboratory testing has to be carried out in neonatal samples within three weeks of life. This presents difficulties as after three weeks of age, virus isolation could be due to acquired infection, which is not usually associated with adverse outcome. Previous studies9, 37 have shown that dried blood spots on Guthrie cards collected at birth for the screening of metabolic disorders have proved a valuable



(a)



(b)



(c)



Figure 69.3 Typical appearances of a Mondini deformity in a child with bilateral severe sensorineural deafness. Axial high resolution CT scan through (a) dilated vestibular aqueducts and (b) a more inferior slice that shows bilateral dysplastic vestibules. (c) Coronal image through the abnormal cochlea showing typical Mondini abnormality with incomplete formation of the apical and middle turns of the cochlea. Images kindly supplied by D Saunders, Consultant Neuroradiologist, Great Ormond Street Hospital, London, UK.



tool for demonstrating CMV DNA for diagnosing congenital infection after months or even years of life. Rubella-specific IgG and IgM synthesized by the foetus are detectable at birth in CRS. However, since maternally derived rubella-specific IgG is also present in infants’ sera, laboratory diagnosis of CRS is almost invariably made by detection of rubella-specific IgM responses. This response is detectable in almost 100 percent of CRS cases up to age three months with the most sensitive antibody-capture assays. The response declines progressively to less than 50 percent at 12 months and is rarely detectable after 18 months. It is therefore suggested that there is value in testing for CRS only if the deafness is identified within the first six months of life.38 GENETIC SCREENING



The disorder DFNB1, caused by mutations in the GJB2 gene (which encodes the protein connexin 26) and the GJB6 gene (which encodes the protein connexin 30), accounts for 50 percent of autosomal recessive nonsyndromic hearing loss. The carrier rate in the general population for a recessive deafness-causing GJB2 mutation is about one in 33. It is current practice that the families of all children with severe/profound hearing loss



852 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY are offered genetic screening for these two mutations. Establishing a molecular diagnosis of GJB2-related deafness is important clinically since these children can avoid further diagnostic tests and are not at increased risk for medical comorbidity. In general, bony abnormalities of the cochlea are not part of the deafness phenotype and developmental motor milestones and vestibular function are normal.4 The hearing loss associated with GJB2related deafness can vary from mild to profound. The diagnostic yield of GJB2 screening is significantly higher in patients with severe to profound sensorineural hearing loss (SNHL) than in all other groups.34 Preciado et al. suggested that a cost-effective approach to screening would be to test children with severe to profound SNHL with a GJB2 screen, as opposed to those with milder SNHL, who should undergo imaging as the initial testing step.39 Screening for the mitochondrial aminoglycoside susceptibility gene mutation A1555G is offered to patients who have developed hearing loss following exposure to aminoglycosides, perhaps on the neonatal intensive care unit. If the gene is identified and subsequent administration of aminoglycosides restricted, then the severity of aminoglycoside-induced hearing loss can be limited. Also, as the gene is inherited from the mother, restricting the use of aminoglycosides in maternal relatives will also limit the prevalence of hearing loss caused by these drugs.40 Numerous genes have been and continue to be identified relating to hereditary hearing loss. DNA testing can be carried out but at present in the UK, except for the connexins and the aminoglycoside susceptibility genes, routine screening is not performed.



BLOOD TESTS



Although routine blood tests such as full blood count (FBC), thyroid function tests, erythrocyte sedimentation rate (ESR), syphilitic blood tests, cholesterol and triglyceride levels, urea and electrolytes are carried out, abnormalities are rarely related to the cause of the hearing loss.34



Genetic counselling Genetic counselling is the process of providing individuals and families with information on the nature, inheritance and implications of genetic disorders to help them make informed medical and personal decisions. The parents of all children with genetic PCHI should be offered the services of a genetic counsellor; they might not be considering further children themselves, but it is important for their present children and their future plans of families of their own. These discussions are likely to be sensitive and the information needs to be shared sensitively with a counsellor present to follow up issues;



discussion of comparative statistical odds can be very confusing and misunderstandings can lead to repercussions later. When the parents of the child are themselves deaf, then there are other issues that have to be taken into consideration. Many deaf people view deafness as a distinguishing characteristic and not as a handicap, impairment or medical condition requiring a ‘treatment’ or ‘cure’, or needing to be ‘prevented’. In fact, having a child with deafness may be preferred over having a child with normal hearing.41 Healthcare professionals need to be sensitive to these feelings and to have a frank discussion with parents to establish their views and aspirations. Hereditary hearing loss may be inherited in an autosomal dominant, an autosomal recessive or an X-linked manner. Mitochondrial disorders with hearing loss also occur. The risk to siblings and risk of inheritance can be predicted from the pattern of inheritance. However, clinical severity and disease phenotype may differ between individuals with the same mutation; thus, age of onset and/or disease progression may not be predictable (www.geneclinics.org/profiles/deafness-overview/details).



AUTOSOMAL DOMINANT HEREDITARY HEARING LOSS



Most children diagnosed as having autosomal dominant hereditary hearing loss have a deaf parent; the family history is rarely negative. However, the disorder may be the result of a de novo gene mutation. The proportion of cases caused by de novo mutations is unknown, but thought to be small. If one of the parents has the mutant allele, the risk to siblings of inheriting the mutant allele is 50 percent. Individuals with autosomal dominant hereditary hearing loss have a 50 percent chance of transmitting the mutant allele to each child. When neither parent has the deafness-causing mutation or clinical evidence of the disorder, it is likely that there has been a de novo mutation. However, possible nonmedical explanations including alternate paternity or undisclosed adoption could also be explored (www.geneclinics.org/profiles/deafness-overview/details).



AUTOSOMAL RECESSIVE HEREDITARY HEARING LOSS



The parents are obligate heterozygotes and therefore carry a single copy of a deafness-causing mutation. Heterozygotes are asymptomatic. At conception, each sibling has a 25 percent chance of being deaf, a 50 percent chance of having normal hearing and being a carrier, and a 25 percent chance of having normal hearing and not being a carrier. For children with GJB2-related (connexin) severeto-profound deafness, siblings with the identical GJB2 genotype have a 91 percent chance of having severe-toprofound deafness and a 9 percent chance of having



Chapter 69 Investigation and management of the deaf child



mild-to-moderate deafness. For children with GJB2related deafness and mild-to-moderate deafness, siblings with the identical GJB2 genotype have a 66 percent chance of having mild-to-moderate deafness and a 34 percent chance of having severe-to-profound deafness (www.geneclinics.org/profiles/deafness-overview/details).



X-LINKED HEREDITARY HEARING LOSS



In this pattern of inheritance, the recessive gene is carried on the X chromosome. Therefore, only males are affected. If pedigree analysis reveals that the deaf male is the only individual in the family with hearing loss, several possibilities regarding his mother’s carrier status need to be considered:  he has a de novo deafness-causing mutation and his mother is not a carrier;  his mother has a de novo deafness-causing mutation, as either: a ‘germline mutation’ (i.e. occurring at the time of her conception and thus present in every cell of her body); or ‘germline mosaicism’ (i.e. present in some of her germ cells only);  his maternal grandmother has a de novo deafnesscausing mutation. A female who is a carrier has a 50 percent chance of transmitting the deafness-causing mutation with each pregnancy. Sons who inherit the mutation will be deaf; daughters who inherit the mutation are carriers and are likely to have normal hearing. If the mother is not a carrier, the risk to siblings is low, but greater than that of the general population because of the possibility of germline mosaicism. Males with X-linked hereditary hearing loss will pass the deafness-causing mutation to all of their daughters and none of their sons (www.geneclinics.org/profiles/ deafness-overview/details).



MITOCHONDRIAL DISORDERS WITH HEARING LOSS AS A POSSIBLE FEATURE



Inheritance is via the mother who usually has the mitochondrial mutation and may or may not have symptoms. Alternatively, there may be a de novo mitochondrial mutation. If the mother has the mitochondrial mutation, all siblings are at risk of inheriting it. All offspring of females with a mutation are at risk of inheriting the mutation. Offspring of males with a mitochondrial DNA mutation are not at risk. UNKNOWN DIAGNOSIS/MODE OF INHERITANCE



The subsequent offspring of a hearing couple with one deaf child and an otherwise negative family history of deafness have an 18 percent empiric probability of



] 853



deafness in future children. If the deaf child does not have DFNB1 based on molecular genetic testing of GJB2 (which codes for the protein connexin 26), the recurrence risk is 14 percent for deafness unrelated to connexin 26. If the hearing couple is consanguineous or comes from a highly inbred community, the subsequent offspring have close to a 25 percent probability of deafness because of the high likelihood of autosomal recessive inheritance. The offspring of a deaf person and a hearing person have a 10 percent empiric risk of deafness. The child of a nonconsanguineous deaf couple in whom autosomal dominant deafness has been excluded has an approximately 15 percent empiric risk for deafness. However, if both parents have connexin 26-related deafness, the risk to their offspring is 100 percent. Conversely, if the couple has autosomal recessive deafness known to be caused by mutations at two different loci, the chance of deafness in their offspring is lower than that of the general population. (www.geneclinics.org/profiles/ deafness-overview/details).



SUPPORT AND EDUCATION The rationale for intervention Deafness has a profound effect on both adults and children and can impose a heavy social and economic burden on individuals, families, communities and countries. In normally hearing children, the language of the home is acquired through the channel of hearing. For profoundly deaf children, this normal acquisition is disrupted, leading to the likelihood of communication, speech and language delay, which may result in underachieving educationally and later in employment. Over 90 percent of deaf children are born to hearing families, where the language of the home will be a spoken language; for deaf children of deaf parents who use sign language, language will be acquired naturally through the visual route. For deaf children of deaf parents then language learning may be easier and more effective. As adults, deafness often makes it difficult to obtain, perform and keep employment. Both children and adults may suffer from social isolation as a result of deafness. There is some evidence that access to services for deafness are dependent on social class and income with those in developing countries and in the poorer sectors of society having difficulty in accessing the services they need. The field of deafness is fraught with controversy for many years and the clinician needs to be sensitive to some contentious issues when working with parents of young deaf children. Deafness can be viewed in differing ways: as a medical condition to be ‘cured’, as a cultural and linguistic identity to be valued rather than ‘cured’ and the



854 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY educational models of deafness and models of intervention often reflect these differences. For parents of newly diagnosed deaf children these issues can be bewildering; although there are differing models of intervention and management, there is little evidence for the efficacy of one over another.42 Today’s technology of early diagnosis, effective hearing aids and cochlear implants emphasizes the medical model of deafness, in which the technology can be viewed as alleviating the effects of deafness and providing the opportunity for ‘normal’ development. For those with the traditional view of the deaf community, these developments can be seen as a threat in which sign language with its own grammar and culture will no longer have a place. In fact it may be that our present situation, offering more effective technology than ever before with better and earlier access to hearing and with a better understanding of the importance of early interaction in language development, offers opportunities to move forward from the old arguments and to offer deaf children real potential to develop spoken language, while not denying the value of sign language too. The interested reader is referred to the review by Marshark and Spencer42 of the historical and theoretical perspectives. The effects of the rapidly changing technology and of our knowledge of child and language development makes it essential that the clinician keeps up to date with current thinking and practice in these areas, and works as part of a multidisciplinary team to ensure that information and discussions are up to date.



Hearing aids Amplification in the form of hearing aids is the mainstay for the treatment of PCHI in most cases. Early audiological assessment and provision of hearing aids at the youngest possible age (as young as four weeks in some cases) makes for the best long-term outcome. Supplying, fitting and after-care of young children with hearing aids is a demanding and skilled professional task and an appropriately trained and experienced audiologist is essential. Behind the ear (BTE) air conduction hearing aids with a soft mould, replaced at regular intervals as the child grows, are most commonly used. The ear mould can be replaced as the child’s ear canal grows. The mould may need to be changed as often as monthly in the first year. Older children, where the ear canal volume is greater, may be able to use in the ear (ITE) hearing aids. The principles that govern choice of hearing aid and on-going support are covered in Chapter 239a, Hearing aids. Children who are unable to use a mould in the ear canal due to conditions such as atresia may be considered for behind the ear hearing aids (Chapter 239a, Hearing aids). As a preliminary measure, while waiting for the



skull to mature sufficiently to enable fitting of a boneanchored hearing aid (BAHAs), a Softband consisting of a microphone mounted on an elastic headband placed around the head may be used (Figure 69.4). The management of conductive or mixed hearing loss is considered in more detail in Chapter 239b, Bone-anchored hearing aids. Children with bilateral hearing loss should be fitted with bilateral aids unless there is a definite medical contraindication, e.g. infected ears. This gives the child binaural hearing with increased volume, better sound localization, reduction of the head shadow effect, improved speech recognition in noise, improvements in spacial balance and listening ease (www.psych-sci.manchester.ac.uk/mchas/). In the UK, the median age at hearing aid fitting is now 16 weeks. The median delay from detection of hearing loss to provision of hearing aids is five weeks.22 The situation in the developing world is far less satisfactory; current annual production of hearing aids is estimated to meet less than 10 percent of global need (whqlibdoc.who.int/publications/2004/9241592435_eng). Cochlear implantation has transformed the rehabilitation of severe and profoundly deaf children in healthcare systems where it is available, and is considered in detail in Chapter 70, Paediatric cochlear implantation.



The education of deaf children The education of deaf children has long been the subject of controversy with the questions of where and how deaf children should be taught being hotly debated throughout the world. Often the debate has not been fuelled with evidence but rather by rhetoric, and decisions about the education of deaf children have often been dependent on



Figure 69.4 Child wearing a BAHAs Softband. Photograph by courtesy of Cochlear Europe Ltd.



Chapter 69 Investigation and management of the deaf child



where the child lived rather than upon their needs or their parents’ wishes. Interestingly, the arguments, choices and debates have taken place throughout the world, and today the advent of modern technologies for deaf children are influencing these decisions. Traditionally, deaf children were educated in special schools, usually residential and often in isolated areas. Large schools for deaf children were established in many countries in the latter half of the nineteenth century and children were taught according to the prevailing philosophy in the school. During the second half of the twentieth century more deaf children were educated in integrated settings, i.e. with their hearing peers, supported by improved technology, such as the fitting of FM systems which enabled teachers to use a microphone to aid communication with the deaf child (Figure 69.5) (Chapter 239a, Hearing aids) coupled with the political will to include more children with disabilities in mainstream education. Now the terms ‘inclusive education’ and ‘least restrictive environment’ are used internationally and supported by the legal requirement in many developed countries for deaf children to have an individual education programme with their needs identified. The options for educational placement can be described as:  mainstream education with varying levels of specialist support: this may be a teacher of the deaf visiting on a regular basis and/or in-class support from a nonspecialist assistant. This may support the development of spoken or sign language, and the inclass support will vary according to the needs of the child; it may be in the form of a note-taker, or an interpreter;  a unit, resource base, or special class in a mainstream school, with specialist teaching always available; this can be a very flexible way of providing access to specialist teacher of the deaf support and the child’s time in mainstream or resource class can be varied according to changing needs over time;



Figure 69.5



A teacher of the deaf using an FM microphone.



] 855



 a school for the deaf, which may be a day school or a residential school. The school for the deaf may have a specialist philosophy, using a sign bilingual approach or an oral approach, for example. It allows deaf children to have a peer group and to access specialist teaching, but can provide limited access to the broader world and, with small numbers, it can be difficult to provide specialist teaching in specialist subjects and to overcome these difficulties. Whatever the educational placement of the child, it is important to ensure that the child’s educational and social needs are met: medical and audiological services have the goal of improving hearing – for the educator there may be differing goals: those of literacy and numeracy and of social and emotional well-being. However, it is also vital that the technology supporting the child’s access to the curriculum is used effectively; the acoustics in schools are often poor, the technology can be difficult to manage, particularly in a mainstream setting, and there may be considerable discrepancy between the child’s functioning in the clinic and in school. A clinician involved in the fitting of hearing aids or implants should ensure that the child’s teacher of the deaf understands the technology, can carry out simple trouble-shooting and has access to spares to ensure that the child is not without their amplification during the day. Accessories such as radio hearing aid systems (or FM) or sound field systems are increasingly common, but need to be managed in the classroom rather than in the clinic. Children and young people are unlikely to use their equipment if it is not functioning properly, and the growing complexity of the technology demands increasing competence from teachers and increasing liaison with the audiology clinic. The other major decision that needs to be made on behalf of deaf children is that of communication mode, and whether to use a signed or oral approach. The Milan conference of 1860 concluded that the deaf were to be taught by oral means with the ‘uncontestable superiority of speech over sign’ and thus began a hundred years of the dominance of oral education over sign. The oral view was challenged strongly by the reports of poor linguistic and educational outcomes, for example,43 and by the increasing voice of the deaf community, promoting its own culture and language.44 Initially, this took the form of total or simultaneous communication with the use of spoken language and signed support. However, sign language has a grammar of its own and cannot be used with spoken language, and an interest in sign bilingual programmes where sign language is used independently of spoken language grew. In the UK, sign bilingualism began in the 1980s, interestingly at the same time as cochlear implantation, with its focus on spoken language. To summarize, communication choices used



856 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY with deaf children can be categorized in three major groups: 1. oral/aural alone; 2. those approaches using speech and sign (total or simultaneous communication); 3. sign bilingualism.45 There is some evidence that the new technologies of implantation and of earlier identification are changing some of these educational decisions and affecting educational outcomes. With regard to educational placement, Archbold et al.46 found that significantly more young deaf children with implants were attending mainstream schools and fewer schools for the deaf than their peers with hearing aids. With regard to mode of communication and communication choice, spoken language is a realistic outcome from cochlear implantation for many,47 but it is interesting to note that the use of sign language appears to be valued by both parents and young people with implants and it does not appear to be the case that early signing will impede the development of spoken language. Rather it appears to be the early development of effective communication which is important in early intervention, whether it is oral or signed.48 Stacey et al.47 showed that children with implants performed at the level of those less deaf than themselves and Thoutenhoofd,49 in a detailed Scottish survey, found that children with implants did better than those without, particularly in mathematics. Vermeulen,50 amongst others, has found that those with implants did better in reading than those without, although not as well as their hearing peers. Recent evidence shows that children with implants are not doing as well as their hearing peers in mainstream schools,51 indicating that although the modern technology of implantation is improving access to spoken language, there remain educational challenges to face. Worldwide, the role of the teacher of the deaf is being challenged in these changing times. Teachers are likely to be involved with deaf children early in infancy and with their families, and require quite a different set of skills to effectively support parents and child at this vital time. The work of Yoshinaga-Itano52 has shown us how vital it is to intervene early in a deaf child’s life. For clinicians, it means we have great responsibilities to liaise with each other, to understand each other’s language and perspective and above all to listen to the parents of the children with whom we work. In a questionnaire survey of adults who had been assessed and treated for deafness as children, it was found that both hearing level and educational environment influence the current mode of communication and use of hearing aids. Familial factors also seem to have an influence. Oral communication is a poor predictor of employment, whereas professional qualifications enhance the chances of finding a job.53



KEY POINTS  One infant per thousand is born with permanent deafness or hearing impairment that significantly affects language and social development. Another one per 1000 has a deafness that has some affect; a further one per 1000 develops permanent deafness during childhood.  Approximately half of permanent childhood hearing impairment is caused by genetic factors.  In the developing world, PCHI is largely a result of preventable causes.  Delayed presentation is commonplace in areas where neonatal screening has not been introduced. Children still frequently present with delayed speech.  Neonatal screening programmes have brought about earlier diagnosis, but some children with progressive hearing loss will not be detected unless surveillance is carried out during early childhood.  Single-sided deafness is detected much earlier as a result of universal screening. The optimum management is unknown, but the condition is associated with significantly greater adverse effects than was previously thought.  Improved imaging and advances in genetic testing have meant that a precise aetiological diagnosis can be made in an increasing proportion of deaf children.  A diagnosis of deafness has profound psychological and social implications for both the parents and the child.  Initial management is usually with hearing aids and referral for cochlear implantation if indicated.  Educational support is an essential part of the management of a deaf child and his/her family.



Best clinical practice [ All children with PCHI should have a full history and clinical examination, with audiological testing as appropriate to the child’s age. Families should be offered diagnostic testing to include imaging and genetic testing, and skilled genetic counselling. [Grade C] [ Optimum initial management is with early fitting of binaural hearing aids. [Grade C]



Chapter 69 Investigation and management of the deaf child



[ If available and if indicated, referral for cochlear implantation should be early. [Grade C/D] [ Clinicians involved in the fitting of hearing aids or implants should ensure that the child’s teacher of the deaf understands the technology, can carry out simple trouble-shooting and has access to spares to ensure that the child is not without their amplification during the day. [Grade C/D] [ Although the optimum management strategy for single-sided deafness is unknown, the child’s teachers should be made aware of this condition. [Grade D]



Deficiencies in current knowledge and areas for future research



$ $ $ $ $



Vaccination programmes, particularly for measles and rubella, will have an important influence on reducing the incidence of PCHI in the developing world. Hearing aid technology continues to improve. These technological improvements need to be accompanied by wider worldwide distribution and availability of hearing aids. Improvements in techniques for making a precise diagnosis in genetic deafness enables focussed genetic counselling and may ultimately facilitate therapeutic interventions. Imaging techniques continue to improve and are helping to facilitate more precise anatomical diagnosis. There is a need for population studies on the effect of single-sided deafness and for a rational approach to treatment and surveillance for these children.



REFERENCES 1. Fortnum HM, Summerfield AQ, Marshall DH, Davis AC, Bamford JM, Davis A et al. Prevalence of permanent childhood hearing impairment in the United Kingdom and implications for universal neonatal hearing screening: questionnaire based ascertainment study. British Medical Journal. 2001; 323: 536. 2. Morton NE. Genetic epidemiology of hearing impairment. Annals of the New York Academy of Sciences. 1991; 630: 16–31. 3. Rehm HL. A genetic approach to the child with sensorineural hearing loss. Seminars in Perinatology. 2005; 29: 173–81. 4. Smith RJ. Clinical application of genetic testing for deafness. American Journal of Medical Genetics A. 2004; 130: 8–12. 5. Davis A, Wood S. The epidemiology of childhood hearing impairment: factor relevant to planning of services. British Journal of Audiology. 1992; 26: 77–90.







] 857



6. Kuriyama M, Tomiwa K, Konishi Y, Mikawa H. Improvement in auditory brainstem response of hyperbilirubinemic infants after exchange transfusions. Pediatric Neurology. 1986; 2: 127–32. 7. Tang HY, Hutcheson E, Neill S, Drummond-Borg M, Speer M, Alford RL. Genetic susceptibility to aminoglycoside ototoxicity: how many are at risk? Genetics in Medicine. 2002; 4: 336–45. 8. Dahle AJ, Fowler KB, Wright JD, Boppana SB, Britt WJ, Pass RF. Longitudinal investigation of hearing disorders in children with congenital cytomegalovirus. Journal of the American Academy of Audiology. 2000; 11: 283–90. 9. Barbi M, Binda S, Caroppo S, Ambrosetti U, Corbetta C, Sergi P. A wider role for congenital cytomegalovirus infection in sensorineural hearing loss. Pediatric Infectious Disease Journal. 2003; 22: 39–42. 10. Fowler KB, Boppana SB. Congenital cytomegalovirus (CMV) infection and hearing deficit. Journal of Clinical Virology. 2006; 35: 226–31. 11. Banatvala JE, Brown DWG. Rubella. Lancet. 2004; 363: 1127–37. 12. WHO Position Papers on Vaccines. Preventing congenital rubella syndrome. Weekly Epidemiological Record. 2000; 75: 290–5. 13. Admiraal RJ, Huygen PL. Changes in the aetiology of hearing impairment in deaf-blind pupils and deaf infant pupils at an institute for the deaf. International Journal of Pediatric Otorhinolaryngology. 2000; 55: 133–42. 14. Saloojee H, Velaphi S, Goga Y, Afadapa N, Steen R, Lincetto O. The prevention and management of congenital syphilis: an overview and recommendations. Bulletin of the World Health Organization. 2004; 82: 424–30. 15. Drake R, Dravitski J, Voss L. Hearing in children after meningococcal meningitis. Journal of Paediatrics and Child Health. 2000; 36: 240–3. 16. Wellman MB, Sommer DD, McKenna J. Sensorineural hearing loss in postmeningitic children. Otology and Neurotology. 2003; 24: 907–12. 17. Koomen I, Grobbee DE, Roord JJ, Donders R, JennekensSchinkel A, van Furth AM. Hearing loss at school age in survivors of bacterial meningitis: assessment, incidence, and prediction. Pediatrics. 2003; 112: 1049–53. 18. Dodge PR, Davis H, Feigin RD, Holmes SJ, Kaplan SL, Jubelirer DP et al. Prospective evaluation of hearing impairment as a sequela of acute bacterial meningitis. New England Journal of Medicine. 1984; 311: 869–74. 19. Olusanya B. Measles, mumps and hearing loss in developing countries. Journal of Community Ear and Hearing Health. 2006; 3: 7–9. 20. Yoshinaga-Itano C, Sedey AL, Coulter DK, Mehl AL. Language of early- and later-identified children with hearing loss. Pediatrics. 1998; 102: 1161–71. 21. Fortnum H, Davis A. Epidemiology of permanent childhood hearing impairment in Trent Region, 1985–1993. British Journal of Audiology. 1997; 31: 409–46. 22. Uus K, Bamford J. Effectiveness of population-based newborn hearing screening in England: Ages of



858 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



23.



 24.



25.



26.



27.



28.



29.



30.



31.



32.



33.



 34.



35.



36.



interventions and profile of cases. Pediatrics. 2006; 117: e887–93. Rance G, Beer DE, Cone-Wesson B, Shepherd RK, Dowell RC, King AM et al. Clinical findings for a group of infants and young children with auditory neuropathy. Ear and Hearing. 1999; 20: 238–52. Lieu JEC. Speech-language and educational consequences of unilateral hearing loss in children. Archives of Otolaryngology, Head and Neck Surgery. 2004; 130: 524–30. Olusanya BO, Luxon LM, Wirz SL. Childhood deafness poses problems in developing countries. British Medical Journal. 2005; 330: 480c. Smith SD, Taggart RT. Genetic hearing loss with no associated abnormalities. In: Toriello HV, Reardon W, Gorlin RJ (eds). Hereditary hearing loss and its syndromes. New York: Oxford University Press, 2004: 37–82. Kurtzer-White E, Luterman D. Families and children with hearing loss: grief and coping. Mental Retardation and Developmental Disabilities Research Reviews. 2003; 9: 232–5. Watkin PM, Beckman A, Baldwin M. The views of parents of hearing impaired children on the need for neonatal hearing screening. British Journal of Audiology. 1995; 29: 259–62. Feher-Prout T. Stress and coping in families with deaf children. Journal of Deaf Studies and Deaf Education. 1996; 1: 155–65. Burger T, Spahn C, Richter B, Eissele S, Lohle E, Bengel J. Parental distress: the initial phase of hearing aid and cochlear implant fitting. American Annals of the Deaf. 2005; 150: 5–10. Young A, Tattersall H. Parents’ of deaf children evaluative accounts of the process and practice of universal newborn hearing screening. Journal of Deaf Studies and Deaf Education. 2005; 10: 134–45. Yoong S, Spencer N. Audit of local performance compared with standards recommended by the national guidelines for aetiologic investigation of permanent childhood hearing impairment. Child: Care, Health and Development. 2005; 31: 649–57. Mafong DD, Shin EJ, Lalwani AK. Use of laboratory evaluation and radiologic imaging in the diagnostic evaluation of children with sensorineural hearing loss. Laryngoscope. 2002; 112: 1–7. Preciado DA, Lim LHY, Cohen AP, Madden C, Myer D, Ngo C et al. A diagnostic paradigm for childhood idiopathic sensorineural hearing loss. Otolaryngology – Head and Neck Surgery. 2004; 131: 804–9. Walsh RM, Ayshford CA, Chavda SV, Proops DW. Large vestibular aqueduct syndrome. ORL. Journal for Otorhinolaryngology and its Related Specialties. 1999; 61: 41–4. Fowler KB, Dahle AJ, Boppana SB, Pass RF. Newborn hearing screening: will children with hearing loss caused by congenital cytomegalovirus infection be missed? Journal of Pediatrics. 1999; 135: 60–4.











37. Barbi M, Binda S, Primache V, Caroppo S, Dido P, Guidotti P et al. Cytomegalovirus DNA detection in Guthrie cards: a powerful tool for diagnosing congenital infection. Journal of Clinical Virology. 2000; 17: 159–65. 38. Reardon W. Syndrome diagnosis and investigation. In: Toriello HV, Reardon W, Gorlin RJ (eds). Hereditary hearing loss and its syndromes. New York: Oxford University Press, 2004: 3–7. 39. Preciado DA, Lawson L, Madden C, Myer D, Ngo C, Bradshaw JK et al. Improved diagnostic effectiveness with a sequential diagnostic paradigm in idiopathic pediatric sensorineural hearing loss. Otology and Neurotology. 2005; 26: 610–5. 40. Fischel-Ghodsian N, Prezant TR, Chaltraw WE, Wendt KA, Nelson RA, Arnos KS et al. Mitochondrial gene mutation is a significant predisposing factor in aminoglycoside ototoxicity. American Journal of Otolaryngology. 1997; 18: 173–8. 41. Arnos KS, Israel J, Cunningham M. Genetic counseling of the deaf. Medical and cultural considerations. Annals of the New York Academy of Sciences. 1991; 630: 212–22. 42. Marschark M, Spencer PE. Spoken language development of deaf and hard-of-hearing children: Historical and theoretical perspectives. In: Spencer PE, Marschark M (eds). Advances in the spoken language development of deaf and hard-of-hearing children. New York: Oxford University Press, 2006: 3–21. 43. Conrad R. The deaf school child. London: Harper & Row, 1979. 44. Nevins N, Chute P. Children with cochlear implants in educational settings. San Diego: Singular Publishing Group, 1996. 45. Lynas W. Communication options. In: Stokes J (ed.). Hearing impaired infants: Support in the first eighteen months. London: Whurr Publishers, 1999: 98–128. 46. Archbold SM, Nikolopoulos TP, Lutman ME, O’Donoghue GM. The educational settings of profoundly deaf children with cochlear implants compared with agematched peers with hearing aids: implications for management. International Journal of Audiology. 2002; 41: 157–61. 47. Stacey PC, Fortnum HM, Barton GR, Summerfield AQ. Hearing-impaired children in the United Kingdom, I: Auditory performance, communication skills, educational achievements, quality of life, and cochlear implantation. Ear and Hearing. 2006; 27: 161–86. 48. Tait M, Lutman ME, Robinson K. Preimplant measures of preverbal communicative behavior as predictors of cochlear implant outcomes in children. Ear and Hearing. 2000; 21: 18–24. 49. Thoutenhoofd E. Cochlear implanted pupils in scottish schools: 4-year school attainment data (2000–2004). Journal of Deaf Studies and Deaf Education. 2006; 11: 171–88. 50. Vermeulen AM, van Bon W, Schreuder R. Reading comprehension of deaf children with cochlear implants.



Chapter 69 Investigation and management of the deaf child Journal of Deaf Studies and Deaf Education. 2007; 12: 283–302. 51. Damen GW, van den Oever-Goltstein MH, Langereis MC, Chute PM, Mylanus EA. Classroom performance of children with cochlear implants in mainstream education. Annals of Otology, Rhinology and Laryngology. 2006; 115: 542–52.



] 859



52. Yoshinaga-Itano C. Benefits of early intervention for children with hearing loss. Otolaryngologic Clinics of North America. 1999; 32: 1089–102. 53. Dauman R, Daubech Q, Gavilan I, Colmet L, Delaroche M, Michas N et al. Long-term outcome of childhood hearing deficiency. Acta Otolaryngologica. 2000; 120: 205–8.



70 Paediatric cochlear implantation JOSEPH G TONER



Introduction Timing of implantation Assessment Preoperative counselling Surgery Special surgical circumstances Postoperative habilitation



860 861 863 864 864 865 865



Controversies in management Quality of life measures Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



866 866 867 867 867 867



SEARCH STRATEGY The data in this chapter are supported by a PubMed search using the key words cochlear implantation and cochlear implants, limited to randomized controlled trials, human, children 0–18 years. The Cochrane and TRIP databases were also consulted.



INTRODUCTION Hair cell loss is the principal cause of senorineural deafness. Sufficient neural elements usually survive and are available for electrical stimulation enabling meaningful activation of the auditory cortex. A cochlear implant is a prosthetic device which replaces the transducer function of damaged hair cells and provide this electrical stimulation.



In 1969, William House recommended clinical use of electrical stimulation in profoundly deaf patients.3 Several were implanted and outcomes evaluated in detail. In the 1980s, commercially available devices were used in routine clinical practice. The Food and Drug Administration (FDA) approved their use in adults in 1984. Increased confidence in results obtained in adults led to more widespread paediatric implantation in the late 1980s. In 1990, the FDA approved the use of cochlear implants in children.



History Current status Electrical stimulation of the ear was attempted as early as the eighteenth century.1 The modern era of cochlear implantation began with Djourno and Eyries2 who inserted a device that produced awareness of sound, but no speech discrimination. Animal models were developed in the 1960s.3, 4 Prototype implants were introduced in the early 1970s and rapid development took place over the next decade.



Almost all children with severe to profound sensorineural hearing loss may now be considered for implantation subject to the provisos outlined below. The potential candidate population of children in the UK is thus some 600 per annum. Most have congenital loss (1 in 1000 births), but a small number have acquired loss mainly due to meningitis.



Chapter 70 Paediatric cochlear implantation



The aetiology of severe congenital hearing loss is considered in Chapter 66, Molecular otology, development of the auditory system and recent advances in genetic manipulation. Genetic disorders account for more than 50 percent of all cases.



] 861



that any one implant device has performance benefits over the others.



TIMING OF IMPLANTATION Prelingual deafness



Devices Depending on the time of onset, deafness can be prelingual or post-lingual, i.e. occurring before or after the development of normal speech and language. The critical period for speech and language development lasts for a significant portion of the first decade of life, but is a continuum rather than a discrete event. Hence there is a transitional category that is referred to as perilingual deafness, i.e. deafness arising during acquisition of key speech and language skills. In outcome, these children are equivalent to those in the prelingual category. Neuroplasticity (the ability of the brain to respond adaptively to behaviorally relevant stimuli) is a feature of both motor and sensory functions. If a congenitally blind child has vision restored after the first few years of life then, although the child is able to ‘see’ objects and different colours, they are unable to recognize them. This is referred to as cortical blindness. The same phenomenon is observed in relation to hearing and speech and language development. Where auditory input is restored to patients only after the first decade of life, they are unable to acquire normal speech and language. The existence of a critical period for language development during the first five years of life is well established.5 Providing auditory stimulation during this period is critical.6 Deafness significantly reduces language development. A normally hearing child



Implantation involves insertion of a receiver/stimulator into a bony recess in the squamous temporal bone. The cochlea is opened (cochleostomy) and an electrode array connected to the receiver is inserted. These constitute the internal device. The internal receiver/stimulator package has an embedded magnet to secure the transmitting coil. A microphone, microprocessor and transmitter coil (external device) is positioned by magnetic coupling over the internal device and delivers the encoded signal (Figure 70.1). The microphone is connected to a speech processor, which digitally processes and then encodes the auditory signal. The signal is then transmitted transcutaneously to the implanted receiver/stimulator, which stimulates the surviving neural elements in the modiolus. This enables selective stimulation of the relevant segments of the surviving ganglion cell population with the tonotopically determined frequency information. All the current cochlear implant devices used in the UK enable good auditory habilitation in most patients. There is still a significant variability in individual results. Although there are many technical differences between the implant devices in relation to speech processing and device specifications, there is no conclusive evidence



3. Transmitter



4. Receiver



1. Microphone



5. Stimulator



2. Speech processor (a)



Figure 70.1



(b)



Implantation of a device. (a) Side view; (b) front view.



862 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY may have a vocabulary of between five and 26 thousand words by the age of five years. A profoundly deaf child of the same age typically has only 200 spoken or signed words.7 Developmental studies support the potential benefits of early implantation. Ryals and colleagues8 have demonstrated that the auditory system requires early stimulation to develop normally and that auditory deprivation may prevent development of normal frequency and place coding. Studies of animal models have shown that electrical stimulation commenced shortly after inducement of deafness enhances neural survival and may prevent changes in the central auditory system. Animal experiments indicate that changes take place both in the brain stem and auditory cortex that produce a neurophysiological barrier to successful implantation after the ‘critical period’.9, 10 There is increasing evidence from functional brain imaging studies with positron emission tomography (PET) to confirm this auditory critical period. In effect, unused brain capacity is reassigned to other tasks if not hard-wired for its primary designated purpose within the first few years of life. This confirms the judgement of experienced clinicians that early implantation, probably before the age of two and possibly significantly earlier, is required to maximize the potential for normal language development in prelingually deaf children. Conversely, if a congenitally deaf child presents late, e.g. after the age of seven, it is likely that the benefit of implantation in relation to speech and language acquisition and recognition will be very limited. Many centres would therefore counsel parents against implantation.



Early implantation The National Institutes of Health (NIH) consensus statement in 199511 advocated implantation in children after the age of two years. However, clinical practice has evolved and implantation is now being performed regularly in children below the age of two years. Implantation below the age of one year usually requires special circumstances, such as rapid obliteration of the cochlea following meningitis. Several studies have begun to confirm the benefit in terms of speech and language outcomes for children implanted at an earlier age.6, 12, 13, 14, 15, 16 Kileny et al.6 studied a group of 48 implant patients at the age of seven years and a further study involving 53 patients who were assessed three years following implantation. The study confirmed that in terms of speech recognition with monosyllables and multisyllabic words and sentences, children perform better with increased duration of use of cochlear implantation and reduced age at implantation. Children implanted at a younger age performed better than children implanted at an older age. Harrison et al.13 reported a retrospective analysis of 70 profoundly deaf children, 78 percent of whom were



congenital. The age range of this group was 2–15 years and they achieved a five-year follow up. Several standard speech perception tasks were administered (phonetically balanced kindergarten (PBK), Glendonald auditory screening procedure (GASP), test of auditory comprehension (TAC), word intelligibility by picture identification (WIPI)). A binary partitioning algorithm was used to divide the dataset on the basis of age and implantation. This allowed the data to be divided into two age categories, those above and those below the age of 4.4 years. Analysis of speech perception performance indicated that there was a statistically significant improvement in scores in those implanted at a younger age compared with those implanted later.13 Govaerts et al.15 analyzed a retrospective longitudinal cross-sectional study of 70 paediatric cochlear implant patients and demonstrated that for children implanted after the age of four years categories of auditory performance (CAP) scores rarely reached normal. However, for patients implanted between two and four years, normal CAP scores were usually reached and two-thirds were integrated into mainstream primary schools. Implantation below the age of two years always resulted in a normal CAP score with 90 percent being integrated into mainstream preschooling. A possible bias in this study related to the fact that children implanted at a young age have normally been identified as having a significant hearing loss at a younger age and therefore received hearing aids at an earlier age. Although the numbers are not sufficiently large to avoid this source of bias, some of the children implanted at an older age had received hearing aids at an early age, but still had outcomes related to their age at implantation.15 All of these studies are cohort studies with no controls. It is extremely unlikely that a study which randomized subjects to either a prolonged hearing aid use group or cochlear implant group would be accepted by parents. So, although Yoshinaga-Itano et al.17 have reported early intervention of whatever kind as being beneficial to the ultimate outcome in terms of language development, a view supported by the study by Moeller,18 the precise reason for this has never been proved. The hypothesis that it may be the habilitative effort that is important in determining the positive outcomes in cochlear implant children is unlikely to be tested.



Post-lingual candidates Post-lingual deafness in children is most frequently a consequence of meningitis or head injury. Although the duration of deafness is a factor in speech perception outcome after cochlear implantation, it is not an overriding issue. In adults with cochlear implants, the trend in most reported series indicates that the longer the duration of deafness, the poorer will be the outcome in relation to speech perception. However, most children who become



Chapter 70 Paediatric cochlear implantation



profoundly deaf after a significant period of normal language development (five to ten years) and who receive a cochlear implant will regain excellent open-set auditoryonly speech discrimination skills.



ASSESSMENT Assessment is a multiprofessional evaluation of hearing, communication ability, cochlear morphology, middle ear/ mastoid status and other relevant general developmental issues.



Audiological tests Many infants will have already undergone electrophysiological tests as part of their diagnostic assessment, usually either auditory brainstem responses (ABR) or electrocochleography (ECochG). While these tests provide some indication of the severity of the hearing loss, behavioural testing when available is preferred as it is a test of the entire auditory pathway (not just the peripheral system and brain stem connections). Behavioural aided and unaided responses are obtained using visual reinforcement audiometry (VRA). In very young children, the level of language competency means that speech discrimination tests are not available. Therefore, audiological candidacy is informed mostly by the aided responses. The criterion adopted now by most UK centres is that if aided thresholds at frequencies above 2 kHz are greater than 50 dBA, then benefit from a cochlear implant is likely. In older children (post-lingual deafness), standard pure tone and aided free field tests are performed, but speech discrimination tests are more critical. Assessment is similar to adults though the test material will need to be ‘language-age’ appropriate. An aided audiogram is a reasonable surrogate predictor of speech discrimination ability. Speech discrimination scores are performed both in the auditory-only and auditory with lip-reading modes. An auditory-only open-set sentence score of below 30 percent is generally regarded as being an indicator of likely benefit from cochlear implantation; however, some centres may implant candidates with 50 percent discrimination scores. Vestibular testing is not routinely performed on children.



Imaging High-resolution computed tomography (CT) scanning shows the bony morphology of the cochlea and labyrinth. Congenital abnormalities include the Mondini dysplasia, common cavity deformity or a large vestibular aqueduct. The internal auditory meatus may be narrow and raise the possibility of auditory nerve aplasia. Meningitis can cause



] 863



significant obliteration of the cochlea. Ossification is usually found in the basal turn of the cochlea near the round window, but sometimes ossification is complete. The accuracy of high-resolution CT scanning to predict patency of the cochlea is extremely variable; highresolution magnetic resonance imaging (MRI) is regarded as being superior to CT in this respect, but it can still underestimate the degree of cochlea ossification.12, 19, 20, 21 Imaging is particularly important in post-meningitic patients in confirming the presence of a fluid-filled basal turn. High resolution three-dimensional (3D) reconstructions may provide information on patency of the individual scalae. In cases where there is no auditory response, high-resolution MRI scans may demonstrate the presence of an auditory nerve. This is particularly important if a CT has demonstrated an abnormally narrow internal auditory canal (IAC). Infants may require sedation or anaesthesia to undergo imaging and therefore it is advantageous to be able to perform MRI and CT together. PET or functional magnetic resonance imaging (fMRI) will increasingly provide functional information in relation to persistent plasticity and in the future may provide valuable information in relation to candidacy.



Medical assessment The trend towards decreasing implantation at a younger age means that anaesthetic and perioperative risk must be evaluated, particularly for children with complex medical conditions. Young22 found that in implanting infants less than 12 months of age there are physiological issues that require the anaesthetic risks to be considered seriously. The benefits of early surgery must be weighed against these risks. He acknowledges that there may be factors such as rapidly progressing obliteration in post-meningitis cases which may alter the risk/benefit balance, but the anaesthetic risks need to be carefully discussed as part of the informed consent process.



Neurodevelopmental issues This is a complex area for implant teams. Candidacy has recently been extended to children with learning disabilities and other developmental problems. Outcomes are more difficult to demonstrate, but benefit in communication is achieved by some. Careful consideration is needed to assess likely benefit to hearing and to quality of life. Problems with lack of reliable reporting of auditory sensations may make devising optimum stimulation parameters difficult, and there is the risk of causing distress due to overstimulation. Some of these issues may be overcome with electrically evoked ABR response testing, now possible on the newest generation of implant devices.



864 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



PREOPERATIVE COUNSELLING The potential implant candidate must not be looked at in isolation. Consideration should be given to the assessment of parental or care-givers attitudes, particularly the ability to assist in the delivery of an intense postoperative programme to enable the child to develop receptive and expressive oral language skills. This will include commitment to an appropriate educational environment to maximize potential. Social workers and/or psychologists may help to ensure the importance of these issues is recognized. Following the assessment process and determination of likely benefit, the parents are counselled regarding the surgery and the habilitative management process. This will include explanation of the benefits and risks of surgery, variability of outcome and the requirement for intensive and prolonged input from the implant centre team and local professionals.



SURGERY A low complication rate is now reported from most established centres, with a ‘learning curve’ in new centres.



Technique The surgery involves several steps.  Incision and flap design. Implant surgery was initially performed via a large incision. Gibson et al.23 were the first to advocate a much smaller post-aural incision and by the late 1990s most surgeons had begun to adopt some version of what resembles a standard post-aural mastoidectomy incision. This reduces flap-related problems and is more acceptable to children and their parents since a large head shave is not required.24 A recent survey of UK implant centres reported a flap-related complication rate of 2.4 percent in children, but that included earlier cases preformed with larger incisions.25  Bony approach to cochlea. A standard mastoidectomy is followed by a posterior tympanotomy preserving the chorda tympani. The round window region is visualized through a posterior tympanotomy. The scala tympani is entered via a cochleostomy anterior to the round window niche. Following electrode insertion, any space between the electrode and the cochleostomy can be sealed with soft tissue, usually fascia or muscle. There is an increasing trend in the UK towards use of a facial nerve monitor.25  Bony recess for and fixation of receiver/stimulator. A device-specific contoured bony recess is fashioned for the receiver/stimulator package. In children this



almost always involves exposure of the dura. Most surgeons prefer to preserve a bony island over the dura if possible. Fixation of the device can be with either absorbable or nonabsorbable sutures. Bone cement is also used for electrode fixation at the mastoid rim or posterior tympanotomy.  Closure and postoperative care. Drains are rarely required. Most patients receive some form of antibiotic cover, though there is no generally accepted regime. Many patients have some form of imaging of the electrode array (most frequently via a modified Stenvers view) to confirm satisfactory position prior to discharge.



Complications THE FACIAL NERVE



The incidence of facial nerve damage is very low. Hoffman and Cohen26 reported 0.58 percent in children (n = 1905); in a UK series, 0.2 percent had permanent facial nerve injury.25 Non-auditory stimulation, usually of the facial nerve, is relatively common following cochlear implantation, particularly in post-meningitic patients. It is thought that bony channels develop, allowing electrical stimulation of the horizontal segment of the facial nerve. Often these problems can be overcome by programming out the relevant electrodes. Care must be taken in these cases as this aberrant effect of stimulation may upset the child and lead to an aversion to device use. Bigelow et al.27 reported an incidence of 14 percent of some form of non-auditory or facial nerve stimulation. In a very small number of cases, if a significant number of electrodes have to be switched off, implant performance may be degraded.



MENINGITIS



There are concerns about a possible increased risk of meningitis – particularly pneumococcal – in children who have had cochlear implantation. Epidemiological studies are complicated by the fact that the cochlear implant population has a higher incidence of meningitis before cochlear implantation than the general population. A survey in the United States showed that of roughly 5000 children under the age of five who had been implanted, 300–400 had had an episode of meningitis prior to the implantation; 17 had had meningitis after implantation. This increased risk of meningitis in deaf individuals may be due to structural abnormalities such as common cavity deformity or enlarged vestibular aqueduct.28 The advice from the Department of Health in the UK is that all patients who have been implanted should be offered pneumococcal vaccination. Prospective patients should be advised regarding immunization prior to



Chapter 70 Paediatric cochlear implantation



implantation. Patients who have implantation and develop otitis media or signs of other febrile illnesses should be urgently assessed by either their general practitioner or the implant centre.



DEVICE EXTRUSIONS



This is one of the most serious complications following implant surgery. Hoffman and Cohen26 reported an incidence in children of 0.16 percent (n = 1905). Gibbin et al.25 reported the experience of UK centres with a similar incidence of 0.25 percent.



DEVICE FAILURES



Earlier implants had various design faults, which provided some device reliability problems. Cumulative survival rates for all the devices in use in the UK are in the range of 97–99 percent.



SPECIAL SURGICAL CIRCUMSTANCES Cochlear anatomical abnormalities Implantation of the malformed cochlea is technically feasible.29 Patients with an enlarged vestibular aqueduct or enlarged vestibule may have ‘gushers’ intraoperatively. Several authors have concluded that outcomes in terms of speech discrimination in these patients are similar to those with no structural abnormalities.



Obliterated cochlea Ossification develops as a consequence of meningitis, temporal bone trauma and some autoimmune disorders. The basal turn of the scala tympani is most frequently affected due to the connection between the subarachnoid space with the scala tympani via the cochlear aqueduct. Scala vestibuli ossification is less common and may take place later in the ossification process.30, 31 Ossification was initially thought to be a contraindication to cochlear implantation, but various surgical procedures have been developed to allow at least partial implantation in those cochleas with significant degrees of ossification. Gantz et al.32 described a drill-out procedure by creating a trough around the modiolus, drilling out the entire basal turn. Balkany et al.33, 34 modified this technique using an intact canal wall procedure. Steenerson et al.31 described scali vestibuli electrode insertion in which the scali vestibuli is opened anterior to the round window by drilling above the spiral ligament. The outcome in patients whose implant was inserted into the scala vestibuli was compared to matched controls who



] 865



had standard scala tympani insertion. The outcome measures including sentence recognition scores and vowel and consonant speech scores were similar in both groups. He concluded therefore that scala vestibuli insertion was an acceptable alternative to scala tympani insertion in partially obliterated cases. Cohen and Waltzman35 described partial insertion through a short inferior tunnel drilled along the basal turn. In some cases, this partial drill out will reach a patent segment of the basal turn and a full insertion is then possible. More recently, the implant manufacturers have produced different electrode configurations to deal with the problem of the ossified cochlea.36, 37 A dual/split array is available; one electrode is inserted into an inferior tunnel drilled along the basal turn and the other electrode is inserted into a separate cochleostomy drilled into the superior limb of the basal turn. A compressed electrode array with an increased number of electrodes along a shorter distance is an alternative in a basal turn drill-out procedure.



Chronic suppurative otitis media and open mastoid cavities Patients with mastoid cavities were initially considered unsuitable for implantation. However, increasing experience has led most centres to offer surgery to these patients. Various techniques are described.38 Surgery is performed as a one- or two-stage procedure. In the twostage procedure, all squamous epithelium is removed from the cavity, which is then obliterated behind a blind sac closure of the external canal. After an interval, the cochlear implant device is inserted at a second-stage operation. In the one-stage technique, the device is inserted following removal of disease and the cavity obliterated and closed.



POSTOPERATIVE HABILITATION This is a multidisciplinary effort involving audiologists, speech and language therapist and specialist educators; it begins with programming the device with stimulation parameters for the electrode array which will provide auditory perception across the speech range. In the very young, this can be assisted by electrophysiological measures which can predict the threshold and comfort levels for electrical stimulation. In older children, behavioural responses can be used to refine the implant stimulation parameters. These parameters can change over time and therefore need to be regularly confirmed at ‘programming’ sessions. The role of the speech and language therapist and specialist educators is to maximize the potential provided by the auditory input to develop speech perception and intelligibility, and thereby to permit access to the educational curriculum as fully as possible.



866 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



CONTROVERSIES IN MANAGEMENT Implantation in older children and adolescents The ideal age for implantation in congenitally deaf children has been discussed. The current recommendation is around or just before the age of two years. Several studies have examined the benefits of implantation in older children and adults. Waltzman et al.39 examined the outcomes in 35 congenitally deaf children who received implants after the age of eight years and 14 congenitally deaf adults. The results indicated that there was improvement in open-set speech perception in children. Adults demonstrated improvement in mean scores for word and sentence recognition, although the improvement was not as significant as in children implanted at a younger age. Results were adversely influenced by increasing duration of deafness and older age at time of implantation.



Bilateral cochlear implantation In the last few years, experience has grown with bilateral implantation and studies in adults report the benefits binaural hearing should provide, including sound localization and enhanced speech recognition in background noise. Muller et al.40 reported a study of nine bilaterally implanted adult patients tested for speech understanding in quiet and in noise. They were tested in three situations, left implant only, right implant only and both implants activated. The speech discrimination tests included monosyllables in quiet and sentences in noise. Results indicated higher speech scores for all subjects with bilateral stimulation. This has encouraged bilateral implantation in children and preliminary results indicate that outcomes measured by auditory perception and speech intelligibility are improved.41 Bilateral implantation will require further longitudinal evaluation of significant numbers of children in order to confirm these early reports indicating that additional benefits accrue from the second implant.



MRI following cochlear implantation The presence of a magnet in the current implant systems has led to MRI being contraindicated in implanted patients. The devices have a removable magnet; a small incision over the posterior half of the receiver/stimulator package will allow the implanted magnet to be removed to enable MRI to be performed. This requires a small surgical procedure with the risk of introducing infection.



A recent study42 shows that it is possible to perform an MRI scan provided the scanner is equipped with a 1-Tesla magnet.



QUALITY OF LIFE MEASURES Cheng et al.43 have reported that cochlear implants in profoundly deaf children have a positive effect on quality of life at reasonable direct costs and appear to result in net savings to society. The cost–utility of paediatric cochlear implantation was analyzed by O’Neill et al.44 They concluded that it was an effective intervention with a quality adjusted life year (QALY) gain of £10.341 ($21.00), and recommended the intervention as effective for profoundly hearing-impaired children. QALYs are a measure of the benefit of a medical intervention. The measure takes into account both quantity and quality of life generated by a healthcare intervention and is based on the number of years of life that would be added by the intervention. Each year in perfect health is assigned the value of 1.0 down to a value of 0 for death. If the extra years would not be lived in full health, for example if the patient were to lose a limb, become blind, or be confined to a wheelchair, then the extra life-years are given a value between 0 and 1 to account for this.45 Francis et al.46 reported that cochlear implantation accompanied by aural rehabilitation increases access to acoustic information of spoken language, leading to higher rates of mainstream placement in schools and lower dependence on special education support services. The cost savings that result from a decrease in the use of support services indicates an educational cost benefit of cochlear implant rehabilitation for many children. However, larger more methodologically robust studies will be required to strengthen evidence and deal with confounding variables such as socioeconomic status, cognitive function, age of diagnosis of deafness and language intervention, and the mode of communication. Although now a decade old, the NIH consensus statement11 is still the outcome of the most extensive systematic review, it concluded that: Cochlear implantation improves communication ability in most adults with severe to profound deafness and frequently leads to positive psychological and social benefits as well. Currently, children at least two years old and adults with profound deafness are candidates for implantation. Cochlear implant candidacy should be extended to adults with severe hearing impairment and open-set sentence discrimination that is less than or equal to 30 percent in the best aided condition. Access to optimal education and (re)habilitation services is important for adults and is critical for children to maximize the benefits available from cochlear implantation.



Chapter 70 Paediatric cochlear implantation



] 867



REFERENCES KEY POINTS  Cochlear implantation is a proven effective treatment for the management of severe to profound sensorineural hearing loss in children.  All infants and children who are diagnosed as having a severe to profound congenital hearing loss should be referred urgently for assessment.  Increasingly, implantation is undertaken between the ages of one and two years.  The results in older children are less good, but assessment by an implant centre is still worthwhile.  Patients in the UK are managed within a multidisciplinary team in a small number of specialized centres.  Complication rates are low.  Studies have confirmed the cost effectiveness of paediatric implantation.



Best clinical practice [ Any child with a hearing loss >70 dB (unaided), [ [ [ [



acquired or congenital, should be referred for assessment. Children with aided levels >50 dB at frequencies of 2 Hz and above should be referred to a specialist centre. Children with significant deafness post-meningitis should be referred urgently because of the risk of progressive obliteration of the cochlea. Age at implantation in the pre- or perilingual deaf has an inverse relationship to outcomes of enhanced oral communication. Implantation should take place in a designated centre, with experienced staff working in a multiprofessional team.



Deficiencies in current knowledge and areas for future research



$ $



The implementation of universal neonatal hearing loss screening with resultant early intervention should lead to improved outcomes for all children with hearing loss, particularily those requiring cochlear implantation. Lack of RCT based evidence is a deficiency but is unlikely to be resolved due to the perceived effectiveness of implantation and therefore difficulty in recruiting to a hearing aid arm of such a trial.



1. Loeb GB. Cochlear prosthetics. Annual Review of Neurosciences. 1990; 13: 357–71. 2. Djourno A, Eyries C. Prothese auditive par excitation electrique a distance du nerf sensoriel a l’aide d’un bobinage inclus a demeure. Presse Me´dicale. 1957; 36: 14–17. 3. Simmons FB. History of cochlear implants in the United States: A personal perspective In: Schindler RA and Merzenich MM (eds). Cochlear Implants. New York: Raven Press, 1985: 1–7. 4. Clarke GM. Responses of cells in the superior olivary complex of the cat to electrical stimulation of the auditory nerve. Experimental Neurology. 1969; 24: 124–36. 5. Cairns H. The acquisition of language. Austin, TX: Pro-ed, 1986. 6. Kileny PR, Zwolan TA, Ashbaugh C. The influence of age at implantation on performance with a cochlear implant in children. Otology and Neurotology. 2001; 22: 42–6. 7. Schabb WA. Effects of hearing loss on education. In Jaffe BF (ed.). Hearing loss in children: A comprehensive text. Baltimore: University Park Press, 1977: 650–4. 8. Ryals BM, Rubel EW, Lippe W. Issues in neural plasticity as related to cochlear implants in children. American Journal of Otology. 1991; 12: 22–7. 9. Harrison RV, Nagasawa A, Smith DW, Stanton S, Mount RJ. Reorganisation of the auditory cortex after neo-natal high frequency cochlear hearing loss. Hearing Research. 1991; 54: 11–9. 10. Larsen SA, Kirchhoff TM. Anatomical evidence of synaptic plasticity in the cochlear nucleii of deaf white cats. Experimental Neurology. 1992; 115: 151–7. 11. NIH Consensus Statement. Cochlear implants in adults and children. 1995; 13: 1–30. 12. Nikolopoulos TP, O’Donoghue GM, Robinson KL, Holland IM, Ludman C, Gibbin KP. Preoperative radiologic evaluation in cochlear implantation. American Journal of Otology. 1997; 18: S73–4. 13. Harrison RV, Panesar J, El-Hakim H, Abdolell M, Mount RJ, Papsin B. The effects of age of cochlear implantation on speech perception outcomes in prelingually deaf children. Scandinavian Audiology Supplement. 2001; 53: 73–8. 14. Illg A, von der Haar-Heise S, Goldring JE, Lesinski-Schiedat A, Battmer RD, Lenarz T. Speech perception results for children implanted with the CLARION cochlear implant at the Medical University of Hannover. Annals of Otology, Rhinology and Laryngology. 1999; 177: 93–8. 15. Govaerts PJ, De Beukelaer C, Daemers K, De Ceulaer G, Yperman M, Somers T et al. Outcome of cochlear implantation at different ages from 0 to 6 years. Otology and Neurotology. 2002; 23: 885–90. 16. Miyamoto RT, Kirk KI, Svirsky M, Sehgal ST. Communication skills in pediatric cochlear implant recipients. Acta Otolaryngologica. 1999; 119: 219–24.



868 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY 17. Yoshinaga-Itano C, Sedey AL, Coulter DK, Mehl AL. Language of early- and later-identified children with hearing loss. Pediatrics. 1998; 102: 1161–71. 18. Moeller MP. Early intervention and language development in children who are deaf and hard of hearing. Pediatrics. 2000; 106: E43. 19. Jackler RK, Luxford WM, Schindler RA, McKerrow WS. Cochlear patency problems in cochlear implantation. Laryngoscope. 1987; 97: 801–5. 20. Nair SB, Abou-Elhamd KA, Hawthorne M. A retrospective analysis of high resolution computed tomography in the assessment of cochlear implant patients. Clinical Otolaryngology. 2000; 25: 55–61. 21. Seidman DA, Chute PM, Parisier S. Temporal bone imaging for cochlear implantation. Laryngoscope. 1994; 104: 562–5. 22. Young NM. Infant cochlear implantation and anaesthetic risk. Annals of Otology, Rhinology and Laryngology. 2002; 111: 49–51. 23. Gibson WPR, Harrison HC, Prowse C. A new incision for placement of the cochlear multi-channel cochlear implant. Journal of Laryngology and Otology. 1995; 109: 821–5. 24. Gibson WPR, Harrison HC. Further experience with a straight, vertical incision for placement of cochlear implants. Journal of Laryngology and Otology. 1997; 111: 924–7. 25. Gibbin KP, Raine CH, Summerfield AQ. Cochlear implantation – UK and Ireland surgical survey. Cochlear Implants International. 2003; 4: 10–21. 26. Hoffman RA, Cohen NL. Complications of cochlear implant surgery. Annals of Otology, Rhinology and Laryngology. 1995; 166: 420–2. 27. Bigelow DC, Kay DJ, Rafter KO, Montes M, Knox TW, Yousem DM. Facial nerve stimulation from cochlear implants. American Journal of Otology. 1998; 19: 163–9. 28. O’Donaghue G, Balkany T, Cohen N, Lenarz T, Lustig L, Niparko J Editorial. Meningitis and cochlear implantation. Otology and Neurotology. 2003; 23: 823–4. 29. Eisenman DJ, Ashbaugh C, Zwolan TA, Zwolan H, Alexander T, Telian SA. Implantation of the malformed cochlea. Otology and Neurotology. 2001; 22: 834–43. 30. Telian SA, Zinnerman-Philips S, Kileny PR. Successful revision of failed cochlear implants in severe labyrinthitis ossificans. American Journal of Otology. 1996; 17: 53–60. 31. Steenerson RL, Gary LB, Wynes MS. Scali vestbuli cochlear implantation for labyrinthine ossification. American Journal of Otology. 1990; 11: 360–3. 32. Gantz BJ, McCabe BF, Tyler RS. Use of multichannel implants in obstructed and obliterated cochleas. Otolaryngology – Head and Neck Surgery. 1988; 98: 72–81.



33. Balkany T, Hodges AV, Luntz M. Update on cochlear implantation. Otolaryngologic Clinics of North America. 1996; 29: 277–88. 34. Balkany T, Luntz M, Telischi FF, Hodges AV. Intact canal wall drill-out procedure for implantation of a totally ossified cochlea. American Journal of Otology. 1997; 18: 558–9. 35. Cohen NL, Waltzman SB. Partial insertion of the Nucleus multichannel cochlear implant: Technique and results. American Journal of Otology. 1993; 14: 357–61. 36. Bredberg G, Lindstro¨m B, Lo¨ppo¨nen H, Skarzynski H, Hyodo M, Sato H. Electrodes for ossified cochleas. American Journal of Otology. 1997; 18: S42–3. 37. Lenarz T, Lesinski-Schiedat A, Weber BP, Issing PR, Frohne C, Buchner A et al. The Nucleus double array cochlear implant: A new concept for the obliterated cochlea. Otology and Neurotology. 2001; 22: 24–32. 38. Babighian G. Problems in cochlear implant surgery. Advances in Otorhinolaryngology. 1993; 48: 65–9. 39. Waltzman SB, Roland JT, Cohen NL. Delayed implantation in congenitally deaf children and adults. Otology and Neurotology. 2003; 23: 333–40. 40. Muller J, Schon F, Helms J. Speech understanding in quiet and noise in bilateral users of the MED-EL COMBI 40/ 40 1 cochlear implant system. Ear and Hearing. 2002; 23: 198–206. 41. Kuhn-Inacker H, Shehata-Dieler W, Muller J, Helms J. Bilateral cochlear implants: A way to optimize auditory perception abilities in deaf children? International Journal of Pediatric Otorhinolaryngology. 2004; 68: 1257–66. 42. Baumgartner WD, Youssefzadeh S, Hamzavi J, Czerny C, Gstoettner W. Clinical applications of MRI imaging in early implant patients. Otology and Neurotology. 2001; 22: 818–22. 43. Cheng AK, Rubin HR, Powe NR, Mellon NK, Francis HW, Niparko JK. Cost–utility analysis of the cochlear implant in children. Journal of the American Medical Association. 2000; 284: 850–6. 44. O’Neill C, O’Donoghue GM, Archbold SM, Normand C. A cost–utility analysis of paediatric cochlear implantation. Laryngoscope. 2000; 110: 156–60. 45. Phillips C, Thompson G. What is a QALY? London: Hayward Medical Communications. Last updated 2003; cited May 2007. Available from: www.evidence-basedmedicine.co.uk/ebmfiles/WhatisaQALY.pdf 46. Francis HW, Koch ME, Wyatt JR, Niparko JK. Trends in educational placement and cost–benefit considerations in children with cochlear implants. Archives of Otolaryngology – Head and Neck Surgery. 1999; 125: 499–505.



71 Congenital middle ear abnormalities in children JONATHAN P HARCOURT



Introduction Classification of congenital ossicular abnormalities Incidence of congenital ossicular abnormalities Clinical presentation of congenital ossicular abnormalities Investigation of congenital ossicular abnormalities Principles of management Management of specific congenital ossicular abnormalities



869 869 870 871 871 871



Nonossicular congenital middle ear abnormalities Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



873 874 875 875 875



871



SEARCH STRATEGY The data in this chapter are supported by a search of the Guidelines databases on www.nelh.nhs.uk, the Cochrane library and TRIP databases using the key words congenital, middle ear and ossicle. This was supplemented by a PubMed search using the key words congenital and middle ear or ear, ossicle.



INTRODUCTION



Definition of congenital ossicular abnormalities



Conductive hearing loss in children is usually acquired. The most common aetiology is otitis media with effusion (OME) but it may be the result of chronic otitis media (mucosal or squamous). In the presence of severe congenital deformity of the external ear, associated abnormalities of the ossicular chain are common but in isolation they are rare and often have a delayed diagnosis, particularly if unilateral. A variety of nonossicular congenital middle ear abnormalities may also be associated with ossicular deformities. They may be symptomatic in themselves or be important aspects of other middle ear conditions and surgery. They include:



Congenital ossicular fixation and defect is defined as a malformation affecting the ossicular chain, present at birth, which leads to a dysfunction of the ossicular mechanism due to immobility or discontinuity of the ossicular chain. Major malformations involve both the tympanic cavity and the external ear (ear canal and pinna) and are described as congenital aural atresia or microtia. In association with these conditions, in either sporadic cases or as part of a syndrome such as Treacher Collins or Goldenhar, there may be a variable degree of ossicular abnormality and there may also be associated inner ear dysplasia. Minor malformations affect the ossicular chain alone and the tympanic membrane and ear canal are normal.



 persistent stapedial artery;  anomalous course of the facial nerve;  congenital perilymphatic fistula;  high jugular bulb;  aberrant internal carotid artery.



CLASSIFICATION OF CONGENITAL OSSICULAR ABNORMALITIES There have been many published classifications based on individual surgeons’ series.1 Cremers’ classification



870 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY describes the largest published series of operated minor congenital ear anomalies (104 cases).2 It has been modified by Tos3 and seems to be the most inclusive and descriptive of all the published schemes (Table 71.1). Minor malformations are divided into four main groups, each of which may be subdivided: isolated stapes ankylosis, ankylosis with other ossicular anomaly, isolated ossicular anomaly and aplasia or severe dysplasia of the oval or round windows. [*]



INCIDENCE OF CONGENITAL OSSICULAR ABNORMALITIES Bergstrøm4 reported that out of a group of 687 children with congenital hearing loss, only eight (1.2 percent) were



found to have isolated middle ear anomalies. There is no systematic report of absolute incidence though Thringer et al.5 reported the incidence of conductive hearing loss, severe enough to require amplification, to be 0.6:1000. These cases were all bilateral. The incidence of unilateral cases is unclear. Cremers and Teunissen’s series2 does provide a relative incidence with stapes ankylosis with another associated ossicular chain anomaly being the most common finding (38 percent) with isolated stapes ankylosis being the other largest group (30 percent). Isolated anomaly of the ossicular chain was found in 22 percent and aplasia or severe dysplasia of the oval or round windows in 10 percent. In a Japanese surgical series, Hashimoto et al.6 reported a much higher rate of incudostapedial joint



Table 71.1 Cremers’ classification of minor congenital anomalies of the ossicular chain in 144 operated ears (modified by Tos). Class



Main anomaly



Subclassification



% 144 ears



1



Isolated congenital stapes ankylosis (or fixation)



30.6



2



Stapes ankylosis associated with another congenital ossicular chain anomaly



3



Congenital anomaly of the ossicular chain but mobile stapes footplate



a. Footplate fixation 1. Normal stapedial arch 2. Monopodial stapedial arch 3. Monocrural stapedial arch b. Stapes suprastructure fixation 1. Elongation of the pyramidal eminence 2. Stapes-pyramidal process bony bar 3. Stapes facial canal bony bar 4. Stapes-promontory wall bony bar 5. More than one bony bar a. Incus and/or malleus deformation or aplasia of the long process of the incus b. Bony fixations of the malleus and/or incus Discontinuity of the ossicular chain Aplasia of the long process of the incus Dysplasia of the long process of the incus Epitympanic fixation Malleus Anterior Superior Lateral Incus body Superior Lateral Medial Short process of the incus In incudal fossa Tympanic fixation Of the malleus handle Of the long process of the incus Aplasia Dysplasia Crossing (prolapsed) facial nerve Persistent stapedial artery



4



Congenital aplasia or severe dysplasia of the oval or round window



Reproduced from Ref. 3, with permission.



38.2



15.3



6.3



6.9 2.1 0.7



Chapter 71 Congenital middle ear abnormalities in children



defects, either in isolation or with stapes fixation, making up nearly 50 percent of cases. The majority of Cremers’ cases were sporadic though in 25 percent they were part of a recognizable congenital syndrome including branchio-oto-renal, hemifacial microsomia, Klippel-Feil, Crouzon and Pfeiffer syndromes. [**]



CLINICAL PRESENTATION OF CONGENITAL OSSICULAR ABNORMALITIES Children with bilateral ossicular abnormalities will often present at a similar age to children with OME because of poor hearing performance and speech delay. It is not unusual that the children are managed with one or more sets of ventilation tubes before the diagnosis is made. The observation of a conductive hearing loss, normal tympanic membrane and normal middle ear pressures should lead to the general diagnosis.



INVESTIGATION OF CONGENITAL OSSICULAR ABNORMALITIES Audiometry There is an average threshold of approximately 50 dB, producing a flat air conduction line, with no low frequency bias as with otitis media. There is an average air–bone gap of 35 dB between 0.5 and 2 kHz, either due to Carhart’s effect or because of the presence of an underlying sensorineural hearing loss. Tympanometry usually demonstrates a normal middle ear pressure with reduced compliance due to fixation of the ossicular chain.



Imaging High resolution computed tomography (CT) scanning remains the primary imaging modality though complementary magnetic resonance imaging (MRI) studies may demonstrate associated labyrinthine and internal auditory meatal abnormalities. CT virtual endoscopy may offer a further mode of presenting the images for preoperative surgical planning though as yet it fails to image satisfactorily the stapes suprastructure.7



Exploratory surgery The diagnosis may only be made during a tympanotomy, though an interesting alternative to lifting the tympanic membrane is to attempt to visualize the ossicles via Eustachian tube and middle ear endoscopy.8 The innate risk of damage to the ossicles and extent of view with this technique is as yet not well defined.



] 871



PRINCIPLES OF MANAGEMENT As with all cases of hearing loss, children and adults with congenital ossicular abnormalities need to be managed with thought to their overall hearing performance and requirements. In the presence of a bilateral moderate hearing loss due to a maximal or near maximal conductive hearing loss, some form of auditory rehabilitation should be recommended. For the majority, this will mean a conventional unilateral or bilateral air conduction hearing aid. With minor malformations there should be a stable external ear canal as a platform for amplification. If the patient develops local complications in the external ear canal, such as recurrent or chronic otitis externa, a boneanchored hearing aid (BAHA) would be a suitable alternative. [*] Unilateral cases are less well defined in terms of best management. In the presence of ipsilateral tinnitus, amplification may act as a tinnitus masker. A hearing aid may improve the patient’s hearing performance in background noise and optimize sound localization. The positive benefits need to be weighed against the potential morbidity of a conventional hearing aid, which includes the occlusion effect, otitis externa and the body image issues involved in wearing hearing aids, particularly amongst children and adolescents. Surgery for congenital ossicular abnormalities should only be undertaken by dedicated otologists with experience of complex middle ear reconstruction. When the diagnosis has been made in childhood, consideration for surgery should be preceded by an adequate trial of amplification. Whatever middle ear surgery is contemplated, but particularly with stapedotomy, there is a significantly higher risk of delayed sensorineural hearing loss due to sporadic episodes of acute otitis media in children under ten years of age. By this time it may be appropriate to involve the child in the decision-making process or to wait until adolescence or adulthood to allow the patient to come to their own decision regarding surgical treatment. A preoperative CT scan would be a mandatory investigation to attempt to identify the ossicular abnormality, but also to visualize any other middle or inner ear abnormality such as an anomalous facial nerve, congenital cholesteatoma, aberrant vascular structures or labyrinthine dysplasia.



MANAGEMENT OF SPECIFIC CONGENITAL OSSICULAR ABNORMALITIES Isolated stapes ankylosis Tos3 subdivided Cremers’ basic classification to separate stapes ankylosis into two main groups, depending on



872 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY whether the footplate or suprastructure is fixed (Table 71.1). In 20 percent of cases the suprastructure may be abnormal, with one crus being absent (monocrural) or there being no recognizable crura but instead a single strut (monopodial).9 Congenital stapes ankylosis at the level of the footplate was first described by Shambaugh10 in 1952. He emphasized the clinical contrast to otosclerosis, in particular that the margins of the congenitally fixed footplate and the annular ligament are difficult to visualize since the footplate bone blends into the bone of the surrounding otic capsule. The pathological processes leading to stapes footplate fixation are unclear. Lindsey et al.11 concluded from a temporal bone study that fixation occurs due to failure of the annular ligament to differentiate from the lamina stapedialis. This leads to continuous ossification from the otic capsule to the footplate. Nandapalan and Tos12 argue, however, that the cause is a subsequent ossification of the already formed annular ligament, around 16 weeks of gestational age, because any arrest of development at this stage would be likely to cause a more widespread inner ear abnormality. An ossified stapedial tendon may develop because of a failure of its precursor to form a tendon and instead become cartilaginous like the neighbouring precursor of the pyramidal eminence.12 This subsequently becomes ossified, fixing the stapes. The other rare forms of fixation of the stapes suprastructure are probably caused by the persistence of the contact between the developing stapes and Reichert’s cartilage, which is usually lost as the stapes forms.



Surgery for congenital stapes footplate fixation Preoperative scanning may demonstrate labyrinthine dysplasia, any degree of which should alert the surgeon to an increased risk of inner ear damage. In particular, a dilated fundus of the internal auditory meatus (IAM) should be sought. This is a feature of X-chromosomelinked progressive mixed deafness with perilymphatic gusher.13 The dilation of the IAM is associated with a defect allowing communication of cerebrospinal fluid (CSF) with the labyrinthine, which is released through a stapedotomy and leads to a high rate of sensorineural hearing loss. The high perilymph pressure may be the cause of the apparent conductive element of the hearing loss, which may mask the principal sensorineural nature of the raised air conduction thresholds. A previously reported association with a ‘patent cochlear aquaduct’ has been shown to have no histological basis in temporal bone studies.14 The aquaduct is normally patent and cases of so called ‘widely-patent’ ducts seems to relate to a flaring on the medial aspect of the duct as it enters the



jugular foramen. There is little variation in the duct as it passes through the otic capsule. The surgical technique is similar to otosclerotic stapes ankylosis. A common feature of congenital fixation is the presence of thick anterior and posterior crura. To reduce the risk of inner ear damage these may be vapourized with a KTP laser.15 This may also be used to aid the formation of a stapedotomy as the footplate is often thick. Reported outcomes of surgery from specialist centres for isolated congenital fixation of the stapes nearly match the outcomes in the literature for adult series for otosclerosis, with more than 70 percent closure of the air–bone gap (ABG) to 20 dB or less.6, 16 However, excellent results, with an ABG of less than 10 dB, seem less common than with surgery for otosclerosis.17 Rates of inner ear damage are also comparable. [**]



Surgery for stapes suprastructure fixations In the very rare cases where the footplate is mobile but the suprastructure is fixed, removal of the bony bar can be achieved using a CO2, argon, KTP or erbium laser or with a microdrill. Traditionally this has been achieved with a curette or by fracturing with good results.12



Isolated non-stapes middle ear anomalies There is a large variety of described anomalies amongst case reports in the literature and these represent the different subclasses in Table 71.1. In addition the stapes itself, though mobile, may be dysplastic. The surgical management is by appropriate tympanoplasty. If there is attic fixation then a separate atticotomy is necessary to expose the problem and allow surgical treatment. If there is a bony bar this can be relieved by laser, microdrill or curette. A similar approach is used with fixation of the malleus handle by an atretic plate. With the absence of the long process of the incus, an interposition prosthesis is indicated.



Stapes ankylosis associated with other deformities The potential anomalies are as described when found in isolation. Stapedotomy is again indicated but it may be necessary to consider a mobilization of the ossicular chain if there is epitympanic fixation or a malleovestibulopexy following removal of the head of the malleus.6 Though surgery is difficult in cases of a deformed incus or fixations of the incus and malleus in the attic, reported results are good.17



Chapter 71 Congenital middle ear abnormalities in children



] 873



Congenital aplasia or severe dysplasia of the oval and round windows The stapes crura are usually poorly developed and do not reach the region of the footplate and may be embedded in the facial nerve. In such cases the course of the facial canal is often anomalous. With such a severe abnormality preventing conduction of sound energy into the inner ear and particularly with the risk of a facial palsy with an anomalous facial nerve, auditory rehabilitation with hearing aids or a BAHA may be most appropriate. The potential surgical solutions have significant drawbacks. A fenestration procedure would leave a cavity requiring long-term care. A neo-oval window operation, in which a de novo entrance into the labyrinth is either drilled on the promontorial or even on the rostral side of the Fallopian canal, has a high risk of inner ear damage.2



SA



Figure 71.1 Finding during stapes surgery – a persistent stapedial artery (PSA).



NONOSSICULAR CONGENITAL MIDDLE EAR ABNORMALITIES Persistent stapedial artery The stapes forms around the stapedial artery, leading to formation of the obturator foramen. By ten weeks of development the artery atrophies, leaving a patent foramen underneath the arch of the stapes. If persistent, it arises from the petrous internal carotid artery (ICA), traverses Jacobsen’s canal for a short segment, exits at the promontory, passes through the stapes obturator foramen and enters the Fallopian canal close to the cochleariform process. The persistent stapedial artery (PSA) passes anteriorly, exiting the canal at the geniculate ganglion and passes into the extradural space of the middle cranial fossa, where it gives rise to the middle meningeal artery. In the presence of a PSA, the foramen spinosum is usually absent and the ICA may have an aberrant course due to its collateral formation secondary to a segmental agenesis of the ICA.18 A PSA may present as a vascular mass within the middle ear,18 found either during otoscopic examination of the ear or during middle ear exploration, but is often asymptomatic or the cause of pulsatile tinnitus and has no relationship to conductive hearing loss. It may also present as a chance finding during stapes surgery (Figure 71.1). It is unclear whether the discovery of a PSA during a stapedectomy indicates a congenital fixation or whether it is coincidental in cases of otosclerotic stapes ankylosis. Traditional teaching has been that the presence of the artery is an absolute contraindication to stapedectomy, partly as it was thought that damage to the artery would lead to ischaemic damage to the facial nerve. Govaerts et al.,19 however, reported 12 cases, including two of their



own, in which the outcome of surgery was satisfactory. In three cases the artery was damaged or clipped without any postoperative complications. [*]



Anomalous course of the facial nerve The Fallopian canal may show dehiscence or have an anomalous course within any part of its course within the temporal bone. This has great surgical significance and may be a hazard in all forms of tympanomastoid surgery. An abnormal course is particularly common with microtia or with dysplasia of the oval and round windows and there should be a high index of suspicion in any surgery for congenital conductive hearing loss and the use of the facial nerve monitor is highly recommended. The Fallopian canal arises from the otic capsule and the second branchial arch and a theory of the cause of an anomalous facial nerve is a failure of fusion of the two. As there is a relatively high association with other ossicular abnormalities other than of the stapes, a local teratogenetic effect may also be responsible. Rohrt and Lorentzen20 classified facial nerve displacement in the middle ear into four groups: 1. facial nerve partially obliterates the stapes footplate; 2. bifurcation of the facial nerve; 3. facial nerve rests on the footplate with deformed stapes or oval window; 4. facial nerve rests on the promontory. Though these abnormalities are often found in association with stapes fixation, it may be normal and mobile, even in the presence of a bifurcated facial nerve (Figure 71.2).



874 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



High jugular bulb



Congenital perilymphatic fistula (PLF) is an abnormal communication between the middle and inner ear. It may be associated with:



A high jugular bulb (HJB) is usually asymptomatic and discovered as an incidental finding on otoscopy or during middle ear surgery. It can cause heavy bleeding from accidental puncture while lifting a tympanomeatal flap or even inserting a ventilation tube. The jugular bulb can be defined as ‘high’ if it reaches the level of the inferior bony annulus and is often covered by thin bone or is dehiscent. In the presence of a plethoric mass within the tympanic cavity, the differential diagnosis will include HJB, aberrant ICA, a PSA and a glomus tympanicum tumour. An HJB can be associated with hearing loss. If it includes a medial portion impinging on the cochlear or vestibular aquaduct, a connection with vestibular symptoms and a sensorineural hearing loss has been suggested. There may be a more direct association with a conductive hearing23 loss due to interference with the ossicles, contact with the tympanic membrane and obstruction of the round window niche. Major surgery to occlude or re-route an HJB is unlikely to be justified by symptoms alone, though there are occasional reports of intrusive pulsatile tinnitus associated with an HJB. It is possible to consider endovascular occlusion in the rare case of persistent haemorrhage, when tympanic cavity surgery is contraindicated.24



 microfissures around the oval and round windows;  labyrinthine or IAM dysplasia.



Aberrant internal carotid artery



Figure 71.2 Stapes may be normal and mobile, even in the presence of a bifurcated facial nerve.



Congenital perilymphatic fistula



The diagnosis is controversial as there is no reliable preoperative test which identifies the condition and management is based on the suggestive diagnosis of a child presenting with progressive or fluctuating sensorineural hearing loss, possibly associated with vertigo. Weber et al.21 define the intraoperative diagnosis as being based on the identification of clear fluid which reaccumulates with anaesthetic Valsalva or Trendelenburg manoeuvre. Beta-transferrin positive samples are consistent with CSF being a constituent of the leak and this will be a feature of some but not all cases of PLF, usually associated with labyrinthine or IAM dysplasia. This is not a real-time test during the surgical procedure and is very specific but not very sensitive for PLF, which may not contain CSF.22 In view of the difficulties of diagnosis, Weber et al.21 suggest the policy of packing temporalis muscle around the oval and round windows in all suspected cases. This is based on the observation that packing does not seem to cause any significant complications (such as a subsequent conductive hearing loss) and he argues that it may be beneficial despite no confirmation of a PLF. In his series of 160 ears with suspected PLF there was a greater than 90 percent rate of stabilization or improvement in hearing in both PLF positive and negative cases, though he acknowledges that in other series there is a similar outcome in nearly half of cases without packing. [**]



This may be associated with other vascular abnormalities such as a PSA and likewise present as a vascular middle ear mass. Associated symptoms include pulsatile tinnitus, which may be objective, and hearing loss. In approximately 20 percent of cases it is bilateral.25 An aberrant ICA is an important differential diagnosis of a glomus tympanicum tumour, which can be resolved by CT scanning. Brisk bleeding, hemiparesis, aphasia, deafness, Horner syndrome and intractable vertigo may result if the vessel is unintentionally injured.25



KEY POINTS  Congenital abnormalities of the middle ear may occur in association with major malformations such as microtia.  Isolated malformations are rare but include anomalies of the ossicular chain and the middle ear vasculature.  Diagnostic delay is common. Deafness due to ossicular pathology is often misdiagnosed as otitis media with effusion. It is not unusual for children to have repeated ventilation tubes before the correct diagnosis is made.



Chapter 71 Congenital middle ear abnormalities in children



Best clinical practice [ Unless deafness is bilateral management can be conservative. [Grade D]



[ If there is bilateral hearing loss, auditory rehabilitation will be required. [Grade D]



[ Surgery is difficult with uncertain outcomes and best considered only by those with specific training and experience. [Grade D] [ Surgery may be best deferred until children can participate in the process of informed consent. [Grade D]



Deficiencies in current knowledge and areas for future research



$ $ $



The incidence, aetiology and pathogenesis of congenital ossicular abnormalities is poorly understood. The role of otoendoscopy in the diagnosis and management of these conditions needs to be better defined. Developments in middle-ear implantation may render corrective surgery almost redundant.



REFERENCES







1. Nandapalan V, Tos M. Isolated congenital stapes suprastructure fixation. Journal of Laryngology and Otology. 1999; 113: 798–802. 2. Cremers CWRJ, Teunissen E. A classification of minor congenital ear anomalies and short- and long-term results of surgery in 104 ears. In: Charachon R, Garcia-Ibanes E (eds). Long-term results and indications in otology and otoneurosurgery. Amsterdam/New York: Kugler Publications, 1991: 11–2. 3. Tos M. Congenital ossicular fixations and defects. In Surgical solutions for conductive hearing loss. New York: Thieme, 2000: 212–39. 4. Bergstrøm L. Assessment and consequence of malformations of the middle ear. Birth Defects. 1980; 16: 217–41. 5. Thringer K, Kankkunen A, Liden G, Nikalson A. Prenatal risk factors in the aetiology of hearing loss in pre-school children. Developmental Medicine and Child Neurology. 1984; 26: 799–807. 6. Hashimoto S, Yamamoto Y, Satoh H, Takahashi S. Surgical treatment of 52 cases of auditory ossicular malformations. Auris, Nasus, Larynx. 2002; 29: 15–8.



] 875



7. Nakasato T, Sasaki M, Ehara S, Tamakawa Y, Muranaka K, Yamaoto T et al. Virtual CT endoscopy of ossicles in the middle ear. Clinical Imaging. 2001; 25: 171–7. 8. Karhuketo TS, Ilomaki JH, Dastidar PS, Laasonen EM, Puhakka HJ. Comparison of CT and fiberoptic videoendoscopy findings in congenital dysplasia of the external and middle ear. European Archives of Otorhinolaryngology. 2001; 258: 345–8. 9. Ombredanne M. Chirurgie des surdites congenitales par malformations ossiculaires de 10 nouveaux cas. Annales d’Oto-laryngologie et de Chirurgie Cervico Faciale. 1960; 77: 423–49. 10. Shambaugh GE. Developmental anomalies of the sound conducting apparatus and their surgical correction. Annals of Otology, Rhinology and Laryngology. 1952; 61: 873–87. 11. Lindsay JR, Sanders SH, Nager GT. Histopathologic observation in so-called congenital fixation of the stapedial footplate. Laryngoscope. 1960; 70: 1587–602. 12. Nandapalan V, Tos M. Isolated congenital stapes ankylosis: an embryologic survey and literature review. Otology and Neurotology. 2000; 21: 71–80. 13. Cremers CWRJ, Hombergen GCHJ, Scaf JJ, Huygen PLM, Volkers WS, Pinckers AJLG. X-linked progressive mixed hearing deafness with perilymphatic gusher during stapes surgery. Archives of Otolaryngology. 1985; 111: 249–54. 14. Jackler RK, Hwang PH. Enlargement of the cochlear aqueduct: fact or fiction? Otolaryngology – Head and Neck Surgery. 1993; 109: 14–25. 15. Nishizaki K, Kariya S, Fukushima K, Orita Y, Okano M, Maeta M. A novel laser-assisted stapedotomy technique for congenital stapes fixation. International Journal of Pediatric Otorhinolaryngology. 2004; 68: 341–5. 16. Welling DB, Merrell JA, Merz M, Dodson EE. Predictive factors in pediatric stapedectomy. Laryngoscope. 2003; 113: 1515–9. 17. De La Cruz A, Angell S, Slattery W. Stapedectomy in children. Otolaryngology – Head and Neck Surgery. 1999; 120: 487–92. 18. Silbergleit R, Quint DJ, Mehta BA, Patel SC, Metes JJ, Noujaim SE. The persistent stapedial artery. American Journal of Neuroradiology. 2000; 21: 572–7. 19. Govaerts PJ, Cremers CWRJ, Marquet TF, Offeciers FE. The persistent stapedial artery: does it prevent successful surgery? Annals of Otology, Rhinology and Laryngology. 1993; 102: 724–8. 20. Rohrt T, Lorentzen P. Facial nerve displacement within the middle ear (report of 3 cases). Journal of Laryngology and Otology. 1976; 90: 1093–8. 21. Weber PC, Bluestone CD, Perez B. Outcome of hearing and vertigo after surgery for congenital perilymphatic fistula in children. American Journal of Otolaryngology. 2003; 24: 138–42. 22. Bluestone C. Implications of beta-2 transferrin assay as a marker for perilymphatic versus cerebrospinal fluid labyrinthine fistula. American Journal of Otology. 1999; 20: 701.



876 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY 23. Weiss RL, Zahtz G, Goldofsky E et al. High jugular bulb and conductive hearing loss. Laryngoscope. 1997; 107: 321–7. 24. Kondoh K, Kitahara T, Mishiro Y, Okumura S, Kubo T. Management of hemorrhagic high jugular bulb with adhesive otitis media in an only hearing ear: transcatheter



endovascular embolization using detachable coils. Annals of Otology, Rhinology and Laryngology. 2004; 113: 975–9. 25. Windfur JP. Aberrant internal carotid artery in the middle ear. Annals of Otology, Rhinology and Laryngology Supplement. 2004; 192: 1–16.



72 Otitis media with effusion GEORGE BROWNING



Definition of otitis media with effusion Aetio-Pathology Epidemiology Diagnosis Outcomes of childhood OME Management



877 878 880 884 891 895



Key points Best clinical practice Deficiencies in current knowledge and areas for future research Acknowledgements References



903 904 905 906 906



SEARCH STRATEGY Different search strategies were used for each section of this chapter, as follows: DEFINITION AND AETIO-PATHOLOGY OF OTITIS MEDIA WITH EFFUSION



Premedline and Medline search using the key word otitis media or otitis media with effusion; and allergy, cleft palate, Down’s syndrome, Eustachian tube, histology, microbiology, pathology, Turner’s syndrome, gastro-oesophageal reflux or GERD or GORD. EPIDEMIOLOGY



Medline was searched using the terms otitis media or otitis media with effusion and prevalence, and epidemiology. DIAGNOSIS



Medline was searched using the terms otitis media or otitis media with effusion and sensitivity or sensitivity and specificity. OUTCOMES OF CHILDHOOD OTITIS MEDIA WITH EFFUSION



Medline was searched using the key words otitis media or otitis media with effusion; and speech and language, and behaviour, and balance and/or equilibrium, and quality of life. MANAGEMENT



Premedline and Medline were searched using the key words otitis media and/or otitis media with effusion; and adenoidectomy, antibiotics, carboxymethylcysteine, hearing aids, and homeopathy. Abstracts were reviewed to identify papers with the highest levels of evidence. These were then read and evaluated.



DEFINITION OF OTITIS MEDIA WITH EFFUSION Otitis media with effusion (OME) is the chronic accumulation of mucus within the middle ear and



sometimes the mastoid air cell system. The time that the fluid has to be present for the condition to be chronic is usually taken as 12 weeks.1 [*] In children, OME usually presents because of the associated hearing



878 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY impairment and sometimes with a preceding history of illness and otalgia consequent on an episode of acute otitis media. Many synonyms have been and are used for this condition. The general public frequently refer to it as ‘glue ear’. Medical terms that are no longer in common usage, because of debate about their appropriateness or tendency to cause confusion, include serous otitis media, secretory otitis media and chronic nonpurulent otitis media. In many children, OME is preceded by an episode of acute otitis media with otalgia and fever. This is particularly so in younger children because of their greater propensity to upper respiratory tract infections. In the majority of children, acute otitis media is considered to have been triggered by a viral upper respiratory tract infection that damages the epithelium of the Eustachian tube, resulting in retention of middle ear fluid. These secretions then become secondarily infected with bacteria – acute otitis media. Once the infection has resolved, it can take time for the epithelium to recover. Hence, OME will be present temporarily in many children after an episode of acute otitis media. However, many children with OME have no recent history suggestive of acute otitis media, albeit they may have had an upper respiratory tract viral infection. This damages the Eustachian tube epithelium with resultant retention of middle ear fluid, which in these children does not become secondarily infected.



AETIO-PATHOLOGY Histology The ciliated, pseudostratified columnar epithelium of the respiratory tract extends up the Eustachian tube as far as the anterior part of the middle ear cavity (Figure 72.1). These cells are capable of producing mucus. In addition there are goblet cells and mucus secreting glands. OME is primarily caused by an inflammation of this epithelium in the Eustachian tube and hypotympanum. The histology has been well described from biopsy specimens of the middle ear mucosa3, 4 and from temporal bone sections.5 Once OME has become established, the normal, flat cuboidal middle ear and mastoid mucosa is patchily replaced by thickened pseudostratified mucus-secreting epithelium with varying degrees of specialization, such as the development of cilia. Goblet cells are frequently present and sometimes mucus-secreting glands are formed. The ciliary lining would appear to be less efficient at moving the secretions into the nasopharynx than normal.6 [***] The submucosa is oedematous and inflamed with dilated blood vessels and an increased number of macrophages, plasma cells and lymphocytes (Figure 72.2).



A



M



P ET H



80%+ 51–80% 10–50%



Figure 72.1 Extent of respiratory-type mucosa within the middle ear represented by the percentage of ciliated mucus-secreting cells per unit area. A, antrum; ET, Eustachian tube; H, hypotympanum; M, mastoid; P, promontory. Redrawn from Ref. 2.



CHARACTERISTICS OF THE EFFUSION



What is frequently debated is the origin of the middle ear fluid in OME. In surgical practice, the fluid is usually characterized by its consistency as being either serous or mucoid. This is obviously a gross simplification of the situation. There will be a full spectrum of fluid types made up of a mixture of the secretions of the epithelial cells, the goblet cells and the mucus glands along with the inflammatory transudate/exudate which comes through the intracellular spaces from the inflammed submucosa. Considerable research time has been spent on analyzing the concentrations of various molecules in middle ear effusions in an attempt to draw conclusions as to why they are there and where they come from. The main finding is that it is the mucins that come from the secretions that are responsible for the variable viscosity of the middle ear fluids.7 [*]



Bacteriology Table 72.1 shows the percentage of ears that had a positive bacteriological culture of a middle ear aspirate of OME fluid (more than two months duration) in children with a wide age spectrum.8 [***] No pathogens were cultured from fluid that had been present for at least six months. The incidence of pathogens was higher in younger children (less than two years) and in those with recurrent upper respiratory infections and recurrent attacks of acute otitis media.



Eustachian tube dysfunction Inevitably, if the epithelium of the Eustachian tube is inflamed, becomes oedematous and loses its cilia, then it



Chapter 72 Otitis media with effusion



] 879



Figure 72.2 Otitis media with effusion – cross section through the temporal bone. Minimal effusion remains post-mortem, but the mucosa (arrowed) of the middle ear over the promontory and in the facial recess is hypertrophied, hyperaemic and slightly polypoid. The bony Eustachian tube (ET) is patent. C, cochlea; EAC, external auditory canal (meatus); ME, middle ear. (Collection MEEI).



Table 72.1



Bacteriology of OME.



Organism cultured Streptococcus pneumoniae Haemophilus influenzae Branhamella catarrhalis Streptococcus pyogenes Other bacteriaa Negative culture Total



Percentage 8 3 7 1 15 66 100



a Staphylococcus aureus, Staphylococcus coagulase negative, Streptococcus viridans, Enterococcus faecalis, diphtheroids. Reprinted from Ref. 8, with permission. n = 141.



will dysfunction as an aerator of the middle ear. Though a viral upper respiratory tract infection is probably the most common cause of damage to the Eustachian tube epithelium, there are other potential reasons. It may be secondary to an allergic reaction (see Allergy) or pollutants such as cigarette smoke. It may be secondary to chronic nasopharyngeal infection in the adenoidal tissue or gastro-oesophageal reflux. It will be rare that abnormal Eustachian tube dysfunction will be due to a disorder of the palatine muscles. The exception to this will be children with a cleft palate.



Craniofacial abnormalities Children with a cleft palate, even if repaired, have deficient palatine muscles and resultant poor Eustachian tube function. As a consequence, OME is virtually universal in infants with a cleft palate and surgical repair of their clefts does not seem to influence the incidence. As the child ages the incidence falls, but there are no cohort studies to give age-related data, though the overall



incidence is at least 40 percent. The disease is extremely persistent and in the longer term gives rise to a high incidence of chronic otitis media of at least 20 percent in ten year olds.9 [***] Whether this is helped or made worse by surgery is debated. Children with a bifid uvula do not appear to have a higher incidence of OME.10 [***] Children with Down or Turner syndrome are also more likely to have OME, but reliable data on the incidence are not available.



Allergy The majority of children with OME do not have allergic symptoms of their upper and lower respiratory tracts or skin. The question is whether allergic children are more likely to develop OME and if they do so whether their OME is more protracted than normal. Comparing the incidence of allergy in OME children attending a particular clinic with the population incidence can be subject to referral bias. This is evident by the considerable variation in the incidence of allergy reported from different clinics that cannot be accounted for by different methods of diagnosis. Population cohorts of children under the age of two years that have investigated atopy/allergy as a risk factor for the development of OME have not found allergy to be significant even in a univariate analysis.11, 12 [***] Similar cohort studies in older children are not available. However, three prospective studies of children covering all ages with OME that looked for risk factors for persistence did not find atopy/allergy to be a significant factor.13, 14, 15, 16, 17 [***] It must be concluded that the best evidence does not support allergy as a risk factor for the occurrence or persistence of OME. [***]



880 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Gastro-oesophageal reflux Radiographic evidence of gastro-oesophageal reflux is common in children and its relevance to many conditions has perhaps been overestimated. However, initial reports on the biochemical analysis of middle ear fluid in OME suggest that pepsin is present in a high proportion of effusions (approximately 80 percent).18 [*] Subsequent prospective reports on children with OME or recurrent otitis media would raise doubts as to whether pepsin is causative as of 14 children with bilateral effusions positive pepsin assays were only present bilaterally in seven children.19 [**] Further prospective investigations are required to clarify the role, if any, of gastro-oesophageal reflux in childhood OME.



16 percent at around five years of age when most children start attending a primary school. In Figure 72.3 there are few data points below the age of two years and hence the exact location of the first peak is liable to errors of interpretation. A more recent study, using a diagnostic algorithm of tympanometry and otoscopy, suggest that the peak is more likely to be around one year of age (Figure 72.4).23 [***] The proportion of unilateral to bilateral OME shown in this figure is discussed below under



40



The epidemiology of OME has been extensively studied in many countries, but mainly in Denmark, Finland, the Netherlands, the United States of America and the United Kingdom. The majority are studies of cohorts of children followed up at regular intervals from birth to two or three years of age. As such, there are many level 2 evidence studies to report from, it being usually impracticable to perform randomized controlled epidemiological trials (see below under Outcomes of childhood OME). No recent systematic review has been identified, that of Daly20 being based on early publications where the study design and method of analysis tends to be poorer than in subsequent papers. Hence, the papers quoted here were mainly published after 1990.



Prevalence The prevalence of a condition is the proportion of a population that has a condition at any one time. Such data are useful as an indicator of the potential clinical workload. The incidence of a condition measures the chances of developing the condition. In childhood OME, the main determinants of the prevalence are the age of the child and the season of the year.



ASSOCIATION WITH AGE



Zielhius et al.21 [***] reviewed a total of 23 studies that all used tympanometry as one of the diagnostic instruments to give age-specific prevalence rates up to ten years of age (Figure 72.3). The prevalence is bimodal with the first and largest peak of approximately 20 percent at two years of age. This is the age at which many children first attend a playgroup or nursery school. Thereafter the prevalence declines, but there is a second peak of approximately



20



10



0



0



20



40



60 80 Age (months)



100



120



Figure 72.3 The prevalence of otitis media with effusion at different ages with 95 percent confidence intervals (CI). Chart shows the combined results of 23 studies. Redrawn from Ref. 21, with permission.



Uni- or bilateral



50



40 Prevalence OME (%)



EPIDEMIOLOGY



Prevalence (%)



30



Bilateral



30



Unilateral



20



10



0 0



3



6



9 12 15 Age (months)



18



21



24



Figure 72.4 Prevalence rates of unilateral, bilateral and unior bilateral OME in a group of healthy-born infants from birth to two years of age. Redrawn from Ref. 22, with permission.



Chapter 72 Otitis media with effusion



Duration and recurrence of episodes in children under the age of three years. By the age of seven to eight years, the prevalence falls, but is still material around 5 percent depending on season (Table 72.2).24



ASSOCIATION WITH SEASON OF THE YEAR



In temperate climates, the effect of the season on the prevalence of OME has been recognized for many years, with around twice as many children being diagnosed with OME in the winter as opposed to the summer months.25, 26, 27, 28 [***] Why there is a difference could be due to many factors. The most likely are the increased frequency of upper respiratory and ear infections in the winter and the greater chance of passing them on between children because of the closer contact in cold weather and when they are not on holiday. Indeed if these factors are controlled for, in some studies the effect of season is no longer as evident,29 [***] but not in others.30, 31 [***] Evidence from nontemperate countries might help to sort out season from upper respiratory tract infections and contact with other children. The prevalence of OME in white children in Mediterranean32, 33 [**] and subtropical countries34, 35 [**] does not appear to be different overall from those in temperate countries. Unfortunately, no seasonal analysis was made. However, a comparison was made36 [***] of the prevalence in a northern tropical country (Vietnam) in the dry (3 percent of ears in April) and the rainy seasons (11 percent of ears in December which is humid and chilly). Their second peak of prevalence was not at five to six years of age, but at over eight years, when many children start work in the paddy fields. More such studies would be of value.



Risk factors for occurrence of OME under three years of age



] 881



commonly used with a type B tympanogram being considered diagnostic of OME. Whilst this may be a sensitive categorization in the diagnosis of the majority of cases, it is not that specific in that there are a good number of false positives (see below under Diagnosis). The risk factors studied are not the same in all studies. The inclusion of the most important ones is vital. This requires a priori knowledge of what these are likely to be. The same factors are unlikely to apply to all age groups, for example, breast feeding. The factors that predict occurrence of OME are not necessarily those that predict the persistence of OME. Many studies have too few subjects with a particular characteristic, e.g. not many five-year-old children do not go to school. A multivariate or a comparable analysis of the data is required to control for confounding variables, for example, if low socioeconomic group is a factor is it the small house, the number of siblings, the smoking, the junk food or the lack of a maternal university education and a car that makes the difference? It is unlikely to be all of them. Surgery in some children complicates the analysis and complicates its interpretation. It is likely that the risk factors for the occurrence of unilateral and bilateral OME will be the same. This has been confirmed by Engel et al.31 [***] Again, on a priori grounds, the risk factors for persistence of OME are likely to be different. This is particularly because of the introduction of disease-related factors such as the degree of the associated hearing impairment that could be related to persistence. Knowledge of risk factors can be of interest in helping to understand why some individuals are more prone to suffer from a condition than others. In a specific individual, the risk of developing the condition can be estimated and any risk factors present modified, if this is possible.



Episodes of acute otitis media Interpreting the literature on risk factors for the occurrence of OME is complicated by the need to use a rapid, simple, readily applicable method for diagnosing OME. Serial tympanometry is the method most Table 72.2 Prevalence of type B tympanogram by age and season using school term as a surrogate. Age (years) Autumn (%) 5 6 7 8 All ages



13 9 4 0 9



Spring (%)



Summer (%)



Mean (%)



20 13 10 6 14



17 8 7 6 10



17 10 7 6



Reprinted from Ref. 24, with permission.



In the age group most studied, i.e. 0–2 years, recurrent episodes of acute otitis media are frequent and this is likely to be the largest single factor for developing OME. Each acute episode has been reported to increase the odds ratio of developing OME by 12 (CI 6, 25) but this risk no longer remains after three months.11 [***] Hence the risk factors for acute otitis media, such as prematurity, frequent upper respiratory infections and nonbreastfeeding, will apply in such children.23 Studies of children in this age group that do not include the frequency of acute otitis media (AOM) must be considered less than optimal. Unfortunately there is no evidence that management of acute otitis media, and in particular with antibiotics, makes any difference to the chances of having subsequent OME.37 [****]



882 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Contact with other children at home and at playgroups Table 72.329 [***] shows the risk factors in addition to AOM in a multivariate analysis for the occurrence of OME. This confirms the increasing effect of age up to two years of age, but also the effect of having siblings and a family history of them having OME. In another study that otoscopically examined family members of children with OME, rather than rely on historical reports, 32 percent of siblings and 19 percent of parents were also classified as otoscopically affected.38 [**] Other multivariate studies have suggested that in addition to the number of siblings, attendance for daycare with four or more other children up to 3.5 years of age can double the risk (CI 1, 3).30 [***] This risk has been confirmed by others.12 [**] Some have advocated that the number of children in daycare settings could be reduced to lessen this risk, but for most parents availability, convenience and cost are limiting determinants.



Hereditability In a same sex twin/triplet prospective cohort study, Casselbrandt et al.39 [***] looked at sets where the zygosity was known. In children who had OME during the first two years of life, there was greater concordance in monozygotic sets in the number and duration of OME episodes than in dizygotic sets. The magnitude of the effect of heredity in comparison to other risk factors for OME was not investigated.



Race Whether the prevalence of OME is different in different races requires control for many other factors. When factors including socioeconomic group and child contacts are controlled for in a multivariate analysis, the prevalence in black children is no different from white children.40 [***] Chinese children may have a lower prevalence, but a multifactorial analysis has not been reported.35 [**] Table 72.3 Risk factors significant in a multivariate model for OME under two years of age. Risk factor AOM Previous AOM Age No. of siblings OME family history



Category



OR



95% CI



Yes/No Yes/No Increase per month Increase per sibling Yes/No



1.6 1.7 1.0 1.6 1.4



1.2, 1.2, 1.0, 1.3, 1.1,



Reprinted from Ref. 29, with permission.



2.1 2.4 1.0 2.0 1.7



It must be concluded that at present there is insufficient evidence to examine any effect of race, apart from black and white where no difference has been shown.



Gender Some multivariate studies report the risk of developing OME to be no different in boys and girls.30, 31, 41 Some report a higher risk in boys11, 38 and others a higher risk in girls.29 It must be concluded that there is likely to be little difference, if any, in the risk for boys compared to girls.



Smoking For public health reasons, parental smoking has frequently been reported as a risk factor for the occurrence of OME. However, in multivariate analysis when other factors have been controlled for, no effect of parental smoking is detected31 [***] or it is present for smokers of up to 20 cigarettes per day, but the risk reduces if it is more than 20 cigarettes per day.30 [***] The conclusion must be that, in comparison to the other material risk factors for the occurrence of OME in children under the age of two years, the effect of parental smoking must be negligible. [***]



Risk factors for occurrence in children older than three years No literature has been identified that reports a multifactorial analysis of risk factors in children older that three years of age. However, episodes of acute otitis media are likely to be less important because of its lower prevalence in this age group.



Duration and recurrence of episodes in children under the age of three years The natural history of the duration, recovery and recurrence of OME in 1328 children from one to two years of age is shown by ear in Figure 72.5.42 [***] The distribution of the duration of episodes is very skewed; the median duration of OME was three months or less, but the 95th percentile was at 12 months. Fifty percent of affected ears had resolved after three months. However, around 50 percent of the ears that resolved had a further episode. What happens in individual infants rather than in ears has been documented by Paradise et al.40 [***] In a large cohort (n = 2253) followed up from 2 months of age to 6, 12, 24 months of age, 49, 79 and 91 percent respectively had at least one episode of OME.



Chapter 72 Otitis media with effusion



] 883



Proportion with glue ear (ears) 80 70



Cumulative incidence



no OME



60 50 40



Starting at 27 months



30 20



Starting at 24 months duration 1st recurrence



10



2nd recurrence



24



27



30



33



36



39



42



45



48



Hogan et al.43 [***] followed up 95 full-term infants monthly for three years. The infants fell into three groups. The majority (65 percent) had OME at less than a third of monthly visits. Of the other 35 percent with OME at more than one visit, half of them had OME in one or both ears for more than 18 (50 percent) of their first 36 months of life. Though the mean duration of OME was longer in those with a propensity for developing OME (unilateral six weeks; bilateral ten weeks), the main difference was time between episodes. On average this was longer in children who had less OME (every 31 weeks compared with every 11 weeks). Overall, infants were twice as likely to develop unilateral as opposed to bilateral OME. One month later, in infants with a unilateral effusion, the majority (50 percent) had resolved, a minority (20 percent) had become bilateral and the other 30 percent remained unilateral. In those with bilateral effusions at the start, one month later the majority (60 percent) remained bilateral. In the others (40 percent), bilateral resolution was more frequent than unilateral resolution (in a ratio of 3:1).



Duration and recurrence of episodes in children older than three years A cohort of 856 British children, aged five to eight years was screened three times a year (approximately every four months) for three successive years to study the prevalence and natural history of OME defined by a type B tympanogram.24 [***] Table 72.2 confirms the already reported age and season effect. There was no gender effect. The proportion of unilateral to bilateral OME was equal. In a subsample of 67 ears with OME, the resolution rates at 4, 8 and 12 months were 52, 78 and 91 percent, respectively. The overall recurrence rate was low (7 percent).



Age (months)



Figure 72.5 Natural history of OME (n = 1217 ears). Shaded areas indicate the presence of OME by episode of occurrence. The breadth of the hatched areas indicates the duration of the episode in various proportions of the total sample of ears. Redrawn from Ref. 42, with permission.



A more detailed monthly study of seven-year-old Danish pupils (n = 387) reported the mean overall duration of an ear episode to be 1.8 months with 12 percent lasting more than 6 months.44 [***] Ears first diagnosed to have OME between September and February persisted longer than those first diagnosed between March and August.



Risk factors for persistence Three studies have been identified that have followed up children to identify risk factors that might be used to predict those that are more likely to have persistent bilateral OME. One study was carried out in primary care and followed up children who were six months or older for three months.13 [***] Another study was of children aged between three and seven years referred from primary care with suspected OME and seen at secondary care on average 13 weeks later.15 [***] The third study was of children of the same age identified at secondary care to have bilateral OME with an associated hearing impairment in both ears of at least 20 dB hearing loss (HL). These children were followed up for three months to identify those in whom the OME persisted bilaterally with the same degree of impairment.14, 16, 17 [***] In these three studies, which all followed up the children for three months, the persistence rates of bilateral OME were 56, 35 and 51 percent, respectively. All studies identified the second half of the year (July–December) as a major risk factor with odds ratios of between 2 and 3. In primary care, whether a child has frequent or upper respiratory symptoms at the time of assessment is also an important determinant of persistence. In secondary care, the degree of associated hearing impairment predicted persistence. The only factor that might have an influence and has the potential to be modified was whether or not



884 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Clinical applicability of best epidemiological evidence The prevalence of OME in childhood is age dependent, with two peaks in the distribution; one centred around one to two years of age and the other around three to seven years of age. [***] In temperate countries, twice as many children have OME in the winter as opposed to the summer. The increased frequency of upper respiratory infections and close contact with other children during the winter months contribute to this association. [***] Under three years of age, episodes of acute otitis media, contact with other children and heredity are factors that increase the risk of occurrence. No easily modifiable factors in the age group have been identified. [***] Under three years of age, unilateral OME is twice as common as bilateral OME. Those with a propensity to develop OME have more frequent episodes rather than longer episodes. Bilateral OME is more likely to persist than unilateral OME. Therefore, most effort should be expended on children with bilateral OME. [***] In primary care, children with bilateral OME and a history of upper respiratory infections are more likely to persist. In secondary care, children with bilateral OME seen in the second half of the year (July–December) with a hearing impairment of 30 dB HL in both ears are more likely to persist. [***] None of the factors for persistence singly or in combination are sufficient to negate a requirement for watchful waiting before considering surgery. [***]



DIAGNOSIS The initial diagnosis in most cases of childhood OME will be by otoscopy, the examiner having been alerted to the possibility by ear or hearing problems. The otoscopic appearances of OME are extremely variable and take experience to reliably detect, but can be aided by the use of a pneumatic otoscope. Hence, in many children a confirmatory method would be helpful. In some ears it will not be possible to see the tympanic membrane because of wax and again an alternative way of diagnosing OME without needing to remove the wax is helpful. The investigations available will be those that are used. In primary care this may be tympanometry and occasionally audiometry. In secondary care, both these investigations should be readily available.



Reference gold standard In most diagnostic situations, the reference gold standard as to whether OME is present or not has to be by some other investigation. In addition, the ears that are being used as the controls are important to define. Ears that have recently had an episode of OME are likely to be different from ears that have never had OME. Because of its availability and semi-objective nature, tympanometry is perhaps the most common reference standard used. In children who are being operated on for OME, the surgical finding of middle ear fluid can be the reference standard. Unfortunately, the surgical series has to have an undesirable ‘dry tap’ rate in order to calculate the ability of the investigation to detect ears that do not have fluid. Even the surgical findings are prone to error as the absence of OME may be due to the anaesthetic gases aerating the middle ear before the myringotomy is performed.



History Though this will often initiate the assessment and be an indicator of previous problems, it is not a reliable indicator of the current presence of OME or the degree of hearing impairment. Several level 1 evidence papers support this statement. In particular, in a cohort study of 216 children followed from birth to 27 months of age at three-monthly intervals, the sensitivity of parental report of current OME and hearing impairment was poor.45 The lack of correlation between parent report of hearing and the current pure-tone thresholds has been further illustrated by Stewart et al.46 (Figure 72.6) [****] In this figure there is no correlation whatsoever between parental reports of their child’s hearing and the child’s hearing thresholds. As might be expected if an infant’s parents give a history of ear problems, recurrent upper respiratory infections, mouth breathing and snoring, such a child is more likely to 6 Parental rating of hearing loss



the mother smoked and this was a significant multivariate factor in the secondary care study.15 [***] In all three studies, the magnitude of the effect of any factor singly or in combination was insufficient to predict with certainty those likely to persist or resolve.



5 4 3 2 1 0 −20



0



20 40 60 Threshold in better-hearing ear



80



100



Figure 72.6 Parental perception of hearing loss (from 1 = definitely no to 5 = definitely yes) versus the actual hearing thresholds in the better ear, n = 84 children. Redrawn from Ref. 46, with permission.



Chapter 72 Otitis media with effusion



have recurrent OME in the future than if there is no such history.23 [**] It remains to be investigated whether children with such a history merit monitoring.



Diagnostic tools



] 885



In some ears the retractions may be more localized than others (in Figures 72.8 and 72.12 there is a Sade grade 3 retraction to the promontory). In Figure 72.10, the pars tensa retraction is particularly evident inferiorly. Colour changes can be more yellow (Figure 72.11), more blue (Figure 72.15) or just clear (Figure 72.7).



OTOSCOPY



Unfortunately the otoscopic appearances of OME are extremely varied. Figures 72.7, 72.8, 72.9, 72.10, 72.11, 72.12 and 72.13 are some examples of what can be seen. Not included are ears with residual fluid following acute otitis media which would add additional variety, particularly in abnormalities of thickness and inflammation of the pars tensa. In such ears, the tympanic membrane may also be bulging rather than retracted. The otoscopic findings in OME are mainly different combinations of retraction of the pars tensa and variations in its colour. Retractions may be evident by indrawing of the handle of the malleus (Figures 72.8, 72.9, 72.10, 72.11, 72.12 and 72.13) or the presence of a neo-annular fold (particularly evident in Figure 72.10, but also in Figures 72.7, 72.8, 72.9, 72.11, 72.14, 72.15).



Figure 72.9 Otitis media with effusion (left). Malleus handle markedly retracted.



Figure 72.7 Otitis media with effusion (left). Malleus handle in normal position.



Figure 72.10 Severely retracted position of malleus handle in otitis media with effusion (left). As the retraction develops a neoannular fold may develop (arrow).



Figure 72.8 Otitis media with effusion (left). Malleus handle slightly retracted.



Figure 72.11 Left tympanic membrane in otitis media with effusion showing yellowish colour.



886 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Figure 72.12 Fluid level in otitis media with effusion (left).



Figure 72.15 Left tympanic membrane in otitis media with effusion showing bluish colour.



doubtful. In secondary care, tympanometry and audiometry will usually be practicable and provide sufficient information to allow decisions to be made. Pneumatic otoscopy



Figure 72.13 Air bubbles in otitis media with effusion (left).



Figure 72.14 Left tympanic membrane in otitis media with effusion showing clear colour.



Fluid levels (Figure 72.12) or air bubbles (Figure 72.13) are relatively uncommon. In some children, otoscopy may not be practicable because the view is obscured by wax. This occurs in approximately 10 percent of ears (Trial of Alternative Regimens of Glue Ear Treatment: TARGET, unpublished data). Whether it is necessary to remove the wax is



This can be carried out with a closed system in a handheld otoscope or with a Siegle’s pneumatic speculum, viewed with headlight illumination or microscope. Though acceptable to most children, the inability to gain a seal with the available speculum can occur in up to 20 percent of children aged over 18 months.47 [***] If the tympanic membrane is seen to move, the lack of a seal is irrelevant as the middle ear is obviously aerated. The problem arises when no movement is seen and air escape is evident. Most reports of the sensitivity and specificity of pneumatic otoscopy do no include such children, nor do they include children in whom the tympanic membrane could not be visualized. Simulated otoscopy using static videos has been compared with simulated pneumatic otoscopy using dynamic videos to compare their accuracy.48 [****] The ‘true’ middle ear status was determined in ‘most ears by tympanometry and myringotomy data’. Overall, there was a relative improvement of 26 percent in the overall score of trainees and consultants with the addition of pneumatic otoscopy. This improvement was greater for the consultants than for the trainees. Whether the results of this simulated study can be generalized to the clinical situation remains to be reported. In practice, many clinicians use pneumatic otoscopy when they have doubts as to whether there is OME rather than routinely. Studies that compare pneumatic otoscopy with tympanometry are available, but studies that compare pneumatic otoscopy with the findings at surgery are chosen as the type of study to report because they are a better reference standard (Table 72.4). The sensitivity of pneumatic otoscopy caried out by trained specialists ranges from 85 to 93 percent and its specificity from 71 to 89 percent. These calculations omit children in whom pneumatic otoscopy was not possible (up to 20 percent of



Chapter 72 Otitis media with effusion Table 72.4 Sensitivity (Sens), specificity (Spec), positive predictive (PPV) and negative predictive values (NPV) of otoscopy in the detection of OME with surgical findings as the reference standard. Author



Toner and Main49 Finitzo et al.50 Vaughan-Jones and Mills51 Nozza et al.52



OME (%)



Pneumatic otoscopy Sens



PPV



Spec



NPV



56 70 68



87 93 90



84 84 88



89 58 75



84 78 78



55



85



78



71



79



Ears in which the tympanic membrane could not be visualized or a seal obtained are not included.



children). If they were to be included, the sensitivities and specificities would alter dramatically. The sensitivity and specificity of pneumatic otoscopy, even where practicable, would be markedly poorer in practitioners whose skills have not been validated.53 Over the years, American clinical practice guidelines have strongly advocated the use of pneumatic otoscopy as the primary diagnostic method for OME.54, 55, 56 It is certainly to be recommended to those in primary care where the examiners are perhaps less experienced than specialists who also will have access to additional methods of assessment, including tympanometry and audiometry. Video otoscopy Video recordings of otoscopy, including pneumatic otoscopy, can be documented and used to monitor changes with time. It can also be used for teaching and research purposes.



] 887



assessed in secondary care using otoscopy, pure-tone audiometry or surgical findings as the reference standard. Though an acoustic seal is sometimes difficult to achieve, bilateral tympanograms should be obtainable in the majority (98 percent) of children between the ages of 3.5 and 7 years,58 in a slightly lower percentage (90–94 percent) of infants 2–11 months of age59, 60 and 78–88 percent of infants 12–24 months of age are included.60 Tympanograms can be classified in multiple ways, the simplest being peaked/no-peaked, usually with the additional subclassification of peaked tympanograms depending on the pressure at which the peak is recorded. Numerical evaluation of various parameters61, 62 has also been advocated. These have the scientific advantages of being able to be analyzed in multiple different ways to produce ROC (receiver operating characteristic) curves. They have the disadvantage of being more difficult to apply in comparison to visually scanning the tympanogram for a peak and noting the pressure at that peak. Peaked versus flat (non-peaked) tympanograms The classification most commonly used is that of Jerger63 modified by Zielhaus et al.21 which uses compliance and pressure as the numerical parameters. The classification can be presented in many ways; visually as examples (e.g. Orchik et al.64), in a summary figure as in Figure 72.16 or in Table 72.5. Analyzing papers with the findings at myringotomy (soon after tympanometry) as the reference ‘gold’ standard, suggest that a type B tympanogram is frequently associated with OME, a type A is infrequently associated with OME and a type C falls in between (Table 72.6). This study64 is particularly informative because



Compliance (ml)



FREE-FIELD VOICE TESTING



In primary care, audiometry is seldom available, particularly for younger children. Under these circumstances, the practitioner could perform free-field voice testing of hearing. Though not reported in a primary care setting, in those over the age of three years the sensitivity of modifications of the voice test (e.g. use of pictures to point at or two-syllable words) is 80–96 percent and the specificity is 90–98 percent.57 [****] Unfortunately, these satisfactory levels are unlikely to be achieved when carried out by less experienced examiners in a primary-care setting.



Type C2



Type C1



Type A



0.2 Type B −400



−200



−100



0



+200



Pressure (mm H2O)



Figure 72.16 Classification of tympanogram types according to Jerger. Modified from Ref. 63, with permission. Table 72.5



Classification of tympanograms. Type



Description



A C1 C2 B



Between 1 200 and –99 daPa Between –100 and –199 daPa Between –200 and –399 daPa No observable peak between 1 200 and –600 daPa



TYMPANOMETRY



Tympanometry with an impedance meter (see Chapter 232, Psychoacoustic audiometry) has been advocated since the 1970s as a reliable method of detecting OME. Since then, there have been many publications on its use as a screening instrument using otoscopy as the reference standard. Its role as a confirmatory test has also been



Peaked



Non-peaked



Reprinted from Ref. 63, with permission.



888 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Table 72.6 Relationship between tympanogram type and effusion at myringotomy. Tympanogram type A B C



% Effusion None



Minimal



Moderate



Impacted



27 0 18



36 12 35



15 13 24



22 75 23



Reprinted from Ref. 64, with permission.



the amount of middle ear fluid is quantified. Unfortunately, in this study type C tympanograms were not subcategorized as type C1 and C2. Taking a variety of level 1 evidence studies (Table 72.7), a type B tympanogram compared with all other types of tympanogram has a sensitivity of between 56 and 73 percent and a specificity of between 50 and 98 percent in detecting OME confirmed surgically. Compared with all other types, a type B plus C2 has a sensitivity of between 73 and 99 percent and a specificity of between 40 and 74 percent (Table 72.8). [****] Why there is such a range of values can only be partly explained by the differing proportion of ears with OME and the fact that the anaesthetic can itself aerate the middle ear giving a ‘false’ dry tap. Pittsburgh algorithm Table 72.962 [****] shows the sensitivity and specificity of various tympanometric measures, including the gradient, which is taken alongside acoustic reflexes (see Acoustic reflexes) appears to give good positive and negative predictive values. Unfortunately, no comparison of this algorithm is made with the easier to apply peak/no-peak distinction. Tympanoscopes Tympanometry can also be performed with a hand-held otoscope that automatically prints out the results. Compared with the more complicated, scientific impedance bridge, these have the advantage of portability and lower cost. Hence, they are commonly advocated for use in primary care. In such a setting and with training,66, 67 if a type B tympanogram is classified with a type C2 tympanogram, the sensitivity and specificity (100 and 75 percent) are comparable to conventional methods.68, 69 Using a type B tympanogram alone to diagnose OME gave poor sensitivity and specificity and is not recommended.



Table 72.7 Sensitivity (Sens), specificity (Spec), positive (PPV) and negative predictive values (NPV) of a Type B tympanogram versus Type A and C in the detection of OME with surgical findings as the reference standard. Author



Orchik et al.64 Toner and Mains 199049 Finitzo et al. 199250 Vaughan-Jones and Mills51 Nozza et al.52 Sassen et al.65 Palmu et al.59



OME (%)



Type B alone Sens



PPV



Spec



NPV



59 56



55 86



88 94



90 93



58 84



70 68



57 67



95 96



94 94



47 58



55 73 74



85 90 70



78 83 93



71 50 98



79 66 94



Table 72.8 Sensitivity (Sens), specificity (Spec), positive (PPV) and negative predictive values (NPV) of a Type B 1 C2 tympanogram versus Type A 1 C in the detection of OME with surgical findings as the reference standard. Author



Finitzo et al. 199250 Vaughan-Jones and Mills51 Sassen et al.65



OME (%)



Type B and C2 Sens



PPV



Spec



NPV



70 68



73 89



87 75



73 40



53 67



73



99



91



74



97



combined sensitivity is 98 percent. When otoscopy suggests no OME and is combined with a type A tympanogram, the specificity is 98 percent. However, in this study the findings were in agreement in only 44 percent of ears. In the other 66 percent, where the results are not in agreement, another indicator of OME such as audiometry would be required.



ACOUSTIC REFLEXES



Two papers70, 71 [****] showed that the results of measuring the acoustic reflexes in association with classification of tympanometry as peaked/non-peaked substantially lowered its specificity whilst adding little to the sensitivity. Such testing is therefore not advocated if this classification is used. On the other hand, if the Pittsburgh algorithm is used it would appear to add value.70 [****]



TYMPANOMETRY IN COMBINATION WITH OTOSCOPY



The results of otoscopy can either be in agreement or disagreement with tympanometry. If in agreement, their diagnostic value can be assessed against the findings at myringotomy.50 [****] Thus, when otoscopy suggests OME and is associated with a type B tympanogram, the



EUSTACHIAN TUBE FUNCTION TESTS



Reliably assessing Eustachian tube function requires a method that does not depend on a child’s ability to perform a task. It also has to be applicable when the



Chapter 72 Otitis media with effusion



] 889



Table 72.9 Sensitivity, specificity and predictive values of various criteria using acoustic emittance measures for the diagnosis of MEE, as determined at myringotomy, in ears with a history of chronic or recurrent otitis media (n = 111). Variables Gr 1 AR Ytm 1 AR AR alone GR alone Ytm alone



Gr and Ytm



Criterion



Sens (%)



Spec (%)



PPV (%)



NPV (%)



Gro0.1, or Gr = 0.2, or 0.3 with AR absent Gr r 0.1, or Gr = 0.2 with AR absent Ytm r 0.2, or Ytm = 0.3, 0.4, or 0.5 with AR absent Ytmo0.3, or Ytm = 0.4 or 0.5 with AR absent AR absent Gr r 0.1 Gr r 0.2 Ytm r 0.1 Ytm r 0.2 Ytm r 0.3 Ytm r 0.4 Gr r 0.1 or Ytm r 0.2



90 88 83 85 88 78 91 31 56 73 82 83



86 86 86 76 85 90 70 97 93 80 63 87



95 95 94 90 94 96 89 96 96 91 86 94



75 71 65 65 72 60 75 34 44 52 56 65



Gr, gradient; AR, Acoustic reflex; Ytm, peak compensated admittance; NPV, negative predictive value; PPV, positive predictive value. Sens, sensitivity; Spec. specificity. Reprinted from Ref. 52, with permission.



tympanic membrane is intact rather than just when a ventilating tube is in place. Sonotubometry is such a method, but needs to be improved further for it to be clinically applicable.72 ACOUSTIC REFLECTOMETRY



This is performed with a hand-held acoustic otoscope that does not require a seal. It is moved around whilst emitting a sound of 80 dB A over a frequency range. The amount of sound reflected back is recorded and analyzed in various ways. Unfortunately, the sensitivity and specificity of the test is poorer than tympanometry when pneumatic otoscopy or the surgical findings are used as the reference standards.73, 74, 75 [****]



AUDIOMETRY



In secondary care, audiometric assessment of the hearing is mandatory in all children referred with a suspected persistent hearing impairment, irrespective as to whether OME is diagnosed at the time. If OME is diagnosed then the laterality and severity of the impairment will dictate management. If there is no OME and if the air-conduction thresholds are elevated, bone-conduction thresholds will detect the 1 percent of children with previously undiagnosed congenital, sensorineural impairments.76 [**] The method used to assess the hearing thresholds depends on the equipment available, the age of the child and the assessor’s preference.77 In essence, it can be performed binaurally free-field or monaurally with headphones. Tones or speech can be presented and the method of response can be by turning to the sound source or by performing a task, which can range from doing something with a toy, to raising a hand or pointing at an object or picture. Conventional air- and bone-conduction



testing is the preferred method as it tests each ear separately and can quantify any conductive component present. In children over 3.5 years of age, air- and boneconduction thresholds are obtainable in the majority of children (95 and 92 percent, respectively) on the first occasion.78 [**] Age is the main determinant of whether thresholds are obtainable, rather than the method of audiometry. Even if the child’s concentration is poor, the results on average are only 5 dB poorer and can still be used to guide management.78 [**] Air–bone gap The presence of an air–bone gap of at least 10 dB is a poor predictor of concurrent OME. Of children referred with a history of OME, 80 percent (1234 of 1551 children) will have an air–bone gap of 410 dB, but only 32 percent will have otoscopic evidence of OME associated with a type B or C2 tympanogram (MRC (Medical Research Council) Multicentre Otitis Media Study Group, unpublished data). Using findings at myringotomy as the reference standard, 37 percent (28 of 75) of ears with an air–bone gap of greater than 10 dB had a dry tap. It was only when the air–bone gap was greater than 30 dB was the dry tap rate reduced to 4 percent (1 of 27).79 [***] The most likely reason for the air–bone gap being such a poor predictor is that children studied prospectively with OME will have a small residual conductive impairment after otoscopic and tympanometric resolution of their OME (personal communication). [***] Carhart notch in the bone-conduction thresholds As with any pure conductive impairment, one would expect OME to be associated with a Carhart notch in the bone-conduction thresholds around 2 kHz. By definition, the notch has to be 10 dB or greater between 0.5 and 4 kHz as any lesser dips could be due to test/retest error.



890 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY As such, accurate bone-conduction thresholds are required and this may be problematic in younger children. Notches are reported to be present in 48 percent of children with OME whose mean age was nine years.80 [**] In those able to perform the audiometry, the positive predictive value of a Carhart notch in diagnosing fluid at myringotomy was 97 percent and the negative predictive value was 87 percent.79 [**]



Ideal diagnostic strategy for OME The ideal strategy for the diagnosis of childhood OME should be applicable to the majority of children being assessed. This rules out myringotomy and means that it has to be based on any combination of otoscopy, tympanometry and audiometry. The first thing to consider is the percentage of children in whom one of these investigations is not possible. In the TARGET study (unpublished data), at the first visit to a specialist clinic of children with suspected OME aged between 3.25 and 7 years, it was not possible in both ears to obtain tympanometry in 1.6 percent, air-conduction thresholds in 3 percent, bone-conduction thresholds in 6 percent and otoscopy in 9 percent of children. The ability to assess the hearing thresholds was age related, ranging from 12 percent in three year olds to 11 percent in seven year olds. The inability to perform otoscopy in such a high proportion of children is partly because aural toilet to visualize the tympanic membrane was not enforced, but this is probably the usual practice of most clinicians for children in this age group. Otoscopy should be attempted in all children not just to diagnose OME, but also acute and chronic otitis media. Pneumatic otoscopy can be of added value where there is uncertainty. In those with OME on otoscopy, audiometry is required to help determine management. Thus, in such children, tympanometry is probably only a confirmatory investigation. In those in whom the tympanic membrane cannot be visualized or the findings are uncertain, then tympanometry is essential. In those with a type B or C2 tympanogram, audiometry will detect those with a bilateral hearing impairment of 20 dB or greater that merit follow up. [Grade D] Unfortunately, a review article of the methods of diagnosing OME81 does not address the question of whether the diagnostic method was practicable and what difference that would make to the comparisons they make. If they were to have done so, they might have proposed the above strategy rather than the use of pneumatic otoscopy.



Assessment of hearing This is important to assess not just for the diagnostic reasons stated above, but in secondary care the current hearing thresholds are the main determinant as to management.



HISTORY



Unfortunately, reliance cannot be placed on parental report of the current hearing, though it may reflect the hearing in the past (see under Diagnosis; and History above). AUDIOMETRY



Routine audiometric testing of the hearing of every child with OME seen at secondary care is recommended, as the associated hearing impairment can vary enormously from negligible to moderately severe (see Outcomes of childhood OME below). TYMPANOMETRY AS A SCREEN FOR AUDIOMETRY



Having excluded a sensorineural impairment by audiometry at a child’s first visit, at subsequent visits tympanometry can be used as a screen to identify those most likely to have a material hearing impairment associated with their OME. This was first proposed by Dempster and McKenzie.82 [****] Since then, others have examined this possibility in more detail using different audiometric cut-offs.58, 71 [****] The results are not materially different between papers, but the later publication58 has analyzed a greater variety of outcomes and cut-offs (Table 72.10). Thus, for example, taking 25 dB HL in the better ear as the level requiring detection, limiting audiometry to those with a bilateral type B reduces the workload by 50 percent, but will detect 90 percent of such impaired children. Limiting audiometry to those with bilateral type B or C2 tympanograms reduces the workload to 69 percent of the sample, yet 95 percent of the impaired children will be detected. Tympanometry can be particularly helpful if resources are limited in choosing children for more timeconsuming audiometry. This should only be done once a coexisting sensorineural impairment has been excluded. [****] OTHER CLINICAL EXAMINATIONS



What other examination is undertaken depends on whether there are any other nonotological symptoms, such as recurrent sore throats, blocked nose and snoring. POPULATION SCREENING FOR OME



As OME is highly prevalent in the childhood population and is frequently asymptomatic, some have suggested that universal screening so that early treatment could be implemented. A Cochrane review83 identified three trials where a population was screened and those with OME were then randomized to ventilation tubes or watchful waiting. No benefit was identified on language development or behaviour, the outcomes that one might hope



Chapter 72 Otitis media with effusion



would benefit from early intervention. Population screening is thus not considered to be efficacious at present. [Grade A]



OUTCOMES OF CHILDHOOD OME Natural history As discussed earlier (see above under Epidemiology), the majority of epidemiological studies are point prevalence studies of a cohort of children at different times rather than longitudinal studies of specific children over time. There are considerably fewer data on the natural history in individual children available, this being the information most relevant to clinical practice. As children with bilateral OME are more likely to have more severe outcomes, studies that report such children are those that are concentrated upon. Hogan et al.43 [***] found that in children under three years of age with a propensity to have bilateral OME, the effusions lasted on average ten weeks, albeit the children were likely to have a further episode within 11 weeks. Older children between the ages of 3.25 and seven years have been studied extensively in the UK MRC-funded



] 891



TARGET study. A total of 3831 children of this age were referred from primary care because of suspected OME. When screened at the trial clinic, only 34 percent (Table 72.11) satisfied the criterion for further study of having bilateral OME associated with a hearing impairment of at least 20 dB HL in both ears.14, 16, 17 [***] This cohort underwent a 12-week watchful waiting period, following which 49 percent of children no longer met the bilateral OME hearing criterion. That is, only 51 percent persisted. At that stage, children with persistent OME were randomized and in those who had nonspecific medical management, three months later only 49 percent were persistent.84 [****] It was only three months later that a group of children was identified that could be truly considered to have persistent bilateral OME with a hearing. To summarize the above data, of the children aged 3.25–7 years referred to secondary care with bilateral OME, if they waited 13 weeks to be seen, underwent a watchful waiting period of three months and then waited a further three months for surgery, by then in 92 percent of them the hearing impairment associated with their OME will have resolved (Table 72.11). Thus, even in secondary care, the vast majority of children with bilateral OME resolve spontaneously over a nine-month period, albeit the OME might recur.



Table 72.10 Sensitivity, specificity and positive predictive value (PPV) as percentages for tympanometry in all children in seven centres. % of sample Type B 1 B tymp versus others



Better ear Better ear Better ear Better ear ABG Z10 ABG Z15 ABG Z20 ABG Z25 ABG Z30



HL HL HL HL



Z15 Z20 Z25 Z30



76 53 34 23 88 75 59 44 32



Type B 1 B or B 1 C2 tymp versus others



Type B 1 B or B 1 C2 or C2 1 C2 tymp versus others



Sensitivity Specificity PPV



Sensitivity Specificity PPV



Sensitivity Specificity PPV



65 81 90 92 58 65 75 83 89



80 90 94 94 72 80 86 91 94



84 92 95 94 77 84 89 93 93



91 81 69 60 92 87 80 72 66



96 82 60 41 98 94 84 70 55



80 62 49 43 84 77 64 54 48



93 72 49 33 97 90 77 61 46



n = 1153. ABG, air–bone gap. Reprinted from Ref. 58, with permission.



Table 72.11 Natural resolution of OME in children (n = 3831) referred to otolaryngology from primary care (GP). Persistence is defined as bilateral OME associated with hearing Z20 dB HL in both ears. Wait



GP to otorhinolaryngologist Otorhinolaryngologist watchful wait Nonsurgical cases Nonsurgical cases



Wait time (months)



Total wait



% persistence



% persistence of original cohort



3 3



3 6



34 51



34 17



3 3



9 12



49 89



8 7



Reprinted from Ref. 14, with permission.



75 54 43 37 79 72 57 48 42



92 69 46 31 96 90 75 58 43



892 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY As might be expected if a less strict definition of what constitutes persistence is taken, such as not having an associated requirement for the degree of hearing impairment, the resolution rates by ear of chronic OME are materially lower; 19, 25, 31 and 33 percent at 3, 6, 12 and 24 months, respectively.85 [****]



Outcomes OTOPATHOLOGY



To answer the question, ‘does OME cause pars flaccida and tensa retractions that can progress to a cholesteatoma?’ would require a cohort of children with documented OME to be followed up and compared with a cohort that do not have OME. Ideally the OME cohort would have no effective therapy for their OME, including the insertion of ventilation tubes as these themselves might cause pathology. The best alternative evidence is from a large cohort (n = 964) of children who were screened for OME every three months between the ages of two and four years and were otoscopically reassessed when they were seven or eight years of age.86 [***] The children were classified as:  never OME (never a type B tympanogram);  persistent OME/no surgery (type B tympanogram on at least two consecutive visits);  transient OME/no surgery (type B tympanograms that were not consecutive);  OME surgery (those with OME that had ventilation tubes inserted; 60 percent had only one tube inserted). Table 72.12 shows the otoscopic sequel in these four groups of children. What one is primarily looking for is a dose–response effect, albeit the OME surgery group would be the most severe group and include any additive effect of ventilation tubes. There is a clear dose–response effect for attic retractions. This is, therefore, a disease effect. There is also a dose–response effect for tympanosclerosis and atrophy of the pars tensa, but with a particularly large increase in the ventilation tube group.



Hence, these are most likely to be disease effects, but with an increased risk following the insertion of ventilation tubes. No dose–response effect is apparent for retractions of the pars tensa. The high incidence (14 percent) of this in the OME surgery group can therefore be mainly attributed to the ventilation tubes. It is concluded [***] that attic retractions are an OME disease effect with an incidence of approximately 14 percent in the most severe children. What proportion of these retractions progress to cholesteatoma is not known. There is no evidence that the risk of attic retraction is altered by the insertion of ventilation tubes. Tympanosclerosis and atrophy of the pars tensa are OME disease effects, the risk of which is increased by the insertion of ventilation tubes to 445 percent and around 70 percent, respectively. Long-term pars tensa retractions are more likely caused by the insertion of ventilation tubes than by the disease process itself with an incidence of 15 percent. CONDUCTIVE HEARING IMPAIRMENT



The hearing impairment associated with OME varies greatly. Table 72.13 shows the percentage of OME children aged between 2 and 12 years with confirmed, bilateral OME who have various pure-tone air-conduction averages in their better hearing ear. Overall, the mean threshold in the better ear was 21 dB HL (s.d. 10), but in 54 percent of those with bilateral OME the pure-tone average in the better ear was better than 20 dB HL.87 [***] The mean threshold in the poorer hearing ears was 31 dB HL (s.d. 13). The speech reception or speech awareness thresholds in infants in this study were of a similar distribution. This suggests that the majority of children with bilateral OME have an insufficient hearing impairment to be materially disabling. This paper also reports a group of children with unilateral OME. In them the mean threshold in the better ear was 11 dB HL (s.d. 7) and 23 dB HL (s.d. 10) in the poorer ear. Coincidentally, ‘normal’ children in this age group on average do not have ‘normal’ thresholds, the mean



Table 72.12 Otopathologic sequelae at seven to eight years of age in four groups of right ears. OME persistence groups



No OME (reference group) (n = 264) Transient OME (n = 251) Persistent OME (n = 219) Ventilation tube (n = 64)



Tympanosclerosis %



Atrophy %



Retractions pars tensa % Attic retraction % stage Z2 stage Z2



0 2.4a 3.7c 47.6d



3.4 4.4 8.8b 68.2d



1.9 2.0 1.9 14.3b



0.4 4.4 7.4 14.3b



Ears classified according to the persistence of OME at preschool age and surgical treatment (96 ears with signs of active middle ear disease, perforation or ventilation tube still present excluded). p-value associated with chi-square test or Fisher’s exact test for the transient OME, persistent OME and ventilation tube groups as related to the no-OME group (i.e. reference group). a o0.05; bo0.01; co0.005; do0.001. Reprinted from Ref. 86, with permission.



Chapter 72 Otitis media with effusion Table 72.13 Pure-tone average air-conduction thresholds in the better hearing ear in 385 children aged 2–11 years with bilateral OME confirmed by a diagnostic algorithm, including pneumatic otoscopy, tympanometry and reflexes. PTA dB HL 0–4 5–9 10–14 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54



Percentage 4 10 20 20 14 13 6 4 1 2 1



Reprinted from Ref. 87, with permission.



air-conduction thresholds in three- to six-year-old children being 3 dB HL.88, 89 [***] Though it might be expected that the viscosity of the middle ear fluid would be a determinant of the hearing, it would not appear to be so. In particular, there is no difference in the hearing levels in ears with serous fluid compared with those with mucoid fluid.90 [***] HEARING DISABILITY SECONDARY TO CONDUCTIVE IMPAIRMENT



The main method of assessing a child’s hearing disability is by questioning the parents. Four of the nine questions in the MRC Reported Hearing Disability (RHD) questionnaire are sufficient to predict the overall disability (Table 72.14) (personal communication). [***] Many hold that audiometric measures such as speechin-noise would be a superior way to measure disability rather than using the pure-tone thresholds. If this were to be the case, such a measure in a specific child might more likely reflect the detrimental effect of OME and predict their likely benefit of management. Research in this area is ongoing.



] 893



Table 72.14 Four questions that predict a child’s hearing disability associated with OME. Question How would you describe your child’s hearing?



Response



Normal Slightly below normal Poor Very poor Not sure Missing Has he/she misheard words when No not looking at you? Rarely Often Always Not sure Missing Has he/she had difficulty hearing No when with a group of people? Rarely Often Always Not sure Missing Has he asked for things to be No repeated? Rarely Often Always Not sure Missing



Weighting 0 0.4 0.6 0.6 0 0 0 0.1 0.6 0.6 0 0 0 0.1 0.5 0.6 0.1 0 0 0.1 0.5 0.6 0 0



Unpublished data.



In a cohort followed from birth to 18 years of age, those who had had otitis media (acute and with effusion) had air-conduction thresholds that were 4 dB on average poorer that those that had not had otitis media.91 [***] Their bone-conduction thresholds were also 2 dB poorer. Whether these poorer thresholds are a disease effect or the result of surgery has yet to be elucidated. High frequency sensorineural damage above 8 kHz has also been reported in a case series of children with OME who had been operated on and followed up for between three and five years.92 [**]



Sensorineural hearing impairment There is no evidence to suggest that OME is associated with a sensorineural hearing impairment in the short term.78 [***] Long-term hearing impairment There is now considerable evidence that following the otoscopic and tympanometric resolution of fluid, ears that have had OME still have a small residual conductive impairment in the order of 10 dB HL (personal communication). [****] The reason for this is not known, but is likely to be due to ossicular chain immobility caused by some residual fluid or mucosal oedema around them.



SPEECH AND LANGUAGE



The development of normal speech and language in a child depends on many interrelated factors, the age of the child being the major factor, but the ethnic background and the degree of interfamily communication, particularly by the mother, are important. Any hearing impairment a child might have due to their OME obviously interplays with these factors. Assessment of speech and language is complicated by it having many components, including speech reception, speech and sound production, expressive language and cognitive understanding. When a parent expresses



894 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY concern regarding their child’s speech and language, this can be partly investigated by comparing the child’s performance with standard milestones. Thus, at 18 months they would be expected to have a vocabulary of ten words with meaning. More formal objective assessments can be undertaken for each of the components of speech and language and related to the results in normal age-matched controls. Total bilateral loss of hearing during the first few years of life will obviously have a profound effect on speech and development. Whether the usually intermittent, mild to moderate conductive impairment that can be associated with otitis media with effusion in the first two to three years of life makes an impact has been extensively studied.93 [****] Prospective cohort studies do indeed suggest that there is an impact of otitis media that is related to the number of days with bilateral fluid during the first two to three years of life. However, follow up of these children when they are seven to eight years of age suggests that by then the children have caught up with their nonaffected peers.94 [***] Thus, overall, the data would suggest at most a mild effect on speech production and reception in early childhood,93, 95 that is not modified by surgical intervention (see Surgery) and spontaneously resolves with age. COGNITION



The factors that interplay and influence a child’s intellectual development are even more complex than those that affect speech and language development. Many objective measures are available with age-related ‘normal’ values, but whether these are affected by a child having had OME is difficult to determine. Cohorts of children can be grouped as to how much OME they have had and this is correlated to a particular outcome, having controlled for demographic and environmental factors. Any effect of OME that has been identified occurs early in a child’s academic development, around the ages of three to four years, and has been minor compared with the demographic and environmental ones. Any effect corrects itself with time and is no longer detectable once the child is seven or eight years of age.96, 97, 98 [***] BEHAVIOUR



Abnormalities of behaviour can be reported in questionnaires by the parents or by the teachers. These can be analyzed to give a global behaviour score or be broken down into various components of behaviour. These vary between questionnaires. Thus, the Rutter score99 can be broken down into antisocial, neurotic, hyperactive and inattentive behaviour. The MRC Behaviour Questionnaire can be broken down into anxiety, aggressive, contextrelated and social immaturity components. Many factors influence a child’s behaviour which have to be controlled for before calculating the effect of the hearing impairment associated with OME.



Whether this is the case has been best reported in OME children being randomized in randomized controlled trials (RCT). Fifty-five percent of three-year-old children with bilateral, three-month persistent OME have abnormal overall Rutter scores. This compares with 10–15 percent of unselected three year olds.100 [***] The overall behaviour scores in children aged between three and seven years with three-month persistent bilateral OME, associated with a hearing impairment of at least 20 dB HL in both ears, is 0.6 s.d. poorer (i.e. moderately) than the population scores (MRC Otitis Media Study Group, unpublished data). There is some evidence that even when a child has reached the age of 15 years, they have still poorer behaviour than non-OME children, particularly in the inattentive and hyperactive aspects.101 [***] BALANCE



Around 30 percent of parents of children with OME report that they are clumsy, often imbalanced and can fall. These symptoms would appear to be more common than in non-OME children.102 [***] The pathophysiology as to why OME might cause such symptoms is difficult to understand. However, laboratory assessments of balance suggest that these are abnormal in children with a history of OME,103 [***] but the clinical significance and relation to symptoms is unclear. It is highly likely that if they are due to vestibular dysfunction they will compensate naturally with time. There are reports that laboratory assessments improve after surgery, but these observations are uncontrolled.104 [**] Improvement in such tests is also likely to occur with practice and is an alternative reason for the improvements reported. It must be concluded that at present there is insufficient knowledge and understanding to guide the assessment and management of balance symptoms in childhood OME. QUALITY OF LIFE



Because there are so many potential outcomes that might be attributed to a child having a mild to moderate hearing impairment due to OME, it would be helpful if an overall OME disease score could be calculated. Two such diseasespecific ‘otitis media’ quality of life questionnaires are available. The OME-6 covers physical suffering, hearing loss, speech impairment, emotional distress, activity limitations and caregiver concern and is applicable to two-month to 12-year-old children with OME or recurrent AOM. This has been shown to improve in children who have had ventilation tubes inserted, but because of the lack of a control group it is not possible to conclude that surgery was beneficial.105, 106 [**] There is also some concern that the OME-6 does not relate to the severity of the hearing impairment that might limit its usefulness.107 A longer questionnaire, the OM7-27, was developed to summarize the outcomes in the MRC TARGET study.108



Chapter 72 Otitis media with effusion



The 27 questions cover the seven domains of hearing difficulty, upper respiratory symptoms, sleep patterns, behaviour problems, ENT-related parental quality of life, global physical health and speech/language impairment. It is applicable to children between the ages of three and seven years with OME. The effect of surgery in comparison to nonsurgical management awaits reporting. Both these questionnaires are potentially useful for research studies, but cannot be described as true qualityof-life instruments. This requires a generic questionnaire that can be used to compare outcomes for different conditions. Many such instruments are available for adults (e.g. the SF-36), but those for children are still in the development stage. A generic Dutch quality-of-life instrument for preschool children (TAPQOL) has been used in a surgical RCT (see Quality of life, under Surgery below) and the communication, positive mood, sleep and aggression domains in this questionnaire relate well to the OME-6 and the Health Utilities Index mark 111.109 To assess the benefit from interventions, the Glasgow Children’s Benefit Inventory has been shown to reflect the parental satisfaction of ventilation tube insertion and allows a comparison to be made between different interventions for different paediatric otolaryngological conditions.110



MANAGEMENT Counselling and hearing tactics Parents of children with OME are often misinformed about many aspects of the condition. In particular, they can have overpessimistic views about its severity and be overoptimistic about the merits of surgery. At their first visit to their general practitioner or to a specialist, time has to be spent explaining that, in general, OME is a benign condition with a high spontaneous recovery rate and no long-term sequelae. In most children, the main concern will be the hearing. It should be explained to the parents that the impairment associated with OME is very variable in degree and mild or moderate at most. It is important that all parents of children with OME receive appropriate general counselling about the natural history and relative benign nature of the condition. Those in contact with the child, including any minder or teacher, should be made aware that the disability can be minimized by hearing tactics, including those shown below:  getting the child’s attention before starting to talk;  reducing the background noise as much as possible by turning off the television, etc.;  facing the child so that they can see you talk;  speaking in a normal voice both in volume, speed and emphasis, as close as possible to the child;  leaflets such as Glue ear explained111 given to the parents can reinforce the above messages.



] 895



Medical Medical management would potentially be of greatest benefit if it could speed the resolution of an episode of OME. Hence, randomized controlled trials carried out in primary care settings would be those most appropriate to consider using resolution of OME as the outcome. Such studies are uncommon and often include heterogeneous groups of children with otitis media rather than OME. The effect of most medications has been systematically reviewed, but the advisability of performing a metanalysis on the data must be questioned because of the heterogeneity of the studies. Most trials follow up children for one to two weeks after therapy. If at this point the therapy is ineffective, there is no reason for further follow up as it is unlikely to be of benefit thereafter. However, if it is effective after one to two weeks, then follow up for the recommended watchful waiting period of 12 weeks is necessary to see if it is of benefit in the longer term and might be used to reduce the proportion of children being considered for surgery. NASAL TOPICAL STEROIDS



Systematic reviews112 [****] of the randomized controlled trials in August 2002 identified one study of topical nasal steroids versus placebo (n = 44) that found no difference in resolution of the OME at three weeks.113 [****] Another study which gave antibiotics in addition to nasal steroids also found no difference.114 [****] Both studies randomized small numbers and no subsequent trials have been identified. It must be concluded that there is insufficient evidence to support the use of topical nasal steroids at present for childhood OME.



SYSTEMIC STEROIDS



Considerable concern has been expressed about the use of oral steroids being given to children for a nonlifethreatening and spontaneously resolving condition. Consequently, it would have to be highly effective in the long term before it could be recommended. There is no evidence to suggest that oral steroids are effective for longer than in the short term (two weeks) even when combined with antibiotics.112, 115, 116, 117 [****] Systemic steroids cannot be recommended at present for childhood OME.



ANTIBIOTICS



Multiple randomized controlled trials of antibiotics have been published in heterogeneous groups of patients. Two metanalyses,118, 119 [****] covering the years 1966–1991 and 1993 respectively, reported that though there might be initial benefit in the first two weeks, there was no evidence of benefit in the longer term (46 weeks). Later literature reviews up to 1999120 [****] and 2006115 [****]



896 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY confirmed that antibiotics had no significant effect on longer-term outcomes. Hence, it is not recommended that antibiotics be used for the longer term (46 weeks) management of childhood OME. NASAL DECONGESTANTS



A meta-analysis of four trials (n = 1202) found that antihistamine/decongestants had no significant effect on the resolution rate of OME.115, 121 [****] Nasal decongestants are not recommended for use in childhood OME. [Grade B]



MUCOLYTICS



A systematic review of six randomized controlled trials of S-carboxymethylcysteine (Mucodyne) published before 1993 could not demonstrate that they had an effect.122 [****] A later trial also showed no significant effect.123 [****] Because of a nonsignificant trend for increased benefit in these small, biased trials, a larger well-designed study is merited.



Other approaches AUTOINFLATION



A systematic review of autoinflation for the treatment of glue ear in children124 identified five randomized controlled trials and considered them to be of variable and low quality. Though a metanalysis of the three studies that investigated Otovent balloons indicated that children allocated to autoinflation were 3.5 times more likely to improve than controls (CI 2.0, 6.1), the authors could not recommend autoinflation for clinical practice. [****] The ability to autoinflate with the balloon is a particular problem in younger children. If some form of nonsurgical management is required during a period of watchful waiting, autoinflation is the therapy with the strongest evidence of efficacy for older children.115 [****] HOMEOPATHY



No randomized controlled trials have been identified. A small, nonblinded study did not show homeopathy to be of benefit.125 [**]



VENTILATION TUBES



Ventilation tubes can be of different materials (teflon, silicone, titanium, gold, etc.) and be coated with materials such as silver oxide. Their shapes vary but can be categorized as grommets or T-tubes. In general, the larger and stiffer the flange that goes in the middle ear, the longer it stays in situ. The longer a tube stays in situ the longer it can be potentially of benefit. On the other hand, the longer a tube is in situ the greater the chance of complications, including infection, granulation tissue, permanent perforation and thinning of the tympanic membrane with possible retraction.127 [****] The object is then to try and estimate the likely duration of the OME and choose the appropriate tube. Unfortunately, prediction in children of the likely duration of the OME is difficult if not impossible. However, in adults T-tubes are justified routinely, as in them OME is likely to be persistent over years rather than months. Though some classify ventilation tubes as short, medium or long term, data are lacking on ‘duration of tube function’ as opposed to the more commonly reported and less relevant ‘duration till extrusion’. An additional problem is that in studies, follow-up intervals are never less than every three months and often every six months. In general, the time of extrusion is taken as the midpoint between visits. This means that duration times are fairly crude estimates. Therefore, most authors prefer to report at set time periods postoperatively. Thus, if one of the most commonly reported tubes is taken, at six months post-operation 55 percent of Shepard teflon ventilation tubes are functioning15 [****] and between 30 and 55 percent are extruded.127, 128, 129 [****] Armstrong tubes have similar extrusion rates, but only 10 percent of T-tubes will have been extruded at six months.127 [****] One potential reason for the range between studies in extrusion times of the same tube is likely to be the varying incidence of recurrent acute otitis media in the children being studied, acute otitis media being more frequent in younger children. Another problem is that even though a tube might have a lower extrusion rate at 24 months (66 percent for Reuter Bobbin compared with 94 percent for Shepard for example), the difference is hardly relevant if 74 percent of the Reuter Bobbins are obstructed.127 [****] Thus, at present there is only sufficient evidence to classify ventilation tubes as short or long term. SURGICAL TECHNIQUE



Site



Surgery MYRINGOTOMY AND ASPIRATION



From three trials, myringotomy with aspiration has not been shown to be effective in restoring the hearing levels in children with OME.126 [****]



Insertion of the ventilation tube posterosuperiorly is not recommended because of the potential for damaging the ossicular chain. It makes no difference to the extrusion rate as to whether the tube is inserted through a radial or circumferential incision130 and whether sited anterosuperiorly rather than antero-inferiorly.131, 132, 133 [****] Placement antero-inferiorly compared with placement



Chapter 72 Otitis media with effusion



postero-inferiorly lengthens the time a ventilation tube is in situ (for a Shepard tube 80 versus 45 percent at 6 months and 30 versus 15 percent at 12 months are in situ).131 [****] To maximize the duration of potential tube function, the preferred insertion site is antero-inferior through a circumferential or radial incision. [****] Associated aspiration Though it is accepted practice to aspirate as much of the middle ear fluid as possible through the myringotomy before inserting a ventilation tube, there is no evidence that this is required. The hearing levels three months following insertion of a ventilation tube was no different in ears that were aspirated compared with those that were not aspirated.134, 135 [****] However, no data are given immediately post-operation when the ears that were aspirated may have had better hearing. There is no necessity to go to great lengths to remove all middle ear fluid when inserting a ventilation tube. [***] Topical preparations These can be used after insertion of the tube to prevent tube blockage with blood and infection in the postoperative period. Their efficacy is discussed under Complications of ventilation tubes. OUTCOMES OF VENTILATION TUBES



The most up-to-date Cochrane systematic review of ‘grommets (ventilation tubes) for hearing loss associated with otitis media with effusion in children’136 [****] is based on a systematic literature search to March 2003 with some subsequent papers to August 2004 included. The RCTs of satisfactory quality identified could be divided into those that randomized within a child, one ear to have ventilation tubes and the other ear not to have ventilation tubes (n = 7) and those where children were randomized to have bilateral ventilation tubes or to watchful waiting (n = 6). Within the first by-ear category, in three trials all children had adenoidectomy and the other four the children were randomized to have adenoidectomy or no adenoidectomy. In the second category, in five trials none of the children had adenoidectomy and in the other trial children were also randomized to have adenoidectomy or no adenoidectomy.



Hearing Ventilation tubes alone will improve the hearing level by 9 dB (CI 4, 14) at six months, 6 dB (CI 3, 9) at 12 months and 4 dB (CI 2, 6) at 24 months. Adenoidectomy has an additional effect of 3–4 dB (CI 2, 5) at six months and 1 dB (CI 0.1, 2.8) at 12 months (Figure 72.17). A subsequent individual patient data metanalysis of three by-ear trials confirmed that the effect of a ventilation tube in those with an impairment of 25 dB HL or greater when randomized was 10 dB at six months and 3 dB at 12 months.141 [****] How this magnitude of improvement translates to reduction in hearing disability is not possible to assess in the ‘by-ear’ studies. Since 1992, the majority of studies have randomized ventilation tubes ‘by child’ so that hearing disability and other child-centred outcomes can be studied in addition to the hearing levels. TARGET is a UK multicentre study that looked at the effect of ventilation tubes alone or with adjuvant adenoidectomy on child-centred outcomes. To be eligible, children had to be aged between 3.5 and 7 years and have bilateral OME, persistent over a 12-week ‘watchful waiting’ period and associated with a hearing impairment in both ears of 20 dB HL or poorer. Children were randomized to one of three arms: no-surgical management, bilateral ventilation tubes (Shepard) or bilateral ventilation tubes with adjuvant adenoidectomy. Figure 72.18 shows the binaural hearing average in the two surgical groups in comparison with the nonsurgical group. The hearing in the nonsurgical group improves with time, mainly due to the natural resolution of the OME. Children randomized to have ventilation tubes had a marked improvement three months following surgery of 12 dB compared with the nonsurgical group.



6 months after V T



B M D



12 months after V T



B M D



Time with effusion In the three studies that provided such data, children who had ventilation tubes without adenoidectomy spent 32 percent (CI 17, 48) less time with effusion during the first year after insertion. After that time most ventilation tubes had become nonfunctioning or had been extruded. In another study137 at nine months post-insertion, those that had had ventilation tubes had 46 percent (CI 32, 61) less OME on otoscopy.



] 897



−10



0



20 10 Mean HL difference (dB)



30



Figure 72.17 Mean improvement in hearing level and 95 percent confidence intervals after ventilation tube (VT) insertion compared to no surgery (NS). B, Black et al.;138 D, Dempster et al.;139 M, Maw and Herod.140 Redrawn from Ref. 126, with permission.



898 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Hearing level dB HL



40 mm VT V T + AD



30 20 10 0 0



3



6



9 12 15 18 Time since randomization



21



It should be noted that in both ventilation tube groups the hearing at three months does not become ‘normal’; that is 0 dB HL. If one excludes the small number of ears in which the ventilation tube is nonfunctioning, there is still a material air–bone gap of 13 dB SD 7.142 [****] At all other postoperative visits, when a tube is functioning, the air–bone gap is of a similar magnitude and, as the middle ear space is well ventilated, is considered due to effusion and oedema around the ossicular chain. The difference in hearing between the two surgical groups and the nonsurgical group becomes almost negligible at 12 months. The reason for the deterioration in the hearing levels over time in those with a ventilation tube is due to a combination of two factors. The proportion of tubes that become nonfunctioning increases with time, though children were eligible to have them reinserted if the hearing entry criteria were resatisfied. The other reason is that in ears where the ventilation tube has extruded and the OME has resolved, there is still a small but material conduction deficit of approximately 14 dB.142 [****] The reason why those in the nonsurgical arm improves over that period is mainly due to the natural spontaneous resolution rate of the OME combined with a much smaller contribution from those that switched management arms to surgery. To summarize from this trial, the improvement in hearing effected by ventilation tubes averaged over the first year was 5.7 dB (CI 3.6, 7.7). In the second year, the difference between the two groups became negligible. Hence, when averaged over two years the benefit of ventilation tubes is reduced to 3.1 dB (CI 1.3, 4.8). These data are not dissimilar to the ‘by-ear’ studies reported earlier. A more recent individual data metanalysis of four trials that randomized children to ventilation tubes or watchful waiting confirmed the benefit of short-term ventilation tube at six months, but not at 12 months, with a magnitude of difference of 4.5 dB (CI 2.5, 6.5).141 [****] It is concluded that short-term ventilation tubes give a benefit to the hearing of 4–5 dB at six months following insertion in those with documented persistent bilateral



24



27



Figure 72.18 Means and 95 percent confidence intervals of the binaural hearing over the two-year follow up in the ventilation tube (VT), ventilation tube with adenoidectomy (VT 1 AD) and nonsurgical/medical management (mm) arms of TARGET. Redrawn from Ref. 142, with permission.



OME over a three-month period and a hearing impairment of at least 20 dB HL in both ears. This short-term benefit has to be balanced against the 5–10 dB poorer hearing in late teenage years in ears that have had OME ventilation tubes inserted having controlled for OME severity (see under Long-term hearing impairment above). Predictive factors for benefit to hearing To date, no individual study has identified factors that predict which children are likely to benefit most from ventilation tubes. To do this requires a larger study than any performed to date. Alternatively, a metanalysis could attempt to do this, albeit the predictive factors that one might wish to investigate might not have been recorded in all studies. Individual child data have been made available for metanalysis from seven trials that randomized children (n = 1232) with 12-week persistent bilateral OME to ventilation tubes or nonsurgical management.141 [****] The only factor predictive of a better hearing outcome at six months was attendance at daycare with an impairment of 20/25 dB in both ears. Those without daycare had no benefit (41 dB) at six months. No effect of age, sex or socioeconomic group was evident. In three trials which randomized children ‘byear’, those with a binaural hearing level of 425 dB HL benefited by 10 dB at six months compared with the 4 dB of those with lesser hearing thresholds. These data suggest that when the aim is improvement in the hearing, then younger children at daycare and those with binaural hearing thresholds poorer than 25 dB HL and persistent over at least 12 weeks will benefit most. Hearing disability In TARGET, the childrens’ hearing disability was assessed by a Reported Hearing Difficulty questionnaire. Using standard deviations as a unit of measurement, the difference between ventilation tubes and no surgery over the first year is ‘large’ even when adjusted for the expectation effect of surgery.142 [****] Over the two-year time period, the reduction in hearing disability with ventilation tubes was ‘modest’.



Chapter 72 Otitis media with effusion



Speech and language Three randomized controlled trials have specifically looked at whether ventilation tubes affect speech and language development. The study of Paradise et al.143 [****] randomized children when they were on average 18 months of age, that of Rovers et al.144 [****] when they were on average 20 months of age and that of Maw et al.137 [****] when they were 36 months of age. All obtained objective measures of speech and language, including receptive and expressive aspects. All assessed the hearing audiometrically. The design of all three studies was to randomize children to ‘early’ or ‘late’ surgery. As with this form of randomization, there is always the potential for those randomized to ‘late’ surgery to have surgery earlier. Paradise et al.143 identified eligible children from a birth cohort of 6350 children. Nine percent were eligible because of the duration of their unilateral or bilateral OME. There was no difference between the early and late surgery groups when followed 18 months after randomization up to when the children were on average three years of age. By then, 34 percent of the early surgery group had had surgery. The children in this trial have now been reported twice subsequently, the last when they were six years of age, which confirmed the lack of benefit of ventilation tubes.145 [****] The negative outcome of this study may in part be due to the inclusion of children with unilateral OME (63 percent of the sample). Unfortunately, the authors did not report the results of the children with bilateral OME separately. Rovers et al.144 identified their eligible children from a Dutch birth cohort of 30,099 children. Those who failed a hearing screen at nine months of age three times (n = 1081) were followed up over four to six months. Those who had persistent bilateral OME (n = 386; 1 percent of total cohort) were eligible for randomization to ‘early’ or ‘late’ surgery. Children were followed up for 12 months at which time, once adjusted for the effects of the mother’s education and the child’s intelligence, there was no effect of early surgery. In this study, only 9 percent of the late surgery group had early surgery. This study would suggest that children aged 20 months with bilateral persistent OME do not benefit, regarding their speech and language development, from ventilation tubes. This does not mean that children with more severe problems do not benefit; it is just that they have not been studied. Maw et al.137 considered children from a birth cohort to be eligible for randomization if they had documented bilateral OME associated with a hearing impairment of more than 25 dB HL persistent over a three-month watchful waiting period. Children also had disruptions to ‘speech, language, learning or behaviour’, though how this was defined was not stated. The mean age of the children at randomization was three years. At nine months following randomization, those that had not had



] 899



surgery were 3.2 months behind in their speech and language, but this difference had disappeared at 18 months. Unfortunately by then 85 percent in the late surgery group had had surgery. Their conclusions were that there is no difference between early and late surgery, but they could not comment upon whether surgery had an effect on speech and language in this age group. A subsequent metanalysis of two of these studies141 [****] and a systematic review of all trials83, 136 [****] confirmed the null effect of ventilation tubes on children in the first four years of life with OME. It must be concluded that, in general, ventilation tubes are not indicated to aid speech and language development in children three years and younger. Randomized controlled trials on severely affected children are merited, but might be difficult to achieve because of their relative rarity. Behaviour The UK randomized controlled trial of surgery, described under Speech and language above, of three-year-old children137 included serial behaviour questionnaires (Richman) and these were reported in a separate publication.100 [****] The mean change in the Richman behaviour score nine months following randomization was no different in the children receiving early ventilation tubes as opposed to the watchful waiting group. However at that time, the proportion of children with extremely poor scores was fewer in the surgery group (30 percent compared with 47 percent). The authors chose to highlight the latter finding, but it could be argued that using the mean scores rather than taking an arbitrary cut off is a more relevant and sensitive analysis. It must be concluded that the effect of bilateral persistent OME associated with a hearing impairment on a child’s behaviour is likely to be small and correct spontaneously with time, irrespective of whether they have surgery. Quality of life The Dutch study described under Speech and language included a generic quality-of-life questionnaire that covers 13 domains, nine of which would be applicable to children with OME (vitality, appetite, communication, motor problems, social, anxiety, aggression, eating and sleeping). No effect of surgery on this measure of quality of life was identified in the 18-month-old (on average) children with persistent bilateral OME.144 [****] Cost–benefit analysis Such an analyis requires the cost of an intervention to be calculated in terms of a change in a generic quality of life outcome. So, though the cost of surgery was obtained in the Dutch study, as there was no change in the quality of life such an analysis was not relevant.146 The TARGET study has yet to report a cost–benefit analysis.



900 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY COMPLICATIONS OF VENTILATION TUBES



At surgery The most common immediate complication at the time of surgery is displacement of the ventilation tube into the middle ear (0.5 percent, CI 0.3, 0.7).147 [****] Efforts should be made to retrieve the tube at the time, but failure to remove it seldom causes problems. Ossicular chain damage will only occur if the myringotomy is placed incorrectly, i.e. posterosuperiorly. Very rarely, a high jugular bulb may be pierced by an inferiorly placed myringotomy. Immediately post-operation Blockage of the tube with blood can be prevented to some extent by aspiration at the time of surgery. Syringing has been suggested but is only likely to be effective before the blood dries. [****] Early infection around the tube occurs in 9 percent of ears (CI 8.5, 9.7).147 [****] Topical antibiotic steroid drops at the time of surgery can reduce the incidence to 1 percent (CI of difference 3, 12).148 [****] Some clinicians prolong topical therapy for three to five days after surgery with similar results.149 [****] The same short-term effect can be achieved with two weeks of prednisolone and trimethoprim,150 [****] but whether this reduction of infection has any effect on longer-term tube function is doubtful.151 [****] Otorrhoea When a ventilation tube is in situ, it can be associated with otoscopic evidence of infection with the production of pus that may be clinically evident as otorrhoea. The reason for the infection could be the presence of the tube acting as a foreign body, an episode of acute otitis media with the middle ear pus coming through the ventilation tube, or a combination of both. Making a distinction is difficult, but the younger the child the higher the chances are that the infection is due to an episode of acute otitis media. In addition, the pus can dry and block the tube, increasing the chances of it being extruded. In a literature review of case-controlled studies and randomized controlled trials of the complications of ventilation tubes inserted in children of all ages and for all indications, approximately 9 percent of parents report



early postoperative otorrhoea (Table 72.15).147 [***] Thereafter, approximately 7 percent report recurrent acute episodes and 3 percent chronic discharge. However, in children aged between 3.5 and 7 years, where the indication for ventilation tubes is persistent OME, parental report of otorrhoea is of the same magnitude but the otoscopic incidence of infection at any time is o1 percent during the two-year follow-up period (personal communication). [****] Management of infection is mainly aural toilet with topical antibiotic steroid drops. [*] Secondary granulation tissue occurs in approximately 1 percent of all cases (CI 0.7, 1.3).147 [***] Granulation tissue is similarly treated with topical preparations. The tube, especially long-term ones, occasionally has to be removed (approximately 4 percent of ears).147 [***] Permanent perforation When a ventilation tube extrudes, an initial pars tensa perforation is inevitable. In the majority, this will heal spontaneously (personal communication). [****] A perforation may then subsequently reoccur at the same site due to a subsequent episode of acute otitis media. Again, in the majority this will heal spontaneously. Thus at any one time following tube extrusion, a proportion of ears will have a perforation which may or may not heal spontaneously. What the proportion is depends on the age of the child and the type of tube inserted. Overall, short-term tubes are associated with a 2 percent incidence of perforations and longer-term tubes an incidence of 17 percent (Table 72.15) an increased relative risk of 3.5 (CI 1.5, 7.1) (Table 72.16). Once a perforation has been documented as chronic, myringoplasty is the management of choice. This is not usually performed until the child has outgrown having recurrent acute otitis media. Tympanosclerosis Hyaline degeneration of the collagen tissue in the fibrous layer of the pars tensa becomes evident otoscopically as localized white patches or plaques of tympanosclerosis. These do not occur in the absence of a history of otitis media. The frequency of tympanosclerosis increases with the frequency of OME (Table 72.12). However, there is a dramatic increased risk difference of 0.33 (CI 21, 45) if a



Table 72.15 Incidence of otorrhoea with indwelling tympanostomy tubes. Type of otorrhoea



Unit of analysis



Incidence Rate %



Early postoperative Recurrent acute Chronic Requiring tube removal



Patients Patients Patients Ears



Reprinted from Ref. 147, with permission.



16.0 7.4 3.8 4.0



95% CI 14.2–17.9 6.0–9.1 2.2–6.0 3.5–4.5



Range 8.8–19.6 0.7–19.6 1.4–9.9 0–34.3



Chapter 72 Otitis media with effusion



] 901



Table 72.16 Incidence of tympanic membrane sequelae after tympanostomy tube extrusion. Tympanic membrane sequelae Tympanosclerosis Atrophy or retraction at short-term tube site Atrophy or retraction at long-term tube site Retraction pocket of pars tensa Chronic perforation, short-term tube Chronic perforation, long-term tube Cholesteatoma, short-term tube Cholesteatoma, long-term tube



No of ears



Incidence %



95% CI



7197 1467 460 543 8107 3356 8231 1899



31.7 25.5 23.3 3.1 2.2 16.6 0.8 1.4



30.6–32.8 24.2–26.8 22.0–24.6 1.8–5.0 1.8–2.5 15.3–17.8 0.6–1.0 0.9–2.0



Reprinted from Ref. 147, with permission.



ventilation tube has been inserted;136 [****] in this instance there being no difference between short- and long-term tubes (Tables 72.15 and 72.16). In a small proportion of ears, tympanosclerosis is a dynamic process with resolution and occurrence occurring with time.152 [**] What effect, if any, tympanosclerosis has on the hearing is debated, but has been reported to be up to 3 dB in the short term.153 [****] In the longer term at the age of 18 years, there is an impairment of 5–10 dB with a 3–4 dB sensorineural component in those who have had ventilation tubes, having controlled for the OME disease load in childhood.154 [***] The more tubes that had been inserted, the greater the impairment. Pars tensa atrophy and retraction Thinning and retraction of the tympanic membrane is itself considered to be a complication of persistent OME. In children initially referred with bilateral OME, pars tensa retraction to the incus or promintory (Sade grade 3/ 4) occurs in 8 percent of the better and 10 percent of the poorer hearing ears.14, 16, 17 [***] Such retractions were not associated with a history of longer ear problems. Such retractions followed up over a 12-week watchful waiting period resolved in 69 percent of the better and 65 percent of the poorer ears. The OME had also resolved in 14 percent of the better and 10 percent of the poorer ears with a pars tensa retraction, respectively.



Pars tensa atrophy also occurs in approximately 3 percent and retraction to the incus or promintory in 2 percent of ears in seven- to eight-year-old children in whom there is no history of OME (Table 72.12). This incidence is similar to that in children with transient episodes of OME. However, the incidence of atrophy, but not of retraction, increases to approximately 9 percent in those with persistent OME. In those who still have a ventilation tube in situ for their OME, the incidence of atrophy increases to approximately 68 percent and retraction to approximately 14 percent. It could be argued that this increase in ears with a ventilation tube still in situ is because these are more persistent cases, rather than because of the ventilation tube itself. This is in part supported by the fact that the incidence is the same with short- and long-term tubes (Table 72.16 and 72.17). What can be conclusively inferred is that ventilation tubes do not prevent the occurrence of atrophy or retraction and should not be inserted for that reason alone. [Grade B] Effect of swimming A review of the literature in 1993155 and a Scottish Intercollegiate Guidelines Network review120 both come to the conclusion that swimming has no effect on the incidence of reported otorrhoea in children with ventilation tubes. This is supported by the one randomized controlled trial of 212 consecutive children. Of the 50



Table 72.17 Meta-analysis of tympanostomy tube sequelae. Outcome assessed Increase Increase Increase Increase Increase a



in in in in in



otorrhoea (long- versus short-term tube) chronic perforation (long- versus short-term tube) cholesteatoma (long- versus short-term tube) atrophy/retraction (tube versus no surgery or myringotomy) tympanosclerosis (tube versus no surgery or myringotomy)



Rate difference (%) (95% CI)a 13.7 7.3 1.3 11.0 29.9



(  0.7–28.0) (1.3–13.3) (0.4–2.2) (2.6–10.3) (21.9–38.0)



Absolute difference in outcomes between groups; po0.05 when the 95% CI does not include zero. Ratio of sequelae incidence between groups; po0.05 when the 95% CI does not include one. Reprinted from Ref. 147, with permission.



b



Relative risk (95% CI)b 2.1 3.5 2.6 1.7 3.5



(1.0–4.1) (1.5–7.1) (1.5–4.4) (1.1–2.7) (2.6–4.9)



902 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY percent followed up for one year, 68 percent in the swimming compared with 60 percent in the noswimming group reported otorrhoea. The mean number of episodes was one per year in both groups.156 [****] It is also unlikely that getting shower or bath water in the ears increases the incidence of otorrhoea. However, there is no substantive literature addressing this question, except laboratory studies that suggest that the pressure would be insufficient for bath water to be forced into the middle ear. There is therefore no reason to advise against getting water in the ear in children who are fitted with ventilation tubes.



ADENOIDECTOMY



Surgical techniques Conventionally, adenoidectomy has been performed by blind curettage. This has the advantage of speed, but the disadvantage of relying on the haemorrhage ceasing spontaneously. Temporary nasal packs can be used to encourage haemostasis. Unfortunately, in the immediate postoperative period, bleeding occurs in approximately four percent of children.157 [**] Suction diathermy ablation, where the tissue is first diathermied then removed by suction, may be associated with less intraoperative blood loss, but haemorrhage in the immediate postoperative period is no less.157 [**] Some do not perform the procedure blindly but inspect the adenoids at intervals with a mirror. This has the potential of allowing more selective removal of adenoid tissue and avoiding palatopharyngeal incompetence by leaving an inferior pad of tissue. Microdebriders have also been advocated as allowing more selective removal of tissue. Though visualization is better with this because the tissue and blood are removed by suction, the intraoperative blood loss is no less than by curettage.158 [***] Outcomes of adenoidectomy The mechanism whereby adenoidectomy resolves the OME in some children is unclear, the most popular unproven hypothesis being that it removes a chronic source of infection in the nasopharynx rather than because it removes tissue that physically obstructs the Eustachian tube.



Figure 72.19 shows the effect of ventilation tubes along with adenoidectomy in the three UK trials and is presented in a similar manner to the results for ventilation tubes alone (Figure 72.17). The additional benefit of adenoidectomy was approximately 2 dB at 6 and at 12 months. In addition, adenoidectomy reduced the necessity for the reinsertion of short-term ventilation tubes where reinsertion is determined by recurrence of a hearing impairment due to OME (see Revision surgery). Follow up of the children in the Maw and Herod study140 for a further 12 years159 [****] showed that the difference of 1–2 dB continues long term. Rather surprisingly, in this study the hearing in all the 12-year-old children who had previous OME was still not normal (approximately 15 dB HL). The TARGET study confirms the additional benefit of adenoidectomy during the second year of that study (Figure 72.18). Over the first year, the magnitude of the adjuvant effect of adenoidectomy was 2.3 dB (CI 0.6, 4.0) and over two years was 3.3 dB (CI 1.7, 4.8). Hearing disability In TARGET, over the two-year follow-up period, the adjuvant benefit of adenoidectomy on the reported hearing difficulty was ‘modest’ (personal communication). [****] Revision surgery The proportion of children meriting reinsertion of the ventilation tubes after extrusion because there was still bilateral OME with a material hearing impairment is significantly reduced by adenoidectomy from 47 to 28 percent (p = 0.01) of children (personal communication). [****] Detailed data regarding revision after two years of follow up are currently unavailable in any RCT. However, 6 months after VT + AD



B M D



12 months after VT + AD B M



Hearing Adenoidectomy alone was the conventional surgical management for many years. Since the introduction of ventilation tubes its use has gradually declined because of its lesser overall effect on the persistence of OME and hearing. In the three UK trials, the overall effect at six months on the hearing of adenoidectomy was 8 dB compared with 12 dB for ventilation tubes. Current practice is to perform adenoidectomy as an adjunct to the insertion of ventilation tubes.



D −10



0



10



20



30



Mean HL difference (dB)



Figure 72.19 Mean improvement in hearing level and 95 percent confidence intervals after ventilation tube with adenoidectomy (VT 1 AD) compared to no surgery (NS). B, Black et al.;138 D, Dempster et al.;139 M, Maw and Herod.140 Redrawn from Ref. 126, with permission.



Chapter 72 Otitis media with effusion



hospital activity data suggest that the children who have had an adenoidectomy are less likely to be admitted to hospital and have further surgery for OME extending for a longer period.160 [***] Upper respiratory tract symptoms Adjuvant adenoidectomy improved the reported upper respiratory tract symptoms, as well as the hearing to a ‘modest’ extent in the TARGET study. The extra benefit is variable and has yet to be shown to be predictable so that adenoidectomy can be offered to children selectively. Interestingly, the mechanism of any benefit is unclear, but it is unlikely to be due to an improvement in Eustachian tube function.161 [****] Indicators of benefit Many authors have attempted to analyze which children in their studies are most likely to benefit from adjuvant adenoidectomy. A metanalysis has yet to be reported. Complications of adenoidectomy Haemorrhage in the postoperative period is the most feared complication, the incidence depending on how haemorrhage is defined. This can be any blood however small (0.5–8 percent) or the numbers requiring intervention such as blood transfusion or return to theatre (0.5 percent).157, 162, 163 [***] Transient velopharyngeal insufficiency can occur in up to 5 percent, but more permanent insufficiency is uncommon (o0.1 percent).164 [****]



Hearing aids While hearing aids have been used in children with persistent OME, to avoid repeated surgery their use as the preferred initial management has not been extensively reported. In case series,165 aids were acceptable to the majority of parents and gave aided thresholds with a mean of 17 dB. [**] The unaided thresholds were not reported, but this improvement is at least in the same range as expected of ventilation tubes. The main concern is potential noise trauma if the aid continues to be worn after the OME has resolved.



BONE-CONDUCTION HEARING AIDS



These have not been considered a viable option until the production of a Softband bone-conduction aid which incorporates a BAHAs (bone anchored hearing aid) in a head-band. This method of amplification has the potential advantage over an air-conduction aid in that should the hearing impairment improve during the period of aid issue due to resolution of the OME, then potentially damaging noise trauma would not occur. No report has been identified that has looked at the benefit from the Soft-band aid.



] 903



MANAGEMENT OF OME IN CHILDREN WITH CRANIOFACIAL ABNORMALITIES



Children with a cleft palate even subsequent to its surgical repair have a high incidence of OME which in some is a constant condition. As such, how to alleviate their hearing impairment without increasing the risk of developing chronic otitis media has increasingly been achieved with the use of hearing aids rather than ventilation tubes.166, 167 [***] The same attitudes apply to the management of OME in children with Down syndrome.168 [**]



KEY POINTS  Otitis media with effusion (OME) is the chronic accumulation of mucus within the middle ear and sometimes the mastoid air cell system.  In children, the prevalence is bimodal with the first and largest peak of 20 percent at two years of age and a second peak of 16 percent at around five years of age. [***]  Natural resolution of the OME is the most likely outcome, albeit there may be further episodes. [***]  In the majority of children their OME causes only a mild hearing impairment that is unlikely to be disabling. [***]  Any effect of this hearing impairment on speech and language and cognition is small, of minor significance and likely to spontaneously correct with age. [***]  The effect of persistent OME on behaviour, though evident, is small. [***]  Follow-up studies suggest that children who have had OME will still have a small conductive impairment (10 dB) once the OME has resolved. [***]  A small conductive impairment is even evident when in their late teens, albeit whether this is a disease effect or the effect of surgery at this age has yet to be clarified. [***]  Attic retractions are an OME disease effect with an incidence of 14 percent in the most severely affected children. [***] What proportion of these retractions progress to cholesteatoma is not known.  There is no evidence that the risk of developing an attic retraction and the potential to develop a cholesteatoma is altered by the insertion of ventilation tubes. [***]  Tympanosclerosis and atrophy of the pars tensa are OME disease effects, the risk of which is increased by the insertion of ventilation tubes to 30 and 25 percent respectively. [***]



904 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY  Long-term pars tensa retractions are more likely caused by the insertion of ventilation tubes than by the disease process itself, with an incidence of 3 percent. [***]  The benefit of ventilation tubes on children under three years of age with significant language, development or behaviour problems requires investigation.  In children over the age of three years, the binaural hearing average can be taken as a surrogate of hearing disability, behaviour and impact upon quality of life. In children of this age, OME is unlikely to have a major effect on language or cognition. [***]  The arbitrary cut-off in hearing thresholds that is taken to indicate a potentially disabling effect that could be improved by ventilation tubes is now taken as 25 dB HL or poorer in both ears. [****]  All children with hearing thresholds at these levels or poorer should have their hearing documented over at least a 12 week watchful waiting period before any decisions regarding surgery are made. [***]  Children whose hearing thresholds do not satisfy the hearing criterion of 25 dB HL bilaterally may or may not be followed up 12 weeks later with repeat audiometry.  Children who persist with OME and a bilateral 25 dB HL over a 12-week-period can benefit from the beneficial effects of ventilation tubes on the hearing over the next nine months. However, the effect is small and is maximal at three months of a magnitude of 12 dB. [****]  Even when a ventilation tube is functioning, there is still a residual conductive impairment of around 13 dB. [****]  This relatively small benefit to the hearing has to be balanced against the longer term deleterious effects of ventilation tubes on the tympanic membrane which are associated with poorer hearing. This balance of benefit against harm has to be fully discussed with and appreciated by the parents or carers.  There is no evidence that a pars tensa retraction in association with OME is an indicator of chronicity or greater benefit from ventilation tubes. [***]  If ventilation tubes are to be inserted there is a strong argument that adenoidectomy should be performed at the same time, particularly if there is concern regarding the general health of the upper respiratory tract. [***]  Adenoidectomy extends the period of benefit to the hearing of short-term ventilation tubes up to two years by a magnitude of around 2 to 3 dB. [****] It also reduces the proportion



of children eligible for reinsertion of shortterm tubes from around 50 to 25 percent. [****]  No predictors of benefit from adenoidectomy have been identified. Whether adjuvant adenoidectomy is merited for this small gain in hearing, given the risk of postoperative bleeding in 4 percent of children, is unresolved.



Best clinical practice In primary care [ OME should be suspected in children who have



[ [ [



[



[ [



[ [



[



recently recovered from an episode of acute otitis media and in those presenting because of parental concern regarding their hearing. [Grade B] Attention should be paid to parental concern regarding their child’s hearing, particularly if there is a history of recurrent acute otitis media. [Grade B] All children should have otoscopy to confirm OME and exclude AOM or COM. [Grade C] Unless specifically trained in otoscopy, this is likely to be insufficiently sensitive or specific even if performed with a pneumatic otoscope to reliably make the diagnosis of OME. [Grade B] Tympanometry should perhaps be more available in primary care. In primary care the unreliability of otoscopy and nonavailability of tympanometry should not cause concern as it is the disability of a bilateral hearing impairment persistent for longer than 12 weeks that is the indicator for referral for audiometric assessment, rather than suspicion of OME. Free-field voice testing is therefore a more meaningful assessment to perform. [Grade B] No medical therapy has been shown to be effective longer term (more than six weeks) including antibiotics [Grade A], topical nasal steroids, systemic steroids, nasal decongestants and mucolytics. [Grade B] In the older child, auto-inflation with the help of ‘Otovent’ balloons may be beneficial. [Grade B] It is important that all parents of children with OME receive appropriate general counselling about the natural history and relative benign nature of the condition. [Grade D] This includes advice regarding hearing tactics (see above under Counselling and hearing tactics). If there is persistent concern regarding the hearing in any child, then referral for hearing assessment is indicated. This will detect previously unsuspected sensorineural impairments as well as quantify any hearing impairment associated with OME. [Grade D]



Chapter 72 Otitis media with effusion



[ In most cases, such referral should be for hearing



[ In children under the age of three years, ventilation



assessment and a further period of audiometric monitoring, rather than for consideration of ventilation tubes. [Grade B]



In secondary care [ Pure-tone air-conduction thresholds should be assessed



[ [



[



[



[ [ [ [ [



[



[



in all children referred with suspected OME and a hearing impairment at their first visit. Not masked bone-conduction thresholds should be obtainable in the majority of 3.5 years or older children. This will aid management of those with OME and detect the 1 percent with previously undiagnosed sensorineural hearing impairments. [Grade B] Otoscopy and parental reports of hearing are insufficient to predict current hearing thresholds. [Grade B] Otoscopy, supplemented by tympanometry if there is any doubt as to whether there is OME, should be performed in all children. This, along with the hearing thresholds, should enable the children with OME to be categorized as having unilateral or bilateral disease, and if bilateral whether the impairment in both ears is 4 = 25 dB HL. This distinction is relevant to future management in secondary care. [Grade B] Thereafter, all parents of children with OME should receive appropriate general counselling about the natural history and relative benign nature of OME. [Grade D] The hearing levels associated with the child’s OME should be discussed with the parents and advice about hearing tactics given to help the parents mitigate any disability (see above under Counselling and hearing tactics). [Grade D] There is insufficient evidence to support the use of topical nasal steroids at present for childhood OME. [Grade B] Systemic steroids cannot be recommended at present for childhood OME. [Grade B] It is not recommended that antibiotics be used for the longer term (more than six weeks) in the management of childhood OME. [Grade A] If some form of nonsurgical management is required, auto-inflation is the therapy with the strongest evidence of efficacy for older children. [Grade B] As there is no evidence that medical therapy is beneficial, clinicians in secondary care have the task of identifying those that might benefit from surgery after a period of watchful waiting. Children with unilateral OME are unlikely to develop bilateral OME. [***] In general, children with unilateral OME do not merit follow-up at secondary care and should not have ventilation tubes inserted. [Grade A] Myringotomy with aspiration has not been shown to be effective in restoring the hearing levels in children with OME and is not recommended without the insertion of ventilation tubes. [Grade A]



] 905



[ [



[ [



[ [



[



tubes are not indicated even in cases with persistent bilateral OME if there are no language, development or behaviour problems. [Grade A] They therefore do not merit routine follow-up at secondary care. In general, ventilation tubes are not indicated to aid speech and language development in children aged three years and younger. [Grade A] The effect of bilateral persistent OME associated with a hearing impairment on a child’s behaviour is likely to be small and correct spontaneously with time. Correspondingly, a child’s behaviour is not an indication for surgery. [Grade B] Ventilation tubes do not prevent the occurrence of atrophy or retraction and should not be inserted for that reason alone. [Grade B] Short-term ventilation tubes are of proven benefit to the hearing up to nine months following insertion in children with bilateral OME documented to have been persistent for at least 12 weeks. The magnitude of the effect is around 4 to 5 dB. It is recommended that such ventilation tubes be limited to those most likely to benefit; children with a hearing impairment of at least 25 dB HL in both ears audiometrically documented to have persisted for at least three months. [Grade A] To maximize the duration of potential tube function, the preferred insertion site is antero-inferior through a circumferential or radial incision. [Grade A] In arriving at a decision regarding surgery, the deleterious effects of ventilation tubes in the longer term on tympanic membrane scarring (RR 0.33), thinning and poorer hearing (5 to 10 dB) must be discussed with the parents to enable any consent to be fully informed. [Grade D] There is no reason to give advice against getting water in the ears in children who have ventilation tubes. [Grade D]



Deficiencies in current knowledge and areas for future research



$ $ $ $



Otitis media with effusion and acute otitis media in children is the most researched area in otology. As such, a high proportion of the important clinical questions on otitis media with effusion have already been satisfactorily addressed for the mild to moderately affected child. What requires investigation are children with the extremes of persistence of OME, such as occurs in children with craniofacial abnormalities. The ideal therapy would be a preventative one, such as immunization to reduce the frequency of upper respiratory viral and bacterial infections.



906 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



$ $ $



For those who develop OME, medical therapy which manages the mucosal aspects of the condition are more likely to be beneficial than surgical therapy. It is important to have confirmation by better controlled studies of the more serious otological damage, such as retraction that occurs to the tympanic membrane following ventilation tube insertion. How to best change the attitudes to ventilation tubes of surgeons and parents given their paucity of benefit for any length of time and their permanent damaging effect is perhaps the aspect most requiring research in OME.



10.



11.



12.



13.



ACKNOWLEDGEMENTS 14.



Figures 72.7 to 72.15 are reprinted from Wormald PJ and Browning GG. Otoscopy: a structured approach. London: Hodder Arnold, 1996.



15.



REFERENCES



16.



1. Bluestone CD. State of the art: definitions and classifications. In: Liu DJ, Bluestone CD, Klien JO, Nelson JD. (eds). Recent advances in otitis media with effusion. Proceedings of the 3rd International Conference. Ontario: Decker and Mosby; 1984 2. Shimada T, Lim DJ. Distribution of ciliated cells in the human middle ear. Electron and light microscopic observations. The Annals of Otology, Rhinology, and Laryngology. 1972; 81: 203–11. 3. Lim D. Normal and pathological mucosa of the middle ear and Eustachian tube. Clinical Otolaryngology. 1979; 4: 213–34. 4. Tos M. Pathogenisis and pathology of chronic secretory otitis media. Annals of Otology, Rhinology and Laryngology. 1980; 89: 91–7. 5. Ishii T, Toriyama M, Suzuki JI. Histopathological study of otitis media with effusion. Annals of Otology and Laryngology. 1980; 89: 83–6. 6. Takkeuchi K, Majima Y, Hirata K, Hattori M, Sakakura Y. Quantitation of tubotympanal mucociliary clearance of otitis media with effusion. Annals of Otology, Rhinology, and Laryngology. 1990; 99: 211–4. 7. Kubba H, Pearson JP, Birchall JP. The aetiology of otitis media with effusion: a review. Clinical Otolaryngology. 2000; 25: 181–94. 8. Jero J, Karma P. Bacteriological findings and persistence of middle ear effusion in otitis media with effusion. Acta Otolaryngologica. 1997; 529: 22–6. 9. Sheanan P, Blaney AW, Sheanan JN, Earley MJ. Sequelae of otitis media with effusion among children with cleft lip



17.



18.



19.



20. 21.



22.



23.



24.



and/or palate. Clinical Otolaryngology. 2002; 27: 494–500. Rivron RP. Bifid uvula: prevalence and association in otitis media with effusion in children admitted for routine otolaryngological operations. Journal of Laryngology and Otology. 1989; 103: 249–52. Alho OP, Oja H, Koivu M, Sorri M. Risk factors for chronic otitis media with effusion in infancy. Archives of Otolaryngology – Head and Neck Surgery. 1995; 121: 839–43. Marx J, Osguthorpe JD, Parsons G. Day care and the incidence of otitis media in young children. Otolaryngology – Head and Neck Surgery. 1995; 112: 695–9. van Balen FAM, de Melker RA. Persistent otits media with effusion; can it be predicted? A family practice follow-up study in children aged 6 months to 6 years. Journal of Family Practice. 2000; 49: 605–11. Medical Research Council Multicentre Otitis Media Study Group. Risk factors for persistence of bilateral otitis media with effusion. Clinical Otolaryngology. 2001; 27: 147–56. Medical Research Council Multicentre Otitis Media Study Group. Selecting persistent glue ear for referral in general practice; a risk factor approach. British Journal of General Practice. 2002; 52: 549–53. Medical Research Council Multicentre Otitis Media Study Group. Pars tensa and pars flaccida retraction in persistent otitis media with effusion. Otology and Neurotology. 2001; 22: 291–8. Medical Research Council Multicentre Otitis Media Study Group. Surgery for persistent otitis media with effusion: generalizability of results from the UK trial (TARGET). Clinical Otolaryngology. 2001; 26: 417–24. Tasker A, Dettmar PW, Panetti M, Koufman JA, Birchall JP, Pearson JP. Reflux of gastric juice and glue ear in children. Lancet. 2002; 359: 493. Lieu JEC, Nuthappan BS, Uppaluri R. Association of reflux with otitis media in children. Otolaryngology – Head and Neck Surgery. 2005; 133: 357–61. Daly KA. Epidemiology of otitis media. Otolaryngologic Clinics of North America. 1991; 24: 775–86. Zielhuis GA, Rach GH, van den Basch A, van den Broek P. The prevalence of otitis media with effusion: a critical review of the literature. Clinical Otolaryngology. 1990; 15: 283–8. Engel J, Anteunis L, Volovics A, Hendriks J, Marres E. Prevalence rates of otitis media with effusion from 0 to 2 years of age: healthy-born versus high-risk born infants. International Journal of Pediatric Otolaryngology. 1999a; 47: 243–51. Engel JAM, Anteunis LJC, Volvics A, Hendricks JJT, Manni JJ. Chronic otitis media with effusion during infancy, have parent-reported symptoms prognostic value? A prospective longitudinal study from 0 to 2 years of age. Clinical Otology. 1999; 24: 417–23. Williamson IG, Dunleavey J, Bain J, Robinson D. The natural history of otitis media with effusion – a three-year



Chapter 72 Otitis media with effusion



25.



26.



27.



28.



29.



 30.



 31.



32.



33.



34.



35.



36.



37.



study of the incidence and prevalence of abnormal tympanograms in four South West Hampshire infant and first schools. Journal of Laryngology and Otology. 1994; 108: 930–4. Tos M, Holm-Jensen S, Sorensen CH, Mogensen C. Spontaneous course and frequency of secretory otitis in 4-year-old children. Archives of Otolaryngology. 1982; 108: 4–10. Rovers MM, Straatman H, Zielhuis GA, Ingels K, van der Wilt GJ. Seasonal variation in the prevalence of persistent otitis media with effusion in one-year-old infants. Paediatric and Perinatal Epidemiology. 2000; 14: 268–74. Rovers MM, Straatman H, Ingels K, van der Wilt GJ, van den Broek P, Zielhuis GA. The effect of ventilation tubes on language development in infants with otitis media with effusion: A randomized trial. Pediatrics. 2000; 106: E42. Midgley EJ, Dewey C, Pryce K, Maw AR. ALSPAC study team. The frequency of otitis media with effusion in British pre-school children: a guide for treatment. Clinical Otology. 2000; 25: 485–91. Sassen ML, Brand R, Grote JJ. Risk factors for otitis media with effusion in children 0 to 2 years of age. American Journal of Otolaryngology. 1997; 18: 324–30. Dewey C, Midgeley E, Maw R. The ALSPAC study team. The relationship between otitis media with effusion and contact with other children in a British cohort studied from 8 months to 3.5 years of age. International Journal of Pediatric Otorhinolaryngology. 2000; 55: 33–45. Good multivariate analysis. Engel J, Anteunis L, Volovics A, Hendriks J, Marres E. Risk factors of otitis media with effusion during infancy. International Journal of Pediatric Otolaryngology. 1999b; 48: 239–49. Good multivariate analysis including 19 pertinent risk factors. Apostolopoulos K, Xenelis J, Tzagaroulakis A, Kandiloros D, Yiotakis J, Papafragou K. The point prevalence of otitis media with effusion among school children in Greece. International Journal of Pediatric Otorhinolaryngology. 1998; 44: 207–14. Marchisio P, Principi N, Passali D, Salpietro DC, Boschi G, Chetri G et al. Epidemiology and treatment of otitis media with effusion in children in the first year of life. Acta Otolaryngologica. 1998; 118: 557–62. Saim A, Siam L, Siam S, Ruszymah BHI, Sani A. Prevalence of otitis media with effusion amongst pre-school children in Malaysia. International Journal of Pediatric Otorhinolaryngology. 1997; 41: 21–8. Rushton HC, Tong CF, Yue V, Wormald PJ, van Hasselt CA. Prevalence of otitis media with effusion in multicultural schools in Hong Kong. Journal of Laryngology and Otology. 1997; 111: 804–6. Son DH, Son NT, Tri L, Phuong TK, Mai TTT, Tien VM et al. Point prevalence of secretory otitis media in children in Southern Vietnam. Annals of Otology and Laryngology. 1998; 1107: 406–10. Del Mar C, Glasziou P, Hayem M. Are antibiotics indicated as initial treatment for children with acute otitis media? A



38.



39.



40.



41.



42.



43.



44.



45.



46.



47.



48.



49.



50.



51.



] 907



meta-analysis. British Medical Journal. 1997; 314: 1526–9. Daly KA, Rich SS, Levine S, Margolis RH, Le CT, Lindgren L et al. The family study of otitis media: design and disease and risk factor profiles. Genetic Epidemiology. 1996; 13: 451–68. Casselbrant ML, Mandel EM, Fall PA, Rockette HE, KursLasky M, Bluestone CD et al. The heritability if otitis media; a twin and triplet study. Journal of the American Medical Association. 1999; 282: 2125–30. Paradise JL, Rockette HE, Colburn DK, Bernard BS, Smith CG, Kurs-Lasky M et al. Otitis media in 2253 Pittsburgh-area infants prevalence and risk factors during the first two years of life. Pediatrics. 1997; 99: 318–33. Rovers MM, Hofstad EA, vann den Brand KI, Ingels K, van den Wilt GJ, Zielhius GA. Prognostic factors for otitis media with effusion. Clinical Otolaryngology. 1998; 23: 543–6. Zielhius GA, Rach GH, van den Broek P. Screening for otitis media with effusion in preschool children. Lancet. 1989; 1: 311–3. Hogan SC, Stratford KJ, Moore DR. Duration and recurrence of otitis media with effusion in children from birth to 3 years: prospective study using monthly otoscopy and tympanometry. British Medical Journal. 1997; 314: 350–3. Lous J, Fiellau-Nikolajsen M. Epidemiology of middle ear effusion and tubal dysfunction. A one-year prospective study comprising monthly tympanometry in 387 nonselected 7-year-old children. International Journal of Pediatric Otorhinolaryngology. 1981; 3: 303–17. Anteunis LJC, Engel JAM, Hendriks JJT, Manni JJ. A longitudinal study of the validity of parental reporting in the detection of otitis media and related hearing impairment in infancy. Audiology. 1999; 38: 75–82. Stewart MG, Friedman EM, Sulek M, Duncan NO, Fernandez AD, Bautista MH. Is parental perception an accurate predictor of childhood hearing loss? A prospective study. Otolaryngology and Head and Neck Surgery. 1999; 120: 340–4. Cavanaugh RM. Obtaining a seal with otic specula: Must we rely on an air of uncertainty? Pediatrics. 1991; 87: 114–6. Jones WS, Kaleida PH. How helpful is pneumatic otoscopy in improving diagnostic accuracy? Pediatrics. 2003; 112: 510–3. Toner JG, Mains B. Pneumatic otoscopy and tympanometry in the detection of middle ear effusion. Clinical Otolaryngology. 1990; 15: 121–3. Finitzo T, Freil-Patti S, Chinn K, Brown O. Tympanometry and otoscopy prior to myringotomy; issues in diagnosis or otitis media. International Journal of Pediatric Otorhinolaryngology. 1992; 24: 101–10. Vaughan-Jones R, Mills RP. The Welch Allyn audioscope and microtymp: their accuracy and that of pneumatic otoscopy, tympanometry and pure tone audiometry as



908 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



52.



53.



 54. 55.



 56. 57.



58.



59.



60.



61.



62.



63. 64.



65.



66.



predictors of otitis media with effusion. Journal of Laryngology and Otology. 1992; 106: 600–2. Nozza RJ, Bluestone CD, Kardatzke D, Bachman R. Identification of middle ear effusion by aural acoustic admittance and otoscopy. Ear and Hearing. 1994; 15: 310–23. Steinbach WJ, Sectish TC, Benjamin DK, Chang KW, Messner AH. Pediatric residents’ clinical diagnositic accuracy of otitis media. Pediatrics. 2002; 109: 993–9. American Academy of Pediatrics. Otitis media with effusion; clinical practice guideline. Pediatrics. 2004; 113: 1412–29. OME guidelines. Rosenfeld RM, Culpepper L, Doyle KJ, Grundfast KM, Hoberman A, Kenna MA et al. American Academy of Pediatrics Subcommittee on Otitis Media with Effusion, American Academy of Family Physicians, American Academy of Otolaryngology – Head and Neck Surgery. Clinical practice guideline: Otitis media with effusion. Otolaryngology and Head and Neck Surgery. 2004; 130: S95–118. Preston K. Pneumatic otoscopy: a review of the literature. Issues in Comprehensive Pediatric Nursing. 1998; 21: 117–28. Good literature review. Pirozzo S, Papinczak T, Glasziou P. Whispered voice test for screening for hearing impairment in adults and children: systematic review. British Medical Journal. 2003; 327: 967–70. Medical Research Council Multicentre Otitis Media Study Group. Sensitivity, specificity and predictive value of tympanometry in predicting a hearing impairment in otitis media with effusion. Clinical Otolaryngology. 1999; 24: 294–300. Palmu A, Puhakka H, Rahko T, Takala AK. Diagnostic value of tympanometry in infants in clinical practice. International Journal of Pediatric Otorhinolaryngology. 1999; 49: 207–13. Engel J, Anteunis L, Chenault M. Otoscopic findings in relation to tympanometry during infancy. European Archives of Otorhinolaryngology. 2000; 257: 366–71. American Speech-Language-Hearing Association. Guidelines for screening for hearing impairments and middle ear disorders. ASHA. 1989; 31: 71–7. Nozza RJ, Bluestone CD, Kardatzke D, Bachman R. Towards the validation of aural acoustic immittance measures for diagnosis of middle ear effusion in children. Ear and Hearing. 1992; 13: 442–53. Jerger J. Clinical experience with impedance audiometry. Archives of Otolaryngology. 1970; 92: 311–24. Orchik DJ, Dunn JW, McNutt L. Tympanometry as a predictor of middle ear effusion. Archives of Otolaryngology. 1978; 104: 4–6. Sassen ML, Van Aarem A, Grote JJ. Validity of tympanometry in the diagnosis of middle ear effusion. Clinical Otolaryngology. 1994; 19: 185–9. De Melker RA. Diagnositic value of microtympanometry in primary care. British Medical Journal. 1992; 304: 96–8.



67. Van Balen FAM, Aarts AM, de Melker RA. Tympanometry by general practitioners: reliable? International Journal of Pediatric Otorhinolaryngology. 1999; 48: 117–23. 68. Mooler H, Tos M. Daily impedance audiometric screening of children. Scandinavian Audiology. 1992; 21: 9–14. 69. Fields MJ, Corwin P, White PS, Doherty J. Microtympanometry, microscopy and tympanometry in evaluating middle ear effusion prior to myringotomy. New Zealand Medical Journal. 1993; 106: 386–7. 70. Roush J, Drake A, Sexton JE. Identification of middle ear dysfunction in young children: a comparison of tympanometric screening procedures. Ear and Hearing. 1992; 13: 63–9. 71. Kazanas SG, Maw R. Tympanometry, stapedius reflex and hearing impairment in children with otitis media with effusion. Acta Otolaryngologica. 1994; 114: 410–4. 72. Van der Avoort SJC, van Heerbeek N, Zeilhuis GA, Cremers CWRJ. Sonotubometry: Eustachian tube ventilatory function test; A state-of-the art review. Otology and Neurotology. 2005; 26: 538–43. 73. Pellett FS, Cox LC, MacDonald CB. Use of acoustic reflectometry in the detection of middle ear effusion. Journal of the American Academy of Audiology. 1997; 8: 181–7. 74. Kemaloglu YK, Sener T, Beder L, Bayazit Y, Goksu N. Predictive value of acoustic reflectometry and tympanometry. Interantional Journal of Pediatric Otorhinolaryngology. 1999; 48: 137–42. 75. Oyiborhoro JMA, Olaniyan SO, Newman CW, Balakrishnan SL. Efficacy of acoustic otoscope in detecting middle ear effusion in children. Laryngoscope. 1987; 97: 495–8. 76. Vartiainen E. Otitis media with effusion in children with congenital or early-onset hearing impairment. Journal of Otolaryngology. 2000; 29: 221–3. 77. McCormick B. Behavioural hearing tests 6 months to 5 years. In: McCormack B (ed.). Paediatric audiology: 0-5 years. London: Whurr Publisher, 1989: 97–116. 78. Medical Research Council Multicentre Otitis Media Study Group. Influence of age, type of audiometry and child’s concentration on hearing thresholds. British Journal of Audiology. 2000; 34: 231–40. 79. Kumar M, Meheshwar A, Mahendran S, Oluwasamni A, Clayton MI. Could the presence of a Carhart notch predict the presence of glue at myringotomy? Clinical Otolaryngology. 2003; 28: 183–6. 80. Ahmad I, Pahor AL. Carhart’s notch: a finding in otitis media with effusion. International Journal of Pediatric Otorhinolaryngology. 2002; 64: 165–70. 81. Takata GS, Chan LS, Morphew T, Mangione-Smith R, Morton SC, Shekelle P. Evidence assessment of the accuracy of methods of diagnosing middle ear fluid in children with otitis media with effusion. Paediatrics. 2003; 112: 1379–87. 82. Dempster JH, MacKenzie K. Tympanometry in the detection of hearing impairments associated with otitis media with effusion. Clinical Otolaryngology and Allied Sciences. 1991; 16: 157–9.



Chapter 72 Otitis media with effusion 83. Butler CC, van der Linden MK, MacMillan H, van der Wouden JC. Screening in the first four years of life to undergo early treatment for otitis media with effusion. Cochrane Database of Systematic Reviews. 2003: CD004163. 84. Browning GG. Watchful waiting in childhood otitis media with effusion. Clinical Otolaryngology. 2001; 26: 1–2. 85. Rosenfeld AR, Kay D. Natural history of untreated otitis media. Laryngoscope. 2003; 113: 1645–57. 86. Schilder AGM, Zielhuis GA, Haggard MP, van den Brook P. Long-term effect on otitis media with effusion: otomicroscopic findings. American Journal of Otology. 1995; 16: 365–72. 87. Fria TJ, Cantekin EI, Eichler JA. Hearing acuity of children with otitis media with effusion. Archives of Otolaryngology. 1985; 111: 10–6. 88. Rakho T, Karma P. Pure-tone hearing thresholds in otologically healthy 5-year-old children in Finland. Archives of Otorhinolaryngology. 1989; 246: 137–41. 89. Hunter LL, Margolis RH, Giebink GS. Identification of hearing loss in children with otitis media. Annals of Otology, Rhinology, and Laryngology. 1994; 163: 59–61. 90. Bluestone CD, Beery QC, Paradise JL. Audiometry and tympanometry in relation to middle era effusions in children. Laryngoscope. 1973; 83: 594–604. 91. Brechtje A de Beer, Graamans K, Snik AFM, Ingels K, Zeilhuis GA. Hearing deficits in young adults who had a history of otitis media in childhood: Use of personal stereos had no effect on hearing. Pediatrics. 2003; 111: e304–8. 92. Hunter LL, Margolis RH, Rykken JR, Le CT, Daly KA, Geibink GS. High frequency hearing loss associated with otitis media. Ear and Hearing. 1996; 17: 1–11. 93. Roberts JE, Rosenfeld RM, Zeisel SA. Otitis media and speech and language: a meta-analysis of prospective studies. Pediatrics. 2004; 113: e238–48. 94. Butler CC, MacMillan H. Does early detection of otitis media with effusion prevent delayed language development? Archives of Disease in Childhood. 2001; 85: 96–103. 95. Roberts J, Hunter L, Gravel J, Rosenfeld AR, Berman S, Haggard M et al. Otitis media; hearing loss, and language learning: controversies and current research. Journal of Developmental and Behavioral Pediatrics. 2004; 25: 110–22. 96. Lous J. Otitis media and reading achievement: a review. International Journal of Pediatric Otorhinolaryngology. 1995; 32: 105–21. 97. Johnson DL, Swank PR, Owen MJ, Baldwin CD, Howie VM, McCormick DP. Effect of early middle ear effusion on child intelligence at three, five, and seven years of age. Journal of Pediatric Psychology. 2000; 25: 5–13. 98. Augustsoon I, Engstand I. Otitis media and academic achievements. International Journal of Pediatric Otorhinolaryngology. 2001; 57: 31–40. 99. Rutter M. A children’s behaviour questionnaire for completion by teachers: preliminary findings. Journal of



100.



101.



102.



103.



104.



105.



106.



107.



108.



109.



110.



111. 112.



113.



114.



] 909



Child Psychology and Psychiatry, and Allied Disciplines. 1967; 8: 1–11. Wilks J, Maw R, Peters TJ, Harvey I, Golding J. Randomised controlled trial of early surgery versus watchful waiting for glue ear: the effect on behavioural problems in pre-school children. Clinical Otolaryngology. 2000; 25: 209–14. Bennett KE, Haggard MP, Silva PA, Stewart IA. Behaviour and developmental effect of otitis media with effusion in the teens. Archives of Disease in Childhood. 2001; 85: 91–5. Golz A, Angel-Yeger B, Parush S. Evaluation of balance disturbances in children with middle ear effusion. International Journal of Pediatric Otorhinolaryngology. 1998; 43: 21–6. Casselbrant ML, Furman JM, Mandel EM, Fall PA, KursLasky M, Rockette HE. Past history of otitis media and balance in four-year-old children. Laryngoscope. 2000; 110: 773–8. Waldron MNH, Matthews JNS, Johnson IJM. The effect of otitis media with effusions on balance in children. Clinical Otolaryngology. 2004; 29: 318–20. Rosenfeld RM, Baya MH, Bower CM, Brookhouser PE, Casselbrant ML, Chan KH et al. Impact of tympanostomy tubes on child quality of life. Archives of Otolaryngology – Head and Neck Surgery. 2000; 126: 585–92. Richards M, Giannoni C. Quality-of-life outcomes after surgical intervention for otitis media. Archives of Otolaryngology – Head and Neck Surgery. 2002; 128: 776–82. Kubba H, Swan IRC, Gatehouse S. How appropriate is the OM6 as a discriminative instrument in children with otitis media. Archives of Otolaryngology – Head and Neck Surgery. 2004; 130: 705–9. Haggard MP, Smith SC, Nicholls EE. Measurement of quality of life outcomes in otitis media. In: Rosenfeld RM, Bluestone CD (eds). Evidence-based otitis media, 2nd edn. Hamilton, Ontario: BC Decker, 2003: 401–29. Kubba H, Swan IRC, Gatehouse S. Measuring quality of life in preschool children with sore throats and otitis media using the TAPQOL questionnaire. Otolaryngology and Head and Neck Surgery. 2005; 132: 647–52. Kubba H, Swan IRC, Gatehouse S. The Glasgow children’s benefit inventory: a new instrument for assessing healthrelated benefit after an intervention. Annals of Otology and Laryngology. 2004; 113: 980–6. Deafness Research UK. Glue ear: A guide for parents. London: Deafness Research UK, 2003. Butler CC, van der Voort JH, Oral or topical nasal steroids for hearing loss associated with otitis media with effusion. Cochrane Database of Systematic Reviews. 2002: CD001935. Shapiro GG, Bierman CW, Furukuwa CT. Treatment of persistent Eustachian tube dysfunction with aerosolized nasal dexamethasone phosphate versus placebo. Annals of Allergy. 1982; 49: 81–5. Tracy JM, Demain JG, Hoffman KM, Goetz DW. Intranasal beclomethasone as an adjunct to treatment of chronic



910 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



115. 116.



117.



118.



119.



120.



121.



122.



123.



124.



125.



126.



127.



128.



129.



middle ear effusion. Annals of Allergy, Asthma and Immunology. 1998; 80: 198–206. Williamson I. Otitis media with effusion. Clinical Evidence Concise. 2006; 16: 245–7. Mandel EM, Casselbrant ML, Rockette HE, Fireman P, KursLasky M, Bluestone CD. Sytemic steroid for chronic otitis media with effusion in children. Pediatrics. 2002; 110: 1071–80. Thomas CL, Simpson S, Butler CC, van der Voort JH. Oral or topical nasal steroids for hearing loss associated with otitis media with effusion in children. Cochrane Database of Systematic Reviews. 2006; 3: CD001935. Rosenfeld RM, Post JC. Meta-analysis of antibiotics for the treatment of otitis media with effusion. Otolaryngology and Head and Neck Surgery. 1992; 106: 378–86. Williams RL, Chalmers TC, Stange KC, Chalmers FT, Bowlin SJ. Use of antibiotics in preventing recurrent acute otitis media and treating otitis media with effusion. Journal of the American Medical Association. 1993; 270: 1344–51. Scottish Intercollegiate Guidelines Network. Diagnosis and management of childhood otitis media in primary care. A national clinical guideline. No 6, February 2003. Griffin GH, Flynn C, Bailey RE, Schultz JK. Antihistamines and/or decongestants for otitis media with effusion (OME) in children. Cochrane Database of Systematic Reviews. 2006; 4: CD003423. Pignataro O, Pignataro LD, Gallus G, Calori G, Cordaro CI. Otitis media with effusion and S-carboxymethylcysteine and/or its lysine salt: a critical overview. International Journal of Pediatric Otorhinolaryngology. 1996; 35: 231–41. Commins DJ, Koay BC, Bates GJ, Moore RA, Sleeman K, Mitchell B et al. The role of Mucodyne in reducing the need for surgery in patients with persistent otitis media with effusion. Clinical Otology. 2000; 25: 274–9. Reidpath DD, Glasziou PP, Del Mar C. Systematic review of autoinflation for treatment of glue ear in children. British Medical Journal. 1999; 318: 1177–8. Harrison H, Fixsen A, Vickers A. A randomised comparison of homeopathy and standard care for treatment of glue ear in children. Complementary Therapies in Medicine. 1999; 7: 132–5. Freemantle N, Sheldon TA, Song F, Long A. The treatment of persistent glue ear in children. Effective Health Care Bulletin No. 4. York: University of York, NHS Centre for Reviews and Dissemination, 1992. Weigel MT, Parker MY, Goldsmith MM, Postma DS, Pilsbury HC. A prospective randomised study of four commonly used tympanostomy tubes. Laryngoscope. 1989; 99: 252–6. Richards SH, Kilby D, Shaw JD, Campbell H. Grommets and glue ears: A clinical trial. Journal of Laryngology and Otology. 1971; 85: 17–22. Shone GR, Griffith IP. Titanium grommets: a trial to assess function and extrusion rates. Journal of Laryngology and Otology. 1990; 104: 197–9.



130. Guttenplan MD, Tom WC, DeVito MA, Handler SD, Wetmore RF, Potsic WP. Radial versus circumferential incision in myringotomy and tube placement. International Journal of Pediatric Otorhinolaryngology. 1991; 21: 211–5. 131. Heaton JM, Bingham BJG, Osbourne J. A comparison of performance of Shepard and Sheehy collar button ventilation tubes. Journal of Laryngology and Otology. 1991; 105: 896–8. 132. Walker P. Ventilation tube duration versus site of placement. Australian and New Zealand Journal of Surgery. 1997; 67: 571–2. 133. Hern JD, Jonathan DA. Insertion of ventilation tubes: does the site matter? Clinical Otology. 1999; 24: 424–5. 134. Youngs RP, Gatland DJ. Is aspiration of middle ear effusions prior to ventilation tube insertion really neccesary? Journal of Otolaryngology. 1988; 17: 204–6. 135. Egeli E, Muzaffer K. Is aspiration necessary before tympanostomy tube insertion? Laryngoscope. 1998; 108: 443–4. 136. Lous J, Burton MJ, Felding JU, Ovesen T, Rovers MM, Williamson I. Grommets (ventilation tubes) for hearing loss associated with otitis media in children. Cochrane Database of Systematic Reviews. 2005: CD001801. 137. Maw R, Wilks J, Harvey I, Golding J. Early surgery compared with watchful waiting for glue ear and effect on language development in preschool children: a randomised trial. Lancet. 1999; 353: 960–3. 138. Black NA, Sanderson CF, Freeland AP, Vessey MP. A randomised controlled trial of surgery for glue ear. British Medical Journal. 1990; 300: 1551–6. 139. Dempster JH, Browning GG, Gatehouse S. A randomised study of the surgical management of children with persistent otitis media with effusion associated with a hearing impairment. Journal of Laryngology and Otology. 1993; 107: 284–9. 140. Maw R, Herod F. Otoscopic, impedance and audiometric findings in glue ear treated by adenoidectomy and tonsillectomy; a prospective randomised trial. Lancet. 1986; 1: 1399–402. 141. Rovers MM, Black N, Browning GG, Maw R, Zielhius GA, Haggard MP. Grommets in otitis media with effusion: an individual patient data meta-analysis. Archives of Disease in Childhood. 2005; 90: 480–5. 142. Medical Research Council Multicentre Otitis Media Study Group. The role of ventilation tube status in the hearing levels in children managed for bilateral persistent otitis media with effusion. Clinical Otolaryngology. 2003; 28: 146–53. 143. Paradise JL, Feldman HM, Campbell TF, Dollaghan CA, Colburn DK, Bernard BS et al. Effect of early or delayed insertion of tympanostomy tubes for persistent otitis media on developmental outcomes at the age of three years. New England Journal of Medicine. 2001; 344: 1179–87. 144. Rovers MM, Krabbe PFM, Straatman H, Ingels K, van der Wilt G-J, Zielhuis GA. Randomised controlled trial of the



Chapter 72 Otitis media with effusion



145.



146.



147.



148.



149.



150.



151.



152.



153.



154.



155.



effect of ventilation tubes (grommets) on quality of life at age 1–2 years. Archives of Disease in Childhood. 2001; 84: 45–9. Paradise JL, Campbell TF, Dollaghan CA, Feldman HM, Bernard BS, Colburn DK et al. Developmental outcomes after early or delayed insertion of tympanostomy tubes. New England Journal of Medicine. 2005; 353: 576–86. Hartman M, Rovers MM, Ingels K, Zeilhius GA, Severns JL, van der Wilt GJ. Economic evaluation of ventilation tubes in otitis media with effusion. Archives of Otolaryngology – Head and Neck Surgery. 2001; 127: 1471–6. Kay DJ, Nelson M, Rosenfeld RM. Meta-analysis of tympanostomy tube sequelae. Otolaryngology and Head and Neck Surgery. 2001; 124: 374–80. Shinkwin CA, Murty GE, Simo R, Jones NS. Per-operative antibiotic/steroid prophylaxis of tympanostomy tube otorrhoea; Chemical or mechanical effect? Journal of Laryngology and Otology. 1996; 110: 531–3. Garcia P, Gates GA, Schechtman KB. Does topical antibiotic prophylaxis reduce post-tympanostomy tube otorrhoea? Annals of Otology, Rhinology, and Laryngology. 1994; 103: 54–8. Daly KA, Giebink GS, Lindgren B, Margolis RH, Westover D, Hunter LL et al. Randomized trial of the efficacy of trimethoprim-sulfamethoxazole and prednisolone in preventing post-tympanostomy tube morbidity. Pediatric Infectious Disease Journal. 1995; 14: 1068–74. Pearson CR, Thomas MR, Cox HJ, Garth JN. A cost–benefit analysis of the post-operative use of antibiotic ear drops following grommet insertion. Journal of Laryngology and Otology. 1996; 110: 527–30. Tos M, Stangerup S-E, Larsen P. Dynamics of eardrum changes following secretory otitis media. Archives of Otolaryngology – Head and Neck Surgery. 1987; 113: 380–5. Johnston LC, Feldman HM, Paradise JL, Bernard BS, Colburn K, Casselbrant ML et al. Tympanic membrane abnormalities and hearing levels at the ages of 5 and 6 years in relation to persistent otitis media and tympanostomy tube insertion in the first 3 years of life: A prospective study incorporating a randomised controlled trial. Pediatrics. 2004; 114: e58–67. De Beer BA, Schilder AGM, Ingels KI, Snik AF, Zielhuis GA, Graamans K. Hearing loss in young adults who had ventilation tube insertion in childhood. Annals of Otology, Rhinology, and Laryngology. 2004; 113: 438–44. Pringle MB. Grommets, swimming and otorrhoea – a review. Journal of Laryngology and Otology. 1993; 107: 190–4.



] 911



156. Parker GS, Tami TA, Madox MR, Wilson JF. The effect of water exposure after tympanostomy tube insertion. American Journal of Otolaryngology. 1994; 15: 193–6. 157. Walker P. Pediatric adenoidectomy under vision using suction-diathermy ablation. Laryngoscope. 2001; 111: 2173–7. 158. Murray N, Fitzpatrick P, Guarisco JL. Powered partial adenoidectomy. Archives of Otolaryngology – Head and Neck Surgery. 2002; 128: 792–6. 159. Maw R, Bawden R. Spontaneous resolution of severe chronic glue ear in children and the effect of adenoidectomy, tonsillectomy and insertion of ventilation tubes (grommets). British Medical Journal. 1993; 306: 756–60. 160. Coyte PC, Croxford R, McIsaac W, Feldman W, Friedberg J. The role of adjuvant adenoidectomy and tonsillectomy in the outcome of the insertion of tympanostomy tubes. New England Journal of Medicine. 2001; 344: 1188–95. 161. Dempster JH, Browning GG. Eustachian tube function following adenoidectomy: an evaluation by sniffing. Clinical Otology. 1989; 14: 411–4. 162. Leighton SEJ, Rowe-Jones JM, Knight JR, Moore-Gillon VL. Day case adenoidectomy. Clinical Otology. 1993; 18: 215–9. 163. Ahmed K, McCormick MS, Baruah AK. Day-case adenoidectomy – is it safe? Clinical Otolaryngology and Allied Sciences. 1993; 18: 406–9. 164. Paradise JL, Bluestone CD, Colburn DK, Bernard BS, Smith CG, Rockette HE et al. Adenoidectomy and adenotonsillectomy for recurrent acute otitis media: parallel randomised clinical trials in children not previously treated with tympanostomy tubes. Journal of the American Medical Association. 1999; 282: 945–53. 165. Jardine AH, Griffiths MV, Midgley E. The acceptance of hearing aids for children with otitis media with effusion. Journal of Laryngology and Otology. 1999; 113: 314–7. 166. Maheshwar AA, Milling MAP, Kumar M, Clayton AT. Use of hearing aids in the management of children with cleft palate. International Journal of Pediatric Otorhinolaryngology. 2002; 66: 55–62. 167. Sheahan P, Blayney AW. Cleft palate and otitis media with effusion. Revue de Laryngologie Otologie Rhinologie. 2003; 124: 171–7. 168. Iino Y, Imamura Y, Harigai S, Tanaka Y. Efficacy of tympanostomy tube insertion for otitis media with effusion in children with Down syndrome. International Journal of Pediatric Otorhinolaryngology. 1999; 49: 143–9.



73 Acute otitis media in children PETER REA AND JOHN GRAHAM



Introduction Definition Diagnosis Aetiology Epidemiology Management options Outcomes Complications



912 912 913 914 917 917 921 921



Conclusion Key points Best clinical practice Deficiencies in current knowledge and areas for future research Acknowledgements References



925 925 925 926 926 926



SEARCH STRATEGY The data in this chapter are supported by searches of Medline and the Cochrane Controlled Trials Register, using the term acute otitis media. Reference lists were reviewed for further articles, and authors of recent presentations contacted personally for their reference lists.



INTRODUCTION For such a common childhood infection, acute otitis media (AOM) remains something of an enigma. It is hard to diagnose accurately and on existing evidence, as opposed to custom and tradition, there is still a high level of uncertainty over how it should best be treated. This is against a background of increasing bacterial resistance to antibiotics. There is plenty of evidence in the literature of the relative frequencies of viral and bacterial pathogens in AOM, but this is often of little help to the clinician on the spot in an individual case. Some of the epidemiological evidence is also relatively ‘soft’, since it is based on the flawed premise that AOM can accurately be diagnosed from the history and otoscopy alone, unsupported by tympanometry or tympanocentesis. In this chapter, the authors have tried to thread their way through the often conflicting evidence about the practical management of AOM at the same time as covering what is known of the pathology, epidemiology and complications of this commonest of childhood illnesses.



We have found ourselves uncomfortably often using expressions such as ‘uncertainty,’ ‘insufficient evidence,’ or ‘limited information,’ and quoting [****] and [***] levels of evidence. Even a metanalysis, the supposed gold standard of evidence, is only as good as the studies it covers. It must be remembered that the prevalence of otitis media with effusion (OME), ‘glue ear,’ only began to be widely appreciated after the development of the twin tools of universal hearing screening and the tympanometer in the 1960s and early 1970s, respectively. The accuracy of correct reporting of cases of AOM lags behind that of OME mainly because in the majority of large population studies of AOM the essential presence of fluid in the middle ear is not confirmed by tympanometry and audiometry.



DEFINITION AOM may be defined clinicopathologically as inflammation of the middle ear cleft of rapid onset and infective



Chapter 73 Acute otitis media in children



origin, associated with a middle ear effusion and a varied collection of clinical symptoms and signs. It is synonymous with acute suppurative otitis media. It normally develops behind an intact tympanic membrane, but may include acute infections arising in the presence of ventilation tubes or existing tympanic membrane perforations. The requirement to confirm a middle ear effusion, and the nature of the symptoms and signs, vary between authors.1, 2 The literature supports four broadly defined subgroups of AOM. 1. Sporadic episodes occurring as infrequent isolated events, typically occurring with upper respiratory tract infections. 2. Resistant AOM: persistence of symptoms and signs of middle ear infection beyond three to five days of antibiotic treatment. 3. Persistent AOM: persistence or recurrence of symptoms and signs of AOM within six days of finishing a course of antibiotics. 4. Recurrent AOM: either three or more episodes of AOM occurring within a six-month period, or at least four or six episodes within a 12 month period (no consensus has been reached on the latter). Groups two and three appear similar at first glance and this distinction may be questioned. It is included to maintain some consistency with the wider literature. Grading of the severity of an episode has been attempted and has merit both clinically and for research. Pyrexia from 37.5–391C, vomiting and severity of otalgia have been used.3, 4 [*]



DIAGNOSIS Diagnosis is based on the combination of often nonspecific symptoms, evidence of inflammation of the middle ear cleft and, by some authors, by the additional confirmation of a middle ear effusion. Diagnostic difficulty has affected the quality of research into AOM. There may well not be a clear history of a crescendo of otalgia in a coryzal child, followed by rapid symptomatic relief associated with tympanic membrane perforation and associated blood-stained otorrhoea. The difficulty in establishing clear diagnostic guidelines has been highlighted in an analysis of 80 studies of AOM.5 In diagnosing AOM, only 52.5 percent of the studies cited middle ear effusions, 32.5 percent included symptoms and signs of inflammation and 2.5 percent considered the rapidity of onset. Clinicians recognize this difficulty. A large multinational study rated clinicians diagnostic certainty in children under one year of age at only 58 percent, rising to 73 percent in those over 31 months.6



] 913



Symptoms Diagnosis by symptomatology alone is inaccurate because of the young age of most patients, and the nonspecific nature of the symptoms. One-third of children may have no ear-related symptoms. Two-thirds may be apyrexial.7 Symptoms suggestive of AOM include rapid onset of otalgia, hearing loss, otorrhoea, fever, excessive crying, irritability, coryzal symptoms, vomiting, poor feeding, ear-pulling and clumsiness (Table 73.1). AOM most commonly develops three to four days after the onset of coryzal symptoms. The otalgia will settle within 24 hours in two-thirds of children without treatment.9 The otorrhoea, if present, is mucopurulent and may be blood-stained. Symptomatic relief is obtained without treatment in 88 percent by day four to seven. The hearing loss, caused by the middle-ear effusion, occurs early in the illness and may persist at greater than 20 dB for one month in over 30 percent, and two months in 20 percent of children. [**/*]



Signs The child may appear unwell, and may rub his or her ear. The diagnosis is often confirmed, rightly or wrongly, by an attempt at otoscopic assessment of the tympanic membrane. However, a poorly functioning otoscope, the moving target of a child’s head, the narrow ear canal of an infant, the natural redness of the tympanic membrane of a screaming child, wax and, above all, the inaccuracy of an untrained (or even trained) eye, straining to interpret a two-dimensional image, all combine to make otoscopy an imprecise art. Since trained observers have been shown to have only an 85 percent accuracy in otoscopic diagnosis,10 it would not be surprising for a sensible primary care physician to rely more on history and the general aspect of a child than on otoscopic findings. With these reservations, diagnosis may be supported by otoscopic assessment of tympanic membrane colour, position and mobility (Figure 73.1). In AOM the tympanic membrane is usually opaque. It is most commonly yellow, or Table 73.1 Relation of reported symptoms to presence of AOM in 302 children under four years of age. Symptom



Earache Restlessness Rhinitis Cough Fever



Sensitivity Specificity Positive predictive value



Negative predictive value



60 64 96 83 69



78 68 74 61 53



92 51 8 17 23



Reprinted from Ref. 8, with permission.



83 46 41 40 38



914 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY with recurrent infection of ventilation tubes may also merit investigation for primary ciliary dyskinesia, particularly if nasal and pulmonary symptoms coexist. [**/*]



Differential diagnosis Pain may be referred from tonsillitis, teething, temporomandibular joint disorders or simply be the result of an uncomplicated upper respiratory tract infection. In a screaming child, the tympanic membrane may well appear red. Diagnostic confusion may occur with acute mastoiditis, otitis media with effusion, otitis externa, trauma, Ramsey Hunt syndrome and bullous myringitis. Very rarely, AOM may be the first indication of serious underlying disease, such as Wegener’s granulomatosis or leukaemia.



Figure 73.1



Otoscopic findings in early AOM.



yellowish pink in colour, being red in only 18–19 percent.7, 10 The position of the tympanic membrane reliably predicts AOM only when it is bulging. Hypomobility of the drum demonstrated by pneumatic otoscopy has been shown to aid diagnosis10 and is felt essential in some countries,1 although others including the Dutch2 take a more pragmatic view and do not include this in their diagnostic criteria. Should the drum have perforated, or a ventilation tube be in situ, mucopurulent otorrhoea will be seen. [**/*]



Investigations Tympanometry may be used to establish the presence of a middle ear effusion, but is not usually available. Tympanocentesis and culture of middle ear effusion have been used in a number of studies assessing diagnostic accuracy of clinical signs, and establishing the organisms prevalent in a community. It is rarely required to make the diagnosis, though may be considered in high risk children such as the immunocompromised, an unwell neonate, those that fail to respond to conventional treatment and children who are seriously ill or have complications of AOM. Taking a bacterial swab of persistent otorrhoea following perforation is recommended. Nasopharyngeal swabbing for bacterial culture has been assessed but the correlation with middle ear organisms has been too weak to recommend it clinically.11 Specific investigation may be prompted by recurrent AOM not responsive to conventional treatment. Both iron deficiency anaemia and white blood cell disorders have been associated with AOM, so a full blood count is indicated. Immunoglobulin assay may be appropriate: Ig A, G (with subclasses) and M are typically assessed. Children



AETIOLOGY Microbiological, anatomical and environmental factors combine with altered host defence mechanisms to predispose to infection. Genetic predisposition to recurrent AOM is being increasingly cited in the literature.



Infective agents AOM results from infection of the middle ear cleft. Both bacterial and viral infections are implicated. These infections may occur in isolation or combination.



VIRUSES



Clinically it is apparent that AOM is commonly associated with viral upper respiratory tract infections. As our ability to identify these improves, the role of viruses in the aetiology of AOM is becoming clearer. Increasing use of polymerase chain reaction assays for respiratory viruses suggests 60–90 percent of cases of AOM may be associated with viral infection.12 In one study, a specific viral cause of upper respiratory tract infection was shown in 41 percent of children with AOM.13 The viruses most commonly associated with AOM vary between studies, but in decreasing frequency include:     



respiratory syncytial virus (RSV); influenza A virus; parainfluenza viruses; human rhinovirus; adenoviruses.



This heterogeneity is important when considering vaccination against viruses as a prophylactic measure. The mechanism by which they give rise to AOM is likely to vary between viruses. Viral material has been demonstrated in the middle ear aspirates of children with



Chapter 73 Acute otitis media in children



AOM in 48–71 percent of cases.12 The viral material may arrive either passively along the Eustachian tube along with other nasopharyngeal secretions or may actively invade the middle ear cleft possibly by haematogenous spread. These alternative routes of entry are suggested by the wide variation in rates of isolation of specific viral strains in the middle ear during systemic infection, ranging from 4 to 74 percent of cases dependent upon the specific virus. If all arrived passively, similar rates of isolation would be expected. This implies some viruses may be actively invading the middle ear cleft, and may be contributing directly to mucosal inflammation. RSV invaded the middle ear most frequently.13 In contrast, those arriving passively appear to cause AOM by virtue of their action on the Eustachian tube, on bacterial adherence, and on host immunity. There is good clinical and animal evidence that viral infection affects Eustachian tube function.12 At a cellular level there is release of multiple inflammatory mediators from cells within the nasopharynx. Ciliated epithelial cells numbers decline, mucus production increases in the Eustachian tube and negative middle ear pressure results. This is likely to predispose to AOM. Alteration of host immunity has been documented after viral infections, increasing susceptibility to bacterial infections. Cell-mediated immunity has been shown to be affected by RSV infection, and neutrophil function altered by influenza viruses. In a study of children with bronchiolitis caused by RSV, 62 percent developed AOM. Bacteria were isolated from the middle ear in all these children. The ability of bacteria to colonize and adhere to the nasopharyngeal epithelium appears to be increased by certain viral infections. Increased colonization by pathogenic bacteria may predispose to AOM. Viral and bacterial infection coexist in the middle ear cleft in AOM in as many as two-thirds of cases where viruses have been identified. This is important as clinical studies show that children who have both viruses and bacteria in their middle ear are very much more likely to have a poor response to antibiotics when compared to those with bacteria only (33 versus 3 percent failure respectively, in one study14). Why this should be is unclear, but may be related to the greater concentrations of inflammatory mediators in ears in which both bacteria and viruses are present. [**/*] BACTERIA



The bacteria isolated from the middle ear in AOM are shown in Table 73.2.15, 16 In persistent or recurrent bacterial AOM, repeat culture of middle ear aspirates has failed to grow pathogenic bacteria in 30–50 percent of patients, implying that inflammation may persist despite the eradication of the infecting organism. The spectrum of organisms is similar to that in isolated episodes. In the 1980s H. influenzae was the most common organism identified



Table 73.2



Bacteria associated with AOM.



Bacteria Haemophilus influenzae Moraxella catarrhalis Streptococcus pyogenes Staphylococcus aureus a



] 915



Incidence (%) 16–37a 11–23 Up to 13 Up to 5



There are some 90 serotypes.



in persistent or recurrent AOM, but this has been replaced by drug-resistant Pneumococcus. After antibiotic treatment for recurrent AOM it is now estimated that 50 percent of H. influenzae are beta-lactamase producing. A similar proportion of pneumococci are penicillin resistant.16 Penicillin resistance in pneumococci results from decreased penicillin binding protein on the bacterial cell walls, so reducing the affinity for penicillin-related drugs, but means that resistance may often be overcome by increasing drug dosage. This is not the case with beta-lactamase producing organisms. Most Moraxella catarrhalis are now beta-lactamase producing. Studies on HIV-positive children suggest a similar spectrum and prevalence of infecting organisms as occurs in immunocompetent children, except where the child is severely immunosuppressed, when a higher percentage of Staphylococcus aureus has been reported. [**]



Routes of spread of infection Three potential routes are described: the Eustachian tube, tympanic membrane perforations or grommets, and haematogenous. The Eustachian tube is traditionally assumed to be the main route by which organisms reach the middle ear, though there are relatively few studies to confirm this. It is speculated that negative middle ear pressure may facilitate the movement of bacteria up the Eustachian tube.17 Circumstantial evidence is also gained from similarities in organisms cultured from the post-nasal space and the middle ear cleft in AOM. Whether anatomical or physiological differences predispose to AOM is unclear. Studies of Native Americans, who are prone to otitis media, suggest their Eustachian tubes are shorter, straighter and more patulous than in whites, but also that they have a low passive tubal resistance.18 Research has found no difference in tubal dimensions in otitis prone and non-prone children. However, altered tubal function may play a role. Specifically, otitis-prone children have been shown to have significantly poorer active tubal function (muscular opening function). Pathogen entry through tympanic membrane perforations or ventilation tubes is most commonly associated with water exposure.



916 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Haematogenous spread is inferred from the evidence provided by studies of viral identification in the blood and middle ear as described previously. It was shown that the wide variation in rates of identification of specific viral strains from the middle ear could not be explained by passive Eustachian tube transport alone.12 [**/*]



Risk factors GENETIC FACTORS



There is growing evidence that recurrent AOM is largely genetically determined. It is likely many genes are involved. There are numerous studies suggesting a familial association. A metanalysis of risk factors has shown that when one family member had AOM the risk increased for other family members (relative risk 2.63).19 Racial differences are well described with increases in American Indians, Eskimos and Australian Aboriginals. However, environmental factors, such as poor economic status, may contribute to the increased risks in these groups. The most powerful evidence comes from twin studies, in particular comparison of monozygotic and dizygotic twins in whom the occurrence of AOM was compared.18 Many immune related mechanisms, which are likely to have a genetic basis, have been proposed. Certain human leukocyte antigen (HLA) classes have been shown to be significantly associated with increased risk of AOM. Maternal blood group A is reported to an independent risk factor (relative risk 2.82). Atopy has also been associated with increased risk of developing AOM. [****/***/**]



IMMUNE FACTORS



Our understanding of the immune response to AOM remains incomplete. However, a number of specific associations have been identified which suggest that certain defective or immature pathways may predispose to infection. Low levels of IgG2 subclasses have been reported in several studies to be more common in otitisprone children. Those with IgG2 deficiency were shown to be three times more likely to develop post-ventilation tube insertion otorrhoea for example. Delayed maturation of anti-pneumococcal antibodies (IgG1 and IgG2 were studied) does appear to predispose to AOM. This may explain in part why children grow out of AOM as immunity matures. Defective complement-dependent opsonization has been associated with recurrent AOM and diarrhoea in infancy.18 This is caused in some examples by low concentrations of mannose-binding protein which acts as an opsonin. This appears to be a common defect with over 20 percent of children with recurrent AOM affected in some studies. This may be particularly important in infancy when the antibody repertoire is limited.



Aberrant expression of critical cytokines, such as tumour necrosis factor and interleukins, resulting in suboptimal host defence, has been postulated as a cause for persistent infection. Expression of mucin genes, at least nine of which have been identified, may differ in those predisposed to AOM. Middle ear mucosa expresses specifically the MUC5B gene. Mucin genes regulate the production of mucin. Limited evidence is beginning to emerge that over-expression may alter the mucociliary transport system.18 A number of studies on children with HIV infection have yielded conflicting results. Advanced disease associated with low CD4 counts does seem to be associated with an increased incidence of AOM. [**/*]



ENVIRONMENTAL FACTORS



There are many reports on the relative contribution of environmental factors. These are important as it may be possible to modify them. The most important is almost invariably stated to be day-care attendance outside the home. The larger the number of children in the group, the greater the risk. Day care outside the home carries a relative risk (rr) for AOM of 2.45, compared to a risk of 1.59 for children cared for in their own home. The incidence of AOM appears to follow that of seasonal upper respiratory tract infections (URTI) in the winter months. Breastfeeding for three months is protective (rr, 0.87). Use of a pacifier (dummy) carries a relative risk of 1.45.19 Poor socioeconomic status associated with poor housing and overcrowding has been reported to be associated with AOM (overcrowding: rr, 5.55 in a Greenlandic population, for example). Passive smoke exposure from parental smoking is weakly associated (rr, 1.0–1.6). There is more limited evidence to support the role of dietary factors, in particular cow’s milk allergy, in predisposing to AOM. [****]



SYNDROMIC ASSOCIATIONS



Syndromes associated with abnormalities of skull base anatomy are well recognized as being associated with chronic middle ear disease, but less is published on associations with AOM. Children with Turner’s syndrome do suffer more frequent episodes of AOM. Down syndrome predisposes to middle ear disease, including AOM. In cleft palate there is only a minor increase reported. Certainly Eustachian tube dysfunction in these groups predisposes to middle ear effusion, but it is not clear whether it is this dysfunction or an increase in risk secondary to subtle immunological factors that predisposed to infection. No increase is found in children with primary ciliary dyskinesia if grommets are not inserted, or cystic fibrosis. A direct association between iron deficiency anaemia, and the degree of anaemia and frequency of AOM has been reported. [**]



Chapter 73 Acute otitis media in children



EPIDEMIOLOGY AOM is one of the commonest illnesses of childhood. It accounts for approximately 25 percent of all prescriptions for children under ten years of age in the USA, for example. Its incidence appears highest in the first year of life, more specifically the second six months of life in most studies, and gradually reduces with increasing age. This progression was shown by Strangerup and Tos20 who reported an incidence of a first episode of AOM in 22 percent in the first year of life, 15 percent in year two and 10 percent in year three, falling to 2 percent by year eight (Figure 73.2). Epidemiological studies have been compromised by difficulty in achieving accuracy in diagnosis when large numbers of children are being assessed, hence there are wide variations in reported numbers. Incidences of over 60 percent are stated in some reports of infants up to age one year. By age three years, some 50–70 percent of all children will have had at least one episode of AOM, and at least 75 percent by the age of nine years. The incidence of AOM certainly varies with the seasonal incidence of viral upper respiratory infections. There are reports that it is increasing over a period of years. Possible reasons include increased day nursery attendance and changes in diagnostic awareness. Recurrent AOM has been reported in 5 percent of children under two years of age. Others have reported that by age three, half of children will have had at least three episodes. An important indicator of future problems is a first episode before nine months of age: these children have a one in four risk of developing recurrent AOM. In the first two years of life, AOM occurs bilaterally in 80 percent of cases. After six years of age, it is unilateral in 86 percent.20



MANAGEMENT OPTIONS Most children with AOM will get better quickly and without treatment. Some will not. A very small number 90 80 % of children



70 60 50 40 30 20 10 0 1



2



3



4 5 6 Age (years)



7



8



9



Figure 73.2 Cumulative incidence of acute otitis media (i.e. how many children have had at least one episode in their life). Redrawn from Ref. 20, with permission.



] 917



may develop potentially serious complications. Current debate questions whether and for whom treatment is required, and the role of prophylactic strategies. As serious complications are rare it can be difficult to obtain high quality evidence of how effective are prophylaxis and the treatment of episodes of AOM in preventing such complications.



Management of acute episodes CONSERVATIVE TREATMENT



Most children will benefit from simple analgesics and anti-pyrexials, in a quiet supportive environment. Paracetamol and ibuprofen are most commonly used in the UK. There is limited experimental animal evidence showing that ibuprofen provides additional benefit by reducing mucosal inflammation when taken in combination with amoxicillin.



MEDICAL TREATMENT



Antibiotics Uncertainty over the use of antibiotics is reflected in wide variations in usage between countries, ranging from 31 percent in the Netherlands to 98 percent in the USA.9 There is, however, an increasingly good evidence-base for the most appropriate management in children over two years of age. In children under two, the evidence-base is weaker. Antibiotics, if not prescribed initially, should be given to a child who fails to improve after two to three days of ‘watchful waiting’, and to all children with an ‘irregular’ illness course. They should also be given to ‘high risk’ children, defined by the Dutch as children with craniofacial abnormalities, Down syndrome, immunodeficiencies and those under the age of two years suffering a recurrent episode of AOM.2 Guidelines must be adapted to suit local experience, for example in regions of the world with a high incidence of complications of AOM where all patients may be regarded as high risk. A recent metanalysis9 has addressed the question of whether antibiotics should be given at initial consultation. Two-thirds of children recovered within 24 hours of the start of treatment, and 80 percent by days two to seven, with or without antibiotics. This, of course, raises the question as to whether the children did indeed have AOM in the first place. Antibiotics did lead to 5 percent fewer children overall having pain between days two and seven. That equates to 17 children needing to be treated to prevent one child experiencing pain during days two to seven. Relatively few data were available on hearing loss at one and two months post-infection, but no differences were found between those who received antibiotics and those who did not. No significant differences were found



918 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY on progression of disease or relapse of symptoms. Similarly, no differences were found in complications of AOM. However, those taking antibiotics suffer nearly double the side effects, such as diarrhoea, than those who do not, and run a greater risk of developing antibiotic resistant bacteria. The length of treatment was addressed in a separate metanalysis.21 Short (five day) and long (ten day) courses of treatment were compared. At 8–19 days the weighted mean failure rate in the short course was 19 percent and 13.7 percent in the long course. By days 20–30, this had converged to 15.7 versus 12.5 percent, which was nonsignificant. It implied that 17 children would need to receive the long course to avoid one treatment failure. The authors concluded that five days of treatment was appropriate in uncomplicated infections in low risk children over two years of age without recurrent AOM or tympanic membrane perforation. Under two years of age evidence is weaker that short course treatment is adequate. Attempts have been made to identify a subgroup of children who may benefit from antibiotics. Younger age may be an important determinant, but good evidence is lacking because of diagnostic difficulties in this group. However, while published data show only modest benefit from antibiotic treatment between six and 24 months of age22 (number needed to treat (NNT) = seven for one symptomatic improvement at day four), most would recommend treatment below two years of age. Using short-term symptomatic outcome markers at day three, it has been shown that immediate antibiotics may benefit those children presenting with higher temperatures (437.51C) or vomiting (NNT = 3–6).4 AOM occurring in the presence of a tympanic membrane perforation or ventilation tubes may be treated equally successfully with oral or topical antibiotics. The potential ototoxicity of topical aminoglycoside ear drops in these cases is well recognized and dose dependent, therefore prolonged topical treatment should be avoided. [****] Which antibiotic? This should be determined by national recommendations. Amoxicillin remains the first choice in most centres, but at higher than previously recommended doses (80 mg/kg/day) if drug-resistant pneumococci are common in a particular country or region, or macrolides for penicillin-sensitive patients. For persistent or resistant episodes, national policies should be sought depending on the prevalence of beta-lactamase-producing organisms and culture results if available. Options include amoxicillin-clavulonate or cefuroxime axetil orally, or intra-muscular ceftriaxone (US Centre for Disease Control and Prevention). Antihistamines and decongestants A metanalysis of the use of oral or intranasal antihistamines and/or decongestants concluded that their use could



not be supported, and that medication side effects were higher when they were used together. While combining the two treatments was shown to slightly reduce persistent AOM at two weeks (NNT = 10.5) the result may have been biased by the design of the studies.23 [****]



SURGICAL TREATMENT



Surgery has a limited role in the treatment of an uncomplicated episode of AOM. Myringotomy was practised in the pre-antibiotic era, and indeed was continued until the late 1980s in some countries as a first-line treatment for AOM. However, there are now a number of good studies showing that myringotomy plus antibiotics offers no advantage over antibiotics alone. Myringotomy alone has a worse outcome than either of the antibiotic groups.24 Myringotomy is reserved for severe cases where complication is present or suspected, to relieve severe pain or when microbiology is strongly required. [****]



Management of recurrent acute otitis media ALTERATION OF RISK FACTORS



It may be possible to alter many of the environmental risk factors discussed previously. Parents should be reassured of the benign natural history of AOM, as these children have been shown to be more demanding than those without recurrent disease, and their mothers more anxious about their care. The most readily modifiable risk factor is exposure to other children. AOM increases with the number of children in day care, the length of time a child spends in day care each week, how young a child is when introduced into day care, the presence of children under two years of age in the day-care setting and having a sibling in day care. Advice should include sitting a child semi-upright if bottle-fed and avoiding passive smoke inhalation. Restricting the use of pacifiers, particularly after infancy, should be recommended for otitis-prone children. The mother may be advised to continue breastfeeding for at least six months after future pregnancies and increasing vitamin C intake and avoiding alcohol in the third trimester, both of which have been weakly associated with AOM. The role of food allergies, in particular cow’s milk, is still unclear. No effective role has as yet been shown for homeopathic remedies. [**]



MEDICAL PROPHYLAXIS



Antibiotics Antibiotic prophylaxis has the potential to cause problems, but should be considered for recurrent AOM. Many organisms need to be covered so a broad-spectrum drug is required. Risks for the development of resistant



Chapter 73 Acute otitis media in children



organisms increase, adverse drug reactions may occur and active disease may be masked. Studies of prophylaxis of recurrent AOM invariably treat each individual recurrence with additional antibiotics. Trials therefore compare antibiotic prophylaxis versus placebo between acute episodes. The natural history of recurrent AOM is reassuring. Over 50 percent of children having no treatment between attacks will not suffer a further episode in the following six months. Indeed, only one in eight continues to suffer recurrent AOM (i.e. three or more episodes) during the trials, if treated only for acute episodes. However, metanalysis does show a benefit of antibiotic prophylaxis equating to a reduction of approximately 1.5 episodes per 12 months of antibiotic treatment given, above that expected from the natural history. So one child would need eight months of treatment to avoid one episode of AOM. There was a trend for those treated with sulfisoxazole to do better than those treated with amoxicillin, both being used at half the therapeutic dosage.25 Recommendations are for six months of treatment through the winter months in children who do not have background OME. Those studies that assessed the length of time a child has OME in association with AOM showed that while antibiotic prophylaxis may reduce the incidence of AOM, it does not reduce the length of time with OME. This is important as antibiotic prophylaxis may therefore be most appropriate for children not prone to OME, while ventilation tubes may be indicated for those prone to OME. Most trials exclude ‘high risk’ children, who are often most in need of treatment. On the basis of the metanalysis described above, which shows a modest benefit with the use of antibiotic prophylaxis, there may also be a place for its use in the management of high-risk children with recurrent AOM, despite the absence of specifically targeted studies. [****] Xylitol Xylitol is a commonly used sweetener that inhibits pneumococcal growth and the attachment of pneumococci and Haemophilus to nasopharyngeal cells. Studies in daycare nurseries using chewing gum or syrup have suggested reductions of 30–40 percent in the occurrence of AOM. However, this translates to 1–1.5 episodes per year.26 It is ineffective if used in acute upper respiratory infections. Given the very large quantities that must be consumed, and potential concerns over the safety of such consumption, its use cannot yet be recommended. [***] Vaccination Vaccines have been used effectively against most common childhood infections caused by single specific organisms such as mumps, measles and rubella. The concept of vaccinating against AOM therefore is an attractive one that is being actively explored. Potential obstacles include the wide range of causative organisms, both bacterial and viral, the varied serotypes, technical difficulties in



] 919



producing an effective immune response, obtaining an immune response before six months of age, parental resistance to multiple vaccination and the possibility that the successfully targeted pathogens will simply be replaced by others. Vaccination against viruses Since 60–90 percent of episodes are initially associated with viral infections (see above under Viruses) viral vaccination seems the most logical first step. AOM secondary to infection by the measles virus is now relatively uncommon in industrialized countries, for example. Influenza A vaccination is currently the only commercially available preparation for the prophylaxis of viral upper respiratory infections. Three trials of children in daycare have shown its efficacy in preventing AOM, resulting in 30–36 percent fewer episodes during a subsequent influenza epidemic, and reducing influenzaassociated AOM by 83–93 percent.27 This is not an absolute reduction in episodes, and this indication is not yet within the UK Department of Health guidelines.28 This does show, however, that preventing viral URTI can be an effective method of reducing AOM. [***] RSV vaccines are undergoing clinical trials for lower respiratory tract infection, but as yet do not seem to be providing significant protection. No studies are under way assessing a role in AOM. Parainfluenza virus vaccines have been evaluated in animals and need to target types 1, 2 and 3 viruses. Limited human studies demonstrate relative safety and immunogenicity, but efficacy studies are not available.29 Vaccination against bacteria Vaccination against Streptococcus pneumoniae, nontypeable Haemophilus influenzae and Moraxella catarrhalis is made difficult by the low immunogenicity of the polysaccharide capsule of these bacteria in young children and infants. Success against Haemophilus influenzae type B (which causes epiglottitis and meningitis) using a polysaccharide-protein conjugated vaccine provides one potential solution. Streptococcus pneumoniae vaccination is particularly challenging because of the 90 serotypes of the bacteria. However, as only a small number of these cause most pneumococcal AOM, and it has been shown that anticapsular antibodies can prevent pneumococcal AOM, progress has been made. Early attempts with unconjugated pneumococcal polysaccharide vaccines proved unsuccessful in children under two years of age. However, a heptavalent conjugated vaccine (Prevenar, WyethLederle Vaccines) has been shown to be highly effective in preventing invasive pneumococcal disease and modestly successful in reducing AOM in two major studies. Immunization occurs at two, four, six and 12, and in one study also 15 months of age. Episodes of AOM from any cause were reduced by 6 and 7 percent, respectively, and pneumococcal AOM by 34 percent.29 To improve cover



920 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY during the critical first six months of life, trials are under way to see if immunizing the mother in the third trimester of pregnancy is effective. Immunization after two years of age is with 23-valent pneumococcal polysaccharide vaccine. Whilst vaccination is recommended in certain ‘at risk’ children,28 its place in the management of AOM is not yet clear. It may be, for example, that vaccinating a child presenting with recurrent AOM will be ineffective because colonization of the upper respiratory tract has already occurred. Publications on this are pending. Debate on the quality of systematic reviews on this topic continues and clear guidance cannot yet be given.30 [***] Non-typeable Haemophilus influenzae vaccines are being developed by several techniques.27 Phase I clinical trials using a conjugated vaccine are under way. Animal experiments using a chinchilla model have shown some reduction in AOM with this type of vaccine. Moraxella catarrhalis vaccine research is at a preclinical stage, but products are under development. Experts in the field hope vaccination against all three of these bacteria may be possible within a decade.29 Special attention should be drawn to children with, or awaiting, cochlear implants. Concern has been raised about a number of cases of meningitis. Whether it is implant related or reflects inner ear abnormalities in many of these children is unclear. All such children are recommended to have the heptavalent pneumococcal vaccine before the age of two years and the 23-valent pneumococcal polysaccharide vaccine at the age of two or over (see Chapter 70, Paediatric cochlear implantation). Hib conjugate vaccine is recommended for all children up to four years of age. Immunoglobulins The importance of immunological immaturity in the occurrence of recurrent AOM has been emphasized. Intramuscular pooled gamma-globulin in otitis-prone children has been shown not to reduce the incidence of AOM. However, in a Japanese study, intravenous immunoglobulin (GB-0998) in IgG2-deficient infants has been shown to be an effective prophylaxis for AOM, as well as pneumonia in this specific group of children. [**] Benign commensals A recent paper considered whether spraying benign commensals (alpha streptococci) into the nose to recolonize the nasopharynx following antibiotics might reduce AOM by inhibiting the growth of pathogenic bacteria. A significant reduction was reported.31 A separate smaller study, which did not pretreat with antibiotics, showed no difference. [***] SURGICAL PROPHYLAXIS



In contrast to the large number of trials comparing antibiotic treatments, there are relatively few addressing



surgical prophylaxis. Surgery is potentially attractive, however, in that it may reduce problems of antibiotic resistance and also treat subsequent OME. Ventilation tubes A recent metanalysis of five trials concluded that the presence of ventilation tubes versus no tubes yielded a relative decrease in episodes of AOM of 56 percent, equivalent to an absolute reduction of 1.0 episode per child per year.32 The effect occurred mostly in the first year of follow up, presumably as this covered the period when the tubes were in place. Of equal importance is the reduction in the prevalence of OME by 115 days per child-year. Seventy-nine percent were reported to have an improved quality of life. Side effects included recurrent otorrhoea in 7 percent and chronic otorrhoea in 4 percent. Other studies have shown a higher incidence of tympanosclerosis and focal areas of tympanic membrane atrophy of questionable significance in the ventilation tube groups. These findings need careful interpretation. One of the studies compares antibiotic prophylaxis with amoxicillin to tubes to placebo.33 The amoxicillin group had a significant reduction in episodes of AOM. The tube and placebo group did not. However, when AOM occurred in the placebo group, it was more distressing than when otorrhoea occurred with AOM in the group with tubes in place. Also, over a two-year period, the surgical group had 26 and 61 days fewer with OME than the antibiotic and placebo groups, respectively. It is difficult to draw conclusions about the role of ventilation tubes. On the evidence available they may be considered for children with recurrent AOM, but no persistent effusion, in whom medical strategies have failed. There may be a greater role for them in preference to, or following failure of, medical prophylaxis in the child with recurrent AOM and persistent OME. [****] Adenoidectomy and adenotonsillectomy The limited evidence-base for best practice is most striking when considering adenoidectomy. Two papers are particularly worth discussing, both with a cohort of children from Pittsburgh. Randomization methods have been questioned in these studies, as has follow up. The first concluded that adenoidectomy may be beneficial in children who had previously had ventilation tube insertion and suffered subsequent AOM. AOM was reduced by 31 percent relative to the control group in a two-year follow up (or 0.32 episodes per child-year), and subjects spent 42 percent less time with OME. Additionally, the need for further tubes was reduced by 50 percent.34 Their second trial was of children who had not previously had ventilation tube insertion. Considering children without overt adenotonsillar disease, a modest reduction in the number of episodes of AOM was recorded in the first year after surgery from 2.1 to 1.4



Chapter 73 Acute otitis media in children



following adenotonsillectomy, but not adenoidectomy. Similarly, OME was reduced from 30 to 19 days in year one in the adenotonsillectomy group, and to 22 days in the adenoidectomy group. The effect was not apparent after the first year. Drop out from the trial was particularly high in the adenoidectomy group, and the results should be viewed cautiously. For children with adenotonsillar symptoms, no AOM benefit was reported from adenotonsillectomy. The authors concluded the risks of surgery were not warranted in children who had not previously had ventilation tube insertion. In summary, there is little evidence to support adenotonsillectomy. Adenoidectomy may be considered in those children who have failed medical therapy and had further AOM following ventilation tube insertion. The presence of OME increases the benefit of adenoidectomy. [***]



OUTCOMES An episode of acute otitis media may resolve rapidly with or without antibiotics; it may prove resistant to first-line antibiotics; it may persist or recur shortly after a course of antibiotics has finished; it may subsequently recur; or it may progress to tympanic membrane perforation or other complication of infection. Here we consider the medium- and long-term consequences of infection: the natural history of AOM, middle ear effusions, auditory functioning and speech and language development.



] 921



Caution should be attached to these findings. Though pooled numbers are large, high risk children and those with baseline OME were generally excluded.



MIDDLE EAR EFFUSIONS



Middle ear effusions are an important outcome of AOM. Looking again at those children randomized to placebo, pooled data show rates of OME of 63 percent two weeks after AOM, 40 percent at one month and 26 percent at three months. Antibiotics did not appear to have any effect.9, 25



AUDITORY FUNCTIONING



What little work has been carried out on short-term audiometric outcomes suggests that approximately one in three children will have an air–bone gap greater than 20 dB at one month after infection, and one in five at three months. There is limited evidence to suggest that AOM may reduce long-term audiometric thresholds. Several studies following cohorts of children have reported small but significant loss of very high frequency hearing (11–16 kHz) in those with many episodes of AOM. There is a suggestion that this may be more a consequence of disturbed middle ear mechanics than cochlear damage. The significance for auditory functioning as the child grows older is not established.



SPEECH AND LANGUAGE DEVELOPMENT



Natural history Data come from the control arms of randomized controlled trials, and hence usually exclude high-risk children, complicated cases, and those under two years old. Without antibiotic treatment, symptomatic relief from pain and fever occurs in approximately 60 percent of children within 24 hours of diagnosis, in over 80 percent by day two to three, and 88 percent by days four to seven.9, 25 These data do not equate with complete resolution, for example otorrhoea may still be present without pain or fever, and only 73 percent reach the stage of complete resolution by day seven to fourteen. In all studies, those with resistant or persistent disease will have received antibiotic treatment. For recurrent AOM the prognosis is also generally favourable. Following study entry, and with only acute episodes treated, recurrence rates fell to 0.13 episodes per child per month in the subsequent 6–24 months – approximately 1.6 episodes per year. Indeed, over half had no further attack in the following six months, and only one in eight continued to satisfy the diagnostic criteria for recurrent AOM.25 Other work has shown that even in early recurrences of infection three to four weeks after a previous episode, a new organism is usually involved.



It is difficult to separate the literature on AOM and OME outcomes. Little is written on speech production or reception. In children with OME, a significant effect seems to occur in the early years of life on expressive language development, but not receptive language. A small number of studies point to persisting effects on expressive language in school-age children. There is little evidence showing different cognitive development in school-age children with a history of otitis media in the first three years of life. There are suggestions that poor behavioural traits may be more common by school age, but more work is required before conclusions are drawn.



COMPLICATIONS Extracranial TYMPANIC MEMBRANE



Tympanic membrane perforation is considered a complication of AOM. It is the commonest complication of infection and is reported in 0–10 percent of episodes. Perforation is associated with a purulent or bloody otorrhoea and immediate relief of pain. It typically occurs



922 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY in the posterior half of the pars tensa, and is associated with loss of the fibrous middle layer of the drum. This may predispose to future posterior retraction pockets. Four outcomes of perforation may result. In most cases the perforation heals spontaneously and the infection resolves. Second, the infection may resolve, but the perforation persists. This may predispose the ear to future AOM or chronic suppurative otitis media. Third, the perforation and otorrhoea may persist, manifesting as chronic suppurative otitis media. ‘Chronicity’ is generally deemed to have occurred by three months. Fourth, a further complication may arise. The long-term outcomes were assessed in a cohort of otitis-prone children followed up from 3 to 14 years of age. By the end of the study 7 percent had collapse of the posterior superior tympanic membrane, chronic suppurative otitis media, or central perforation.35 Scarring or tympanosclerosis was present in 27 percent, although several studies report that ventilation tubes increase this risk. ACUTE MASTOIDITIS



Four classes of mastoiditis are defined. During episodes of acute otitis media, infection and inflammation may naturally extend into the mastoid cavity, and be visualized radiologically. This is not associated with the typical signs of acute mastoiditis and is not considered a complication of AOM. Infection may spread to the mastoid periosteum by emissary veins: acute mastoiditis with periosteitis. At this stage no abscess is present but the post-auricular crease may be full, the pinna may be pushed forward and there may be mild swelling, erythema and tenderness of the post-aural region. When acute mastoid osteitis develops, the infection has begun to destroy the bone of the mastoid air cells and a subperiosteal abscess may develop. Signs may be similar to those when periosteitis is present. A subperiosteal abscess develops most commonly in the post-auricular region. A zygomatic abscess may develop above and in front of the pinna. A Bezold’s abscess may result from perforation of the medial mastoid cortex, tracking down the sternomastoid to the posterior triangle. Pus tracking down peritubal cells may result in a retropharyngeal or parapharyngeal abscess (Figure 73.3). A fourth stage may be reached, subacute (‘masked’) mastoiditis, in incompletely treated AOM after 10–14 days of infection. Signs may be absent, but otalgia and fever persist. This stage can also progress to serious complications.3 In the pre-antibiotic era, mastoiditis was a common and serious complication of AOM. In a study in 1954 the control group was reported to have developed mastoiditis in 17 percent of cases.9 In some developing countries rates of 5 percent are still quoted. In the 1970s it was estimated that 0.004 percent of cases of AOM resulted in surgery for



Figure 73.3 (a) and (b) Acute mastoiditis.



mastoiditis. The incidence is reported by several authors to be increasing gradually again. The incidence varies between countries. In the UK, Canada, Australia and the United States, where antibiotic prescription rates are over 96 percent, the incidence ranges from 1.2 to 2.0 per 100,000 population per year. In Norway, Denmark and The Netherlands (prescription rates 67, 76 and 31 percent, respectively) rates are higher at 3.5, 4.2 and 3.8 respectively.36 Acute mastoiditis is a disease of childhood. A large multicentre study found 28 percent to be in children less than one year of age, 38 percent in one to four year olds, 21 percent in four to eight year olds, 8 percent in 8–18 year olds and 4 percent in those over 18 years of age.37 This higher incidence in younger children reflects the peak ages for AOM. Traditional teaching was that acute mastoiditis is preceded by 10–14 days of middle ear symptoms. However, in many papers the short length of middle ear symptoms prior to presentation is noteworthy. For example, in one large study approximately 32 percent had one to two days symptoms, 34 percent had three to six days, 26 percent seven to fourteen days and 8 percent over 14 days.37 Prior antibiotic treatment of the infection is common, reported in 22–55 percent of children. Clearly antibiotics do not fully protect against mastoiditis. Symptoms are of otalgia and irritability in most children. Pyrexia is less common in those treated with antibiotics. Otorrhoea is present in only approximately 30 percent. Clinically, a red or bulging tympanic membrane will often be seen. A normal drum is reported in a very variable proportion of cases, but certainly does not exclude the diagnosis and is believed to result from resolution of the mesotympanic infection following antibiotic treatment, while the osteitis in the mastoid progresses. Retro-auricular swelling is seen in approximately 80 percent and retro-auricular erythema in 50–84 percent (less in previously treated children). Tenderness is typically sited over MacEwen’s triangle (on palpation



Chapter 73 Acute otitis media in children



through the conchal bowl). Pinna protrusion is present in two-thirds of cases. Sagging of the posterior wall of the external auditory canal, resulting from subperiosteal abscess formation, should be looked for, but is quoted as an uncommon finding. Few patients will have a ‘full house’ of the classic signs. A somewhat different incidence of organisms has been identified from those gained from culture in AOM. Around 20 percent of samples do not grow bacteria. Streptococcus pneumoniae, Streptococcus pyogenes, Pseudomonas aeruginosa and Staphylococcus aureus are the most commonly reported. Haemophilus influenzae is less commonly reported, and Moraxella catarrhalis, Proteus mirabilis and Gram-negative anaerobes rarely. Recommended investigations vary between institutions. A full blood count, C-reactive protein (CRP), and blood cultures are often obtained. A CT scan of the mastoid is recommended when intracranial complications are present or suspected (though MRI may be more helpful in identifying specific intracranial pathology), when mastoidectomy is to be performed and in those not improving on antibiotic treatment. A CT may show evidence of osteitis, abscesses and intracranial complications. Differential diagnosis includes AOM, otitis externa, furunculosis and reactive lymphadenopathy. Rarely, undiagnosed cholesteatoma, Wegener’s granulomatosis, leukaemia and histiocytosis may first present with AOM, hence tissue should be sent for histology if mastoidectomy is performed. Myringotomy with or without ventilation tube placement, culture of the aspirate and high-dose intravenous antibiotics is the most commonly recommended initial treatment in acute mastoiditis. This is adequate in 75 percent of cases. Failure to improve, subperiosteal abscess formation (occurring in 10–30 percent) or development of complications merits at least abscess drainage with or without cortical mastoidectomy. This can be challenging surgery for the less experienced as the mastoid is often full of granulations and the facial nerve superficial in the young child. A most important message is that intracranial complications from acute mastoiditis develop in 6–17 percent of cases, and many of these may develop during hospitalization. Although acute mastoiditis may be less common than in the past, its severe complications still occur. [**] PETROSITIS



Infection may extend to the petrous apex. The classic features of Gradenigo’s triad (VI nerve palsy, severe pain in the trigeminal nerve distribution and middle ear infection) are not always present. Patients commonly present with other intracranial complications. Recent papers recommend high-dose broad spectrum antibiotics and a variety of mastoidectomy, from cortical to radical, though drainage of the petrous apex is no longer felt necessary. [*]



] 923



FACIAL NERVE PALSY



In the pre-antibiotic era, it was estimated that acute lower motor neurone facial palsy complicated 0.5 percent of episodes of AOM (see Chapter 80, Facial paralysis in childhood). It is now quoted at 0.005 percent.38 Most are related to bacterial infection but case reports with viral AOM exist. Approximately four out of five children present with a partial paralysis. The case series in the literature report that approximately 80 percent of palsies respond well to ventilation tube insertion and intravenous antibiotics. The remainder undergo cortical mastoidectomy. Advice is conflicting about when and in whom mastoidectomy is required and the role of facial nerve decompression. As recovery is generally so good, a more conservative approach without facial nerve decompression seems appropriate. Most children achieve rapid restoration of normal facial function, with a mean time to complete recovery of four months. Those with a total paralysis at presentation have a recovery stretching over many months. Sixth nerve palsy in the absence of petrositis has also been reported. It is speculated this may stem from phlebitis spreading along the inferior petrosal sinus from the lateral sinus. [*] LABYRINTHITIS



Round window permeability changes during acute infection are important as these may allow entry of bacterial toxins. There is some experimental evidence that permeability can be increased by streptococcal toxins. Preformed channels for bacterial entry may also exist, such as surgical or congenital perilymph fistulae. These may allow infection to spread directly to the subarachnoid space causing meningitis. Particular concern arises in children with congenital inner ear abnormalities and those with cochlear implants. Three types of labyrinthitis are recognized. Perilabyrinthitis is not associated with AOM. Serous labyrinthitis is inflammation of the labyrinth without pus formation, and is characterized by recovery of auditory and vestibular function. Suppurative labyrinthitis may result from spread of infection from the mastoid or middle ear. Severe vertigo, nausea, vomiting, nystagmus and permanent hearing loss result. Suppurative labyrinthitis is rare, and the treatment of cases presented in the literature ranges from ventilation tube insertion and aggressive antibiotic use, to tympanomastoidectomy and cochleotomy. [*]



Intracranial In the pre-antibiotic era, intracranial complications of AOM were more common and mortality rates of over 75 percent are presented. Published mortality rates from intracranial complications now average approximately 5 percent in industrialized countries.



924 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Persistent headache and fever are the most common early symptoms of an intracranial complication. In half of cases there may be signs only of AOM and not mastoiditis. Frequently two or more complications coexist. Early diagnosis is important for improving outcomes. Seven classical intracranial suppurative complications of AOM are described.3 1. Meningitis is usually cited as the commonest intracranial complication of AOM, accounting for 54–91 percent of cases. In contrast, studies assessing the aetiology of meningitis are conflicting. One of the largest recent studies found no association between bacterial meningitis and AOM, while another found an antecedent history of AOM in 29 percent,39 though this does not equate to a causal relationship. Special mention has already been made of possible associations between congenital inner ear malformations such as cochlear dysplasia, cochlear implants and meningitis. Younger children, average age two years, are most commonly infected. Studies focus almost exclusively on bacterial aetiologies. The rate of Haemophilus influenzae type B meningitis has dropped dramatically since vaccination was introduced. Streptococcus pneumoniae is the causal agent in a greater proportion because of this reduction.39 A second intracranial complication should be looked for in any infant with meningitis with MR scanning. Myringotomy may help to establish the infective agent if evidence has not been obtained from lumbar puncture. Treatment is medical. If mastoid surgery is required, it is usual to try and wait for an improvement in the medical condition of the child first if possible. [*] 2. Extradural abscess is the next commonest intracranial complication. It is more commonly associated with chronic disease. Pus collects between dura and bone, usually after bone erosion. If lying in the posterior fossa medial to the sigmoid sinus, it is termed an extradural (epidural) abscess, if within the split of dura enclosing the sigmoid sinus it is called a peri-sinus abscess. It may be discovered only at mastoidectomy, but may be suspected in the patient with persistent headache and fever or severe otalgia. Treatment is surgical drainage. [*] 3. Subdural empyema is a collection of pus between the dura and arachnoid membranes and is termed a subdural empyema. It is rare. It develops by direct extension of infection or thrombophlebitis. In addition to headaches and pyrexia, focal neurological signs, seizures and signs of meningeal irritation may be present. Paranasal



sinusitis is reported to be a much commoner cause than AOM. Surgical drainage of the abscess through burr holes or craniectomy may be indicated. Mastoidectomy may sometimes be required, though many cases cited in the literature were treated medically. [*] 4. Sigmoid sinus thrombosis most commonly results from erosion of the bone over the sinus from mastoiditis and may also be associated with other complications. However, it occurs in association with otitis media alone in 43 percent of cases. Infected thrombus develops within the sinus and may then extend proximally and distally to the internal jugular vein and superior vena cava, entering the systemic circulation and causing septicaemia. In addition to headache and otorrhoea, a spiking pyrexia may develop. Griesinger’s sign is mastoid tenderness and oedema secondary to thrombophlebitis of the mastoid emissary vein. MRI is the imaging of choice showing high signal intensity in the sigmoid sinus on both T1-and T2-weighted images and absent flow. If caused only by otitis media, myringotomy and antibiotics may suffice. However, in the presence of mastoid infection, it is more usual to perform a canal wall up mastoidectomy, needle the sinus to assess blood flow and occasionally to remove infected thrombus. As persistent sepsis and distant thrombosis are uncommon, the role of anticoagulation is unclear in the literature.40 Serial imaging to look for propagation of thrombus has been recommended. [**/*] 5. Focal otitic encephalitis (cerebritis). Focal inflammation and oedema of brain tissue may occur independent of, or in association with, any suppurative complication of AOM. Intensive antibiotic treatment is required. [*] 6. Brain abscess is more commonly associated with chronic ear disease but may occur in association with AOM and its complications (Figure 73.4). Brain abscess forms a larger proportion of complications in developing countries. It may develop in both the temporal lobe and cerebellum. Persistent headaches are the commonest symptom. Initial symptoms may be of encephalitis, but these often settle as the abscess organizes over days or weeks. Eventually, signs of raised intracranial pressure, focal neurology and infection develop. Invstigations include CT imaging followed by lumbar puncture, if safe. In the early stages of cerebritis, neurosurgical drainage may be avoided but will be required if the abscesses are expanding. Brain abscesses carry a potentially high mortality rate, although in industrialized countries the few large



Chapter 73 Acute otitis media in children



] 925



KEY POINTS  Acute otitis media is one of the commonest illnesses of childhood.  Diagnosis can be difficult particularly in very young children.  Management recommendations vary widely between countries.  A range of modifiable risk factors should be addressed.  Evidence is emerging to support new prophylactic strategies.  Intracranial complications are still seen despite prior antibiotic treatment.



Best clinical practice [ AOM is one of the commonest illnesses of childhood.



[ Figure 73.4



Brain abscess secondary to AOM.



[ series now quote rates of below 10 percent. One large review found the mortality from otogenic causes, at 3.8 percent, was much lower than from other causes. [*]. 7. Otitic hydrocephalus is a complication of AOM manifesting as raised intracranial pressure in the absence of any space-occupying lesion, and without obstruction to the flow of cerebrospinal fluid (CSF). Benign intracranial hypertension is a synonym. The aetiology is obscure. Headache is the predominant symptom. It is commonly associated with sigmoid or transverse sinus thrombosis and so MRI is an important investigation. Lumbar puncture will show raised CSF pressure, but normal CSF composition. A number of medical treatments may be tried and liaison with a paediatric neurologist is recommended. [*].



[



[



[



[ [



CONCLUSION It can be seen from this chapter that there are deficiencies in our current knowledge of both the diagnosis and aetiology of AOM, and uncertainties in management strategies. However, we have been able to describe a number of potentially exciting developments that have occurred in the past few years.



[



Accurate diagnosis is notoriously difficult. A high index of suspicion is required in the unwell child. [Grade B] The clinician should distinguish between sporadic, resistant, persistent or recurrent AOM as management strategies differ. [Grade A] Two-thirds of children recover within 24 hours with or without treatment, so a period of watchful waiting may be reasonable in uncomplicated AOM. [Grade A] Antibiotics should not be withheld in severe or irregular infections, should be given if a child fails to improve within two to three days of the onset of AOM and should be given in sufficiently high doses. [Grade A] Pyrexia (437.51C), severe otalgia, vomiting, age under two years and ‘high risk’ children have all been used as indicators to use antibiotics sooner rather than later. [Grade B/C] Practitioners should be aware of local bacterial antibiotic resistance patterns and prescribing policies. Broad spectrum antibiotics are not generally required as first-line therapy. [Grade A] In otherwise healthy children over two years of age, five days of antibiotics is usually adequate. [Grade A] For persistent or resistant AOMs it should be noted that whilst pneumococcal drug resistance can usually be overcome by increased antibiotic doses, Haemophilus may be beta-lactam producing, so broader spectrum antibiotics may be required. [Grade B] Modifiable risk factors should be discussed with parents. These include nursery attendance, parental smoking, breastfeeding and the use of pacifiers. [Grade C]



926 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY [ In the management of recurrent AOM, on average



[



[ [ [



eight months of antibiotic prophylaxis would be needed to prevent one episode of AOM. This strategy may be preferred in the absence of effusions between episodes of AOM. [Grade A] Ventilation tube insertion reduces the number of episodes of AOM by over 50 percent. This option may be preferred when effusions persist between episodes of AOM. [Grade A] Additional adenoidectomy may further reduce the number of episodes of AOM. [Grade A] The benefits of tonsillectomy on episodes of AOM are not sufficient to warrant its use in the management of recurrent AOM. [Grade A] Vigilance should be maintained for complications of AOM, the most common symptoms being persistent pyrexia and headache. [Grade C].



ACKNOWLEDGEMENTS We would like to thank Dr Anne Schilder from the University Medical Center, Utrecht, for her data on management practices in the Netherlands.



REFERENCES







Deficiencies in current knowledge and areas for future research The following list summarizes what we feel this chapter should be able to report on when the next edition is being prepared. Greater standardization and reproducibility of diagnostic criteria is required to compare trials. As most children appear to recover without treatment, better characterization of those who may have initial antibiotic treatment withheld is needed. Trials should be set up to study high risk groups of children who are currently excluded from most studies: those with conditions predisposing them to acute otitis media and children under 18 months of age. These are the children likely to benefit the most from our intervention. We know many important risk factors, but trials are needed to show whether attempts to modify them actually help. The potential benefit of vaccination to older children with recurrent infection needs exploring. As more vaccines are developed, we must know to whom we should be recommending them. The number and quality of trials of surgical intervention does not allow confident guidance to be given as to the long-term benefits or consequences of ventilation tube insertion. More data are required on the long-term consequences of recurrent infection in terms of altered audiometric thresholds, quality of life and language and cognitive development.



$ $ $



$ $ $ $







1. Diagnosis and Treatment of Otitis Media in Children. Bloomington (MN): Institute for Clinical Systems Improvement (ICSI); 2001 July. http://www.guideline.gov 2. Appleman CLM, van Balen FAM, van de Lisdonk EH, van Weert HCLM, Eizinga WH. NGH-standaard Otitis Media Acuta (eerste herziening). Huisarts en Wetenschap. 1999; 42: 362–6, and www.artsennet.nl/nhg 3. Bluestone CD, Gates GA, Klein JO, Lim DJ, Mogi G, Ogra PL et al. Definitions, terminology, and classification of otitis media. Annals of Otology, Rhinology and Laryngology. 2002; 188: 8–18. Up-to-date and succinct account of terminology in otitis media. 4. Little P, Gould C, Moore M, Warner G, Dunleavey J, Williamson I. Predictors of poor outcome and benefits from antibiotics in children with acute otitis media: pragmatic randomised trial. British Medical Journal. 2002; 325: 22–5. 5. Chan LS, Takata GS, Shekelee P, Morton SC, Mason W, Marcy SM. Evidence assessment of management of acute otitis media: II. Research gaps and priorities for future research. Pediatrics. 2001; 108: 248–54. 6. Froom J, Culpepper L, Grob P, Bartelds A, Bowers P, Bridges-Webb C et al. Diagnosis and antibiotic treatment of acute otitis media: report from International Primary Care Network. British Medical Journal. 1990; 300: 582–6. 7. Schwartz RH, Stool SE, Rodriguez WJ, Grundfast KM. Acute otitis media: towards a more precise definition. Clinical Pediatrics. 1981; 20: 549–54. 8. Heikkinen T, Ruuskanen O. Signs and symptoms predicting acute otitis media. Archives of Pediatrics and Adolescent Medicine. 1995; 149: 26–9. Well-researched systematic review on the role of antibiotics. 9. Glasziou PP, Del Mar CB, Sanders SL, Hayem M. Antibiotics for acute otitis media in children (Cochrane Review). The Cochrane Library. Oxford: Update Software Ltd, 2002. 10. Karma PH, Pentilla MA, Sipila MM, Kataja MJ. Otoscopic diagnosis of middle ear effusion in acute and non-acute otitis media. I. The value of different otoscopic findings. International Journal of Pediatric Otorhinolaryngology. 1989; 17: 37–49. 11. Gehano P, Lenoir G, Barry B, Bons J, Boucot I, Berche P. Evaluation of nasopharyngeal cultures for bacteriologic assessment of acute otitis media in children. Pediatric Infectious Disease Journal. 1996; 15: 329–32. 12. Heikkinen T. The role of respiratory viruses in otitis media. Vaccine. 2001; 19: S51–5.



Chapter 73 Acute otitis media in children 13. Heikkinen T, Thint M, Chonmaitree T. Prevalence of various respiratory viruses in the middle ear during acute otitis media. New England Journal of Medicine. 1999; 340: 260–4. 14. Chonmaitree T, Owen MJ, Howie VM. Respiratory viruses interfere with bacteriologic response to antibiotic in children with acute otitis media. Journal of Infectious Diseases. 1990; 162: 546–9. 15. Eskola J. Polysaccharide-based pneumococcal vaccines in the prevention of acute otitis media. Vaccine. 2000; 19: S78–82. 16. Pichichero ME, Reiner SA, Brook I, Gooch 3rd WM, Yamauchi T, Jenkins SG et al. Controversies in the medical management of persistent and recurrent acute otitis media. Annals of Otology, Rhinology and Laryngology. 2000; 183: 1–12. 17. Buchman CA, Doyle WJ, Skoner DP, Post JC, Alper CM, Seroky JT et al. Influenza A virus induces acute otitis media. Journal of Infectious Diseases. 1995; 172: 1348–51. 18. Casselbrant ML, Mandel EM. The genetics of otitis media. Current Allergy and Asthma Reports. 2001; 1: 353–7. 19. Uhari M, Mantysarri K, Niemela M. A meta-analysis of the risk factors for acute otitis media. Clinical Infectious Diseases. 1996; 22: 1079–83. 20. Strangerup SE, Tos M. Epidemiology of acute suppurative otitis media. American Journal of Otolaryngology. 1986; 7: 47–54. 21. Kozyrskyi AL, Hildes-Ripstein GE, Longstaff SEA, Wincott JL, Sitar DS, Klassen TP et al. Short course antibiotics for acute otitis media (Cochrane review). In: The Cochrane Library. Oxford: Update Software Ltd, 2002. 22. Damoiseaux RA, van Balen FA, Hoes AW, Verheij TJ, de Melker RA. Primary care based randomised, double blind trial of amoxicillin versus placebo for acute otitis media in children under 2 years. British Medical Journal. 2000; 320: 350–4. 23. Flynn CA, Griffin T, Tudiver F. Decongestants and antihistamines for acute otitis media in children (Cochrane Review). In: The Cochrane Library. Oxford: Update Software Ltd, 2002. 24. Kaleida PH, Casselbrant ML, Rockette HE, Paradise JL, Bluestone CD, Blatter MM et al. Amoxicillin or myringotomy or both for acute otitis media: results of a randomised clinical trial. Pediatrics. 1991; 87: 466–74. 25. Rosenfeld RM. Natural history of untreated otitis media and what to expect from medical therapy. In: Rosenfeld RM, Bluestone CD (eds). Evidence-based otitis media. New York: Decker Inc, 1999: 157–221. 26. Uhari M, Kontiokari T, Niemela M. A novel use of xylitol sugar in preventing acute otitis media. Pediatrics. 1998; 102: 879–84. 27. Giebink GS, Bakaletz LO, Barenkamp SJ, Eskola J, Green B, Gu X-X et al. Recent advances in otitis media: vaccine.



28.



 29.



30.



31.



32. 33.



34.



 35.



36.



37.



38.



39.



40.



] 927



Annals of Otology, Rhinology and Laryngology. 2002; 111: 82–94. Donaldson L, Mullally S, Smith J. Update on immunisations. Department of Health. August 2002: 1-8. http://www.doh.uk/cmo/index.htm Snow JB. Progress in the prevention of otitis media through immunization. Otology and Neurotology. 2002; 23: 1–2. Account of current evidence on the possible preventive role of vaccination. Jefferson T, Demicheli V. Polysaccharide pneumococcal vaccines. Existing guidance is at variance with the evidence. British Medical Journal. 2002; 325: 292–3. Roos K, Hakansson EG, Holm S. Effect of recolonisation with ‘interfering’ alpha streptococci on recurrences of acute and secretory otitis media in children: randomised placebo controlled trial. British Medical Journal. 2001; 322: 210–2. Rosenfeld RM. Surgical prevention of otitis media. Vaccine. 2000; 19: S134–9. Casselbrant ML, Kaleida PH, Rockette HE, Paradise JL, Bluestone CD, Kurs-Lasky M et al. Efficacy of antimicrobial prophylaxis and of tympanostomy tube insertion for prevention of recurrent acute otitis media: results of a randomised clinical trial. Pediatric Infectious Disease Journal. 1992; 11: 278–86. Paradise JL, Bluestone CD, Rogers KD, Taylor FH, Colborn DK, Bachman RZ et al. Efficacy of adenoidectomy for recurrent otitis media in children previously treated with tympanostomy-tube placement. Journal of the American Medical Association. 1990; 263: 2066–73. Stenstrom C, Invarsson L. Late effects on ear disease in otitis-prone children: a long-term follow-up study. Acta-Otolaryngology. 1995; 115: 658–63. Highlights national variations in prescribing patterns for acute otitis media. Van-Zuijlen DA, Schilder AG, Van-Balen FA, Hoes AW. National differences in incidence of acute mastoiditis: relationship to prescribing patterns of antibiotics for acute otitis media? Pediatric Infectious Diseases Journal. 2001; 20: 140–4. Luntz M, Brodsky A, Nusem S, Kronenberg J, Keren G, Migirov L et al. Acute mastoiditis: a multicentre study. International Journal of Pediatric Otorhinolaryngology. 2001; 57: 1–9. Ellefsen B, Bonding P. Facial palsy in acute otitis media. Clinical Otolaryngology and Allied Sciences. 1996; 21: 393–5. Ryan MW, Antonelli PJ. Pneumococcal antibiotic resistance and rates of meningitis in children. Laryngoscope. 2000; 110: 961–4. Bradley DT, Hashisaki GT, Mason JC. Otogenic sigmoid sinus thrombosis: what is the role of anticoagulation? Laryngoscope. 2002; 112: 1726–9.



74 Chronic otitis media in childhood JOHN HAMILTON



Introduction and definitions Histology and pathogenesis of chronic otitis media What may prevent resolution of otitis media? Chronic perforation of the tympanic membrane Atrophy of the pars tensa of the tympanic membrane (including retraction pockets) Cholesteatoma Ossicular erosion in chronic otitis media



928 930 931 932 941 947 954



Conclusions Key points Best clinical practice Deficiencies in current knowledge and areas for future research Acknowledgements References



956 957 958 958 959 959



SEARCH STRATEGY The data in this chapter are supported by Medline searches through the PubMed online database using medical subject headings (inter alia otitis media; otitis media, suppurative; tympanic membrane perforation; tympanoplasty; myringoplasty; cholesteatoma, ossicular prosthesis, etc.). Searches were curtailed by using the ‘child (0–18)’ limiter and by focusing on appropriate subheadings in the medical subject headings database. Free text searches using terms not included in the medical subject headings database such as ‘tympanic membrane atrophy’ and ‘tympanic membrane retraction’ were expanded and refined using Boolean operators. Results were stored using reference manager software (Endnote), which also allowed searches to be combined without duplication of results. The Cochrane database was also consulted. These electronic searches were extended by reference to proceedings of major international symposia, particularly the series of international otitis media conferences, the international cholesteatoma meetings and the middle ear mechanics series.



INTRODUCTION AND DEFINITIONS



complicate its teaching. The following problems may arise.



In any specialized field new terminology is introduced to facilitate dialogue between interested parties. These specialized terms require precise definition in order that communication and comparison between centres can have any value at all. However, in any developing area of human endeavour a rudimentary subject may be found to harbour several new aspects of interest. This development not only leads to the introduction of new terms in the field of study but also the reclassification of the parent subject. In this way, the progressive change caused by increasing understanding of a subject paradoxically may



 Older classifications and definitions may become obsolete.  Some terms may become repeatedly defined and therefore ambiguous.  Increasing complexity of classifications, whilst of value to those developing the subject, may hinder their widespread use. The balance between precision and utility may have no perfect solution. The study of chronic otitis media has resulted in all these problems.



Chapter 74 Chronic otitis media in childhood



WHO definition of chronic otitis media, 1996 A WHO/CIBA Foundation Workshop1 in 1996 produced the following definitions:  Chronic suppurative otitis media: ‘a stage of ear disease in which there is chronic infection of the middle ear cleft, i.e. Eustachian tube, middle ear and mastoid, and in which a non-intact tympanic membrane (e.g. perforation or tympanostomy tube) and discharge (otorrhoea) are present’. The same report qualifies this definition by stating that the otorrhoea should be present for two weeks or longer.  Chronic otitis media: ‘includes both chronic suppurative otitis media and chronic perforation of the tympanic membrane’.



] 929



exhaustive and hierarchical classification of otitis media (see Table 74.1). Nonetheless, the new classification continues to have difficulty with ‘chronic otitis media’, not only denying the term its significance as an immediate subclassification of otitis media but also refuting the validity of the term. The panel report states that: ‘the term chronic otitis media is confusing and potentially misleading and should not be used, since some consider it to mean chronic otitis media with effusion, others think it means chronic suppurative otitis media, and still others include cholesteatoma under this term’. The panel prefers the term ‘chronic suppurative otitis media’ and define this to mean ‘a chronic perforation with chronic otitis media’.



Resolution: new classification International Symposium on Recent Advances in Otitis Media, 1978 A task force was appointed on behalf of this symposium2 to provide definitions of terms relevant to otitis media, as well as a classification of the subject. It proposed inter alia:  chronic otitis media: ‘a chronic discharge from the middle ear through a perforation of the tympanic membrane’.  chronic otitis media and ronic suppurative otitis media were defined as synonymous. However, the definitions continued that ‘suppurative refers to active clinical infection’. Furthermore, ‘a perforation without discharge can be an inactive stage of the infection’. As the term ‘chronic suppurative otitis media’ appears here to be considered equal to ‘active chronic otitis media’ there is not only some inconsistency within this set of definitions but a clear path for this evolution of these terms. As a result, the terms ‘active chronic otitis media’ and ‘inactive chronic otitis media’ have also enjoyed widespread use. The unambiguous part of the original definition was the perforation. Unfortunately, this elevated the presence of a perforation as sufficient to establish the existence of chronic otitis media. Given that tympanic membrane perforations need not be caused by chronic otitis media, this was manifestly wrong. This is particularly true in the paediatric population in which a common cause for tympanic membrane perforation is acute otitis media. In most such cases, the perforation heals with resolution of the acute infection without a phase of chronic infection.



International Symposium on Recent Advances in Otitis Media, 1999 A panel reporting to the Post-Symposium Research Conference of this Symposium has issued a revised classification of otitis media.3 The panel produced an



Hitherto, the definition of chronic otitis media has been either that of a clinical syndrome, or as a mixture of pathological entity and clinical syndrome. The tension seems most easily resolved by considering chronic otitis media to be a purely pathological entity.  ‘Chronic otitis media’ is a pathological term only, indicating chronic inflammation affecting the middle ear cleft.  Chronic otitis media may be attended by various pathological complications affecting the structures of the middle ear cleft, including perforated tympanic membrane, tympanic membrane retraction pockets, cholesteatoma, ossicular defects, tympanosclerosis and myringitis.  The patient may be troubled by symptoms and signs caused by a mixture of these pathologies. This framework (see Figure 74.1) provides the basis for the organization of this chapter, since this model predicts:  multiple complications of chronic otitis media may present simultaneously (see Figure 74.1 and 74.2);  treatment of an ear affected by chronic otitis media may require attention to each of many complications;  treatment of complications of chronic otitis media does not necessarily influence the severity of the underlying chronic otitis media;  the treated ear remains at risk of further complications of chronic otitis media. The last of these points is of particular importance when considering chronic otitis media in children. If otitis media in children is more vigorous than in adults, it is possible to predict further that the complications of chronic otitis media are more prevalent in children than in adults; and the outcomes of the treatment of ears affected by chronic otitis media in children, particularly in the long term, are worse than in adults. These two assertions clearly justify a greater understanding of this condition in children and are examined throughout this chapter.



930 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Table 74.1



Recent advances in classification of otitis media.



Classification Otitis media Eustachian tube dysfunction Intratemporal (extracranial) complications



Acute otitis media Otitis media with effusion Perforation of tympanic membrane



Mastoiditis Apical petrositis Facial paralysis Labyrinthitis



Atelectasis of middle ear



Intracranial complications



Aural acquired cholesteatoma Cholesterol granuloma Ossicular discontinuity Adhesive otitis media Tympanosclerosis Ossicular fixation Meningitis Extradural abscess Subdural ernpyema Encephalitis Brain abscess Dural sinus thrombosis Hydrocephalus



Acute perforation Chronic perforation



Acute otitis media Chronic otitis media (and mastoiditis; chronic suppurative otitis media)



Acute Chronic Acute Chronic Acute Serous Suppurative Chronic Localized Generalized Without infection With infection



Modified from Ref. 3, with permission. This classification is widely recognized, but the concepts which underpin it are not used in this chapter. Please refer to Resolution: new classification and Figure 74.1.



HISTOLOGY AND PATHOGENESIS OF CHRONIC OTITIS MEDIA Histological analysis of human temporal bones with various types of otitis media has revealed that although the clinical subgroups of otitis media may be considered to be distinct, the histology of these groups overlaps considerably.4, 5  All forms of otitis media display submucosal inflammatory infiltrates and mucosal metaplasia with the development of glandular structures, mucusproducing cells and ciliated cells on histological examination.  Chronic inflammation is characterized by tissue destruction, as well as attempts at healing.



 The general histological features include mononuclear cell infiltrates, submucosal fibrosis, the formation of highly vascular granulation tissue and osteitis.  Mucosal complications such as cholesterol granuloma and tympanosclerosis may be seen.  Changes in special structures associated with the middle ear cleft, such as the tympanic membrane, are also seen.  It is of particular importance that histologically discernible chronic otitis media occurs quite frequently in the absence of tympanic membrane perforation.4  Animal experiments indicate that features of chronic inflammation appear within 14 days if acute otitis media fails to resolve.6



Chapter 74 Chronic otitis media in childhood



Underlying pathology



Chronic otitis media



Effusion Consequent pathology (complications)



Clinical symptoms



] 931



Tympanic membrane pathology: perforation, atrophy and cholesteatoma



Ossicular pathology



Otorrhoea, hearing loss



Other pathology



Figure 74.1 Model for the classification of chronic otitis media. In this model, chronic otitis media is regarded as a pathological entity only. The arrows indicate a one-way relationship. Thus, chronic otitis media may be complicated by a wide variety of largely independent pathological complications and the clinical presentation is governed by which complications are present. This model predicts that treatment of a complicating pathology will not influence the nature of the underlying chronic otitis media, a problem of considerable significance in the management of this disease in children.



tube. The following mechanisms are suggested to enhance transfer between the nasopharynx and middle ear.  In the presence of an intact tympanic membrane, inflammation of the pharyngeal end of the Eustachian tube may prevent gas exchange along the Eustachian tube. Isolation of the middle ear allows diffusion effects with the middle ear mucosa to dominate the physiology of the middle ear cleft gas pocket. Over long periods this results in lowering of the middle ear pressure. In turn, this may predispose to aspiration of nasopharyngeal microbes into the middle ear cleft on the rare occasion when the Eustachian tube does manage to open.  In the presence of a perforated tympanic membrane, the middle ear air cushion no longer impedes the movement of gas from the nasopharynx into a closed box, predisposing to reflux from the nasopharynx to the middle ear cleft.7 Figure 74.2 Chronic otitis media and its complications. In this ear chronic otitis media has been complicated by tympanic membrane perforation, tympanic membrane sclerosis, tympanic membrane atrophy, ossicular erosion and erosion of the posterior annulus. Although gross inspection suggests the middle ear mucosa to be healthy, published histological survey has established that microscopic analysis will confirm the presence of chronic inflammation.



Repeated infection from the external ear canal



WHAT MAY PREVENT RESOLUTION OF OTITIS MEDIA?



In the presence of a perforated tympanic membrane, microbes can be transported in fluid media from the ear canal into the middle ear. In particular, water that gains access to the external ear canal can flow without impediment into the middle ear through the perforation. In addition, mucus from the middle ear may straddle the tympanic membrane and provide a vehicle for the transfer of microbes from the external ear into the mesotympanum.



Repeated infection from the nasophaynx



Persistent colonization by bacterial biofilms



Acute upper respiratory tract infections may spread to the middle ear cleft from the nasopharynx via the Eustachian



There is evidence that many chronic infections are caused by the ability of bacteria to alter their form to create



932 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY nonmotile communities adherent to mucosa and protected by polysaccharide matrix. Such biofilms are able to resist most forms of host resistance as well as antibiotics and are not detected by conventional microbiological assays. The biofilms often consist of many microbial strains and are capable of converting back to more conventional planktonic bacterial populations. Biofilms have long been recognized in chronic infections.8 Recent microscopic evidence of biofilms in otitis media has indicated that this mechanism may be important in maintaining chronic ear infections.9



CHRONIC PERFORATION OF THE TYMPANIC MEMBRANE Definitions According to the panel reporting to the Seventh PostSymposium Research Conference of the International Symposium on Recent Advances in Otitis Media a perforation of the tympanic membrane is deemed to be chronic if present for three months.3 Chronic suppurative otitis media is defined by otorrhoea of at least six weeks duration in the presence of a chronic tympanic membrane perforation.3 By contrast, the WHO defines chronic suppurative otitis media as ‘a stage of ear disease in which there is chronic infection of the middle ear cleft, i.e. Eustachian tube, middle ear and mastoid, and in which a non-intact tympanic membrane (e.g. perforation or tympanostomy tube) and discharge (otorrhoea) are present’.1 The same report qualifies this definition by stating that the otorrhoea should be present for two weeks or longer.1



Chronic otitis media and the pathology of tympanic membrane perforation Tympanic membrane perforation usually occurs secondary to acute otitis media. The process by which acute traumatic tympanic membrane perforations heal has been studied in animals. These investigations indicate that healing is mediated by hyperplasia, proliferation and migration of the outer keratinizing squamous epithelium.10 This advances ahead of connective tissue rich in fibroblasts. Failure of this process has been attributed primarily to persistent infection. One study examined 30 entire perforated tympanic membranes taken from patients receiving tympanic allografts for the treatment of their perforations. All showed evidence of inflammation, epithelial hyperplasia and fibrosis as if still trying to heal.11 A second study of the margins of surgically treated chronic perforations indicated that the connective tissue was inadequately vascularized and extracellular matrix substances and



growth factors were deficient. In 5 of the 30 cases in the former study, squamous epithelium extended medially more than 0.7 mm from the perforation edge.11 A tympanic membrane perforation associated with chronic otitis media is not just a defect in an otherwise normal structure. Not only is the tympanic membrane participating in the chronic inflammation, but the endoepithelial junction need not be located at the perforation edge. Chronic otitis media need not be associated with tympanic membrane changes.12 By contrast, inflammatory changes in the tympanic membrane are almost universally associated with chronic otitis media.12



Classification Older textbooks distinguish between tubotympanic disease, which is a defect of the anterior and inferior tympanic membrane and atticoantral disease, which is a defect of the posterior and superior tympanic membrane. The former group is also termed safe and the latter unsafe. Older textbooks also distinguish between central and marginal perforations. These classifications do not clearly distinguish tympanic membrane perforation from other tympanic membrane pathology, such as retraction pockets or cholesteatoma. By grouping a number of complications of chronic otitis media, these classifications draw conclusions about their combined behaviour instead of informing us about perforations alone. There is no evidence that true perforations (perforations with no retraction) in the epitympanum and posterior mesotympanum behave differently from other tympanic membrane perforations. However, tympanic membrane disease affecting the epitympanum and posterior mesotympanum is more likely to be cholesteatoma or a retraction pocket. These do behave differently to perforations. There is no evidence that true marginal perforations behave differently from other tympanic membrane perforations. However, marginal tympanic membrane disease is more likely to be cholesteatoma or a retraction pocket. These do behave differently to perforations. As both perforations and retractions can affect any part of the tympanic membrane, a classification based on tympanic membrane locations is not an accurate means of distinguishing between them. These classifications attempt on an anatomical basis to distinguish tympanic membrane pathologies which behave differently from one another. It is more germane to distinguish between tympanic membrane perforation and tympanic membrane retraction pockets (including cholesteatoma). This distinction is not difficult to make with the help of the otological microscope. There is evidence that granulation tissue, which is a feature of all forms of chronic otitis media, can also give



Chapter 74 Chronic otitis media in childhood



rise to acute intracranial infections. Thus even the distinction between cholesteatoma and perforations is not a valid means of identifying safe and unsafe disease.13



Epidemiology The epidemiology of chronic ear discharge associated with chronic ear perforations (chronic suppurative otitis media) has been widely studied in children. The WHO recognize a prevalence of chronic suppurative otitis media of more than 1 percent of children as presenting an ‘avoidable burden’ which can be managed in the ‘general health care context’1 and more than 4 percent as representing a ‘massive’ public health problem requiring urgent attention.1 The populations with the highest prevalence of childhood chronic suppurative otitis media are found among the Inuit and Australian Aboriginals. The most extreme rates of tympanic membrane perforation in a series have been recorded at 67 percent in Australian Aboriginals and 46 percent in the Inuit.14, 15 Apache and Navaho Indians, rural Maori and Solomon Islanders as well as some rural Indian and African populations have been found to have a point prevalence of chronic suppurative otitis media of over 5 percent. Other Native American, Indian and African populations have prevalence rates of 1–2 percent.1 Prospective studies of large populations of urban children in Saudi Arabia and Brazil found point prevalence of chronic suppurative otitis media of 1.3 and 0.9 percent, respectively.16, 17 A well-planned national study in South Korea found a rate of less than 1 percent in the paediatric population.18 No such thorough or representative population screening of paediatric populations has been performed recently in Europe or the USA. Audiometric screening studies of small populations suggests a point prevalence of less than 1 percent in these countries. By contrast, the UK National Study of Hearing found a somewhat higher rate (4.1 percent) of tympanic membrane perforation in the British adult population.19



Risk factors A firm understanding of the causes of the exceptionally high prevalence of chronic suppurative otitis media in Inuit and Australian Aboriginals remains elusive. In the past focus has been directed at the biological antecedents of disease. For instance, the racial variations in Eustachian tube anatomy at one time attracted attention as a possible cause for the early onset and persistence of otitis media. Subsequently, nasopharyngeal colonization by a wide variety of otological pathogens became recognized as a more pivotal risk factor.20 Early nasopharyngeal colonization is associated with early acute otitis media,21 which in turn is associated with the early onset of tympanic



] 933



membrane perforation. More recently, the social factors underlying the biological parameters have been determined. It has been established that these populations acquire pathogens early because of high rates of cross-infection from infected siblings in overcrowded conditions.22, 23 This problem also helps to perpetuate the carriage of pathogens, a process assisted by factors that reduce the child’s ability to clear the disease (see Table 74.2).



AGE



Data from aboriginal populations indicate a high prevalence of tympanic membrane perforation in very young children with a somewhat lower prevalence in adulthood. The highest rate of perforation is seen in the two- to four-year-old age group, at which stage the rate of perforation is roughly three times the rate seen in adulthood (see Figure 74.3).24, 29 By contrast, in South Korea, a rising prevalence with age is found (see Figure 74.4).18 No information was presented regarding whether these perforations were symptomatic or not, so these data may reflect a steady reduction in perforation rates in childhood over the last 50 years.



TYMPANOSTOMY TUBES



Tympanostomy tube insertion is a recognized cause of subsequent tympanic membrane perforation. A recent review of 62 studies revealed that chronic perforation of the tympanic membrane occurred in 964 of 20,222 ears (4.8 percent).30 Short-term tubes caused chronic perforations in 175 of 8107 ears (2.2 percent). Long-term tubes resulted in chronic perforation in 556 of 3356 ears (16.6 percent). More detailed metaanalysis from eight studies providing separate outcomes for both long-term and short-term tubes indicated that long-term tubes increased the relative risk of chronic perforation by 3.5 (CI: 1.5–7.1)30 compared with short-term tubes. Table 74.2 Risk factors proposed for high rates of chronic suppurative otitis media in indigenous children. Risk factor Early nasopharyngeal acquisition of otological pathogens due to high rates of cross-infection20, 22, 23 More siblings under the age of five24 More crowded accommodation23 Higher number of siblings with a history of ear inflammation23 Prolonged carriage rates of nasopharyngeal pathogens21, 25 Age at first episode of acute otitis media26 Poorer nutritional status27 Reduced exposure to medical services and supportive therapies Other biological factors Eustachian tube anatomy Differences in immune genes28



934 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



19.2 17.9 16



Percentage



15



14.2 12.5 10.4



10.1



10 8.8 7.4 5.5 4 2.8



0−2



2−4



4−6



6−8



8−10 10−12 12−14 14−16 16−18 18−20 20−30 30−40 40−50 50−60



Age



Prevalence rate (%)



4 3 2 1 0 1500 g;  extubation failure secondary to laryngeal pathology;



Figure 89.3 The anterior cricoid split procedure.



 no assisted ventilation for ten days before evaluation;  supplemental O2 requirement o35 percent;  no congestive heart failure for one month prior to evaluation;  no acute upper or lower respiratory tract infection at the time of evaluation;  no antihypertensive medication for ten days before evaluation. The procedure involves an anterior incision of the trachea from the second tracheal ring, up through the cricoid and into the lower third of the thyroid ala, just below the insertion of the anterior commissure. The child is then left intubated for ten days, with the neck wound left at least partly open to minimize the risk of subcutaneous air build-up. A thyroid alar interposition graft is a modification that permits earlier extubation. Laryngotracheal reconstruction: anterior cartilage graft Mild-to-moderate subglottic stenosis is well managed with costal cartilage grafting to the anterior cricoid (Figure 89.1).11 The anterior airway is split from the tracheotomy site to the lower aspect of the thyroid cartilage. An age-appropriate sized endotracheal tube or suprastomal stent is then inserted, and a measurement taken of the size of graft needed to comfortably close the deficit in the anterior airway. A costal cartilage graft is then harvested and carved to allow a boatshaped and perichondrium-lined insert to distract the anterior cricoid. An outer flange prevents graft prolapse into the airway. This technique is also useful for managing suprastomal collapse or narrowing of the upper trachea.



Chapter 89 Laryngeal stenosis



Laryngotracheal reconstruction: posterior cartilage graft Costal cartilage grafting of the posterior cricoid for subglottic stenosis may be performed in an identical fashion as for posterior glottic stenosis. If the anterior cricoid can then close comfortably over an appropriate sized endotracheal tube or stent, the additional anterior grafting is not required. Laryngotracheal reconstruction: anterior and posterior cartilage grafts If the anterior cricoid cannot comfortably close over an appropriately sized endotracheal tube or stent, then an



Figure 89.4



Cricotracheal resection (continued over).



] 1163



additional anterior graft is required, as previously described. This is necessary for most grade III and all grade IV stenoses. Cricotracheal resection Cricotracheal resection has an increasing role in the management of subglottic stenosis. This procedure requires the removal of the subglottic scar tissue, with the anastomosis of healthy trachea to a healthy larynx. This is a technically more challenging operation than laryngotracheal reconstruction with cartilage grafts (Figure 89.4).



1164 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Figure 89.4



Cricotracheal resection (continued).



The success of CTR in infants and children has been documented in two large series of patients.12, 13 [**/*] The results reported are superior to those of the laryngotracheal reconstruction procedures for similar indications and stenosis grades. CTR with primary anastomosis is a safe and effective procedure for the treatment of severe subglottic stenosis in infants and children. Diagnostic precision is essential, operative timing should be judged carefully, and operative technique must be precise. The reasons for the high success rate include the complete resection of the stenotic segment with restoration of a lumen using a normal tracheal ring, the preservation of normal laryngotracheal support structures without disruption of the cartilaginous framework, and full mucosal lining on both sides of the anastomosis, thus minimizing or preventing granulation tissue and restenosis. The complications of the surgery, however, should not be underestimated. The first is injury to the recurrent laryngeal nerve. Typically, the recurrent laryngeal nerve is not identified during the procedure as is lies posterior to the cricothyroid joint and the resection margin is anterior to the cricothyroid joint. The second complication is dehiscence of the anastomosis, which is most likely



to happen if there is tension at the site of the anastomosis or because of forceful reintubation if the endotracheal tube becomes dislodged. In the paediatric larynx the technique of laryngeal release is not required in most cases of CTR. If mobilization of the trachea is difficult or tracheal resection is extensive (more than five tracheal rings), then laryngeal release should be performed to minimize the risk of dehiscence. If granulation tissue is allowed to grow there is also a risk of restenosis, and for this reason sutures through the cartilage should emerge submucosally at the edge of the anastomosis. Experimental data in primates show that when CTR and primary tracheal anastomosis are performed with a good initial result, the thyroid cartilage and tracheal ring sutured together continue to grow normally. A further advantage of the CTR technique may be that voice quality should not deteriorate because the anterior commissure of the larynx is maintained in its initial position and there is no widening of the larynx posteriorly with an interposition graft. However, in older children, CTR may limit the ability to tilt the thyroid cartilage on the cricoid cartilage, in turn restricting cord tensioning and leading to a loss of vocal range.



Chapter 89 Laryngeal stenosis



CONTRAINDICATIONS All contraindications to airway reconstruction are relative. Usually, airway reconstruction is not attempted unless decannulation is the goal. Gastro-oesophageal reflux disease and eosinophilic oesophagitis should be controlled preoperatively, and pulmonary function optimized. Operating on a child requiring pulmonary pressure support to ventilate adequately is unwise. Single-stage reconstruction is inadvisable in a child who is difficult to intubate. In children with a history of sedation problems, past failure of airway reconstruction or multiple levels of airway pathology, single-stage reconstruction should be approached with caution. In children undergoing CTR, the risk of anastomotic dehiscence seems higher in the presence of Down syndrome, MRSA or a past history of distal tracheal surgery. The greatest disservice to a child is for airway reconstruction to cause or exacerbate ongoing aspiration.



COMPLICATIONS Complications may be subdivided into intraoperative, early postoperative and late postoperative.14 Intraoperative complications include bleeding, pneumothorax and loss of the airway with resultant hypoxia. Early postoperative complications include infection, air leakage from the operative site, dehiscence of an anastomosis and loss of a graft. The risk of air leakage and graft loss is highest when systemic steroid use is continued beyond two or three peri-extubation doses. With singlestage procedures there is a risk of accidental extubation, and risks associated with paralysis or sedation. Extubation may be compromised because of glottic oedema and granulation caused by the endotracheal tube. The most significant long-term complication is failure of the reconstruction with restenosis of the subglottis, with an incidence between 10 and 20 percent in most series.



] 1165



KEY POINTS  The narrowest point in the infant airway is the cricoid ring. The only fixed ring within the airway, it is the most vulnerable point for iatrogenic damage caused by intubation.  Acquired subglottic stenosis secondary to prolonged endotracheal intubation remains the most frequent cause of laryngeal stenosis.  Laryngeal stenosis continues to cause significant morbidity and mortality. The commonest primary aim of intervention is decannulation or preventing the need for tracheotomy. In selected patients, voice restoration or provision of a safer airway is the primary consideration, with decannulation a secondary goal.  Evaluation of the paediatric airway should not be considered a minor procedure, particularly when performed in a child with an unstable airway. Extreme care must be taken not to exacerbate the child’s condition.  Rigid endoscopy remains the gold standard for paediatric airway evaluation.  Laryngeal and upper tracheal reconstruction may be challenging; no single operation can address all types of laryngeal stenosis. It is prudent to evaluate each child on an individual basis.  The most significant long-term complication of surgery for laryngeal stenosis is failure of the reconstruction with restenosis of the subglottis, with an incidence between 10 and 20 percent in most series.  CTR has an increasing role in the management of subglottic stenosis. CTR is particularly effective as a salvage procedure following failed laryngotracheal reconstruction.  The greatest disservice to a child is for airway reconstruction to cause or exacerbate ongoing aspiration.



Revision airway surgery/the recalcitrant airway Best clinical practice Failure of laryngotracheal reconstruction or CTR does not preclude further attempts at reconstruction, but may complicate further reconstructive efforts. In revision airway surgery, particular care should be taken to optimize the outcome by careful preoperative evaluation of the patient and their airway. Failed expansion cartilage grafting may still be amenable to either further cartilage grafting or resection, while failed resection may still be amenable to further resection or cartilage grafting.



[ In the neonate who has failed extubation, the anterior cricoid split procedure is an alternative to tracheotomy. [Grade C] [ Preoperative evaluation of gastro-oesophageal reflux should be mandatory in a child with an active larynx or recalcitrant airway stenosis following previous reconstruction. [Grade B] [ Functional endoscopic evaluation of swallowing can provide valuable information about the risk and



1166 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



[ [



[



[ [



[



mechanism of aspiration as well as the presence or absence of normal laryngeal sensation. [Grade B] In a child who has not been evaluated previously, flexible nasopharyngoscopy with the child awake should be performed prior to rigid endoscopy. [Grade D] In children with significant lung disease, decannulation may be imprudent. Children with progressive neuromuscular disorders, diaphragmatic weakness or central hyperventilation syndrome, may not be candidates for decannulation. [Grade C] In a child with a compromised airway who does not have a tracheotomy tube, preoperative administration of dexamethasone, 0.5 mg/kg up to a maximum of 20 mg, is a prudent precaution. [Grade B] Spontaneous ventilation offers the best dynamic assessment of the airway and is thus recommended. [Grade D] In children in whom a stenting period of less than two weeks is anticipated, consider intubation, with the endotracheal tube acting as the stent (see Singlestage reconstruction). [Grade C] A prerequisite for single-stage laryngotracheal reconstruction is an excellent ICU in which staff are familiar with the management of airway patients. [Grade D]



2.



3.



4.



5.



6.



7.



8.



Deficiencies in current knowledge and areas for future research



$ $ $ $ $ $



While laryngotracheal reconstruction and CTR have become accepted management techniques for subglottic stenosis, supraglottic airway management remains poorly understood and managed. Preoperative evaluation and optimization of patients prior to reconstructive airway surgery still requires refinement. Recent technological advances, such as impedance probe evaluation of gastro-oesophageal reflux, need critical evaluation. Paediatric voice research may be an area for future endeavour. An airway grading system that is not limited to just the subglottis needs to be developed. Tissue engineering techniques may offer alternatives to current methods of airway reconstruction.



9.



 10.



11.



 12.



 13. REFERENCES







1. Myer III CM, O’Connor DM, Cotton RT. Proposed grading system for subglottic stenosis based on endotracheal tube sizes. Annals of Otology, Rhinology, and Laryngology.



14.



1994; 103: 319–23. Description of the most widely accepted grading system for subglottic stenosis. Koufman JA. The otolaryngologic manifestations of gastroesophageal reflux disease (GERD): a clinical investigation of 225 patients using ambulatory 24-hour pH monitoring and an experimental investigation of the role of acid and pepsin in the development of laryngeal injury. Laryngoscope. 1991; 101: 1–78. Wenzl TG. Evaluation of gastroesophageal reflux events in children using multichannel intraluminal electrical impedance. American Journal of Medicine. 2003; 115: 161S–5S. Johnson LB, Rutter MJ, Putnam PE, Cotton RT. Airway stenosis and eosinophilic esophagitis. Transactions American Bronchoesophagological Association. Nashville, TN, 2003, 137. Zalzal GH, Cotton RT, McAdams AJ. The survival of costal cartilage grafts in laryngotracheal reconstruction. Otolaryngology and Head and Neck Surgery. 1986; 94: 204–11. Zalzal GH. Rib cartilage grafts for the treatment of posterior glottic and subglottic stenosis in children. Annals of Otology, Rhinology and Laryngology. 1988; 97: 506–11. Gustafson LM, Hartley BE, Liu JH, Link DT, Chadwell J, Koebbe C et al. Single-stage laryngotracheal reconstruction in children: a review of 200 cases. Otolaryngology and Head and Neck Surgery. 2000; 123: 430–4. Cotton RT. The problem of pediatric laryngotracheal stenosis: a clinical and experimental study on the efficacy of autogenous cartilaginous grafts placed between the vertically divided halves of the posterior lamina of the cricoid cartilage. Laryngoscope. 1991; 101: 1–34. Rutter MJ, Cotton RT. The use of posterior cricoid grafting in managing isolated posterior glottic stenosis in children. Archives of Otolaryngology – Head and Neck Surgery. 2004; 130: 737–9. Cotton RT, Myer III CM, Bratcher GO, Fitton CM. Anterior cricoid split 1977–1987: evolution of a technique. Archives of Otolaryngology – Head and Neck Surgery. 1988; 114: 1300–2. First description of the cricoid split procedure. Cotton RT, Gray SD, Miller RP. Update of the Cincinnati experience in pediatric laryngotracheal reconstruction. Laryngoscope. 1989; 99: 1111–6. Monnier P, Lang F, Savary M. Partial cricotracheal resection for pediatric subglottic stenosis: a single institution’s experience in 60 cases. European Archives of Oto-rhino-laryngology. 2003; 260: 295–7. Outcome analysis of cricotracheal resection. White DR, Cotton RT, Bean JA, Rutter MJ. Pediatric cricotracheal resection: surgical outcomes and risk factor analysis. Archives of Otolaryngology–Head and Neck Surgery. 2005; 131: 896–9. Update on the Cincinnati experience with cricotracheal resection. Cotton RT. Management of subglottic stenosis. Otololaryngologic Clinics of North America. 2000; 33: 111–30.



90 Paediatric voice disorders BEN HARTLEY



Introduction Growth and development of the larynx Assessment of the child’s voice Specific disorders Key points



1167 1167 1168 1169 1171



Best clinical practice Deficiencies in current knowledge and areas for future research References



1171 1172 1172



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words paediatric, voice and dysphonia.



INTRODUCTION There have been tremendous advances in the understanding of voice disorders and their management in recent years. The subspecialty of phoniatrics has evolved principally in adults and this knowledge may now be applied to voice disorders in children. Disorders of voice – where the predominant symptom is hoarseness or ‘dysphonia’ – must be distinguished from disorders of speech, articulation and language. Speech and articulatory disorders are characterized by difficulty in producing speech sounds, often in the presence of normal laryngeal function. In language disorders, the child uses words and sentences inappropriately, again usually despite normal laryngeal function. Voice disorders are largely a manifestation of laryngeal pathology. Six to 23 percent of 5–18 year olds have some form of voice ‘problem’.1, 2 In the majority, neither the child nor their parents perceive the voice as a concern and they do not present for medical evaluation. The spectrum of voice disorders seen in a specialist paediatric voice clinic is very wide. Patients range from child performers who have a normal conversational voice but whose parents have been concerned about a loss of the upper range of their singing voice, to children with severe congenital or acquired laryngeal disease with no voice at all.



With careful assessment, voice therapy and occasionally surgical intervention, most paediatric dysphonias can be corrected or improved.



GROWTH AND DEVELOPMENT OF THE LARYNX The paediatric larynx is quite different from that of the adult (Figure 90.1 and Table 90.1). The embryonic development of the larynx is described in detail in Chapter 162, Anatomy of the larynx and tracheobronchial tree. This section focusses on those aspects relevant to voice disorders and on normal growth and development after birth. Current knowledge stems largely from the detailed studies of Hirano et al.3, 4



Elongation of the vocal folds Hirano found that up to the age of ten years, the length of the vocal fold was very similar in both males and females (6–8 mm). At puberty there is a substantial increase in growth, much more marked in males with the membranous vocal fold increasing to 14.8–18 mm (more than double). In females the increase is to 8.5–12 mm – an increase in length of a third. The cartilaginous portion of



1168 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY The five layers of the vocal fold are the epithelium, superficial layer of lamina propria (Reinke’s space), the intermediate and deep layers of the lamina propria and the muscle layer. The first two are referred to the as the ‘cover’ and move freely over the deeper layers which form the ‘ligament’ and ‘body’ of the vocal fold. This laminar structure is not present at birth, but starts to differentiate over the first few months of life. It becomes more developed throughout childhood and the adult form is quite easily recognizable by puberty. One surgical implication of this is that microflaps are more difficult to raise in early childhood due to a less welldeveloped plane for dissection in the superficial lamina propria.



Changes in pitch



Figure 90.1 Normal larynx in a six-month-old child. Table 90.1 Feature



Differences between the larynx of an adult and child. Difference



Size Position Shape Mucosa



The paediatric larynx is relatively smaller The child’s larynx is relatively higher Curled epiglottis, shorter vocal folds More reactive and prone to airway obstruction. Croup is uncommon in adults The laminar vocal fold Immature in young children structure



the vocal fold also grows with age but less rapidly, with a relative increase in the ratio of the membranous to cartilaginous vocal fold from 1.5 in the newborn to 4.0 in the adult female and 5.5 in the adult male. Hirano devised the concept of the respiratory glottis posteriorly with a wider aperture and the phonatory glottis anteriorly. These concepts are useful in planning surgical procedures. For example, in bilateral vocal cord paralysis it is possible to increase the airway by performing an ablative procedure, such as a laser arytenoidectomy. If this procedure is kept to the posterior part of the larynx then the anterior membranous vocal fold can be preserved for phonation.



Changes in vocal fold laminar structure Much has been written about the layered structure of the vocal folds and its importance both to the understanding of disease and the development of phonosurgical treatment.



An important feature of the paediatric voice is its pitch. This drops throughout infancy and childhood in males and females, with a marked change at puberty, particularly in males. This change in pitch corresponds to the anterior growth of the thyroid cartilage in response to testosterone and coincides with the development externally of the thyroid prominence or Adam’s apple. The fall in pitch is approximately proportional to the growth of the membranous vocal fold.



ASSESSMENT OF THE CHILD’S VOICE Children with voice disorders can either be seen in a general clinic or in a voice clinic. An otolaryngologist and speech and language therapist should staff paediatric voice clinics. Videostroboscopic equipment and the expertise to use it should be available. The history and basic otolaryngological examination is the same in both settings. Whether laryngeal videostroboscopy and speech therapy assessment is required will depend on the clinical situation.



History and examination Mild or moderate dysphonias that have been present for some time tend to be accepted as part of the child’s personality and not a medical disorder. Often, an incidental remark years later leads to the parents seeking medical assessment to exclude an underlying disorder. Some children and families see no problem with their dysphonia. Alternatively, some children and families find a mild dysphonia very troublesome, particularly if they are performers or have aspirations to perform. Information should be sought from both the child and parents. If the problem has been present from birth, a congenital lesion is likely. However, a history of endotracheal intubation around the time of birth or around the



Chapter 90 Paediatric voice disorders



time of onset of symptoms may suggest laryngeal stenosis, cricoarytenoid joint fibrosis, intubation granuloma or cyst formation. Much more commonly, symptoms start with an upper respiratory tract infection that has been accompanied by laryngitis, a situation made worse by habitual patterns of voice misuse. The severity of the disorder may range from a loss of singing voice to complete loss of conversational voice. The time course of the dysphonia is also important. For example, dysphonia is often persistent with discrete vocal fold lesions and rarely returns to normal, although it may fluctuate and fatigue during the day. Enquire about symptoms suggestive of gastro-oesophageal reflux, as this may irritate the larynx and cause dysphonia. Post-nasal drip associated with allergic rhinitis will do the same, particulary if there is a habit of forceful or constant throat clearing. Cough is quite harsh on the vocal mechanism and constant coughing associated with respiratory disease may lead to hoarseness. Restrictive respiratory disease may cause reduced infraglottic pressure and subsequent dysphonia. The use of corticosteroid inhalers can also cause dysphonia and this can be helped by modification of drug regime or possibly inhaler technique and the use of spacer devices.5 Hearing loss may make the child shout. This can be the underlying cause of a dysphonia secondary to vocal misuse. Voice misuse – shouting – is common in children and may lead to disorders of hyperfunction such as nodules. Other abusive behaviours, such as smoking and alcohol, may occasionally be relevant. It is extremely important to enquire about exercise intolerance and stridor, symptoms and signs that may be caused by laryngeal stenosis. Swallowing problems or choking may be the first indication of laryngeal paralysis. A general otolaryngological examination, including assessment of the ears and hearing, should be performed. Important clues can be gained by carefully listening to the quality of the child’s voice during history taking. Laryngoscopy should be undertaken in all cases.



Most focal lesions can be excluded by awake fibreoptic laryngoscopy. This can be performed on a child of almost any age. It is usually quite straightforward, performed transnasally with a 2.2-mm fibreoptic endoscope. The optics of the larger fibreoptic endoscopes are better and, if possible, a 4-mm endoscope should be used. From age one to five years, compliance is limited and general anaesthesia and microlaryngoscopy may need to be considered. Some images are of sufficient quality to permit stroboscopy with examination of the mucosal wave. Rigid laryngoscopy requires significant cooperation, which can only be obtained in children over six years of age. High quality images of the larynx combined with stroboscopy give unparalleled information on vocal fold movement and structure. They also provide an important educational tool for parents and older children. Paralysis of a vocal cord is generally obvious but this technique gives insight to more subtle mobility disorders such as limited posterior glottic closure (glottic chink) and supraglottic constriction. This degree of information is useful and relevant to planning voice therapy.



SPECIFIC DISORDERS Vocal nodules and functional voice disorders Vocal nodules – now regarded as an organic manifestation of laryngeal hyperfunction – are the commonest cause of dysphonia in children (Figure 90.2).6 The mainstay of treatment is voice therapy to correct the hyperfunction. While in adolescents this therapy is similar to that employed in adults, different strategies are necessary in younger children. The traditional view, based on clinical experience, was that most vocal nodules in children could be expected to improve at puberty.7 Puberty is a time of great change in the larynx, with tremendous growth of the membranous



Laryngoscopy The objectives of laryngoscopy are two-fold, first to identify any structural lesion, such as a vocal nodule or papilloma; second, to assess laryngeal mobility during phonation. Older children may be cooperative enough for indirect laryngoscopy but examination of the paediatric larynx has traditionally been performed under general anaesthetic. Detailed structural information can be obtained but little information is gained with regard to mobility. It is usual to watch vocal mobility on awakening from general anaesthesia. This can provide information with regard to vocal paralysis, but it should be recognized that by today’s standards this is a very crude method of assessment. Awake laryngeal and voice examination should be the standard of care in a compliant child.



] 1169



Figure 90.2



Vocal cord nodules.



1170 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY vocal fold in the male and, to a lesser extent, in the female. One can see how the dynamics of vocalization might change with potential improvement in the nodules. Mori,8 however, reported that 12 percent of nodules did not improve at puberty. In the same study it was found that those treated with vocal hygiene alone (i.e. general advice regarding voice care) did not improve. Those given voice therapy under the supervision of a speech and language therapist did tend to improve and the degree of improvement was related to the number of therapy sessions. [**] The essentials of voice therapy include reduction of vocal strain. Various techniques can be employed to reduce shouting, whispering, coughing and throat clearing and to encourage the use of a smooth easy voice. Periods of quiet play are recommended after noisy activity (e.g. football) to allow vocal fold recovery. Surgery for nodules is rarely recommended in children. Concerns about the potential effects of scarring are very real. On occasion it can be justified when prolonged voice therapy has failed. Bouchayer and Cornut9 reported cysts, polyps and sulci in children previously diagnosed as having nodules when examined at microlaryngoscopy. Some children with nodules have vocal cord cysts with a nodule on the contralateral cord and this might explain why some fail to respond to voice therapy. The use of videostroboscopic techniques in the voice clinic should help diagnose these conditions earlier. Cysts are treated by microsurgical excision. A wide range of functional voice disorders can be demonstrated in children without demonstrable nodules. These children may present with dysphonia, or even aphonia, secondary to underlying psychological factors. For these children, voice therapy may need to be combined with a psychological assessment. Aphonia in the presence of a normal laryngoscopy evaluation is suggestive of psychological disturbance. A normal cough or laugh adds support to this contention. Puberphonia, when the prepubertal voice persists into adolescence or adulthood, is another condition associated with psychological disturbance. Highly specialized voice therapy, possibly in conjunction with a psychologist, is important to help these children develop an adult voice.



commend it. After all, surgery cannot cure papillomatosis. Medical treatments have been advocated including intra-lesional acyclovir with some reported success.11 [**] In theory, any medical adjunct that reduces the number of surgical procedures would be beneficial in terms of voice preservation, although voice outcomes following these therapies have yet to be reported. Cidofovir is showing some promise and its use is considered in Chapter 91, Juvenile-onset recurrent respiratory papillomatosis.



Intubation injuries and voice Much has been written about acquired subglottic stenosis consequent on prolonged endotracheal intubation in premature infants. Surgical techniques have evolved to correct this abnormality and allow eventual decannulation.12, 13, 14 Voice problems following this surgery have been documented and are not uncommon.15 Subglottic stenosis is just one example of laryngeal injury caused by intubation. More minor scarring may affect the voice but not the airway (Figures 90.3 and 90.4). Damage to the vocal fold or underlying lamina propria causes dysphonia. Cricoarytenoid joint fixation and posterior glottic scarring may also be caused by endotracheal intubation and reduce vocal fold mobility. It is easy to confuse these conditions with vocal cord paralysis. Voice therapy can be successful in some children. Surgical treatment of voice problems caused by intubation trauma remains a significant challenge. There is potential for the introduction of medialization techniques in this group of children, but there must be an adequate airway before this can be considered and this situation is uncommon.



Laryngeal papillomatosis A detailed account of laryngeal papillomatosis is given elsewhere (Chapter 91, Juvenile-onset recurrent respiratory papillomatosis). From the voice perspective, there are a number of important considerations. The emphasis of modern surgical techniques is to minimize damage to the underlying lamina propria and to restrict surgery to the papillomas. In this way, voice quality will be optimized in the long term. Glottic scarring and webbing are well-known complications of surgical intervention with the laser.10 The use of a microdebrider and cold steel instruments has much to



Figure 90.3 Intubation granulomas.



Chapter 90 Paediatric voice disorders



] 1171



carry out these procedures as nerve grafting in general is more successful in this age group. It is worth restating that it should never be assumed that an immobile vocal cord is paralyzed. Vocal cord immobility in previously intubated children may be due to joint fixation or posterior glottic scarring. The distinction between these diagnoses is extremely important, particularly when vocal cord immobility is bilateral. The management of paralysis is conservative with a tendency towards spontaneous recovery. Fixation and scarring do not improve and children may be subject to years of unnecessary waiting with a tracheotomy before a surgical lateralization procedure is contemplated.



Tracheotomy and voice Figure 90.4 Vocal cord polyp following prolonged endotracheal intubation.



Vocal cord paralysis In general terms, children with bilateral vocal cord paralysis have a good voice or cry. It is the airway that gives concern. Spontaneous recovery of function takes place in some, but after two to three years of observation this is unlikely to happen.16 A number of lateralization procedures can be performed to develop a wide posterior glottis for respiration while preserving the anterior glottis for phonation. Either laser arytenoidectomy17, 18 or endoscopic suture lateralization can achieve these goals. In contrast to bilateral cord paresis, children with unilateral cord paralysis present with aspiration and a weak cry. The airway is relatively spared. Most improve with time due to compensation from the contralateral nonparalyzed vocal fold or to spontaneous recovery. Medialization thyroplasty can be considered for those that do not improve and continue to have a significant voice problem. However, there is limited published information on this and it should be remembered that its efficacy to diminish aspiration has been limited.19 The concept of placing a silastic implant before puberty has raised concerns that this might interfere with the anticipated physiological growth spurt. It is the author’s practice to consider medialization using a fat injection in dysphonic, prepubertal children and to restrict silastic thyroplasty techniques for dysphonic post-pubertal patients. [*] The concept of reinnervation is attractive. In a series of eight children with unilateral vocal cord paralysis, three underwent implantation with ansa cervicalis/strap muscle pedicles. This, together with postoperative voice therapy, gave ‘good’ results.20 There has to be some scepticism about these procedures as re-innervation techniques are capable of restoring laryngeal tone but have had limited success in achieving voluntary movement. Nevertheless, children seem to be the ideal population on which to



Children with tracheotomies will often have impaired voice. This is only partly due to the tracheotomy tube diverting air from the glottis. The main factor is usually the laryngeal lesion that required tracheotomy formation in the first place. Children with a healthy larynx can obtain good voice by occluding the tube on expiration (with a finger or speech valve) and projecting air around the tube and up through the glottis. To facilitate this, the tube needs to have sufficient space around it and may need to be ‘downsized’. This is usually possible without restricting the airway. Speech valves are frequently used in the author’s practice for children. As well as helping with speech they help with secretion management and swallowing. [*]



KEY POINTS  Voice disorders in children are common; few require medical intervention.  A diagnosis can usually be made by careful history taking and examination to include laryngoscopy and ideally videostroboscopy.  Laryngeal muscle hyperfunction with or without nodules constitutes the commonest group of disorders.  These disorders respond well to skilled voice therapy.



Best clinical practice [ Laryngeal microflaps are more difficult to raise in early childhood due to a less well-developed plane for dissection in the superficial lamina propria. [Grade A] [ An immobile vocal cord is not always paralyzed. Consider joint fixation and posterior glottic scarring, especially if the child has been intubated. [Grade B]



1172 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY [ The term ‘vocal abuse’, often used to describe adult dysphonias, is best replaced by ‘voice misuse’ in the paediatric setting. Beware the sensitivity of recording the word ‘abuse’ in children’s records. [Grade D] [ Advances in endoscopic equipment are such that awake laryngeal and voice examination should now be the standard of care in a compliant child. [Grade C] [ Due to the potential for permanent scarring, surgery for nodules is rarely recommended in children. It should be considered only as a last resort when prolonged and skilled voice therapy has failed. [Grade B]



Deficiencies in current knowledge and areas for future research



$



$



$



Laryngeal transplantation for voice. Perhaps the most challenging of paediatric voice conditions is the child with no voice due to complete laryngeal stenosis which is beyond surgical reconstruction. Such children are encountered after severe burns or caustic ingestion. They are tracheotomy dependent and communicate with signing and non-oral devices. Laryngeal replacement using transplantation carries the best hope for these children and although there has been one successful adult patient, this has yet to be repeated. Reinnervation of the paralyzed larynx. Phonosurgery can be helpful for the voice in laryngeal paralysis with medialization using injection or thyroplasty techniques. However, the biggest concern for children with laryngeal paralysis is not voice but aspiration. We do not have a universally accepted solution for aspiration due to laryngeal incompetence. Withdrawal of oral feeding and gastrostomy remains the mainstay. Fortunately, compensation frequently takes place for unilateral lesions but often not bilateral problems. Laryngeal reinnervation may hold the key but is still in its infancy. Repair of damaged vocal cords with synthetic ‘SLP’. Vocal cords damaged by intubation or surgery have lost the mucosal wave due to destruction of the important superficial layer of the lamina propria (SLP layer). Medialization of damaged vocal cords has had only limited success and attempts are being made to create a synthetic substitute for this important gel layer for potential replacement by injection into Reinke’s space.



REFERENCES 1. Aran D, Ekelman B, Nation J. Preschoolers with language disorders – 10 years later. Journal of Speech and Hearing Research. 1984; 27: 232–44.



  











2. Silverman EM. Incidence of chronic hoarseness among school age children. Journal of Speech and Hearing Disorders. 1975; 40: 211–5. 3. Hirano M, Kurita S, Nakashima T. In: Bless DM, Abbs JH (eds). Vocal fold physiology: contemporary research and clinical issues. San Diego: College Hill Press, 1983: 37–56. 4. Hirano M, Kurita S, Kiyokawa K, Sato K. Posterior glottis: morphological study in excised human larynges. Annals of Otology, Rhinology and Laryngology. 1986; 95: 576–81. 5. Roland NJ, Bhalla RK, Earis J. The local side effects of inhaled corticosteroids. Chest. 2004; 126: 213–9. 6. Koufman JA, Blalock PD. Functional voice disorders. Otolaryngologic Clinics of North America. 1991; 24: 1059–73. 7. Toohill RJ. The psychosomatic aspects of children with vocal nodules. Archives of Otolaryngology. 1975; 101: 591–5. 8. Mori K. Vocal nodules in children, preferable therapy. International Journal of Pediatric Otorhinolaryngology. 1999; 49: S303–6. 9. Bouchayer M, Cornut G. Microsurgical treatment for benign vocal fold lesions, indications, technique, results. Folia Phoniatric (Basel). 1992; 44: 155–84. 10. Crockett DM, McCabe BF, Shive CJ. Complications of laser surgery for respiratory papillomatosis. Annals of Otology, Rhinology and Laryngology. 1987; 96: 639–44. 11. Pransky SM, Brewster DF, Magit AE, Kearns DB. Clinical update on 10 children treated with intralesional acyclovir injections for severe recurrent respiratory papillomatosis. Archives of Otolaryngology – Head and Neck Surgery. 2000; 126: 1239–43. 12. Hartnick CJ, Hartley BEJ, Lacy PD, Liu J, Willging JP, Myer CM et al. Surgery for pediatric subglottic stenosis disease – specific outcomes. Annals of Otology, Rhinology and Laryngology. 2001; 110: 1109–13. 13. Rutter MJ, Hartley BEJ, Cotton RT. Cricotracheal resection in children. Archives of Otolaryngology – Head and Neck Surgery. 2001; 127: 322–4. 14. Hartley BEJ, Cotton RT. Paediatric airway stenosis: laryngotracheal reconstruction or cricotracheal resection? Clinical Otolaryngology and Allied Sciences. 2000; 25: 342–9. 15. Clary RA, Pengilly A, Bailey M, Jones N, Albert D, Comins J et al. Analysis of voice outcomes in pediatric patients following surgical procedures for laryngotracheal stenosis. Archives of Otolaryngology – Head and Neck Surgery. 1996; 122: 1189–94. 16. Tucker HM. Vocal cord paralysis in small children: principles and management. Annals of Otology, Rhinology and Laryngology. 1986; 95: 618–21. 17. Worley G, Bajaj J, Cavalli L, Hartley BEJ. Laser arytenoidectomy in children with bilateral vocal cord immobility. Journal of Laryngology and Otology. 2007; 121: 25–7.



Chapter 90 Paediatric voice disorders 18. Hollinger LD, Lusk RP, Green CG. Pediatric laryngology and bronchoesophagology. New York: Lippincott–Raven Publishers, 148. 19. Link DT, Rutter MR, Liu JH. Pediatric Type 1 thyroplasty: an evolving procedure. Annals of Otology, Rhinology and Laryngology. 1999; 108: 1105–10.



] 1173



20. Senturia BH, Wilson FB. Otorhinolaryngic findings in children with voice deviations. Preliminary report Annals of Otology, Rhinology and Laryngology. 1968; 77: 1027–41.



91 Juvenile-onset recurrent respiratory papillomatosis MICHAEL KUO AND WILLIAM J PRIMROSE



Introduction Aetiology Epidemiology Clinical presentation Diagnosis Staging Treatment Natural history



1174 1174 1175 1175 1175 1176 1176 1179



Tracheobronchial disease Tracheostomy Malignant disease Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



1179 1179 1180 1180 1180 1180 1181



SEARCH STRATEGY The data in this chapter are supported by a PubMed (NLM) search using the key words respiratory papillomatosis and laryngeal papillomatosis.



INTRODUCTION



AETIOLOGY



Recurrent respiratory papillomatosis (RRP) is a potentially life-threatening disease characterized by the development of papillomas anywhere in the respiratory tract from the nasal vestibules to the terminal bronchi. Predominant sites are where there is a change of epithelium, e.g. from squamous to ciliated, and especially the tonsillar pillars, uvula, vocal folds and laryngeal commisure. There is a bimodal age distribution with a juvenile-onset peak occurring at three to four years of age and an adult-onset peak occurring at 20–30 years of age. Boys and girls appear to be nearly equally affected in juvenile-onset RRP (JORRP). This contrasts with adult-onset RRP, which is a disease transmitted via sexual contact or via indirect contact with anogenital lesions, preferentially affecting men by a ratio of approximately three to two. Although adult-onset respiratory papillomatosis and juvenile-onset respiratory papillomatosis share many features, this chapter addresses only JORRP.



Juvenile respiratory papillomatosis was first described by Morrell Mackenzie in 1880. By 1923, Ullmann had demonstrated an infective aetiology by injecting homogenized papilloma from a child’s larynx into his own forearm and inducing local growth of papillomas. Further direct evidence of the association between human papilloma virus (HPV) and RRP came from the identification of HPV DNA within laryngeal papillomas by Southern blot hybridization and the subsequent recognition in papillomas of HPV types 6 and 11.1 Human papillomavirus is a naked, double-stranded, icosahedrally-shaped virus with circular supercoiled double-stranded DNA genome surrounded by an outer capsid of protein that belongs to the Papovavirus family. There are 90 known subtypes of HPV – although many vary only slightly in their DNA sequence – but only types 6, 11 and rarely 16 are associated with RRP. HPV types 6 and 11 are also associated with condyloma acuminata (genital warts). Types 16 and 18 have been implicated in



Chapter 91 Juvenile-onset recurrent respiratory papillomatosis



carcinogenesis, particularly in the uterine cervix and in squamous cell carcinoma of the head and neck. HPV is thought to first enter traumatized epithelium and reside in the basal layer of the mucous membrane, where it replicates by a process known as episomal maintenance. This replication interferes with the normal process of cell maturation causing epithelial proliferation and neovascularization. Conversely, the virus may lie dormant causing subclinical infection and can often be recovered from apparently normal tissue adjacent to papillomas. Viral protein, DNA synthesis and virion assembly only takes place in the granular and cornified layers of the terminally differentiated epithelium.



] 1175



controls, leading to speculation that the clinical severity of disease is due not to a failure of T-cell proliferative response to HPV, but a delay in that response due to a low frequency of HPV-specific T cells or modulation of the T cell response by immunoregulatory networks.5 Bonagura et al.4 showed that HLA-DRB1*0301-positive patients exhibited reduced interferon-gamma expression. While the associations between JORRP and HLA-genotype are being increasingly well recognized, the underlying mechanism for evasion of the cellular immune response is far from clear.



CLINICAL PRESENTATION EPIDEMIOLOGY Juvenile-onset RRV is an uncommon condition, with a prevalence of only four in 100,000 children. The oftquoted triad of susceptibility factors for JORRP – young mother, vaginal delivery and low maternal socioeconomic status – is of limited predictive usefulness.2 Latent or active HPV has been detected in cervical swabs from 10–25 percent of women of childbearing age. The associations between HPV, JORRP and a history of maternal genital warts are well established but what is uncertain is the influence, if any, these associations should have on obstetric management. HPV DNA has been found in one-third to one-half of aerodigestive tract swabs of children born to affected mothers. However, the majority of these children do not develop disease. Calculations suggest that only one in 400 infants delivered to women with genital warts subsequently develop JORRP. This relative risk is much lower than that for sexually transmitted diseases. The discrepancy between the incidence of HPV colonization of the aerodigestive tract of the infant and the incidence of RRP implies that host factors must be involved in determining susceptibility. Although a history of maternal genital warts is not universal, a large retrospective study has shown that children born to mothers with genital warts carry a 231 relative risk of developing JORRP. A prolonged delivery time (exceeding ten hours) conferred a two-fold increased risk but delivery by Caesarean section did not appear to reduce that risk.3 The evidence for protection from vertical transmission of HPV into the baby’s upper respiratory tract by Caesarean section is conflicting and therefore the decision on method of delivery must be made on a case-by-case basis. Current evidence does not warrant Caesarean section as a prophylaxis against RRP. Variation in the susceptibility of the host to viral infection may be associated with specific HLA polymorphisms, several of which have been reported. Two recent, large independent studies have shown an association between HLA-DRB1*0301-and severe disease.4, 5 T-cell responses were shown to be the same by one group between HLA-DRB1*0301-positive patients and negative



Although respiratory papillomas can arise in any respiratory mucosa, their initial presentation is usually in the larynx. The diagnosis requires the surgeon to have an awareness of the condition as the presenting symptoms can be variable and mimic many common laryngeal and respiratory pathologies in children. In addition to hoarseness and stridor, children may present with a chronic cough, paroxysms of choking, recurrent respiratory infections or failure to thrive. These latter symptoms may lead to a misdiagnosis of asthma, laryngitis, bronchitis or croup and a delay of diagnosis of JORRP of up to eight years.1



DIAGNOSIS If at all possible, the clinical diagnosis should be established with awake fibreoptic nasolaryngoscopy (using an infant 2.2 mm endoscope) because the difficulties presented to the anaesthetist by unexpected laryngeal papillomas prolapsing into and obstructing the glottis cannot be overstated.6, 7, 8 The preferred anaesthetic technique in our institution is that of spontaneous respiration without endotracheal intubation. General anaesthesia is induced either by intravenous propofol or, more frequently, by inhalation of sevoflurane in oxygen. The larynx is topically anaesthetized with 2 percent lidocaine and anaesthesia maintained by sevoflurane in oxygen through a nasopharyngeal airway, of a size and length appropriate for the child’s age. This allows excellent surgical access to the airway but carries the disadvantage that the lower airway is not directly protected from bleeding. Therefore, meticulous haemostasis is required during the procedure using topical epinephrine (1:10,000 applied on neurosurgical patties). Regardless of the surgical technique, it is important for the child to recover from anaesthesia with humidified oxygen. Many units use preoperative dexamethasone to reduce laryngeal oedema but some surgeons feel that antiinflammatory agents are best avoided during active manipulation of papilloma tissue. Laryngopharyngeal exposure to gastric acid is increasingly being recognized



1176 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY as a cause of laryngeal pathology. This is particularly relevant during the immediate postoperative period where laryngeal mucosa has been breached.9 Therefore, it is our practice to give prophylaxis against gastrooesophageal reflux with an H2-antagonist or a proton pump inhibitor for 48 hours after all laryngeal surgery including surgery for JORRP. [*] Macroscopically, papillomas can be pedunculated or sessile, spread over the mucosal surface of the larynx (Figure 91.1). They tend not to be friable and can be grasped using microlaryngeal instruments and excised for histological examination. Microscopically, the papillomas appear as exophytic projections of keratinized squamous epithelium overlying a fibrovascular core, with varying degrees of dyskeratosis, parakeratosis and dysplasia. Koilocytes (vacuolated cells with clear cytoplasmic inclusions) are often seen indicating viral infection.



STAGING Staging of disease is not universally undertaken even in units with a large case-load of RRP. The staging system proposed and modified by Derkay10, 11 is based on a combination of the anatomical distribution and extent of lesions and their clinical effects on voice and the child’s airway. It is increasingly used by members of the American Society of Pediatric Otolaryngologists (ASPO). Widespread adoption of a universally agreed staging system would enable comparison of results between centres and facilitate multi-centre trials, particularly of adjuvant treatments.



mucosa and vocal cords. In children requiring repeated extirpations of extensive papillomas, especially if they predominantly occur in the larynx, it may be ultimately impossible to achieve normal voice. It is in these patients that the surgical technique, balance of the extent of lesion removal against interval timing, and application of adjuvant therapies must be combined and finely judged to ensure minimal dysphonia and airway stenosis, both of which can be extremely difficult to manage.



Surgical treatment POWERED MICRODEBRIDER



The use of the powered microdebrider is a relatively new development in the surgical removal of laryngeal papillomas, but one which has become the gold standard for papilloma removal in the larynx (Figure 91.2).12 The laryngeal debrider blade allows gentle but comprehensive removal of papillomas with minimal contamination of the lower respiratory tract with blood or papillomas. There is no thermal trauma and using direct endoscopic control it is extremely precise with minimal mucosal damage. Small retrospective studies demonstrate that, compared with the CO2 laser, patients undergoing laryngeal papilloma debridement have good disease clearance, require a shorter procedure and experience less postoperative pain. One retrospective study showed no incidence of delayed soft tissue complications using the microdebrider.13 Long-term results of this are still awaited, but early results are extremely promising.14, 15, 16 [**]



TREATMENT The aim of surgical treatment for recurrent respiratory papillomatosis is the removal of papillomas and restoration of a normal airway while minimizing trauma to the



COLD STEEL SURGERY



Advocates of papilloma removal using microlaryngeal instruments claim that the use of a microflap technique minimizes trauma to the vocal fold while satisfying disease clearance. Thermal damage to neighbouring tissue is also avoided as is the vapour plume. Cold steel surgery has the distinct disadvantage of having no direct haemostasis when dealing with a very vascular lesion. Nevertheless, such surgery is successful in the hands of the exponents of this technique. [**]



CO2, KTP, ND:YAG AND PULSED-DYE LASER



Figure 91.1 larynx.



Recurrent respiratory papillomatosis in the



The carbon dioxide (CO2) laser has been, for many years, the mainstay of surgical management of JORRP. It remains the treatment of choice for many surgeons because of its ability to ablate the papillomas with minimal bleeding and its ease of use with a microscope and micromanipulator. However, the frequency of late soft tissue complications, such as vocal fold fibrosis, interarytenoid fibrosis and stenosis, glottic webbing and



Chapter 91 Juvenile-onset recurrent respiratory papillomatosis



(a)



Figure 91.2



] 1177



(b)



(a) Microdebrider; (b) close-up of blade.



arytenoid fixation, has been reported to be 13–45 percent.6, 17, 18 This increases with frequency of treatment and the number of laser ablations the child has received. Soft tissue complications can be minimized by appropriate laser settings and careful assessment of depth of ablation. A Japanese group has suggested, based upon a single case, a two-stage procedure whereby the bulk of the papillomas are removed at the first operation, which is then followed by a repeat endoscopy and laser ablation ten days later, by which time the coagulum and carbonized tissue have resolved, allowing a more precise laser clearance of residual papillomas.19 It is also important to have an effective plume extractor. This serves to allow a clearer image of the larynx, but also removes the potentially infectious laser vapour plume which carries a significant viral load. The latter may be simple conjecture as there has only been one reported case of a surgeon developing laryngeal papillomas with a history of repeated laser treatment to a patient with anogenital condylomas, in situ hybridization showing HPV types 6 and 11 in both the patient’s and his lesions.20 The KTP laser and the Nd:YAG laser are as effective as the CO2 laser in papilloma ablation and haemostasis but, in addition, can be delivered through an optical fibre. Fibre-delivered laser systems play a role predominantly in the treatment of tracheal and bronchial papillomas. A new fibre-guidance system with a bendable distal tip developed for the Nd:YAG laser achieves a 501 range of directional manoeuvrability with minimal power loss.21 In adult patients, the delivery of pulsed-dye laser through a flexible bronchoscope can render it an outpatient procedure.22 The pulsed-dye laser has attracted much interest in recent years. The advantages of the pulse-dyed laser are that it can be fibre-delivered and causes minimal vocal fold fibrosis and consequently minimal voice damage. However, it also distinguishes itself from the other forms of laser treatment in its mode of action. Rather than direct vaporization of tissue, it has been proposed that the 585 nm pulsed-dye laser is a vascular laser which causes photoangiolysis of sublesional microcirculation, denaturation of epithelial basement membrane linking



proteins, and cellular destruction.22 Therefore, it is less effective against large exophytic lesions and can be used as an adjunct to surgery either before or after the laser application.23 The application of the pulsed-dye laser in the paediatric population is still under evaluation, but the results in adults with recurrent respiratory papillomatosis are very encouraging. [**] PHOTODYNAMIC THERAPY



Photodynamic therapy (PDT) relies upon the observation that rapidly proliferating tissue selectively takes up a number of photosensitizing agents when administered intravenously, and that these agents release tumoricidal oxygen derivatives when activated by laser light of the appropriate wavelength. A non-blinded randomized prospective trial using dihaematoporphyrinether (DHE) at two different doses in combination with 50 J of 630 nm argon laser light was compared with laser treatment alone. The patients remain photosensitive for six to nine months and may experience skin erythema, blistering and ocular discomfort. Patients on the higher dose of DHE (4.25 mg/kg body weight) were reported to show a significantly larger decrease in papilloma growth rate but, despite that, only approximately half of the 48 patients receiving DHE showed a response, and no response was seen in patients previously treated with PDT.24 A further randomized trial by the same group using meso-tetra (hydroxyphenyl) chlorin (mTHPC) as a photosensitizer showed reduction of severity of laryngeal papillomas in the mTHPC group, but this was not maintained and there was no effect on tracheal disease.25 It should be noted that this trial only recruited 23 patients of whom only 15 patients completed the study. In both trials, there was a combination of juvenile-onset and adult-onset patients. [****]



Adjuvant therapy Adjuvant medical therapies can be broadly divided into antiviral therapies and drugs with antiproliferative or



1178 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY immunomodulatory properties. Vaccines and gene therapy remain in the early experimental stages.26, 27, 28 The decision to implement adjuvant therapy must depend on a careful consideration of the benefits against potential adverse effects of the therapy.29 Adjuvant therapies which have been described in isolated case reports have been included for reference because while their use in ‘routine’ cases may not be justified, one may wish to consider their use when other better-tested avenues are exhausted.



INTERFERON-a



Interferons are naturally produced by human leukocytes, although as a pharmaceutical, interferon is now produced via recombinant DNA technology. Interferon-a can claim to have antiviral, antiproliferative and immunomodulatory properties. Interferons exert an indirect antiviral action by interfering with normal host cell translation mechanisms and by inducing synthesis of intracellular enzymes that act to control viral growth. By depleting essential metabolites in papilloma cells, interferon-a increases the length of their multiplication cycle, thereby slowing target cell growth. Interferon-a also facilitates recognition of papilloma cells by circulating leukocytes by enhancing expression of cell surface antigens. It is administered systemically by subcutaneous injection at a dose of 2–5 MU/m2 of body surface area.30 It is also the most well-studied of the adjuvant therapies for JORRP. A large randomized trial of 123 patients showed significant reduction in papilloma growth rate within the interferon arm. However, this was only significant for the first six months and the difference was not statistically significant during the second six months.30 In another randomized crossover trial of 66 children, there was significant reduction in disease bulk in the interferon arms of the trial.31 This was extended to longer follow-up, at which data on 60 children were still available, showing complete remission in 22 children, partial remission in 25 patients and no response in 13.32 [****] The main problem preventing more widespread use of interferon-a is that there are many serious, idiosyncratic and unpredictable side effects including pancytopenia, hepatorenal failure and cardiac dysfunction. There is also a rebound phenomenon associated with withdrawal of the drug therapy. There is anecdotal evidence of the use of intralesional interferon in combination with laser debulking of JORRP, but this has not found widespread use.33



cytomegalovirus in human immunodeficiency virus (HIV)-infected patients by intravenous injection. This mode of administration and high doses is associated with neutropenia and nephrotoxicity. Intralesional injection of cidofovir into JORRP is not associated with similar side effects. A canine model has shown that local irreversible soft tissue damage could be avoided in twice weekly cidofovir injections if the dose is limited to below 40 mg/mL.34, 35 Based on their animal work, Chhetri and Shapiro36 have proposed a schedule for treatment of JORRP based on intralesional injections of cidofovir at a concentration of 1 mg/mL. Injections were given at two-weekly intervals for four treatments and then the interval between treatments extended by one week after each and every subsequent treatment. Concomitant laser surgery was reserved for bulky lesions. Five patients were treated with this schedule with a mean follow-up time of 66 weeks. The mean papilloma stage decreased from 9.2 at initial presentation to 3.4 within two weeks of the first injection, and continued to decrease for the remainder of the follow-up period. After nine weeks of treatment, no patients required further laser surgery. Other studies have used cidofovir at the higher dose of 5 mg/mL with varying schedules of administration. Pransky et al.’s study37 with a long follow-up (51.6 months) of 11 patients who had all previously required surgical debulking at two-to-six-week intervals showed a reduction in papilloma score in five, complete disease remission in five and one patient who continued to require treatment. In a more recently reported study with a mean follow-up of 30 months, 11 children were treated with intralesional cidofovir of whom five required no further treatment, four had an initial remission but relapsed and two had no apparent response. The authors counsel caution that the potential long-term carcinogenic effects of cidofovir are unknown and a ‘response’ may be related to the natural history of the disease.38



RIBAVIRIN



Ribavirin is a synthetic nucleoside which has activity against a broad spectrum of viruses, but which is principally used as an aerosol in the treatment of respiratory syncytial virus pneumonia and systemically in the treatment of hepatitis C. There have been reports of its use both as an aerosol and systemically.39, 40, 41 However, these remain anecdotal reports and ribavirin is not widely used as an adjuvant treatment of JORRP. [*]



CIDOFOVIR



Cidofovir is an acyclic nucleoside phosphonate which is active against a broad spectrum of DNA viruses including cytomegalovirus, Epstein–Barr virus and HPV. Its mechanism of action is by inhibition of viral DNA polymerases essential for viral replication. The principal application of this drug has been in the treatment of



ACYCLOVIR



The evidence on the efficacy of acyclovir is weak and conflicting.42, 43 Although HPV is a DNA virus and acyclovir has a medium spectrum of antiviral activity, it does not directly inhibit HPV. Acyclovir is a nucleoside



Chapter 91 Juvenile-onset recurrent respiratory papillomatosis



analogue, which inhibits thymidine kinase, which is present in herpes simplex viruses (HSV) but not HPV. Adult patients, but not paediatric patients, with RRP have been shown to have molecular evidence of coinfection with other viruses, particularly HSV which may have a potentiating effect on HPV. It has been suggested that the mechanism of action of acyclovir is to eradicate HSV, thus removing this synergism. Side effects are rare and include nausea, vomiting, diarrhoea, fatigue and headache. [*]



] 1179



Indole-3-carbinol is a substance derived from cruciferous vegetables (e.g. cabbage, broccoli, Brussels sprouts), which has been shown to alter growth patterns of JORRP cell cultures in vitro. It affects oestrogen metabolism, shifting production to antiproliferative oestrogen. A prospective observational study with a mixed adult and paediatric population who received indole-3-carbinol as an adjunctive treatment to surgical removal showed partial or total responses in 21 of 33 patients. Within the paediatric subgroup, four out of nine showed partial or total response with no evident side effects.44 [**]



month intervals during periods of disease activity. Severe disease may necessitate weekly surgical intervention to prevent airway obstruction from rapidly growing papillomas. The median number of debulking procedures required in a patient is reported to be 7–13.47, 48 Not surprisingly, the repeated debulking of laryngeal papillomas results in chronic voice changes. Objective assessment of voice by GRBAS scale and Visi Pitch II 3000 acoustic analysis has shown significant difference between JORRP patients in remission and normal controls, but a Voice-Related Quality of Life Questionnaire (V-RQoL) showed that the dysphonia does not have any impact on quality of life.49 There is a tendency for remission in the early teenage years and this has been attributed to hormonal changes occurring with the onset of puberty. Interestingly, women with adult-onset RRP commonly experience severe exacerbations of their disease during the hormonal fluctuations of pregnancy. What is clear, however, is that remission is not related to the clearance of HPV from the mucosa as viral DNA is detected in previously affected mucosa in patients in remission as well as in normal mucosa of patients with active disease.50



CIMETIDINE



TRACHEOBRONCHIAL DISEASE



Cimetidine – a histamine receptor type 2 (H2) antagonist – has been reported as a useful treatment for cutaneous warts. It has also been successfully used in treatment of an 11-year-old boy who had an eight-year history of diffuse conjunctival papillomas. The mechanism for this is attributed to immunomodulatory side effects of cimetidine at high doses. There is a single case report of very advanced JORRP with tracheo-bronchial-pulmonary involvement being treated successfully with adjuvant cimetidine at a dose of 40 mg/kg for four months with remarkable improvement.45 [*]



Extralaryngeal spread of JORRP occurs in approximately one-third of patients with tracheal involvement in approximately one-quarter of patients. Tracheal involvement may appear as cobblestoning of the mucosa coupled with the presence of papillomas. Factors predisposing to tracheal spread include the presence of subglottic papillomas, presence of a tracheostomy and a long duration of disease.51 More distal bronchopulmonary involvement is reported in 4–11 percent of children with longstanding disease and may result in obstructive pneumonias.30 Patients may develop cavitary pulmonary lesions leading to fever, sepsis and pulmonary atelectasis. Radiographically, these lesions may appear as solid or cystic pulmonary masses (Figure 91.3).52 A high index of suspicion must be maintained for malignant degeneration of bronchopulmonary lesions.



INDOLE-3-CARBINOL



NATURAL HISTORY The natural history of RRP is extremely variable. This makes it difficult for the surgeon to be able to reliably counsel the anxious parents on how their child’s pathology will behave or at what age they might expect remission. Pathologically, severe disease is strongly associated with HPV-11 infection and thus patients should have viral typing as part of their initial pathology work-up. Poor prognostic signs include onset of disease before the age of three years, and birth by Caesarean section. In the United States, Medicaid insurance – often seen as a proxy measure of low socioeconomic status – is also associated with severe disease.46 However, there are no other studies showing correlation between socioeconomic status and disease severity. Most affected children require debulking of papillomas at two-to-three



TRACHEOSTOMY The indications for tracheostomy in patients with JORRP continue to divide clinicians. The source of this debate is that tracheostomy essentially constitutes an iatrogenic squamociliary junction and may present an additional area of predilection for papillomas.53 However, it is not universal that all patients with a tracheostomy develop stomal papillomas. A tracheostomy in patients with severe disease may be life-saving and may also facilitate a longer interval between debulking procedures to restore some normality to the child’s life.



1180 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY malignant transformation.54 HPV expression may be lost in malignant tranformation but this may not be a sufficiently robust clinicopathological predictor.



KEY POINTS



Figure 91.3 Chest radiograph showing extensive pulmonary involvement with papilloma.



 Juvenile respiratory papillomatosis can affect any part of the respiratory tract.  While it is uncommon, its management can constitute a major surgical burden to the otolaryngologist.  The mainstay of treatment is by surgical debulking.  The role of adjuvant therapy is uncertain but supported by level 3 evidence.  In the majority of cases, the disease goes into spontaneous remission.  Some children develop uncontrolled disease, which may be fatal.



MALIGNANT DISEASE Malignant degeneration of papillomas is a rare but devastating sequel. It is universally fatal.52 Irradiation of papillomas dramatically increases the risk of malignant transformation but as this is no longer used as a treatment modality, it is only of historical relevance. Most instances of malignant transformation have been reported in adult patients and have been associated with other risk factors including tobacco use and longstanding disease. Malignant transformation appears to be more likely with HPV 16, an unusual cause of JORRP, but HPV 6 and 11 have been shown to oncogenically transform cell culture lines in vitro. In adults, malignant degeneration usually involves the larynx, unlike children where cancer usually develops in the bronchopulmonary tree. Approximately 20 paediatric cases of malignant degeneration have been reported, all of which have been fatal. The currently proposed mechanism of malignant transformation involves oncoproteins E6 and E7. HPV types 6 and 11 produce transforming oncoproteins E6 and E7 that have been implicated in growth dysregulation through their ability to inactivate the tumour suppressor proteins p53 and the retinoblastoma tumour-suppressor gene product (pRb). The inactivation of the tumour suppressor genes results in a loss of control over proliferation and cell division and contributes to the development of the malignant phenotype. It is also becoming clear that the E6 and E7 proteins function to promote tumorigenesis through direct interactions with cell-cycle regulatory proteins. Unfortunately, apart from heightened vigilance, there is little to aid the clinician in predicting the rare cases of malignant transformation. There is no evidence of a papilloma-carcinoma sequence while p53 over-expression is variable and not a marker of



Best clinical practice [ Current evidence does not warrant Caesarean section as a prophylaxis against RRP. [Grade B]



[ Ideally, the clinical diagnosis should be established with awake fibreoptic nasolaryngoscopy prior to general anaesthesia. [Grade D] [ Patients should have viral typing as part of their initial pathology work-up. [Grade C/D] [ The powered microdebrider is quickly becoming the ‘gold standard’ for papilloma removal in the larynx. [Grade C] [ Spread to the tracheobronchial tree is associated with a greatly worsened prognosis. [Grade C]



Deficiencies in current knowledge and areas for future research



$



$ $



JORRP has a great impact on the child’s life during active disease and constitutes a big surgical load for the surgeon. While the majority of patients enjoy spontaneous remission, a small number continue to develop distal disease which may be fatal. Improvement in surgical technologies have reduced operative time and improved operative morbidity. The powered microdebrider has been a significant advance and better and more precise microdebriders seem likely to be developed. There is consistent level 3 evidence that cidofovir extends the treatment interval and promotes



Chapter 91 Juvenile-onset recurrent respiratory papillomatosis



$ $



$



remission of disease. Consolidation of this evidence will require multicentre collaboration in order to standardize protocols and increase patient recruitment. Such a multicentre initiative has already been proposed to seek critical genes in the pathogenesis of RRP.55 The future of the management of RRP lies in a better understanding of its pathogenesis. A vaccine against HPV has become available just as we go to press. Its main use will be in the prevention of cervical cancer in women but at least one type of vaccine (‘quadrivalent vaccine’) is active against HPV types 6, 11, 16 and 18 and may have a role in the prevention of RRP.56 Trials are likely to get underway in the next few years with the added possibility of a therapeutic role for the vaccine in children with established RRP. Despite extensive and sophisticated molecular immunology research, the precise relationship between the human papilloma virus, host immunity and the development of papillomas eludes us. The elucidation of this relationship holds the key to conquering this disease.



8.



9.



10.



 11.



12.



13.



REFERENCES 1. Mounts P, Shah KV, Kashima H. Viral etiology of juvenileand adult-onset squamous papilloma of the larynx. Proceedings of the National Academy of Sciences of the United States of America. 1982; 79: 5425–9. 2. Kashima HK, Shaf F, Lyles A, Glackin R, Muhammad N, Turner L et al. A comparison of risk factors in juvenileonset and adult-onset recurrent respiratory papillomatosis. Laryngoscope. 1992; 102: 9–13. 3. Silverberg MJ, Thorsen P, Lindeberg H, Ahdieh-Grant L, Shah KV. Clinical course of recurrent respiratory papillomatosis in Danish children. Archives of Otolaryngology – Head and Neck Surgery. 2004; 130: 711–6. 4. Bonagura VR, Vambutas A, DeVoti JA, Rosenthal DW, Steinberg BM, Abramson AL et al. HLA alleles, IFN-gamma responses to HPV-11 E6, and disease severity in patients with recurrent respiratory papillomatosis. Human Immunology. 2004; 65: 773–82. 5. Gelder CM, Williams OM, Hart KW, Wall S, Williams G, Ingrams D et al. HLA class II polymorphisms and susceptibility to recurrent respiratory papillomatosis. Journal of Virology. 2003; 77: 1927–39. 6. Saleh EM. Complications of treatment of recurrent laryngeal papillomatosis with the carbon dioxide laser in children. Journal of Laryngology and Otology. 1992; 106: 715–8. 7. Stern Y, McCall JE, Willging JP, Mueller KL, Cotton RT. Spontaneous respiration anesthesia for respiratory



14.



15.



 16.



17.



18.



19.



20.



21.



] 1181



papillomatosis. Annals of Otology, Rhinology and Laryngology. 2000; 109: 72–6. Theroux MC, Grodecki V, Reilly JS, Kettrick RG. Juvenile laryngeal papillomatosis: scary anaesthetic. Paediatric Anaesthesia. 1998; 8: 357–61. Holland BW, Koufman JA, Postma GN, McGuirt Jr. WF. Laryngopharyngeal reflux and laryngeal web formation in patients with pediatric recurrent respiratory papillomas. Laryngoscope. 2002; 112: 1926–9. Derkay CS, Malis DJ, Zalzal G, Wiatrak BJ, Kashima HK, Coltrera MD. A staging system for assessing severity of disease and response to therapy in recurrent respiratory papillomatosis. Laryngoscope. 1998; 108: 935–7. Derkay CS, Hester RP, Burke B, Carron J, Lawson L. Analysis of a staging system for prediction of surgical interval in recurrent respiratory papillomatosis. International Journal of Paediatric Otorhinolaryngology. 2004; 68: 1493–8. Update on staging, demonstrating its importance and suggesting a system which has high validity and reliability. Tasca RA, McCormick M, Clarke RW. British Association of Paediatric Otorhinolaryngology members experience with recurrent respiratory papillomatosis. International Journal of Pediatric Otorhinolaryngology. 2006; 70: 1183–7. El-Bitar MA, Zalzal GH. Powered instrumentation in the treatment of recurrent respiratory papillomatosis: an alternative to the carbon dioxide laser. Archives of Otolaryngology – Head and Neck Surgery. 2002; 128: 425–8. Patel N, Rowe M, Tunkel D. Treatment of recurrent respiratory papillomatosis in children with the microdebrider. Annals of Otology, Rhinology and Laryngology. 2003; 112: 7–10. Patel RS, MacKenzie K. Powered laryngeal shavers and laryngeal papillomatosis: a preliminary report. Clinical Otolaryngology and Allied Sciences. 2000; 25: 358–60. Pasquale K, Wiatrak B, Woolley A, Lewis L. Microdebrider versus CO2 laser removal of recurrent respiratory papillomas: a prospective analysis. Laryngoscope. 2003; 113: 139–43. A comparative study looking at the relative merits of the CO2 laser and the microdebrider. Ossoff RH, Werkhaven JA, Dere H. Soft-tissue complications of laser surgery for recurrent respiratory papillomatosis. Laryngoscope. 1991; 101: 1162–6. Crockett DM, McCabe BF, Shive CJ. Complications of laser surgery for recurrent respiratory papillomatosis. Annals of Otology, Rhinology and Laryngology. 1987; 96: 639–44. Sakoh T, Fukuda H, Sasaki S, Sakaquchi R, Shiotani A, Kawaida M et al. Laryngomicrosurgery with carbon dioxide laser for laryngeal papillomatosis: application of a twostage operation. Auris, Nasus, Larynx. 1993; 20: 223–9. Hallmo P, Naess O. Laryngeal papillomatosis with human papillomavirus DNA contracted by a laser surgeon. European Archives of Otorhinolaryngology. 1991; 248: 425–7. Janda P, Leunig A, Sroka R, Betz CS, Rasp G. Preliminary report of endolaryngeal and endotracheal laser surgery of



1182 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



22.



23.



24.



25.



26.



27.



28.



29.



30.



31.



32.



33.



juvenile-onset recurrent respiratory papillomatosis by Nd:YAG laser and a new fiber guidance instrument. Otolaryngology – Head and Neck Surgery. 2004; 131: 44–9. Zeitels SM, Franco Jr. RA, Dailey SH, Burns JA, Hillman RE, Anderson RR. Office-based treatment of glottal dysplasia and papillomatosis with the 585-nm pulsed dye laser and local anesthesia. Annals of Otology, Rhinology and Laryngology. 2004; 113: 265–76. Franco Jr. RA, Zeitels SM, Farinelli WA, Anderson RR. 585nm pulsed dye laser treatment of glottal papillomatosis. Annals of Otology, Rhinology and Laryngology. 2002; 111: 486–92. Shikowitz MJ, Abramson AL, Freeman K, Steinberg BM, Nouri M. Efficacy of DHE photodynamic therapy for respiratory papillomatosis: immediate and long-term results. Laryngoscope. 1998; 108: 962–7. Shikowitz MJ, Abramson AL, Steinberg BM, DeVoti J, Bonaqura VR, Mullooly V et al. Clinical trial of photodynamic therapy with meso-tetra (hydroxyphenyl) chlorin for respiratory papillomatosis. Archives of Otolaryngology – Head and Neck Surgery. 2005; 131: 99–105. Moffitt Jr. OP. Treatment of laryngeal papillomatosis with bovine wart vaccine: report of cases. Laryngoscope. 1959; 69: 1421–8. Pashley NR. Can mumps vaccine induce remission in recurrent respiratory papilloma? Archives of Otolaryngology – Head and Neck Surgery. 2002; 128: 783–6. Sethi N, Palefsky J. Treatment of human papillomavirus (HPV) type 16-infected cells using herpes simplex virus type 1 thymidine kinase-mediated gene therapy transcriptionally regulated by the HPV E2 protein. Human Gene Therapy. 2003; 14: 45–57. Shykhon M, Kuo M, Pearman K. Recurrent respiratory papillomatosis. Clinical Otolaryngology and Allied Sciences. 2002; 27: 237–43. Healy GB, Gelber RD, Trowbridge AL, Grundfast KM, Ruben RJ, Price KN. Treatment of recurrent respiratory papillomatosis with human leukocyte interferon. Results of a multicenter randomized clinical trial. New England Journal of Medicine. 1988; 319: 401–7. Leventhal BG, Kashima H, Levine AS, Levy HB. Treatment of recurrent laryngeal papillomatosis with an artificial interferon inducer (poly ICLC). Journal of Pediatrics. 1981; 99: 614–6. Leventhal BG, Kashima HK, Mounts P, Thurmond L, Chapman S, Buckley S et al. Long-term response of recurrent respiratory papillomatosis to treatment with lymphoblastoid interferon alfa-N1. Papilloma Study Group. New England Journal of Medicine. 1991; 325: 613–7. Herberhold C, Walther EK. Combined laser surgery and adjuvant intralesional interferon injection in patients with laryngotracheal papillomatosis. Advances in Oto-RhinoLaryngology. 1995; 49: 166–9.







34. Chhetri DK, Jahan-Parwar B, Hart SD, Bhuta SM, Berke GS, Shapiro NL. Local and systemic effects of intralaryngeal injection of cidofovir in a canine model. Laryngoscope. 2003; 113: 1922–6. 35. Jahan-Parwar B, Chhetri DK, Hart S, Bhuta S, Berke GS. Development of a canine model for recurrent respiratory papillomatosis. Annals of Otology, Rhinology and Laryngology. 2003; 112: 1011–3. 36. Chhetri DK, Shapiro NL. A scheduled protocol for the treatment of juvenile recurrent respiratory papillomatosis with intralesional cidofovir. Archives of Otolaryngology – Head and Neck Surgery. 2003; 129: 1081–5. 37. Pransky SM, Albright JT, Magit AE. Long-term follow-up of pediatric recurrent respiratory papillomatosis managed with intralesional cidofovir. Laryngoscope. 2003; 113: 1583–7. 38. Chung BJ, Akst LM, Koltai PJ. 3.5-Year follow-up of intralesional cidofovir protocol for pediatric recurrent respiratory papillomatosis. International Journal of Pediatric Otorhinolaryngology. 2006; 70: 1911–7. 39. Kimberlin DW. Current status of antiviral therapy for juvenile-onset recurrent respiratory papillomatosis. Antiviral Research. 2004; 63: 141–51. 40. Morrison GA, Kotecha B, Evans JN. Ribavirin treatment for juvenile respiratory papillomatosis. Journal of Laryngology and Otology. 1993; 107: 423–6. 41. Balauff A, Sira J, Pearman K, McKiernan P, Buckels J, Kelly D. Successful ribavirin therapy for life-threatening laryngeal papillomatosis post liver transplantation. Pediatric Transplantation. 2001; 5: 142–4. 42. Kiroglu M, Cetik F, Soylu L, Abedi T, Aydogan B, Akcali C et al. Acyclovir in the treatment of recurrent respiratory papillomatosis: a preliminary report. American Journal of Otolaryngology. 1994; 15: 212–4. 43. Morrison GA, Evans JN. Juvenile respiratory papillomatosis: acyclovir reassessed. International Journal of Pediatric Otorhinolaryngology. 1993; 26: 193–7. 44. Rosen CA, Bryson PC. Indole-3-carbinol for recurrent respiratory papillomatosis: long-term results. Journal of Voice. 2004; 18: 248–53. 45. Harcourt JP, Worley G, Leighton SE. Cimetidine treatment for recurrent respiratory papillomatosis. International Journal of Pediatric Otorhinolaryngology. 1999; 51: 109–13. 46. Wiatrak BJ, Wiatrak DW, Broker TR, Lewis T. Recurrent respiratory papillomatosis: a longitudinal study comparing severity associated with human papilloma viral types 6 and 11 and other risk factors in a large pediatric population. Laryngoscope. 2004; 114: 1–23. A longitudinal study looking at prognostic factors in RRP. 47. Lindeberg H. Laryngeal papillomas: histomorphometric evaluation of multiple and solitary lesions. Clinical Otolaryngology and Allied Sciences. 1991; 16: 257–60. 48. Morgan AH, Zitsch RP. Recurrent respiratory papillomatosis in children: a retrospective study of management and complications. Ear, Nose and Throat Journal. 1986; 65: 19–28.



Chapter 91 Juvenile-onset recurrent respiratory papillomatosis 49. Lindman JP, Gibbons MD, Morlier R, Wiatrak BJ. Voice quality of prepubescent children with quiescent recurrent respiratory papillomatosis. International Journal of Pediatric Otorhinolaryngology. 2004; 68: 529–36. 50. Steinberg BM, Topp WC, Schneider PS, Abramson AL. Laryngeal papillomavirus infection during clinical remission. New England Journal of Medicine. 1983; 308: 1261–4. 51. Weiss MD, Kashima HK. Tracheal involvement in laryngeal papillomatosis. Laryngoscope. 1983; 93: 45–8. 52. Bauman NM, Smith RJ. Recurrent respiratory papillomatosis. Pediatric Clinics of North America. 1996; 43: 1385–401. 53. Shapiro AM, Rimell FL, Shoemaker D, Pou A, Stool SE. Tracheotomy in children with juvenile-onset recurrent respiratory papillomatosis: the Children’s Hospital of







] 1183



Pittsburgh experience. Annals of Otology, Rhinology and Laryngology. 1996; 105: 1–5. 54. Go C, Schwartz MR, Donovan DT. Molecular transformation of recurrent respiratory papillomatosis: viral typing and p53 overexpression. Annals of Otology, Rhinology and Laryngology. 2003; 112: 298–302. 55. Buchinsky FJ, Derkay CS, Leal SM, Donfack J, Ehrlich GD, Post JC. Multicenter initiative seeking critical genes in respiratory papillomatosis. Laryngoscope. 2004; 114: 349–57. A review of current thinking and an outline of a proposed study to further unravel the aetiology and pathogenesis of this complex disease. 56. Freed GL, Derkay CS. Prevention of recurrent respiratory papillomatosis: role of HPV vaccination. International Journal of Pediatric Otorhinolaryngology. 2006; 70: 1799–803.



92 Foreign bodies in the ear and the aerodigestive tract in children A SIMON CARNEY, NIMESH PATEL AND RAY CLARKE



Introduction Foreign bodies in the ear canal Nasal foreign bodies Ingested foreign bodies Key points



1184 1184 1186 1186 1191



Best clinical practice Deficiencies in current knowledge and areas for future research References



1191 1191 1191



SEARCH STRATEGY The data and opinions presented in this chapter are supported by a Medline search using the key words child, foreign body, ear, nose, throat, larynx, trachea, bronchus, pharynx and oesophagus.



INTRODUCTION Nature determined that we possess seven orifices. The otolaryngologist deals with five. Children are naturally curious about their surroundings and about these orifices! They are inclined to place toys, foodstuff and household articles in the ear, nose or the oral cavity. Sometimes the culprit is a sibling or a playground or nursery chum. Foreign bodies lodged within the ear, nose, larynx, trachea, pharynx or oesophagus may present as a minor irritation or a life-threatening problem.



FOREIGN BODIES IN THE EAR CANAL An aural foreign body is an object not derived from the individual’s own external ear, lying within the external auditory canal. Although the pinna is a part of the external ear, diseases of the pinna are covered in Chapter 236, Conditions of the pinna and external auditory canal; Chapter 236g, Acquired atresia of the external ear; Chapter



236i, Perichondritis of the external ear; and Chapter 236m, Haematoma auris; and foreign bodies of the pinna (e.g. ear-rings and studs) are not discussed in this chapter. Foreign bodies within the external auditory canal can be classified as inanimate or animate (e.g. live insects). Inanimate objects can be inert or corrosive/irritant, organic or nonorganic and hydrophobic or hydrophilic.



Presentation Although it is usually children that present with aural foreign bodies, self-instrumentation of the ear canal not infrequently brings the embarrassed adult to the otolaryngologist. Inert foreign bodies may be completely asymptomatic and go unnoticed for many weeks, months or even years. Many patients complain of irritation, deafness, otorrhoea or tinnitus. Gradual wax impaction medial to a small foreign body may lead to a later presentation due to the disruption of the physiological keratinocyte migration.



Chapter 92 Foreign bodies in the ear and the aerodigestive tract in children



The isthmus is the narrowest part of the external auditory canal and objects frequently impact at this point. Objects lateral to the isthmus are usually more readily removed. In a review of 191 aural foreign bodies, 74 percent were in children aged seven or under, with beans, pebbles, beads (Figure 92.1), insects and small toys being the most common objects.1, 2 Although parents and primary care health personnel are often worried about the possibility of residual perforation of the eardrum, this is uncommon unless there is unskilled instrumentation in attempted removal. Button batteries are arguably the most dangerous foreign body in the ear, producing tissue necrosis as the alkaline fluid seeps out of them. The tissue necrosis then produces granuloma formation. Symptoms which can mimic malignant otitis externa can also develop.3 Insects in the ear can produce very distressing symptoms and the removal of live, moving insects can exacerbate oedema and trauma (especially where multiple bites and/or stings occur).4 [**]



Management In a cooperative awake patient, the majority of foreign bodies can be removed with ease. General anaesthesia may be required for removal of up to 30 percent of objects, especially in the paediatric population.1 Removal of foreign bodies in an uncooperative child can result in trauma, such as lacerations or tympanic membrane perforation which can occur in between 10–47 percent of cases.2, 5 If an object is not removed easily on the first



Figure 92.1



Foreign body (bead) in the ear canal.



] 1185



attempt, referral to an otolaryngologist is essential to minimize further trauma to the child.5, 6 In a large review of 603 patients, the presence of a spherical foreign body, an object near the tympanic membrane and delay of 24 hours following insertion were identified as high risk factors for poor outcomes.7 In a review of the management of aural foreign bodies in an emergency department, when an object was graspable (e.g. a piece of plastic), emergency staff could remove it in 64 percent of cases with a complication rate of 14 percent, whereas if the object was non-graspable (e.g. a bead), only 45 percent could be removed (with 65 percent being referred to an otolaryngologist) with a 70 percent complication rate.8 Where objects are not removed rapidly otitis externa can develop and, if this is not treated, complications can rarely include mastoiditis and deep neck space infections.9 Techniques for removal include irrigation, suction, instrumentation or a combination of the three.6 The use of irrigation is naturally contraindicated for extracting hydrophilic items such as peas, beans, other vegetable matter or tissue paper. The use of the otomicroscope has been shown to produce a higher success rate for foreign body removal.7 The technique of extracting a spherical or ovoid item from the canal is to slide a wax hook over the top of it and then rotate the hook to engage the object and pull it out of the canal. Do not attempt removal by forceps as this results in it slipping further away. ‘Superglue’TM (cryanoacrylate) impregnated into cotton-buds can be used to aid aural foreign body removal,10 although it is significantly slower than using hooks and other standard instruments.11 Superglue probes are particularly useful for the management of smooth foreign bodies such as beads or beans impacted at the isthmus.12 Unwanted superglue in the external auditory canal is a particularly troublesome problem. Although some individuals will require surgical removal via an end-aural incision,13 there are now case reports of the use of acetone14 and warm 3 percent hydrogen peroxide15 which suggest these conservative techniques may be worth trying before subjecting a child to general anaesthesia. Repeated unsuccessful attempts at foreign body removal by an inexperienced practitioner will distress an anxious child and may compromise the options open to an otolaryngologist. If there is any doubt whatsoever about the potential success of an attempted extraction, then referral to a specialist is indicated. The otolaryngologist should have a low threshold for arranging a semi-urgent admission for general anaesthetic removal if conservative methods fail. Urgent removal can only be justified for battery removal, or where severe oedema and symptoms exist. Rarely, an end-aural incision may be necessary to remove the foreign body.16 Studies comparing management options for aural foreign bodies are few in number. For alive, animated foreign bodies, an in vitro study involving 17 test preparations showed that 95 percent



1186 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY ethanol killed all test insects in the fastest time (median 27 seconds). Oils performed poorly (51 seconds) and water and saline were particularly slow (180 seconds).4 [***]



NASAL FOREIGN BODIES Presentation and management Nasal foreign bodies are most common between the ages of two and five.17 They may be inert, hygrophilic or corrosive. Soft objects such as sponge fragments or tissue paper are the commonest to be found in the nose, in contrast to the hard objects inserted into the ears.2 As is the case in the ear, small button batteries can produce devastating tissue damage and necrosis including septal perforation in a short time.18 A nasal foreign body may remain in situ for weeks and only present with a unilateral nasal discharge, often with a pronounced vestibulitis (Figure 92.2). Rigid endoscopes provide excellent visualization and can potentially minimize mucosal trauma. Although it has been claimed that a 4 mm rigid nasendoscope can be passed with ease in children as young as two years of age without decongestion,19 with the increased availability of smaller 2.7 mm endoscopes, it is the author’s practice to use decongestion and as small an endoscope as possible in the paediatric population. The use of nebulized adrenaline provides excellent nasal decongestion and can greatly facilitate removal of foreign bodies, either by nose blowing or using instrumentation.20 In general, the clinician has to make a clinical judgement as to what is going to be the best method for removal of a foreign body,17 bearing in mind that a child is unlikely to tolerate repeated manipulation and the doctor will only have one attempt at using a method that is going to cause any pain whatsoever. Magnets may be of use21 and nasal washing has been proposed as a method of removing foreign bodies, but is not widely practised.22 Superglue can also be effective.23 The use of oral positivepressure techniques has now been shown to be an effective way of removing anterior nasal foreign bodies.



An oral Ambubag can be used24 (three patients, 100 percent success) but the ‘parent’s kiss’, where the carer blows into the open mouth of the child whilst occluding the contralateral nostril, is probably less traumatic for the child. Reported success rates vary from 79–100 percent.25, 26 The use of a ‘Fogarty’ embolectomy balloon catheter is another excellent way to easily remove foreign bodies from the nasal cavity and in one series of 25 patients, successfully removed 23 foreign bodies (92 percent) without the need for a general anaesthetic.27 The possibility of inhalation into the tracheobronchial tree28 needs to be borne in mind when managing a child with a nasal foreign body. [**] This is remarkably uncommon and is probably only a significant risk in the neurologically compromised child who has a poor gag reflex. If a nasal foreign body slips back into the nasopharynx it will usually be swallowed or expectorated. No complication of barotrauma to either the ears or lower airway has been reported in the literature search employed for this chapter.



Rhinolith A rhinolith is a partially or totally calcified mass of tissue in the nasal cavity which may form around a foreign body nidus or can develop de novo.29, 30 They usually have a laminated structure, suggestive of a pathophysiological mechanism that involves layers of mucin aggregating around the foreign body. Each mucin layer subsequently becomes calcified, perhaps aided by the presence of turbulent air currents. They are classified as endogenous when they form around normal body material, such as blood clots, misplaced tooth remnants or bony sequestra. Exogenous rhinoliths form around foreign bodies inserted into the nose – usually of nonhuman material.29, 30 They can cause considerable symptoms, including nasal obstruction, purulent nasal discharge, rhinosinusitis and septal perforation.29, 30 They frequently require a general anaesthetic and transnasal removal, with or without fragmentation of the rhinolith.31



INGESTED FOREIGN BODIES



Figure 92.2 body.



Nasal vestibulitis due to a neglected foreign



These may be swallowed and impact in the pharynx or oesophagus, or inhaled and obstruct the larynx or tracheobronchial tree. Either may prove fatal. Ingested foreign bodies in children represent a major global public health problem.32, 33 Coins, toys and food particles are the principal dangers, although regional and cultural factors in part dictate the frequency with which different objects occur.34 Fish bones in the pharynx are commonly encountered in the Far East and in Greece. In Turkey watermelon seeds are the most frequently aspirated objects.35 Coins are a universal danger particularly for pharyngo-oesophageal impaction



Chapter 92 Foreign bodies in the ear and the aerodigestive tract in children



and nuts for tracheobronchial aspiration.32, 36 Nonspherical objects equal to or less than 1.5 inches (38.10 mm), and particularly spherical objects equal to or less than 1.75 inches (44.50 mm) in diameter, are especially dangerous. Consumer product legislation should take account of these data.33 The main risks are to children under three years. In this age group the second molars have not yet developed, the child’s grinding and swallowing mechanisms are poor and glottic closure is immature.37



Pharynx and oesophagus When considering foreign bodies in the pharynx and oesophagus, whether the object is potentially penetrating (sharp or corrosive) or nonpenetrating is probably the most pertinent consideration. The vast majority of swallowed objects pass through the digestive tract and are excreted, often undetected.38 A small (and statistically undeterminable) proportion will impact in the pharynx or oesophagus. There is a clear history of ingestion in 96 percent and the median age was three years in a retrospective ten-year review of 327 patients. Complications, including respiratory symptoms and abscess formation, occured in 7.6 percent of cases, the latter always associated with sharp foreign bodies and the former more common in the younger age group.38



(a)



] 1187



Management A plain radiograph will often show the position of the foreign body, provided the child is stable enough to undergo an x-ray. If there is doubt, a lateral view will help to determine if the object is in the pharynx or the airway (Figure 92.3). Bear in mind that an object in the upper oesophagus can cause airway obstruction by pressing on the adjacent trachea. Once the presumptive diagnosis of a pharyngo-oesophageal foreign body has been made, arrangements should be made for early removal as oedema and mucosal swelling will make retrieval more difficult. The use of rigid angled nasendoscopes and curved forceps designed for fish bone removal has greatly facilitated fish bone removal from the oropharynx under local anaesthetic in adults,39 but general anaesthesia is more likely to be needed in a child. Other original techniques, such as using nasal suction catheters in conjunction with nasendoscopy, have also been described.40 Small fish bones may be undetectable and there is some evidence that, even without removal, they may absorb without further sequelae.41 This fact must not lull the surgeon into a false sense of security as prompt management of all pharyngo-oesophageal foreign bodies results in better outcomes and reduced complications.34, 38 Major complications include retropharyngeal and mediastinal abscess, migration of the foreign body into deep structures, oesophageal perforation (from either the



(b)



Figure 92.3 (a) Foreign body (coin) in the oesophagus. The lateral view (b) confirms the position.



1188 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY foreign body or endoscopic procedure) and luminal stenosis.34, 42 Again, alkaline batteries are particularly dangerous because the tissue necrosis can be devastating and fatalities have been described.43 For oesophageal foreign bodies the choice between flexible and rigid endoscopy remains controversial.34 Rigid endoscopy is historically associated with a perforation rate of 0.2–1.2 percent compared to rates of 0.02–0.05 percent with flexible endoscopy;34 however, in retrospective series looking specifically at endoscopy for foreign body removal, no difference in complication rates was found.38 Rigid endoscopy gives a much better view of the hypopharynx, cricopharyngeus and the first few centimetres of the cervical oesophagus, whereas a flexible endoscope gives an excellent view in the thoracic oesophagus and oesophago-gastric junction.34 In a review of 5240 patients with ingested foreign bodies from Hong Kong, a management algorithm with flexible oesophagoscopy only used for patients (of all ages) with symptoms distal to the sternal notch resulted in 7.7 percent requiring rigid pharyngo-oesophagoscopy and 1.5 percent having flexible endoscopy (with over 90 percent of patients being managed conservatively). Complications from the foreign body or endoscopy occurred in 0.6 percent.34 The delicate structures of the oesophagus in the child mean that instrumental perforation of the pharynx or oesophagus is an ever-present danger. For the investigation and management of suspected perforation of the oesophagus the reader is referred to Chapter 156, Oesophageal diseases.



Larynx Glottic impaction of a foreign body often leads to laryngospasm and sudden complete airway obstruction and death with peak incidence at one to three years.44 Large, thin objects, such as artificial nails,45 may be more prone to lodge in the larynx and indeed in almost all cases of successful surgical removal of a foreign body, the object is small and thin.46 Larger objects are either removed by emergency clearing procedures or else death is rapid, unless a surgical airway is created within seconds.46 It has been estimated that in the 1970s, almost 600 children per year died in the USA from airway obstruction.47 However, public education and awareness of first-aid procedures for acute foreign-body ingestion into the airway has resulted in a nine-fold reduction in mortality rates. The ‘Heimlich manouevre’, which involves compression of the upper abdomen to encourage expulsion of a foreign body, was introduced in 1975.48 In infants, lying the child on its back on the adults knee and pressing firmly on the upper abdomen (Figure 92.4) is the preferred manouevre. In recent years, audit has shown that 85 percent of airways are cleared before emergency teams arrive, with 38 percent being cleared by the children themselves.49



Younger children are less likely to clear their own airway. Emergency practitioners use this fact to argue for the widespread teaching of airway-clearing manoeuvres to new parents.49 Older methods previously taught for airway-clearing, such as ‘finger-sweeping’ can in fact result in the subglottic impaction of foreign bodies and are no longer recommended.50 [**/*][Grade C/D] If the history is at all suggestive of an inhaled foreign body in a child, even if examination is normal, seek the advice of an experienced otolaryngologist.



Tracheo-bronchial tree The peak incidence of inhaled foreign bodies is between the ages of one and three years, with a male:female ratio of 2:1.49, 51 Only 12 percent will impact in the larynx with most passing through the cords into the tracheobronchial tree.52 In contrast to adults, where objects tend to lodge in the distal bronchi or right main bronchus, in children they tend to lie more centrally within the trachea (53 percent) or just distal to the carina (47 percent).53, 54 The typical history is of a choking episode while the child feeds or while (s)he is playing with a toy or small object. Vigorous coughing ensues. The parents then find to their alarm that the object has disappeared. This can be followed by a relatively symptom-free period as the object lodges in the lower airway. Partial obstruction of one of the main stem bronchi causes the characteristic wheeze over one side of the chest on auscultation and the hyperinflation of one lung evident on chest x-ray, although these classical findings are by no means universal. The hyperinflation occurs due to a ‘ball-valve’ effect where the negative intrathoracic pressure on inspiration dilates the bronchial lumen around the foreign body. Upon expiration, the positive pulmonary pressure compresses the main bronchi, occluding the airway around the offending object, preventing expulsion of the air. Although most patients present with a history of wheeze or cough, up to 20 percent may present after several days due to secondary respiratory complications.51 A high index of suspicion and early consideration of bronchoscopy are essential. If the child is well enough, a plain chest x-ray may show the characteristic changes of ‘obstructive emphysema’ as air is trapped beyond a partly occluded bronchus (Figure 92.5). The use of CT to aid diagnosis of tracheo-bronchial foreign bodies has not been a great advantage. Peanuts do not show up well (sensitivity of o35 percent), although objects such as LEGOTM can be detected easily with a sensitivity and specificity in excess of 90 percent.55 Better public awareness of the dangers of small objects and legislation to control their use in toys and household objects has meant that in western communities even highly specialized otolaryngologists in large tertiary institutions such as Johns Hopkins Hospital may only see an average of 5.9 cases per year. Residents may only



Chapter 92 Foreign bodies in the ear and the aerodigestive tract in children



] 1189



(b) (a)



(c)



(d)



Figure 92.4 Emergency manoeuvres to encourage expectoration of a laryngeal foreign body in children: (a) back blows to an infant; (b) abdominal thrusts to an infant; (c) back blows to a small child; (d) Heimlich manoeuvre in a standing child. Redrawn with permission from Mackway-Jones K (ed). 2004. Advanced paediatric life support: the practical approach, 4th edn. London: BMJ Books.



see or perform between one and eight cases during their training.56 Teaching on animal models and/or manikins is mandatory for trainees to develop the skills required to cope with this most difficult airway emergency in a particularly vulnerable patient population. If the foreign body goes untreated, mediastinal shift, pneumothorax or pneumonia are common sequelae. Smaller foreign bodies may induce granuloma formation and an aggressive search for the foreign body must be made in new cases of bronchial granuloma.57 If a foreign body is left untreated for many years, lung resection for bronchiectasis may eventually be required.58



In the acute situation, whilst emergency preparations for surgical removal are underway, high flow oxygen is essential and a helium-oxygen mixture (Helioxs) may help reduce the work of breathing.59 Even after successful removal, atelectasis, pneumonia, retained fragments, airway spasm or airway oedema may still occur and the child should be monitored closely,56 ideally in a highdependency or intensive-care facility. Once a foreign body is removed, a meticulous examination of the tracheobronchial tree must be undertaken to exclude further foreign bodies/fragments or other abnormalities.60, 61 Following rigid bronchoscopy,



1190 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



(a)



(b)



Figure 92.5 Chest x-ray showing hyperinflation of the right lung due to a foreign body in the right main stem bronchus.



some institutions use systemic steroids and nebulized vasoconstrictors to reduce airway oedema and the incidence of postoperative complications. In a series of 250 patients with tracheobronchial foreign bodies, there was a clear history of ingestion in only 38 percent of cases and over 99 percent of cases were successfully treated endoscopically.62 In 95 percent of patients in one series, a history of choking, coughing or wheeze was present. When bronchoscopy was performed, foreign bodies were identified and removed in 80 percent of cases.63 However, in a series of 235 cases, the sensitivity of choking and coughing (82 and 80 percent) was high. The sensitivity of a chest radiograph was 66 percent with a specificity of 51 percent. The sensitivity of auscultation was 80 percent with a specificity of 72 percent. The combination of history, signs and radiological abnormalities is more useful than any one separately and a high index of suspicion is essential.64 In a large series of 500 patients, the mortality rate in 1977 was 1.8 percent,65 but in a more recent large series the mortality was only 0.8 percent.35 The choice of either using a flexible or rigid endoscope remains controversial. Otolaryngologists traditionally believe rigid endoscopes to be the optimal instrument for tracheo-bronchial foreign bodies.66 The airway is controlled, there is a large spectrum of sizes and instrumentation available, excellent visualization via rigid optical endoscopes and the ability to use the scope itself as a conduit for removal of the foreign body.61 However, there are certain objects that may be more suitably removed with flexible fibreoptic instruments, or a combination of rigid and flexible techniques.67 In one series of 26 patients (including two where rigid endoscopy had failed due to the distal location of the objects), using ureteral stone baskets and forceps, flexible bronchoscopy



(c)



Figure 92.6 (a) Optical forceps with an integral telescope for foreign body retrieval. (b) The blades of the optical forceps project through a ventilating bronchoscope allowing safe retrieval under direct vision. (c) A peanut in the bronchus.



was able to remove 100 percent of objects.61 In most cases, the airway was secured via an endotracheal tube or laryngeal mask, although in the cases of failed rigid endoscopy, the rigid scope served as the conduit for the flexible instrument.61 Where standard 3.6-mm paediatric flexible bronchoscopes are used, the biopsy channel can be used for instrumentation, but if the ultrathin (2.2 mm) scope is used in smaller children, baskets and forceps need to be passed down the side of the flexible scope.61 Bleeding may occur during the removal of a foreign body, but this can usually be controlled using adrenaline-soaked balls or patties.61 [**/*] The standard paediatric flexible scope requires an endotracheal tube with a minimum internal diameter of 4.5 mm. Even then, delicate instrumentation may be impossible without periods of apnoea. As is often the case in medicine, the decision to use flexible or rigid



Chapter 92 Foreign bodies in the ear and the aerodigestive tract in children



instrumentation depends on a variety of things, including the availability of equipment, experience of personnel, the age and medical status of the child, nature of the object and length of time since impaction. Ideally, a coordinated team of surgeons and physicians, trained in both rigid and flexible endoscopy, who can perform the removal of the foreign body in one procedure undertake this work and accept patients from a designated catchment area.61 Skilled paediatric anaesthesia is vital and close cooperation between the anaesthetist and bronchoscopist is essential to ensure the child’s alveolar ventilation is maintained throughout the procedure. The newer optical grabbing forceps contain an integrated telescope and can be passed through most rigid ventilating bronchoscopes (size 3.5 and above). They give a superb view of the trachea and the main stem bronchi (Figure 92.6). This enables the operator to grasp an object such as a peanut under direct vision. These optical instruments have made the management of tracheobronchial foreign bodies much easier and safer, although not all institutions have such instrumentation. Fogarty catheters and other improvised equipment can be used to successfully extract bronchial foreign bodies68 although, even in experienced centres, a small percentage of cases (1.8 percent in one large series)35 will need thoracotomy to remove the foreign body and extra-corporeal membrane oxygenation can be used to ‘buy time’ and prevent mediastinal compression in extreme circumstances.69



KEY POINTS  ‘Button’ batteries are potentially corrosive. Whether in the ear canal, the nose or the aerodigestive tract they should be removed as soon as possible.  Instrumental perforation is an ever present danger when removing oesophageal foreign bodies in children.  Bronchoscopy for foreign body extraction is a highly skilled technique. Teaching and learning this skill is difficult in communities where foreign body aspiration is now rare.  A high index of suspicion is essential in suspected foreign body inhalation in children. This is a potentially lethal condition.



Best clinical practice [ Spherical objects, items in the deep meatus and foreign bodies present for over 24 hours are more likely to require a general anaesthetic for removal. [Grade B]



] 1191



[ Multiple, ill-prepared attempts at removal should be



[ [ [



[



[



avoided. Early referral to an otolaryngologist must be considered if there is any doubt as to the success of an extraction attempt. [Grade C] Otomicroscopy, micro-instruments, suction, irrigation, glue-tipped probes and open surgery all have their place. [Grade B] Ninety-five percent ethanol is most effective for killing lodged insects prior to their removal. [Grade B] If general anaesthesia is required, it is reasonable to wait for a suitable elective operating list, except for corrosive items and live animals/insects which require urgent removal. [Grade C] Once the presumptive diagnosis of a pharyngooesophageal foreign body has been made, arrangements should be made for early removal as oedema and mucosal swelling will make retrieval more difficult. [Grade C] If the history is at all suggestive of an inhaled foreign body in a child, even if examination is normal, seek the advice of an experienced otolaryngologist. [Grade C/D]



Deficiencies in current knowledge and areas for future research



$ $ $



A global campaign focussing on consumer legislation to monitor the size of small objects, toys and household goods would greatly reduce mortality from inhaled foreign bodies in children. Increased parent and carer awareness of the dangers of small objects which children can swallow or inhale should also help reduce mortality. Re-organization of otolaryngology services and restucturing of training in ORL needs to take account of the need for centralized and skilled care for the removal of inhaled foreign bodies in children.



REFERENCES 1. Ansley JF, Cunningham MJ. Treatment of aural foreign bodies in children. Pediatrics. 1998; 101: 638–41. 2. Balbani AP, Sanchez TG, Butugan O, Kii MI, Angelico Jr. FV, Ikino CM et al. Ear and nose foreign body removal in children. International Journal of Pediatric Otorhinolaryngology. 1998; 46: 37–42. 3. Bhisitkul DM, Dunham M. An unsuspected alkaline battery foreign body presenting as malignant otitis externa. Pediatric Emergency Care. 1992; 8: 141–2.



1192 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY







4. Antonelli PJ, Ahmadi A, Prevatt A. Insecticidal activity of common reagents for insect foreign bodies of the ear. Laryngoscope. 2001; 111: 15–20. 5. Bressler K, Shelton C. Ear foreign-body removal: a review of 98 consecutive cases. Laryngoscope. 1993; 103: 367–70. 6. Fritz S, Kelen GD, Sivertson KT. Foreign bodies of the external auditory canal. Emergency Medicine Clinics of North America. 1987; 5: 183–92. 7. Schulze SL, Kerschner J, Beste D. Pediatric external auditory canal foreign bodies: a review of 698 cases. Otolaryngology – Head and Neck Surgery. 2002; 127: 73–78. Excellent review demonstrating which features are particularly associated with complications and suggesting referral guidelines for primary care clinicians. 8. DiMuzio Jr. J, Deschler DG. Emergency department management of foreign bodies of the external ear canal in children. Otology and Neurotology. 2002; 23: 473–5. 9. Jones RL, Chavda SV, Pahor AL. Parapharyngeal abscess secondary to an external auditory meatus foreign body. Journal of Laryngology and Otology. 1997; 111: 1086–7. 10. Benger JR, Davies PH. A useful form of glue ear. Journal of Accident and Emergency Medicine. 2000; 17: 149–50. 11. McLaughlin R, Ullah R, Heylings D. Comparative prospective study of foreign body removal from external auditory canals of cadavers with right angle hook or cyanoacrylate glue. Emergency Medicine Journal. 2002; 19: 43–5. 12. Pride H, Schwab R. A new technique for removing foreign bodies of the external auditory canal. Pediatric Emergency Care. 1989; 5: 135–6. 13. White SJ, Broner S. The use of acetone to dissolve a Styrofoam impaction of the ear. Annals of Emergency Medicine. 1994; 23: 580–2. 14. Abadir WF, Nakhla V, Chong P. Removal of superglue from the external ear using acetone: case report and literature review. Journal of Laryngology and Otology. 1995; 109: 1219–21. 15. Persaud R. A novel approach to the removal of superglue from the ear. Journal of Laryngology and Otology. 2001; 115: 901–2. 16. Engelsma RJ, Lee WC. Impacted aural foreign body requiring endaural incision and canal widening for removal. International Journal of Pediatric Otorhinolaryngology. 1998; 44: 169–71. 17. Kadish HA, Corneli HM. Removal of nasal foreign bodies in the pediatric population. American Journal of Emergency Medicine. 1997; 15: 54–6. 18. Brown CR. Intranasal button battery causing septal perforation: a case report. Journal of Laryngology and Otology. 1994; 108: 589–90. 19. Kubba H, Bingham BJ. Endoscopy in the assessment of children with nasal obstruction. Journal of Laryngology and Otology. 2001; 115: 380–4. 20. Douglas AR. Use of nebulized adrenaline to aid expulsion of intra-nasal foreign bodies in children. Journal of Laryngology and Otology. 1996; 110: 559–60.











21. Douglas SA, Mirza S, Stafford FW. Magnetic removal of a nasal foreign body. International Journal of Pediatric Otorhinolaryngology. 2002; 62: 165–7. 22. Lichenstein R, Giudice EL. Nasal wash technique for nasal foreign body removal. Pediatric Emergency Care. 2000; 16: 59–60. 23. Hanson RM, Stephens M. Cyanoacrylate-assisted foreign body removal from the ear and nose in children. Journal of Paediatrics and Child Health. 1994; 30: 77–8. 24. Finkelstein JA. Oral Ambu-bag insufflation to remove unilateral nasal foreign bodies. American Journal of Emergency Medicine. 1996; 14: 57–8. 25. Backlin SA. Positive-pressure technique for nasal foreign body removal in children. Annals of Emergency Medicine. 1995; 25: 554–5. 26. Botma M, Bader R, Kubba H. ‘A parent’s kiss’: evaluating an unusual method for removing nasal foreign bodies in children. Journal of Laryngology and Otology. 2000; 114: 598–600. 27. Nandapalan V, McIlwain JC. Removal of nasal foreign bodies with a Fogarty biliary balloon catheter. Journal of Laryngology and Otology. 1994; 108: 758–60. 28. Cohen HA, Goldberg E, Horev Z. Removal of nasal foreign bodies in children. Clinical Pediatrics. 1993; 32: 192. 29. Balatsouras D, Eliopoulos P, Kaberos A, Economou C. Rhinolithiasis: an unusual cause of nasal obstruction. Rhinology. 2002; 40: 162–4. 30. Ezsias A, Sugar AW. Rhinolith: an unusual case and an update. Annals of Otology, Rhinology and Laryngology. 1997; 106: 135–8. 31. Celikkanat S, Turgut S, Ozcan I, Balyan FR, Ozdem C. Rhinolithiasis. Rhinology. 1997; 35: 39–40. 32. Reilly BK, Stool D, Chen X, Rider G, Stool SE, Reilly JS. Foreign body injury in children in the twentieth century: a modern comparison to the Jackson collection. International Journal of Pediatric Otorhinolaryngology. 2003; 67: S171–4. A detailed account of the spectrum of objects commonly ingested. 33. Milkovich SM, Rider G, Greaves D, Stool D, Chen X. Application of data for prevention of foreign body injury in children. International Journal of Pediatric Otorhinolaryngology. 2003; 67: S179–82. An analysis of size and shape which should inform consumer legislation. 34. Lam HC, Woo JK, van Hasselt CA. Management of ingested foreign bodies: a retrospective review of 5240 patients. Journal of Laryngology and Otology. 2001; 115: 954–7. 35. Eren S, Balci AE, Dikici B, Doblan M, Eren MN. Foreign body aspiration in children: experience of 1160 cases. Annals of Tropical Paediatrics. 2003; 23: 31–7. 36. van As AB, du Toit N, Wallis L, Stool D, Chen X, Rode H. The South African experience with ingestion injury in children. International Journal of Pediatric Otorhinolaryngology. 2003; 67: S175–8. 37. Morley RE, Ludemann JP, Moxham JP, Kozak FK, Riding KH. Foreign body aspiration in infants and toddlers: recent trends in British Columbia. Journal of Otolaryngology. 2004; 33: 37–41.



Chapter 92 Foreign bodies in the ear and the aerodigestive tract in children 38. Singh B, Kantu M, Har-El G, Lucente FE. Complications associated with 327 foreign bodies of the pharynx, larynx, and esophagus. Annals of Otology, Rhinology and Laryngology. 1997; 106: 301–4. 39. Savage J, Brookes N, Lloyd S, Mackay I. Fish bones in the vallecula and tongue base: removal with the rigid nasal endoscope. Journal of Laryngology and Otology. 2002; 116: 842–3. 40. Viney R, Reid A. An alternative approach to fishbone extraction. Journal of the Royal College of Surgeons of Edinburgh. 2002; 47: 515. 41. Canbay E, Prinsley P. The case of the disappearing fish bone. Journal of Otolaryngology. 1995; 24: 375–6. 42. Osinubi OA, Osiname AI, Pal A, Lonsdale RJ, Butcher C. Foreign body in the throat migrating through the common carotid artery. Journal of Laryngology and Otology. 1996; 110: 793–5. 43. Blatnik DS, Toohill RJ, Lehman RH. Fatal complication from an alkaline battery foreign body in the esophagus. Annals of Otology, Rhinology and Laryngology. 1977; 86: 611–5. 44. Byard RW. Mechanisms of unexpected death in infants and young children following foreign body ingestion. Journal of Forensic Sciences. 1996; 41: 438–41. 45. Bhat NA, Oates J. An unusual foreign body in the larynx: a case report. Journal of Laryngology and Otology. 1996; 110: 1164–5. 46. Brama I, Fearon B. Laryngeal foreign bodies in children. International Journal of Pediatric Otorhinolaryngology. 1982; 4: 259–65. 47. America NSCo. Accident Facts. 1980; 7. 48. Ross GL, Steventon NB, Pinder DK, Bridger MW. Living on the edge of the post-nasal space: the inhaled foreign body. Journal of Laryngology and Otology. 2000; 114: 56–7. 49. Andazola JJ, Sapien RE. The choking child: what happens before the ambulance arrives? Prehospital Emergency Care. 1999; 3: 7–10. 50. Sharma HS, Sharma S. Management of laryngeal foreign bodies in children. Journal of Accident and Emergency Medicine. 1999; 16: 150–3. 51. Burton EM, Brick WG, Hall JD, Riggs Jr. W, Houston CS. Tracheobronchial foreign body aspiration in children. Southern Medical Journal. 1996; 89: 195–8. 52. Cohen SR, Herbert WI, Lewis Jr. GB, Geller KA. Foreign bodies in the airway. Five-year retrospective study with special reference to management. Annals of Otology, Rhinology and Laryngology. 1980; 89: 437–42. 53. Baharloo F, Veyckemans F, Francis C, Biettlot MP, Rodenstein DO. Tracheobronchial foreign bodies: presentation and management in children and adults. Chest. 1999; 115: 1357–62. 54. Banerjee A, Rao KS, Khanna SK, Narayanan PS, Gupta BK, Sekar JC et al. Laryngo-tracheo-bronchial foreign bodies in children. Journal of Laryngology and Otology. 1988; 102: 1029–32.



] 1193



55. Applegate KE, Dardinger JT, Lieber ML, Herts BR, Davros WJ, Obuchowski NA et al. Spiral CT scanning technique in the detection of aspiration of LEGO foreign bodies. Pediatric Radiology. 2001; 31: 836–40. 56. Hughes CA, Baroody FM, Marsh BR. Pediatric tracheobronchial foreign bodies: historical review from the Johns Hopkins Hospital. Annals of Otology, Rhinology and Laryngology. 1996; 105: 555–61. 57. Barben J, Berkowitz RG, Kemp A, Massie J. Bronchial granuloma – where’s the foreign body? International Journal of Pediatric Otorhinolaryngology. 2000; 53: 215–9. 58. Cataneo AJ, Reibscheid SM, Ruiz Junior RL, Ferrari GF. Foreign body in the tracheobronchial tree. Clinical Pediatrics. 1997; 36: 701–6. 59. Brown L, Sherwin T, Perez JE, Perez DU. Heliox as a temporizing measure for pediatric foreign body aspiration. Academic Emergency Medicine. 2002; 9: 346–7. 60. Blazer S, Naveh Y, Friedman A. Foreign body in the airway. A review of 200 cases. American Journal of Diseases of Children. 1980; 134: 68–71. 61. Swanson KL, Prakash UB, Midthun DE, Edell ES, Utz JP, McDougall JC et al. Flexible bronchoscopic management of airway foreign bodies in children. Chest. 2002; 121: 1695–700. 62. Abdulmajid OA, Ebeid AM, Motaweh MM, Kleibo IS. Aspirated foreign bodies in the tracheobronchial tree: report of 250 cases. Thorax. 1976; 31: 635–40. 63. Black RE, Johnson DG, Matlak ME. Bronchoscopic removal of aspirated foreign bodies in children. Journal of Pediatric Surgery. 1994; 29: 682–4. 64. Ayed AK, Jafar AM, Owayed A. Foreign body aspiration in children: diagnosis and treatment. Pediatric Surgery International. 2003; 19: 485–8. 65. Aytac A, Yurdakul Y, Ikizler C, Olga R, Saylam A. Inhalation of foreign bodies in children. Report of 500 cases. Journal of Thoracic and Cardiovascular Surgery. 1977; 74: 145–51. 66. Pasaoglu I, Dogan R, Demircin M, Hatipoglu A, Bozer AY. Bronchoscopic removal of foreign bodies in children: retrospective analysis of 822 cases. Thoracic and Cardiovascular Surgeon. 1991; 39: 95–8. 67. Clancy MJ. Bronchoscopic removal of an inhaled, sharp, foreign body: an unusual complication. Journal of Laryngology and Otology. 1999; 113: 849–50. 68. Ross MN, Haase GM. An alternative approach to management of Fogarty catheter disruption associated with endobronchial foreign body extraction. Chest. 1988; 94: 882–4. 69. Goldman AP, Macrae DJ, Tasker RC, Edberg KA, Mellgren G, Herberhold C et al. Extracorporeal membrane oxygenation as a bridge to definitive tracheal surgery in children. Journal of Pediatrics. 1996; 128: 386–8.



93 Tracheostomy and home care MICHAEL SAUNDERS



Introduction and historical perspective Indications for paediatric tracheostomy Techniques of tracheostomy specific to children Tracheostomy care Types of tracheostomy tube Complications of tracheostomy Discharge and home care



1194 1195 1196 1197 1199 1201 1204



Decannulation Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



1205 1207 1207 1208 1208



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words tracheostomy and child. The author has a personal bibliography of key papers on tracheostomy in children.



INTRODUCTION AND HISTORICAL PERSPECTIVE Widely acknowledged as one of the oldest documented surgical procedures, detailed historical accounts of tracheostomy are many and vivid. Widespread use of the procedure in children developed in the nineteenth century after Trousseau used the technique to relieve airway obstruction in diphtheria. Subsequently, the procedure was widely used in the treatment of poliomyelitis. With the introduction of widespread vaccination programmes these diseases have largely disappeared in the Western world. Until the late 1970s many tracheostomies in children were performed to relieve airway obstruction in acute airway infections such as epiglottitis and acute laryngotracheobronchitis (ALTB or croup, see Chapter 87, Acute laryngeal infections).1, 2, 3 As the standard of paediatric intensive care facilities improved and prolonged endotracheal intubation became a practical alternative to surgical tracheostomy, progressively fewer procedures were carried out for these indications. By the time



haemophilus influenza B (HiB) vaccine was introduced in the 1990s, endotracheal intubation rather than tracheostomy had become the accepted mode of airway management for acute bacterial epiglottitis.4 In a series of 153 paediatric tracheostomies, Line et al.2 report that prior to 1980, 38 percent of tracheostomies were performed for acute airway infections whereas after 1980 this figure had dropped to 12 percent. Similar findings were reported by Friedberg and Morrison5 when comparing a series of tracheostomies from 1981 to 1985 to a similar series from the same institution from 1976 to 1980. Crysdale et al.6 also report an overall reduction by half in the incidence of tracheostomy in the same period, attributed to the change in management of epiglottitis. Corbett7 reviewed 116 cases over a ten-year period (1995–2004) and reported a further shift in indications with no tracheostomies for acute airway infections alone and an increasing proportion required for congenital defects such as craniofacial anomalies and major upper gastrointestinal defects. Eighteen children (15.5 percent) required tracheostomy for acquired airway lesions



Chapter 93 Tracheostomy and home care



including subglottic stenosis, vocal cord palsy and respiratory papillomatosis, whilst 14 (12.1 percent) tracheostomies were to facilitate management of airway malacia (laryngotracheal, bronchial or a combination). Tracheostomy was also required for long-term ventilation in patients with neuromuscular disorders (14, 12.1 percent) or ventilator dependency (31, 26.7 percent). Tracheostomy in children is now an uncommon operation. Due to a shift in tertiary paediatric treatment to larger centres in the last decade and the lack of a reliable means of collecting data on a national basis, it is difficult to estimate the true incidence of paediatric tracheostomy. A survey of 2065 tracheostomies across the United States estimates a rate of 6.6 tracheostomies per 100,000 child years, with the highest incidence in the first year of life but with a second peak in incidence in the late teens due to increased risk of serious injury and trauma.8 A more recent study of indications in 362 tracheostomies between 1993 and 2001 has found no real change in incidence over the last decade.9 As a consequence of the relative scarcity of the procedure, the medical literature relating to paediatric tracheostomy is generally related to levels 3 and 4 evidence. There are no significant randomized controlled trials and the majority of publications tend to document the authors’ own series of tracheostomies and their complications in larger children’s hospitals. Reports of changing indications from such institutions may also be skewed by changes in medical practice in individual units.



INDICATIONS FOR PAEDIATRIC TRACHEOSTOMY The general indications for tracheostomy are as follows:    



relieve upper airway obstruction; prevent complications of prolonged intubation; reduce anatomical dead space; allow suction toilet of the trachea.



However, in practice, tracheostomies in children are nearly always performed to relieve upper airway obstruction or to allow or assist with mechanical ventilation.



Obstruction of the upper airway The upper airway (from the lips and anterior nares to the carina) may become obstructed at one or more anatomical levels by a range of pathologies (Table 93.1). If the obstruction is significant and life-threatening and no other means of relieving the obstruction (for example, nasopharyngeal airway or prong) is appropriate then a tracheostomy must be considered. Increasing availability and standard of paediatric intensive care facilities has allowed surgical procedures involving the airway to be undertaken without the need



] 1195



Table 93.1 Examples of obstruction of the upper airway potentially requiring tracheostomy. Anatomical site Oropharynx, tongue base



Nose, nasopharynx Supraglottis Glottis Subglottis Trachea



Example Macroglossia Treacher Collins/Goldenhar syndrome Cystic hygroma Choanal atresia Supraglottic cyst Vocal cord palsy Physical trauma Subglottic stenosis, haemangioma Tracheomalacia High tracheal stenosis



for a covering tracheostomy. Instead, the risk of postoperative airway obstruction is avoided by a period of intubation and ventilation (the ‘single stage’ approach). The relative indications for tracheostomy continue to change. As an example, until the late 1990s, tracheostomy was considered the mainstay of management for obstructing subglottic haemangioma. More recently, tracheostomy is often avoided by open excision.10 Similarly, the introduction of the cricoid split11 and single stage laryngotracheal reconstruction12 can avoid the need for tracheostomy for extubation failure due to subglottic oedema. [**]



Prolonged intubation The long-term complications of prolonged endotracheal intubation are well recognized – ulceration at the level of the glottis and, particularly in children, the subglottis, can lead to cicatrization and stenosis of the airway. [****] Being softer and more flexible than the adult and with correct selection of tube size and appropriate intensive care, the neonatal larynx is able to tolerate prolonged intubation for relatively longer than the adult. There is no clear consensus as to the maximum safe duration of intubation. Premature babies may now be intubated for several weeks before permanent damage becomes a risk. Although practice varies in different units, tracheostomy should normally be considered in older children after two to three weeks of endotracheal intubation.



Long-term and home ventilation An increasing number of children are now surviving previously lethal conditions, resulting in chronic respiratory failure because of the availability of long-term ventilation. Around half of these are ventilated by



1196 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY tracheostomy.13 Indications for long-term ventilation include:



Skin incision



       



The conventionally used skin incision is horizontal, situated halfway between the cricoid and sternal notch. A vertical incision has traditionally been less favoured because of the potentially poorer cosmetic outcome, but after decannulation from a long-standing tracheostomy, the resulting scar is such that it is unlikely that the orientation of the original incision will make much difference. An advantage of a vertical incision is that it facilitates midline dissection through the layers of the neck and is therefore sometimes advocated as the incision of choice in emergency tracheostomy. The subcutaneous fat immediately surrounding the incision may be removed after completing the skin incision. This allows the skin edges to invert slightly so as to line the tracheostome with squamous epithelium. This effect can be increased by suturing the edge of the skin incision to the edge of the tracheal incision (maturation sutures). The resulting tract is felt to be more secure as it is already lined with squamous epithelium, and may be associated with a lower rate of postoperative complications.15



failure of control of breathing; chest wall dysfunction; disorders of lung parenchyma; large airway disease; central sleep apnoea, Ondine’s curse; thoracic dystrophy; bronchopulmonary dysplasia (BPD); tracheobronchomalacia.14



Increasingly, these patients can be ventilated at home although the cost in terms of manpower and equipment is high.



Tracheal toilet In practice, very few children now require tracheostomy for toilet of the airway. Children with intractable aspiration may need regular suction but the presence of a tracheostomy can predispose to aspiration in itself and increase the risk of respiratory tract infection.



Dissection TECHNIQUES OF TRACHEOSTOMY SPECIFIC TO CHILDREN Positioning The infant is positioned supine on the operating table. Neck extension is achieved with a rolled towel or gel pillow under the shoulders. The neck can be fixed in extension and stabilized in the midline using adhesive tape such as Elastoplasts. (Figure 93.1) Theoretically, extension of the neck in infants increases the risk of injury to the great vessels in the root of the anterior neck; in practice, with careful dissection and identification of structures this is rarely a clinical problem.



Figure 93.1 Child positioned on the operating table for tracheostomy.



Dissection using monopolar or bipolar diathermy is advisable in small children to minimize blood loss. Although in adults and larger children the thyroid isthmus is traditionally divided and tied to prevent haemorrhage, in infants it is usually adequate to divide the isthmus of the thyroid with bipolar diathermy. Given the relatively small size of the infant neck and trachea, and secondly the relative proximity of the carotid sheath, it is advisable to palpate the trachea regularly throughout the dissection to ensure that the direction of dissection has not strayed from the midline.



Tracheal incision A vertical incision is made in the midline, usually in tracheal rings 3–4. It has long been established that too high an incision in the trachea predisposes to subglottic stenosis.16 A variety of other incisions has been advocated, including excision of an anterior tracheal window, a superiorly or inferiorly based tracheal flap which is raised and sutured to the skin or, recently, a cruciate incision in the trachea, the tracheal edges being closely apposed to the skin edges.17 The theoretical advantage of most of these techniques is increased stability of the initial tracheostomy tract and therefore greater safety in the event of accidental decannulation. However, although there is no evidence available from randomized clinical trials, animal experiments18 suggest that tracheal flaps may lead to an



Chapter 93 Tracheostomy and home care



increased risk of long-term stenosis and the majority of authors currently favour a simple vertical incision. [**]



Stay sutures Stay sutures are placed in the wall of the trachea on either side of the vertical incision. These are generally a removable suture (e.g. 4/0 PROLENE) and are left in situ until the first tube change. In the event of accidental decannulation, upward and lateral traction on the sutures will open the tracheostomy to make tube reinsertion simpler. The sutures may be taped to the chest wall (Figure 93.2) to prevent accidental removal.



Securing tracheostomy tubes Until the tracheostome has epithelialized and matured, the risks associated with accidental decannulation are more significant. Initially, it is the author’s practice to fix the tube into position in the neck using the inelastic linen tapes supplied with the tube. The tapes are tied in a secure knot, sufficiently tight to allow one finger to be inserted between the tapes and the neck skin (Figure 93.3). The tapes should be tightened with the neck flexed, rather than



Figure 93.2 Stay sutures in place.



] 1197



in the operative position with the neck extended. In the author’s view, suturing the flange of the tracheostomy tube to the skin should be avoided in children. After seven days the first change is undertaken and the linen tapes may be changed for a Velcros fastening which allows for easier changing and is less traumatic to the skin of the neck.



TRACHEOSTOMY CARE Adequate tracheostomy care is critical in the first two to three postoperative days. It is during the formation of the tract of the stoma that the risk of tube displacement is at its highest as the tract can close very quickly making reinsertion difficult. It is difficult to overstate the importance of tracheostomy nursing care in the postoperative period. With meticulous and skilled care, most of the complications of tracheostomy can be avoided. Inexperienced staff with no specific training are often reluctant to intervene. It is therefore essential that hospitals maintain a high standard of internal training in this regard and many units have specific teams and regular training programmes for nursing staff.



Suction Immediately after tracheostomy, the change from air that is warmed and humidified by the upper airway to dry cold air leads to a rapid increase in airway secretions. This gradually reduces after a few weeks. Secretions dry on the inside of the tracheostomy tube and gradually reduce the effective lumen. Humidification of inspired air and regular suctioning will reduce this tendency. Suctioning is required as often as is necessary to keep the tube and airway clear. Overzealous suctioning may lead to mucosal trauma in the distal trachea if the catheter is inserted into the tracheal lumen itself19 and eventually granulation may form at the tip of the tracheostomy tube, which in itself may lead to tube obstruction. It is suggested that the suction tube be inserted as far as the tip of the tracheostomy tube and withdrawn with a finger occluding the side port. The exact distance may be measured and marked on the suction tube. The need for suctioning decreases in frequency over time, although with lower respiratory tract infections, secretions may become thicker and more profuse. On such occasions, irrigation of the tracheostomy tube with sterile saline to loosen secretions prior to suctioning is often advocated but there is little evidence to support this practice and it may increase contamination of the lower airway.20



Humidification



Figure 93.3



Securing the tapes.



Initially humidification should be given via nebulizers and a tracheostomy mask (Figure 93.4). After a week or so, secretions reduce and the level of humidification



1198 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Figure 93.4



Tracheal mask for humidification.



Figure 93.5



Swedish nose.



required is less. A few weeks after tracheostomy, more mobile devices can replace permanent humidification. Longer-term humidification may be achieved by using a Swedish nose or a tracheostomy bib. A ‘Swedish nose’ attachment (Figure 93.5) contains a filter which becomes saturated by the moisture in exhaled air; this in turn humidifies the inhaled air. The tracheal bib works in a similar way. Both devices have the advantage of acting as filters for inspired air.



Skin care The tracheostomy wound itself becomes rapidly infected with skin commensals and is impossible to keep sterile. The wound heals by secondary intention and eventually the tract becomes lined with squamous epithelium and organized scar tissue. Securing the wound edge skin to the trachea with maturation sutures hastens the development



of an epithelialized tract. At this stage the tracheostome is considered mature. Until this point there is usually a considerable discharge from the wound itself and if the skin care in the first few days is not meticulous, skin and wound breakdown will occur. Usually a dry gauze or foam (e.g. Lyofoams) dressing is inserted between the peritrachesotomy skin and the flange of the tube. This is rapidly saturated and needs to be changed regularly. Of course the action of changing the dressing increases the risk of accidental decannulation and there is often reluctance on the part of nursing staff to do this. Adequate training in tracheostomy care is essential in a hospital where regular paediatric airway surgery takes place. Large skin incisions are generally not required in paediatric tracheostomy. If not adequately closed, a large incision will lead to gaping and wound breakdown. It is futile to attempt closure in this instance because of the bacterial colonization of the wound and inevitable infection. Large tracheostomy wounds require careful dressing and packing similar to a healing ulcer and a range of wound products are available. The wound will eventually close by secondary intention. The tapes used to secure the tube in place can lead to ulceration of the neck skin if they are left too tight or for too long. Although the linen tapes supplied with tubes are secure and inelastic, they have a tendency to cut into the skin. Again, meticulous nursing and skin care is vital. The problem can be reduced by tying the tapes inside a sleeve of TubigripTM or part of a large plastic (e.g. endotracheal) tube. When the tracheostomy matures, wider softer bands with Velcro fittings may be used and are less traumatic to the neck skin.



Change of tracheostomy tube The first change of tube is generally undertaken at around the seventh postoperative day. This allows some time for maturation of the stoma but is short enough to reduce the risk of tube obstruction from dried and thickened secretions. The first change should be undertaken by an otolaryngologist; in the relatively rare instance of difficulty reinserting the second tube, an emergency surgical procedure may be required. If oral intubation is difficult or impossible (e.g. retrognathia, laryngeal stenosis) it is advisable to undertake the first change in the operating theatre in case surgical intervention is needed. In children who can be easily intubated, (for example the majority of children tracheostomized for pronged ventilation), it is more usual to undertake the first change on the intensive care unit, with an intubation trolley and senior ITU medical staff on hand should reinsertion be difficult and reintubation required. If this procedure is uneventful, the nursing staff can carry out subsequent changes. If discharge home is



Chapter 93 Tracheostomy and home care



anticipated, the parents must be taught the tube-changing technique in a secure environment. There is no standard proscribed interval at which tubes should be changed; this varies between children and also in the same child given variation in season and in the health of the lower airway. The tube needs to be changed before dried secretions start to reduce the lumen of the tube. The old tube should be inspected after removal to determine the degree of contamination and this will influence the interval until the next change. If a tube visibly contains dried secretion on external inspection, has an audible whistle due to obstruction (‘if you can hear a tube you should change it’) or if the suction catheter cannot be passed due to obstruction, then it should be changed. If the suction catheter does not pass freely after changing the tube, the advice of an otolaryngologist should be sought.



TYPES OF TRACHEOSTOMY TUBE Diameter Modern tracheostomy tubes are sized in relation to the diameter of the lumen in millimetres. Older tubes still use the French gauge system of sizing. The age-appropriate size for a tracheostomy tube can be derived from the guide shown in Figure 93.6. In general terms, smaller sized tubes become obstructed more easily and may impair respiration if too small for the age of the child. A child with a long-standing tracheostomy should undergo regular age-appropriate ‘upsizing’. However, tracheostomy tubes tend to almost completely fill the trachea in very young children making normal speech difficult. In some cases, for example, if the airway above the tracheostome is not completely obstructed, a smaller sized tube will allow air to flow up thought the glottis and aid in normal speech production. Too large a tube will predispose to suprastomal collapse and may increase the risk of granulations and stenosis.



Length When inserted correctly, the end of the tracheostomy tube should sit comfortably proximal to the carina. Too long a tube will abut the carina and lead to mucosal injury and potentially subsequent scarring. Too short a tube increases the risk of accidental decannulation. Ideally, the tube should be at least 2 cm, inside the stoma and 1–2 cm, clear of the carina.21 Shileys and Bivonas tubes commonly in use in the UK are available in neonatal and paediatric lengths. For unusual applications (for example children with very thick necks), tubes with adjustable lengths are available (Figure 93.7). Tube tip position can be assessed on chest



] 1199



x-ray, at regular rigid bronchoscopy or by passing a flexible endoscope down the lumen of the tube to inspect the carina.



Material Silicone is now the most widely used material in paediatric tracheostomy tubes. Their relative flexibility compared to older metal tubes reduces the risk of mucosal trauma during neck movements and the soft flange is less likely to lead to skin injury around the tracheostome. One advantage of metal tubes is that as the stronger metal wall is thinner, it is possible to achieve a smaller outside diameter for the same internal diameter as a silicone tube.



Speaking valve Speaking valves are one-way valves which allow inhalation though the tube but force air upwards through the glottis on exhalation, creating sufficient subglottic pressure to allow phonation. As well as allowing speech, by increasing tracheal pressure on exhalation, these may improve lung function and reduce aspiration. Speaking valves should be used under supervision and not while the child is asleep. Small infants will not always tolerate the valve and quickly learn to blow the valve off the tracheotomy by coughing hard. Many children learn to occlude the end of the tube on exhalation by flexing the neck and occluding the end of the tube with the neck skin to achieve the same result to help phonation.



Fenestration Fenestrated tubes allow air to pass upwards to improve phonation. These are less practical in smaller children as the fenestration tends to become a focus for granulation and mucosal trauma on suctioning. Adequate passage of air upwards into the larynx is usually achieved by selecting a smaller tube diameter and allowing air leakage on expiration.



Inner tube For a tube of any given outer diameter, an inner tube will substantially reduce the diameter of the lumen and is therefore not practical in smaller children. For older children, tubes are available with inner tubes in silicone or metal. The main advantage is that the inner tube can be removed daily for cleaning; secondly, the inner tubes are often available with an integrated speaking valve.



1200 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Preterm − 1 Month



1−6 Months



6−18 Months



18 Months − 3 Years



3−6 Years



6−9 Years



9−12 Years



12−14 Years



5



5.0−6.0



6.0−7.0



7.0−8.0



8.0−9.0



9.0−10



10−13



13



ID (mm)



3.0



3.5



4.0



4.5



5.0



5.5



6.0



7.0



OD (mm)



4.5



5.0



6.0



6.7



7.5



8.0



8.7



10.7



Size



3.0



3.5



4.0



4.5



5.0



5.5



6.0



6.5



ID (mm)



3.0



3.5



4.0



4.5



5.0



5.5



6.0



6.5



7.1



7.7



8.3



9.0



44*



46*



50*



52*



54*



56*



Trachea (Transverse Diameter mm)



Great Ormond Street



Shiley



* Cuffed Tube Available



Portex (Blue Line)



Portex



PLASTIC



(555)



Bivona



All sizes available with Fome Cuff, Aire Cuff, & TTS Cuff.



Bivona Hyperflex



Bivona Flextend



OD (mm)



4.5



5.2



5.9



6.5



Length (mm) Neonatal



30



32



34



36



Paediatric



39



40



41*



42*



Long Paediatric ID (mm)



3.0



3.5



4.0



4.5



5.0



5.0



6.0



7.0



OD (mm)



4.2



4.9



5.5



6.2



6.9



6.9



8.3



9.7



Size



2.5



3.0



3.5



4.0



4.5



5.0



5.5



ID (mm)



2.5



3.0



3.5



4.0



4.5



5.0



5.5



OD (mm)



4.5



5.2



5.8



6.5



7.1



7.7



8.3



Length Neonatal



30



32



34



Paediatric



30



36



40



44



48



50



52



Size



2.5



3.0



3.5



4.0



4.5



5.0



5.5



ID (mm)



2.5



3.0



3.5



4.0



4.5



5.0



5.5



OD (mm)



4.0



4.7



5.3



6.0



6.7



7.3



8.0



Length Neonatal



30



32



34



36



Paediatric



38



39



40



41



42



44



46



ID (mm)



2.5



3.0



3.5



4.0



4.5



5.0



5.5



55



60



65



70



75



80



85



2.5



3.0



3.5



4.0



4.5



5.0



5.5



38



39



40



41



42



44



46



10



10



15



15



17.5



20



20



12−14



16



18



20



22



24



16



18



20



22



24



26



28 28



Usable Length (mm) ID (mm) Shaft Length (mm)



SILVER



Alder Hey Negus



Flextend Length ( mm) FG FG



Chevalier Jackson



FG



14



16



18



20



22



24



26



Sheffield



FG



12−14



16



18



20



22



24



26



ID (mm)



2.9−3.6



4.2



4.9



6.0



6.3



7.0



7.6



Cricoid (AP Diameter)



ID (mm)



3.6−4.8



4.8−5.8



5.8−6.5



6.5−7.4



7.4−8.2



8.2−9.0



9.0−10.7



10.7



Bronchoscope



Size



2.5



3.0



3.5



4.0



4.5



5.0



6.0



6.0



ID (mm)



3.5



4.3



5.0



6.0



6.6



7.1



7.5



7.5



OD (mm)



4.2



5.0



5.7



6.7



7.3



7.8



8.2



8.2



(Storz)



Endotracheal



ID (mm)



2.5



3.0



3.5



4.0



4.5



5.0



6.0



7.0



8.0



Tube (Portex)



OD (mm)



3.4



4.2



4.8



5.4



6.2



6.8



8.2



9.6



10.8



Figure 93.6 Chart for sizing tracheostomy tubes. Reproduced with kind permission from Michelle Wyatt and colleagues at the Department of Otolaryngology, Great Ormond Street Hospital, London, UK.



Chapter 93 Tracheostomy and home care



Cuff The presence of a cuff increases the risk of mucosal ischaemia and subsequent tracheal stenosis, particularly if high cuff pressures are employed. Cuffed tubes are rarely indicated in paediatric practice; until adolescence, a sufficient seal to allow positive pressure ventilation can normally be achieved with an uncuffed tube. There are two specific indications for cuffed tubes in children: firstly where there is a significant risk of aspiration (although a cuff will not completely protect against this) and secondly, where there is a decrease in lung compliance with intercurrent infection in a ventilated child and ventilation pressures need to be raised temporarily. In this instance, the risk of tension pneumothorax is significantly increased. In the author’s institution, children on the home ventilation programme are admitted to hospital for observation if the ventilation pressures become high enough to require a cuffed tube.



Variable/custom-fitted tubes The Bivonas HyperflexTM tube has an adjustable flange which allows change of the effective length of the tracheostomy tube (Figure 93.7). This is useful in



Figure 93.7 Bivonas adjustable tube.



Table 93.2



] 1201



children with prominent soft tissue in whom the prominent neck tissue would obstruct a normal tracheostomy tube, and in children in whom a low tracheal obstruction needs to be bypassed by the tube. The disadvantages of this sort of tube are that it is firstly expensive and secondly, the wall of the tube is thick resulting in a relatively small tube lumen. Custom-made tubes are available for children with specific anatomical difficulties. Tube manufacturers should be approached directly for advice.



COMPLICATIONS OF TRACHEOSTOMY General Tracheostomy complications are more likely in children than in adults, and more common in children under two years,1, 22, 23 particularly preterm infants.24 [**] Overall complication rates are quoted between 25 percent25 and 77 percent.26 There is likely to be considerable variation in the documentation and reporting of minor complications – some authors do not consider granulation to be a reportable complication (Table 93.2). The higher complication rate in smaller children is likely to reflect the relatively small diameter of the airway in small children and the ease with which the airway may be occluded (for example by secretions, granuloma or suprastomal collapse), but also the fact that younger children receiving tracheostomy may remain tracheostomized for a longer period.26 As with adult tracheostomies, emergency procedures are associated with a higher rate of complications27 and longer duration of tracheotomy is associated with a higher risk of long-term complications.28 Given the relative scarcity of paediatric tracheostomy and the limited number of large published series, it is difficult to accurately derive a risk for fatal complications of paediatric tracheostomy. Of larger studies reported since 1980, the mortality related to the tracheostomy tube itself ranges from 0 to 3.6 percent (Table 93.3).25, 29 In nearly all cases, the cause of tracheostomy-related death is tube



Complications of tracheostomy.



General Tube obstruction Accidental decannulation General complications of surgery and anaesthesia Death



Early postoperative (up to one week)



Late postoperative (after one week)



Bleeding; postoperative, wound edge Pneumorthorax Subcutaneous emphysema



Granulation Bleeding Suprastomal collapse



Infection Apnoea



Skin complications Aphonia, speech delay Psychological factors Adverse effects on family



1202 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Table 93.3



Larger (n4100) series of paediatric tracheostomies and complications.



Study and Number of publication year tracheostomies



Line et al., 19862 Crysdale et al., 19886 Carter and Benjamin, 198325 Carr et al., 200126 Prescott, 198930 Carron et al., 200029 Midwinter et al., 200223 Wetmore et al., 19821 Ward et al., 199527 Corbett et al., 20077



Years of study



Overall Early complication rate complication (%) (%)



Late Overall complication mortality (%) (%)



Tracheostomyrelated death (%)



153 319



1970–85 1976–85



38 32



12 9



26 23



22 13.5



3 0.9



164



1972–81



25



5% (est)



19 (est)



10.9



0



142



1990–99



77



14



63



15



0.7



293 218



1980–85 1988–98



32 (est) 44



-



-



10 19



2 3.6



143



1979–99



46



-



-



7



2.8



420



1971–80



49



28.3



52.6



28



2



103



1980–90



45.6



30



15.6



36



2.9



116



1995–2004 45



11.2



44.8



19.6



1.8



Est, figure estimated from text.



obstruction or accidental decannulation. Earlier reports quote higher mortality rates but paediatric otolaryngology practice has changed considerably in the last 30 years and older studies are unlikely to reflect current practice and safety. Mortality from nontracheostomy-related medical conditions, such as respiratory or cardiovascular disease, is consistently high (7–36 percent) in all series,23, 27 reflecting the complex medical conditions of children requiring tracheostomy.



Accidental decannulation If there is little or no natural airway above the tracheostomy or if a child is ventilator dependent, accidental decannulation can be fatal. The risk is increased by insufficiently tight ties, too short a tracheostomy tube and excessive traction on the tube from ventilator tubing. In the first few days before the tract matures, it is likely to be harder to reintroduce the tube if decannulated. Wetmore et al.1 report accidental decannulation in 29 of 420 (6.9 percent) children in the first week. The risk of accidental decannulation may be reduced by meticulous tracheostomy nursing care, and surgical techniques (stay sutures, maturation sutures) may help reduce the morbidity by allowing easier and safer reinsertion of a displaced tube.



Tube obstruction Immediately after tracheostomy the tube is most likely to become blocked with secretions. Regular suction is required. Humidification reduces the rate of secretion and helps to prevent the secretions drying in the lumen of the tube and narrowing the airway. In a mature tracheostomy, the tube is more likely to be obstructed by granulation as the tube tip, either as a result of use of suction catheters or direct trauma from the tube itself. In the event of accidental decannulation, the tube should be reintroduced in a controlled manner to prevent the creation of a false passage.



Pneumothorax, pneumomediastinum, surgical emphysema In the infant the domes of the pleura extend well into the neck. Inadvertently straying from the midline during dissection increases the risk of postoperative pneumothorax. This should be detected immediately postoperatively on a routine chest x-ray. Small pneumothoraces can be treated conservatively while larger ones will require chest drainage. If the tracheostomy wound is closed too tightly around the tube, or the dressings are too tightly applied to the neck skin, air may leak into the soft tissues of the neck



Chapter 93 Tracheostomy and home care



(surgical emphysema) or track down into the mediastinum. In this instance the wound should be reopened to allow air to track back out through the tissues and a corrugated drain should be inserted.



Bleeding Bleeding in the first few days after tracheostomy usually arises as a result of failure to achieve complete haemostasis during surgery. Commonly, bleeding may persist from the wound edge, anterior jugular veins or their tributaries, or the edge of the thyroid isthmus. If direct pressure is not adequate to control haemorrhage, the wound may be carefully packed with haemostatic gauze (Surgicel or Kaltostats). Re-exploration is rarely required. Later, minor bleeding may arise from areas of granulation around the tube. This can normally be controlled with cautery and ongoing medical treatment such as application of steroid and antibiotic ointment (for example, TriadcortylTM).



Tracheal innominate fistula Tracheoinnominate artery fistula is a rare but lethal complication. There is no reliable estimate of the risk in children, which in adults has been estimated as 0.4 percent.31 In some children the innominate artery lies abnormally high in the neck (Figure 93.8). If this finding is made at the time of surgery, the decision to perform a tracheostomy should be reconsidered. If there is no safe alternative, it is acceptable to place the tracheal incision higher than one would normally advocate and accept the risk of subglottic stenosis. An abnormally low tracheostomy will also increase the risk. A fistula into the artery



Figure 93.8



High innominate artery.



] 1203



forms as a result of erosion of the arterial wall by direct pressure from the tube. Although most bleeding coming from the tracheostomy tube itself is likely to represent granulation formation at the tube tip, in all cases the possibility of tracheal innominate artery fistula should be considered. The trachea should be examined by flexible bronchoscopy on the ward or by rigid endoscopy under anaesthesia. If the bleeding appears to arise from the anterior tracheal wall rather than tube tip granulation, the wound must be re-explored immediately, ideally with the assistance of a cardiothoracic surgeon. It may be possible to tamponade the bleeding by using a cuffed tube temporarily and if the laryngeal anatomy permits, endotracheal intubation should be established prior to exploration. The mortality from this complication remains very high.



Granulation The presence of the tracheostomy tube as a foreign body and the persistent presence of bacterial flora in the tract act as an ongoing stimulus for the formation of granulation tissue. Granulation may form at the skin edge of the tract (peristomal granulation) and inside the trachea itself, both on the anterior wall of the trachea above the tube (suprastomal granulation) and also at the tube tip lower in the trachea. Excessive or overexuberant suctioning can lead to more granulation through mucosal trauma and the tube itself can cause mucosal injury. This is generally more common with more rigid tube designs, particularly the silver32 and PVC designs. Granulation tissue can pose a number of problems; on the surface, granulations tend to discharge and bleed and, when severe, can lead to difficulty in changing the tube. More modern tubes made from less reactive silicone are more flexible and softer and are felt to reduce the problem both at the skin and inside the trachea. Peristomal granulations can generally be controlled with steroid/antibiotic preparations (e.g. Triadcortyl ointment). When more severe they may be removed with bipolar diathermy. Caution needs to be exercised when using silver nitrate cautery as the silver nitrate solution can easily enter the trachea itself leading to irritation, coughing and mucosal injury. In the author’s unit, this practice is avoided. Suprastomal granulations are almost universal. In theory, if large they will reduce the lumen of the supraglottic airway above the tube and increase the risks associated with accidental decannulation. Some authors advocate their removal at endoscopy on a regular basis, however, others feel that they are an inevitable consequence of tracheostomy; Rosenfeld and Stool33 describe granulation in 80 percent of 265 tracheostomies at bronchoscopy and advise against interval endoscopy to remove granulation tissue. At microlaryngoscopy, immediately prior to planned decannulation, all



1204 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY granulation should be removed to improve the airway. After decannulation and stomal closure, granulation generally resolves spontaneously.



Removal of suprastomal granulation Suprastomal granulations may if required be removed endoscopically using microlaryngeal instruments, a microdebrider using a Skimmers or Tru-Cuts blade or by KTP or CO2 laser. The KTP laser has the advantage of beam delivery using a flexible optic fibre in the relatively limited confines of the subglottis. Large granulation may be removed using a small sphenoid punch inserted into the tracheostome from externally, under endoscopic guidance at mircrolaryngoscopy.



Suprastomal collapse Suprastomal collapse is distinct from suprastomal granulation although the two conditions often coexist. For reasons that are not completely understood, the anterior tracheal wall immediately superior to the stoma itself softens and prolapses into the lumen of the subglottic trachea (Figure 93.9). This can significantly reduce the available airway, which in turn increases risks associated with accidental decannulation and also leads to decannulation failure. Minor collapse may be left, as it will tend to improve after decannulation. More significant collapse will require surgical treatment. The simplest of these involves excision and transfixion of the tracheostomy tract followed by endotracheal intubation for two to three days to support the trachea as the stoma heals.34 The author’s preference in mild to moderate suprastomal collapse is to explore the neck, identify the area of suprastomal collapse and pass a suture through the cartilage and around the strap muscles to elevate the



Figure 93.9



Suprastomal collapse.



collapsed section. This is carried out with excision and transfixion of the tracheostome skin. Sharp and Hartley35 describe ablation of the collapsed segment with KTP laser and a number of authors have described supporting the collapsed segment with a cartilage graft in more severe collapse.36 The specific procedure will depend on the degree of collapse and the surgeon’s personal preference.



Speech development There is debate about the extent to which tracheostomy affects the development of speech in children. Clearly, normal phonation will be impaired for the duration of a tracheostomy as insufficient subglottic pressure is generated and small infants tend not to tolerate speaking valves well. One difficulty in interpreting such studies is that a significant proportion of tracheostomized children have coexisting developmental abnormalities. If decannulation occurs in the first year to 18 months, before the time at which normal speech patterns begin to develop, the long-term outcome is favourable,37 whereas longer-term tracheostomy may lead to longer-term impairment of speech function.



DISCHARGE AND HOME CARE Discharge Getting home with a tracheostomy is a complex and timeconsuming process. Not all families will have sufficient support or resources at home to care for children with a tracheostomy. Whilst in hospital, the caregivers must be educated to care for the day-to-day eventualities of tracheostomy, including tube changing and the recognition and initial treatment of complications. Generally, two responsible adults are required for tube change; home tracheostomy care is difficult, but not impossible for single carers. Children who are included in home ventilation programmes tend to be more carefully supervised and a national protocol for discharge requirements has been formulated.38 Nonventilated tracheostomized children tend to have less structured support. In some areas, specific local organizations are available (e.g. the lifetime service in Avon and Wiltshire). Generally, home care is shared between hospital and primary care district nurses who have little specific training. With sufficient support and education of teachers and co-workers, tracheostomized children without other significant disabilities can now attend mainstream schooling in the UK, although certain activities must be avoided, particularly swimming, water-based sports and contact sports. The parent and carer support organization Aid for Children with Tracheostomies (ACT) provides a discussion forum, advice and information for caregivers in the UK,39 and caregivers may be given contact details prior to discharge.



Chapter 93 Tracheostomy and home care



Physical requirements



DECANNULATION



Tables 93.4 and 93.5 list the resources required for the child with a tracheostomy at home. In the author’s experience, the ease with which equipment and accessories can be obtained by caregivers is extremely variable in the community as financial constraints in the delivery of care lead to reluctance to supply regular consumables. Prior to discharge, it is essential to communicate with the child’s general practitioner and other primary care workers and establish responsibility for provision of equipment.



Decision to decannulate



Table 93.4



] 1205



Decannulation may be considered when the original condition requiring tracheostomy has improved, however, to make decannulation successful the child must be able to maintain an adequate airway without the tracheostomy in place. The majority of paediatric tracheostomies are short term, as the natural airway tends to improve with overall



Requirements for children at home with tracheostomy.



Caregivers Generally two responsible adults



Physical



Support



Home with adequate space, heating, electricity, telephone, access to transport



District nurse Community paediatrician Health visitor General practitioner Hospital-based support Specific community organizations where available



Table 93.5



Equipment requirements for children at home with tracheostomy with and without home ventilation.



Requirements for children without ventilation Appropriate sized tracheostomy tubes and one a size smaller Neck ties to hold tube in place Scissors for emergency tube change to cut neck ties Lubricant for inserting tube Sterile saline and syringes for saline suction if required Tracheostomy dressing if required Gauze to clean stoma Appropriate sized suction catheters Heat and moisture exchangers/Swedish noses Speaking valves Gloves – nonsterile for procedures and alcohol gel hand rub Plastic aprons and protective goggles Stethoscope



Additional requirements for children on home ventilation Two ventilators/CPAP machines, one of which is portable plus batteries and chargers for use outside the home Disposable ventilator circuits Humidifier for ventilator circuit and water for inhalation to supply humidifier Dry circuit for ventilation when outside the home Heat and moisture exchanger for dry circuit Nebulizer CO2 monitor Rechargeable torch for use at night in the event of a power cut Uninterrupted power supply – battery which powers ventilator in the event of a power cut Suitable trolley in bedroom for equipment Adequate power sockets in house, particularly child’s bedroom Trolley for children with a lot of equipment to transport equipment at nursery/school Larger than normal buggy when baby/toddler to transport child and equipment



Two suction machines, one of which must be portable. Most children keep a third machine at school as spare Saturation monitor and possibly portable saturation monitor for use outside the home Ambu bag Oxygen: concentrator if used on a daily basis, cylinders if used less frequently, portable cylinders Most of the above needs to be duplicated in a portable set. A battery powered suction machine is essential and if the child is oxygen dependent, portable cylinders. Reprinted from Ref. 40, with permission.



1206 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY growth of the child or as a result of corrective surgery such as laryngotracheal reconstruction. The decision to decannulate is a complicated one which needs to be taken by a senior clinician after careful discussion with the parents and other relevant health care professionals. In paediatric otolaryngology practice it is generally considered essential to undertake endoscopic assessment of the airway prior to definitive decannulation.34 Suprastomal collapse and granulation leads to a considerable reduction in the lumen of the subglottic airway in children. Prescott41 suggested that this was the most common cause of decannulation failure in children, finding significant granulation in 50 and significant suprastomal collapse in 52 of 300 tracheostomies. [**] In addition, vocal cord mobility should be assessed at endoscopy. Granulation may be removed at the time of endoscopy using punch forceps or laser ablation. More significant suprastomal collapse requires KTP laser ablation or reconstructive surgery using cartilage grafting if the collapse is greater than 50 percent.35 If the subglottic airway is deemed satisfactory at endoscopy, the child may then proceed to formal decannulation in the next few days, delaying the risk of reformation of granulation. One should also consider comorbidity, such as pulmonary or neurological disease, and lastly the need for further surgery; for example, if a child with Treacher Collins syndrome and an indwelling tracheostomy requires further (e.g. mandibular or palatal) surgical procedures one would normally consider delaying decannulation.



Decannulation technique Removal of a tracheostomy leads to a significant change in the physiology of the upper airway. The dead space is doubled and airway resistance is trebled. With a longstanding tracheostomy, the child may have no memory of mouth and nose breathing and the new sensation may be distressing.



Staged decannulation To effect these changes more gradually, decannulation protocols have been developed which involve tube ‘downsizing’ and reversible capping (Table 93.6).42 To assess whether the child can breathe through the normal anatomical airway, the tube is capped off, either with a button, by taping or by inserting the obturator. However, the tracheostomy tube itself occupies a significant fraction of the tracheal lumen and to try and reduce this effect, the tube size is reduced to a size 3.0 (or size 2.5 in children under 13 months,43 either in stages or in one step. Leaving the small tube in situ allows a certain amount of



Table 93.6 Great Ormond Street protocol for ward decannulation. Day 1 2 3 4 5



Procedure Admission, downsize to 3.0 tube Block for 12 hours from 8 am, if successful continue overnight for a further 12 hours Decannulate, occlude stoma with adhesive tape and dressing. Observe on the ward Observe off the ward Discharge



Reprinted from Ref. 42, with permission.



respiration if required and also prevents the tract from closing down, should decannulation fail.



Immediate decannulation The tracheostomy tube may occupy as much as half of the lumen of the trachea in an infant. If a child can tolerate this degree of obstruction, the airway after decannulation is likely to be more than sufficient. However, some children will not be able to tolerate this degree of tracheal obstruction. In this instance, it may be considered appropriate to simply remove the whole tube and occlude the stoma with a dressing. However, it is vital that this be carried out in a controlled setting (i.e. intensive care) where facilities for intubation are available should decannulation fail and reinsertion of the tracheostomy not be possible. If the nature of the child’s airway obstruction is such that oral intubation is not possible (e.g. some cases of Treacher Collins syndrome) then it is not safe to remove the tube in this manner, and decannulation should be delayed until the child is large enough to tolerate staged decannulation with capping off.



Persistent tracheocutaneous fistula After decannulation, a fistula may persist between the trachea and skin. This may be small and only lead to problems with discharge of tracheal secretions. A larger fistula may continue to function as an alternative airway. The incidence of tracheocutaneous fistula (TCF) is between 19 and 42 percent in various series. Certain factors lead to an increased risk; lower age at initial tracheostomy, duration of tracheostomy and, most importantly, persistent obstruction above the level of the stoma (e.g. inadequate reconstruction of subglottic stenosis). There is no clear consensus as to how long a persistent TCF should be allowed to close before considering surgery. Most authors would allow 6–12 months before formal closure.



Chapter 93 Tracheostomy and home care



Closure of TCF It is essential that the upper airway be reassessed prior to TCF closure to exclude persistent obstruction and tracheal granulation. The persistence of squamous epithelium lining the tracheostome increases the likelihood of a persistent fistula. Simply removing the skin lining the tract and reattempting conventional extubation may lead to satisfactory closure.44 More commonly, the tract of the stoma is dissected down to the level of the trachea and closed with transfixion sutures. The strap muscles can then be reapposed to each other, which tends to fill in the cosmetic defect left after conventional decannulation. Lastly, the scarred skin surrounding the tracheostome can be excised in a fusiform incision with horizontal skin closure. This leads to an excellent cosmetic result. It is the author’s practice to leave a drain in the wound for 24 hours in case of air leak from the closure, which might otherwise lead to surgical emphysema and pneumomediastinum.



Revision of tracheostomy scar After conventional decannulation, the tracheostome heals by secondary intention. The resulting scar is usually unsightly (Figure 93.10) and features a pronounced depression at the line of the tract. Older children may become self-conscious and embarrassed about this. Revision of the scar is usually left until the child is over ten years of age, as cosmetic appearances become more important to the child with age and secondly because there is likely to be less further widening of the subsequent scar with age if the procedure is delayed. Revision of the scar usually involves a fusiform horizontal incision to excise the scarred skin of the tracheostome with wide undermining of surrounding skin to assist in primary closure. The strap muscles should be identified and reapposed in the midline to eliminate the defect in the contour of the neck skin. Deep dermal/ platysmal sutures are used to support the wound and then the skin edges are closed meticulously. Flexing the neck makes it easier to close a large skin defect.



KEY POINTS  Tracheostomies in children are performed in the main to relieve upper airway obstruction or to assist with mechanical ventilation.  A reduction in the incidence of infections – notably HiB epiglottitis – which cause airway obstruction in children, and improved management of airway pathology by anaesthetists and paediatricians have brought about a change in the indications for tracheostomy.  Tracheostomy in children is now an uncommon operation.  Tracheostomy complications are more likely in children than in adults. Preterm infants are at particular risk.  Suprastomal granulations are almost universal in children.  Getting home with a tracheostomy is a complex and time-consuming process. Not all families will have sufficient support or resources at home to care for a child with a tracheostomy.



Best clinical practice [ Endotracheal intubation rather than tracheostomy is [



[ [ [ [



[



[



Figure 93.10



Scar following long-term tracheostomy.



] 1207



the accepted mode of management for acute obstructing airway infection in children. [Grade C/D] Premature babies may be safely intubated for several weeks. Tracheostomy should normally be considered in older children after two to three weeks of endotracheal intubation. [Grade C/D] Large skin incisions are generally not required in paediatric tracheostomy. [Grade C/D] Careful dissection using diathermy is advisable in small children to minimize blood loss. [Grade C/D] Palpate the trachea regularly throughout the dissection to ensure that you have not strayed from the midline. [Grade C/D] A simple vertical incision to open the trachea is associated with the lowest risk of long-term complications. Avoid removing any cartilaginous tissue in children. [Grade C/D] Stay sutures in the wall of the trachea on either side of the vertical incision facilitate reintroduction of the tube in the event of accidental decannulation before a mature track has developed. [Grade C/D] In children who are difficult to intubate (e.g. retrognathia, laryngeal stenosis), it is safer to do the first tube change in the operating theatre in case surgical intervention is needed. [Grade C/D]



1208 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY [ The small diameter of the child’s airway makes [ [ [



[ [



suction and humidification especially important as secretions can quickly occlude the airway. [Grade C/D] Tube obstruction or accidental decannulation may be fatal. [Grade C/D] An ‘inner tube’ reduces the diameter of the lumen and is therefore not practical in small children. [Grade C/D] Fenestrated tubes are impractical in smaller children. The fenestration tends to become a focus for granulation and mucosal trauma on suctioning. [Grade C/D] Cuffed tubes are rarely needed in children. [Grade C/D] Speaking valves should be used under supervision and not while the child is asleep. [Grade C/D]



Deficiencies in current knowledge and areas for future research



$ $ $











Paediatric tracheostomy is now largely undertaken in specialist paediatric units. It is difficult for otolaryngologists in training to get experience in the management of children with tracheostomies outside of these centres. This trend is likely to increase. Research in paediatric airway disorders is focussed on conditions which give rise to the need for tracheostomy, such as laryngotracheal stenosis and major congenital anomalies. There is a need to improve training resources and support in primary care and community settings to enable families to look after tracheostomized children at home.



REFERENCES 1. Wetmore RF, Handler SD, Potsic WP. Pediatric tracheostomy. Experience during the past decade. Annals of Otology, Rhinology and Laryngology. 1982; 91: 628–32. 2. Line Jr. WS, Hawkins DB, Kahlstrom EJ, MacLaughlin EF, Ensley JL. Tracheotomy in infants and young children: the changing perspective 1970-1985. Laryngoscope. 1986; 96: 510–5. 3. Prescott CA, Vanlierde MJ. Tracheostomy in the management of laryngotracheobronchitis. Red Cross War Memorial Children’s Hospital experience, 1980-1985. South African Medical Journal. 1990; 77: 63–6. 4. Benjamin B, O’Reilly B. Acute epiglottitis in infants and children. Annals of Otology, Rhinology and Laryngology. 1976; 85: 565–72. 5. Friedberg J, Morrison M. Paediatric tracheotomy. Canadian Journal of Otolaryngology. 1974; 3: 147–55.







6. Crysdale WS, Feldman RI, Naito K. Tracheotomies: a 10year experience in 319 children. Annals of Otology, Rhinology and Laryngology. 1988; 97: 439–43. 7. Corbett HJ, Mann KS, Mitra I, Jesudason EC, Losty PD, Clarke RW. Tracheostomy: a 10-year experience from a UK pediatric surgical center. Journal of Pediatric Surgery. 2007; 42:1251–4. A recent account of the changing indication for tracheostomy particularly due to acute infection. 8. Lewis CW, Carron JD, Perkins JA, Sie KC, Feudtner C. Tracheotomy in pediatric patients: a national perspective. Archives of Otolaryngology – Head and Neck Surgery. 2003; 129: 523–9. 9. Hadfield PJ, Lloyd-Faulconbridge RV, Almeyda J, Albert DM, Bailey CM. The changing indications for paediatric tracheostomy. International Journal of Pediatric Otorhinolaryngology. 2003; 67: 7–10. 10. Wiatrak BJ, Reilly JS, Seid AB, Pransky SM, Castillo JV. Open surgical excision of subglottic hemangioma in children. International Journal of Pediatric Otorhinolaryngology. 1996; 34: 191–206. 11. Cotton RT, Seid AB. Management of the extubation problem in the premature child. Anterior cricoid split as an alternative to tracheotomy. Annals of Otology, Rhinology and Laryngology. 1980; 89: 508–11. The first account of this important technique to aviod tracheostomy in mild subglottic stenosis. 12. Lusk RP, Gray S, Muntz HR. Single-stage laryngotracheal reconstruction. Archives of Otolaryngology – Head and Neck Surgery. 1991; 117: 171–3. 13. Edwards EA, O’Toole M, Wallis C. Sending children home on tracheostomy dependent ventilation: pitfalls and outcomes. Archives of Disease in Childhood. 2004; 89: 251–5. 14. Amin RS, Fitton CM. Tracheostomy and home ventilation in children. Seminars in Neonatology. 2003; 8: 127–35. 15. Park JY, Suskind DL, Prater D, Muntz HR, Lusk RP. Maturation of the pediatric tracheostomy stoma: effect on complications. Annals of Otology, Rhinology and Laryngology. 1999; 108: 1115–9. 16. Jackson C. High tracheotomy and other errors-the chief causes of chronic laryngeal stenosis. Surgery, Gynecology and Obstetrics. 1921; 32: 392–8. Seminal paper which did much to change surgical practice and reduce the incidence of post-tracheostomy subglottic stenosis. 17. Koltai PJ. Starplasty: a new technique of pediatric tracheotomy. Archives of Otolaryngology – Head and Neck Surgery. 1998; 124: 1105–11. 18. Fry TL, Jones RO, Fischer ND, Pillsbury HC. Comparisons of tracheostomy incisions in a pediatric model. Annals of Otology, Rhinology and Laryngology. 1985; 94: 450–3. 19. Bailey C, Kattwinkel J, Teja K, Buckley T. Shallow versus deep endotracheal suctioning in young rabbits: pathologic effects on the tracheobronchial wall. Pediatrics. 1988; 82: 746–51. 20. Raymond SJ. Normal saline instillation before suctioning: helpful or harmful? A review of the literature. American Journal of Critical Care. 1995; 4: 267–71.



Chapter 93 Tracheostomy and home care 21. Sherman JM, Davis S, Albamonte-Petrick S, Chatburn RL, Fitton C, Green C et al. Care of the child with a chronic tracheostomy. This official statement of the American Thoracic Society was adopted by the ATS Board of Directors, July 1999. American Journal of Respiratory and Critical Care Medicine. 2000; 161: 297–308. 22. Donnelly MJ, Lacey PD, Maguire AJ. A twenty year (1971–1990) review of tracheostomies in a major paediatric hospital. International Journal of Pediatric Otorhinolaryngology. 1996; 35: 1–9. 23. Midwinter KI, Carrie S, Bull PD. Paediatric tracheostomy: Sheffield experience 1979-1999. Journal of Laryngology and Otology. 2002; 116: 532–5. 24. Kenna MA, Reilly JS, Stool SE. Tracheotomy in the preterm infant. Annals of Otology, Rhinology and Laryngology. 1987; 96: 68–71. 25. Carter P, Benjamin B. Ten-year review of pediatric tracheotomy. Annals of Otology, Rhinology and Laryngology. 1983; 92: 398–400. 26. Carr MM, Poje CP, Kingston L, Kielma D, Heard C. Complications in pediatric tracheostomies. Laryngoscope. 2001; 111: 1925–8. 27. Ward RF, Jones J, Carew JF. Current trends in pediatric tracheotomy. International Journal of Pediatric Otorhinolaryngology. 1995; 32: 233–9. 28. Gianoli GJ, Miller RH, Guarisco JL. Tracheotomy in the first year of life. Annals of Otology, Rhinology and Laryngology. 1990; 99: 896–901. 29. Carron JD, Derkay CS, Strope GL, Nosonchuk JE, Darrow DH. Pediatric tracheotomies: changing indications and outcomes. Laryngoscope. 2000; 110: 1099–104. 30. Prescott CA, Vanlierde MJ. Tracheostomy in children – the Red Cross War Memorial Children’ Hospital experience 1980–1985. International Journal of Pediatric Otorhinolaryngology. 1989; 17: 97–107. 31. Cooper JD. Trachea-innominate artery fistula: successful management of 3 consecutive patients. Annals of Thoracic Surgery. 1977; 24: 439–47. 32. Quiney RE, Spencer MG, Bailey CM, Evans JN, Graham JM. Management of subglottic stenosis: experience from two centers. Archives of Disease in Childhood. 1986; 61: 686–90. 33. Rosenfeld RM, Stool SE. Should granulomas be excised in children with long-term tracheotomy? Archives of



34.



35.



36.



37.



 38.



 39. 40. 41.



42.



43.



44.



] 1209



Otolaryngology – Head and Neck Surgery. 1992; 118: 1323–7. Benjamin B, Curley JW. Infant tracheotomy – endoscopy and decannulation. International Journal of Pediatric Otorhinolaryngology. 1990; 20: 113–21. Sharp HR, Hartley BE. KTP laser treatment of suprastomal obstruction prior to decannulation in paediatric tracheostomy. International Journal of Pediatric Otorhinolaryngology. 2002; 66: 125–30. Froehlich P, Seid AB, Kearns DB, Pransky SM, Morgon A. Use of costal cartilage graft as external stent for repair of major suprastomal collapse complicating pediatric tracheotomy. Laryngoscope. 1995; 105: 774–5. Jiang D, Morrison GA. The influence of long-term tracheostomy on speech and language development in children. International Journal of Pediatric Otorhinolaryngology. 2003; 67: S217–20. Jardine E, Wallis C. Core guidelines for the discharge home of the child on long-term assisted ventilation in the United Kingdom. UK Working Party on Paediatric Long Term Ventilation. Thorax. 1998; 53: 762–7. A good review of current indications for home ventilation and discussion of many of the issues this raises. www.actfortrachykids.com. Essential resource for the parents/carers of tracheostomy children. Longdon J. 2005. Personal communication. Prescott CA. Peristomal complications of paediatric tracheostomy. International Journal of Pediatric Otorhinolaryngology. 1992; 23: 141–9. Waddell A, Appleford R, Dunning C, Papsin BC, Bailey CM. The Great Ormond Street protocol for ward decannulation of children with tracheostomy: increasing safety and decreasing cost. International Journal of Pediatric Otorhinolaryngology. 1997; 39: 111–8. Kubba H, Cooke J, Hartley B. Can we develop a protocol for the safe decannulation of tracheostomies in children less than 18 months old? International Journal of Pediatric Otorhinolaryngology. 2004; 68: 935–7. Stern Y, Cosenza M, Walner DL, Cotton RT. Management of persistent tracheocutaneous fistula in the pediatric age group. Annals of Otology, Rhinology and Laryngology. 1999; 108: 880–3.



94 Cervicofacial infections in children BEN HARTLEY



Introduction Assessment of neck swellings Clinical assessment Examination of the neck Head and neck examination Specific conditions



1210 1210 1210 1211 1212 1212



Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



1217 1217 1218 1218



SEARCH STRATEGY The data in this chapter are supported by a Medline search using the key words infection and child, neck, cervical node, cervical adenopathy and focussing on diagnosis and treatment of specific conditions. The Cochrane Library was also consulted.



INTRODUCTION Neck swellings in children are common. Most are due to lymphadenopathy secondary to the common acute upper respiratory infections, notably pharyngitis and tonsillitis, that are a feature of normal childhood. They are rarely indicative of serious pathology. Lymphadenopathy is usually self-limiting but may progress to cellulitis, suppuration and abscess formation. Chronic infections are less common but need to be considered if the swelling persists. Congenital lesions do not always declare themselves at birth; they may present in older children. The onset of the swelling may be precipitated by an acute inflammatory episode. Neck swellings may involve the parotid and submandibular regions and extend on to the face. In a very a small number of children, a neck swelling will be due to a malignancy. These are most often lymphoproliferative or connective tissue tumours. Squamous carcinoma is extremely rare in children. A child with a neck mass requires an entirely different approach to that required in an adult.1 Open biopsy is rarely



needed. Nevertheless, it is important to maintain an index of suspicion for malignancy in all persistent neck swellings in children.



ASSESSMENT OF NECK SWELLINGS There is a broad spectrum of aetiologies for inflammatory lymphadenopathy in children (Table 94.1). Noninflammatory disorders will need to be considered in the differential of a neck mass, especially if there are unusual features. A methodical approach is required with a focus on clinical assessment.



CLINICAL ASSESSMENT History  Age of child: Most acute lymphadenitis presents in children older than six months. Swellings occurring at or shortly after birth are more likely to be congenital or neoplastic.



Chapter 94 Cervicofacial infections in children Table 94.1



] 1211



Inflammatory neck nodes in children. Infective aetiologies



Viral Upper respiratory: rhinovirus, adenovirus, enterovirus Common childhood illnesses: measles, mumps, rubella, varicella Infectious mononucleosis Cytomegalovirus



Bacterial



Fungal



Acute lymphadenitis: Histoplasmosis Streptococcus, Staphylococcus, less commonly Gram-negative organisms Suppurative lymphadenitis with Uncommon fungal deep or superficial neck infections abscess; usually pyogenic (immunocompromised organisms (Streptococcus host) and Staphylococcus) Mycobacteria: tuberculosis or Candida and Aspergillus ‘atypical’ mycobacteria Other granulomatous bacterial infections: cat scratch disease, actinomycosis, brucellosis, tularaemia, bubonic plague, syphilis



Parasitic



Non-infective adenopathy



Toxoplasmosis



Kawasaki syndrome



Filariasis



Sarcoidosis



HIV



 Duration of swelling: A short history (a few days) suggests acute inflammation. After six weeks a swelling is regarded as chronic and further investigation should be considered – even earlier if there are suspicious clinical features.  Size: Very large swellings or swellings that progressively enlarge despite antimicrobial treatment should be investigated.  Associated symptoms: A preceding upper respiratory infection is often a feature of inflammatory lymphadenitis. Fever, rhinorrhoea, sore throat and malaise are common. With chronic swellings, enquire about weight loss, night sweats and swellings elsewhere in the body.  Contacts: Enquire about tuberculosis, other infections and exposure to cats, farm animals and ticks.  Medical history: Identify any known illnesses.  Family and social history: Identify any familial disease or congenital anomalies and any relevant social factors. If a diagnosis of human immunodeficiency virus (HIV) infection is suspected then involve a paediatrician.



EXAMINATION OF THE NECK Site of the swelling The site of the swelling within the neck provides important information about the possible aetiology.



Sinus histiocytosis with massive lymphadenopathy Kikuchi-Fujimoto disease



PFAPA syndrome (periodic fever, apthous stomatitis, pharyngitis, cervical adenitis)



LATERAL NECK SWELLINGS



Lymph nodes are distributed throughout the neck but the most common site is along the superficial and deep cervical chains. These lie deep to sternomastoid in the upper neck and along its anterior border in the lower neck. Enlarged lymph nodes are the most common cause of lateral neck swellings. The principal differential includes congenital anomalies such as branchial cysts, which may also become acutely inflamed. Vasoformative lesions (see also Chapter 99, Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma), haemangiomas and vascular malformations, including lymphatic abnormalities and benign and malignant neoplasms arising from the neural or connective tissue elements present, as well as rare secondary metastases.



CENTRAL NECK SWELLINGS



The principle causes of a neck swelling in the central area of the neck around the midline are thyroglossal duct cyst, lymph nodes and dermoid cyst. Thyroglossal duct or dermoid cysts may become acutely inflamed. Children can also develop inflammatory and neoplastic thyroid disease (see also Chapter 99, Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma). Lymphatic vascular malformations can occasionally involve this region.



1212 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY PAROTID SWELLINGS



Investigation



Acute parotitis (mumps) is common and is due to a selflimiting viral infection. Vaccination for measles, mumps and rubella is now reducing the frequency of mumps and to some degree the clinical awareness of this condition (see also Chapter 97, Salivary gland disorders in childhood).



In many cases no investigation is required. Symptomatic treatment of presumed viral infection or antibiotic treatment of bacterial infection with careful clinical follow-up will result in resolution.



BACTERIAL PAROTITIS



Laboratory and skin tests



This occurs in children and may be recurrent. It is usually characterized by acute painful presentation followed by resolution on antibiotics. Occasionally, chronic inflammatory swelling persists and must be distinguished from neoplasia. Vascular malformations and haemangiomas cause swellings in the parotid region. Occasionally, rhabdomyosarcoma or other connective tissue tumours present. Magnetic resonance scanning can be very helpful with this differential diagnosis (see also Chapter 97, Salivary gland disorders in childhood).



If the child is systemically unwell, a full blood count may demonstrate neutrophilia consistent with bacterial infection. A blood count may also be a screening investigation for suspected haematological malignancy. Consider a Monospot test for infectious mononucleosis (see also Chapter 95, Diseases of the tonsil). Serological tests for toxoplasmosis, bartonella (cat scratch disease), or cytomegalovirus (CMV) should be considered for persistent lymphadenopathy. Mantoux or Heaf tests for tuberculosis may be helpful, particularly in the nonimmunized.



SUBMANDIBULAR SWELLINGS



Enlarged lymph nodes, floor of mouth infections, acute sialadenitis and occasionally lymphatic or vascular malformations and congenital ‘plunging’ ranula all cause swelling in this region (see also Chapter 97, Salivary gland disorders in childhood). POSTERIOR TRIANGLE SWELLINGS



Most commonly these are lymph nodes but branchial anomalies, vascular malformations and neoplasia enter into the differential diagnosis.



Nature of the swelling Classical signs of acute inflammation may be present, e.g. redness, tenderness and heat. Chronic swellings usually do not show these signs. If abscess formation has occurred, the clinical sign of fluctuance may be present and the mass may feel cystic. A classical tuberculous abscess lacks the clinical features of acute inflammation (‘cold abscess’).



HEAD AND NECK EXAMINATION A careful examination for a source of primary infection should be made. This should include examination of the pharynx, nose and ears as well as looking for any cutaneous lesion.



General examination Fever, tachycardia or rash should be noted. A general examination should include a search for any associated lymphadenopathy and hepatosplenomegaly. The otolaryngologist may wish to enlist the help of a paediatrician.



Imaging There is only a limited place for plain radiographs. Films of the chest may be helpful in tuberculosis and a lateral neck view may demonstrate a retropharyngeal mass. Ultrasound examination of a neck mass may be undertaken without sedation or anaesthesia. Ultrasound will help determine if a lesion is cystic or solid. An experienced ultrasonographer can comment on the internal architecture of lymph nodes and may raise suspicion of malignancy. If an abscess is identified, sonography will help with the anatomical relationships of the abscess and with surgical planning. Computed tomography (CT) scanning may require general anaesthesia in small children. Good anatomical detail is provided and it is helpful in surgical planning. Neither ultrasound nor CT2 are absolutely accurate in the diagnosis of abscess and in the presence of strong clinical features surgical exploration should be considered. [**/*] Magnetic resonance (MR) scanning rarely adds useful information in the case of acute inflammatory lesions but is very useful for vascular malformations, salivary gland and soft tissue masses (see also Chapter 102, Imaging in paediatric ENT).



SPECIFIC CONDITIONS Viral infections VIRAL UPPER RESPIRATORY INFECTIONS



Adenovirus, rhinovirus or enterovirus (Coxsackie A and B) may cause reactive lymphadenopathy. This is generally self-limiting.



Chapter 94 Cervicofacial infections in children



] 1213



INFECTIOUS MONONUCLEOSIS



This is an acute infection caused by Epstein–Barr virus (EBV). It occurs mainly in adolescence and is spread by close contact. Fever, fatigue, malaise and an exudative tonsillitis are characteristic. Cervical lymphadenopathy may be massive. Other lymphoid tissue including liver and spleen may be enlarged. There is a characteristic picture on the blood film with the presence of atypical lymphocytes. Serological tests such as Monospot or Paul Bunnel will usually confirm the diagnosis. Although the aetiology is viral, intravenous antibiotics may be needed to treat any coexistent bacterial infection. Ampicillin is contraindicated. In cases where the acute tonsillitis is associated with airway obstruction, steroids may be considered. On occasion, endotracheal intubation may be required to protect the airway until the swelling subsides.3 Cases with hepatosplenomagaly should be managed in cooperation with a paediatrician (see also Chapter 95, Diseases of the tonsil).



(a)



HIV



Infection is associated with repeated opportunistic infections. Most cases of paediatric HIV infection are acquired from the mother by vertical transmission. Acute infection may mimic infectious mononucleosis. Persistent generalized lymphadenopathy including the cervical nodes becomes a feature as the disease progresses. Weight loss and recurrent fevers occur. Following HIV infection it may be years or decades before full acquired immune deficiency syndrome (AIDS) develops. There is significant evidence of increased life expectancy with early retroviral treatment.4 Investigation and management should be in cooperation with a specialist in paediatric infectious diseases.



(b)



Figure 94.1 (a) Child with acute cervical abscess; (b) CT scan of the same child.



Many neck abscesses require surgical drainage. This may be performed by a neck incision, which is ideal for superficial lesions and the majority of deep cervical abscesses. A wide-bore needle aspiration may be adequate if the pus has coalesced and liquefied. Dental abscess



Bacterial infections ACUTE LYMPHADENOPATHY WITH SUPPURATION



Cervical abscesses Bacterial infection within a cervical lymph node may progress to cause local cellulitis and abscess (Figure 94.1). Occasionally, a solid mass of inflammatory tissue forms due to coalescence of a group of lymph nodes and this is clinically referred to as a phlegmon. The distinction between phlegmon and abscess is important, as abscesses usually require surgical drainage whereas phlegmon settle with intravenous antibiotics. The most important assessment is clinical. Abscesses are tender, usually reddened and exhibit the clinical sign of fluctuance confirming their cystic nature. Imaging is helpful but not always required.



An infected molar or premolar may cause extensive swelling extending into the face and neck and should be considered in the differential of a neck abscess (Figure 94.2). Retropharyngeal abscess This occurs typically in the young child (under two years) and is due to suppuration in the loose aggregate of lymph nodes between the pharynx and the prevertebral fascia (Figure 94.3). It is now uncommon but should be recognized because of its potential to cause fatal airway obstruction.5 The child is febrile, drooling and may adopt a characteristic posture with the neck flexed and the head extended. Retropharyngeal abscess is a surgical emergency. The child must be admitted for intravenous antibiotics, rehydration and careful observation. Early drainage under general anaesthesia with the help of a



1214 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Figure 94.2



Dental abscess.



(a)



skilled paediatric anaesthetist is indicated if there is any concern about impending airway deterioration. CT scanning will demonstrate the abscess but may itself require anaesthesia and is contraindicated if there is significant airway compromise. Drainage is via the intraoral route. Lemierre’s syndrome This is a post-tonsillitis septic thrombosis of the internal jugular vein. It is most often caused by the bacteria Fusobacterium necrophorum. It was common in the preantibiotic era. Following an episode of tonsillitis, the patient develops high spiking fevers with tenderness and fullness of one side of the neck. (See also Chapter 95, Diseases of the tonsil.) Pulmonary emboli may occur and, untreated, the mortality is high. Incidence in the UK is rising, perhaps related to changing patterns of antibiotic use.6 Contrast enhanced CT scanning is usually diagnostic and treatment with intravenous antibiotics and drainage of any abscess is usually curative. The role of anticoagulants is controversial and surgical treatment of the suppurating internal jugular vein is rarely required, although has been used in cases of repeated septic emboli.



(b)



Figure 94.3 Retropharyngeal abscess: (a) plain film showing large prevertebral shadow with compression of the airway; (b) CT scan of same child showing swelling filling the retropharyngeal space.



MICROBIOLOGY



Mycobacterial infections



Group A beta haemolytic streptococcus and Staphylococcus aureus are the most common causative organisms for suppuration in the neck. Other bacteria that may be implicated include anaerobes (19 percent); Haemophilus influenzae, Moraxella catarrhalis7 and Beta lactamase positive organisms.



Two groups of mycobacterial infections involve the neck in children. The distinction is important and can be challenging. First there are infections caused by Mycobacterium tuberculosis (TB). Secondly are a group of infections caused by other mycobacteria. Commonly



Chapter 94 Cervicofacial infections in children



referred to as atypical mycobacteria, these are most accurately termed nontuberculous mycobacteria (NTM) (see also Chapter 97, Salivary gland disorders in childhood). They include Mycobacterium avium intracellulare, Mycobacterium scrofulaceum, Mycobacterium fortuitum and Mycobacterium haemophilum.8 The typical presentation is of a painless firm enlarging mass in the neck. In NTM infection the overlying skin is frequently discoloured (Figure 94.4). In children with tuberculosis, weight loss, fever and anorexia may be present. The differential diagnosis includes lymphoma. To exclude this a tissue diagnosis is usually required. Biopsy specimens are sent for both histopathological examination and for microbiology, including staining and culture specific to mycobacterium. The typical histopatholocal appearances of tuberculosis are of caseating granuloma formation. Acid and alcohol fast staining bacilli may be seen. However, it may not be possible histopathologically to distinguish tuberculous from NTM infection. Culture of the mycobacterium may take several weeks. In this situation all patients should have a chest x-ray to identify features of tuberculosis. A tuberculous skin test is often helpful. A positive test in a nonimmunized population (such as the United States or children under 13 in most regions of the United Kingdom) is highly suggestive of tuberculous infection. Unfortunately there is an incidence of positive testing with nontuberculous infection and this test is not absolutely diagnostic for tuberculosis. In cases of doubt, antituberculous medication may be commenced pending the outcome of cultures. Genetic probing of the cultured organisms can also be used to try and make the distinction between tuberculous and NTM infection. The management of NTM lymphadenopathy is controversial. Complete surgical excision of the involved nodes brings about quick disease control and often no further treatment is necessary. [**/*] Prolonged medical therapy is an alternative.9 With medical therapy, it is not unusual for the mass to persist for up to 18 months. Treatment is discussed in Chapter 97, Salivary gland disorders in childhood.



] 1215



(a)



SINUS FORMATION AND DISCHARGE



If the nodes have been breached by the infection and there has been spread into the surrounding tissue, then there is a risk of skin breakdown and formation of a sinus. These sinuses often persist for months with troublesome discharge. Once a sinus has formed, there is a strong indication to perform surgery as the constant discharge is extremely disruptive to the life of an otherwise well child. Curettage has been advocated as a treatment for infection that has spread beyond the lymph nodes and is not amenable to complete excision. This may reduce the duration of the disease.



(b)



Figure 94.4 (a) MR scan showing large left neck mass; histology confirmed NTM; (b) Skin changes in NTM.



Primary medical treatment of nontuberculous neck masses has been advocated using either macrolide antibiotics based on some in vitro experiments10 or antituberculous



1216 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY therapy. This usually involves prolonged treatment as the organisms are notoriously resistant to antituberculous therapy. Hawkins and Clark11 showed resolution of neck mass in four of 18 cases by chemotherapy alone. CAT SCRATCH DISEASE



This is a granulomatous condition characterized by lymphadenopathy with fever and malaise. The disease has been recognized for over 50 years but only recently has the causative organism been identified as the bacteria Bartonella henselae. It is spread by close contact with cats. Patients are mainly under 20 years of age and more than 90 percent have a history of feline contact.12, 13 Serologic testing has a high sensitivity and specificity. Although there is no medical treatment confirmation of the diagnosis, performing a blood test will reassure parents who often become concerned about possible malignancy.



working with livestock or consumption of unpasteurized dairy products. It is caused by several species of the genus Brucella which are small Gram-negative bacilli. There is lymphadenopathy involving the neck and other body regions associated with fever and malaise. Diagnosis is by serology and management is with tetracycline in adults and trimethoprim-sulphamethoxazole in children. SYPHILIS



This is a spirochaete infection, now rare. Primary syphilis is associated with an ulcer (chancre) and local lymphadenopathy. Early presentation with a neck mass may occur in patients with HIV.



Fungal infections HISTOPLASMOSIS



TULARAEMIA



This is caused by infection with the Gram-negative bacillus Francisella tularensis. It rarely infects the neck and is transmitted to humans by rabbits, ticks or contaminated drinking water. Skin or mouth ulcers may be present with associated lymphadenopathy, fevers, malaise and headache. Diagnosis can be difficult but is confirmed by rising serological titres in the convalescent phase of the illness. Treatment is with streptomycin, tetracycline, aminoglycosides or chloramphenicol. ACTINOMYCOSIS



This is caused by Gram-positive non-spore forming bacteria. The most common human pathogen is Actinomyces israelii but there are several other species which can rarely be pathogenic. Approximately 50 percent of cases are cervicofacial. Actinomyces are normal commensals in the oral cavity and infections arise from a breach of the mucosas, e.g. dental extraction. The most common presentation is a slow growing painless mass near the mandible. Local lymph nodes may be involved and in a small number of cases metastasis of disease to liver or brain may occur. Untreated, the mass progresses to fibrosis and chronic suppuration with draining sinuses. A less common presentation is with an acute, warm, tender mass with fever. The presence of sulphur granules on pathological examination is suggestive but not diagnostic. If the diagnosis is suspected, special culture conditions increase the chance of culturing this organism. Most cases are treated by surgical excision followed by prolonged antibacterial therapy, usually penicillin for up to six months.



Histoplasmosis is caused by the fungus Histoplasma capsulatum. It is associated with bird droppings and acquired via airborne spores. Infection in the central USA is very common and typically asymptomatic. In immunocompromised patients, pulmonary and systemic disease may occur. Mucosal head and neck lesions may mimic squamous carcinoma. Biopsy is needed and treatment is surgical excision. CANDIDA AND ASPERGILLUS



Mucosal candidiasis is a common problem in children but neck masses caused by these infections are extremely rare and limited to immunocompromised patients. PARASITOSIS



Toxoplasmosis Infection with the parasite Toxoplasma gondii is usually through the ingestion of poorly cooked meat or ingestion of oocytes excreted in cat faeces. Cervical adenitis occurs in more than 90 percent of clinical cases.14 However, subclinical infection is not uncommon and positive serology may be found in asymptomatic individuals. The lymphadenopathy may persist for months and children may require a biopsy to rule out malignancy. Treatment is rarely required for cervical lymphadenopathy which is self-limiting but the infection responds to sulphonamides and pyrimethamine.



Noninfective inflammatory disorders



BRUCELLOSIS



SARCOID



Brucellosis is a zoonosis, i.e. an infection transmitted from animals to humans. It can be acquired through



Sarcoidosis is a chronic multisystem disorder of unknown aetiology. It mainly occurs in the second decade of



Chapter 94 Cervicofacial infections in children



life and is rare in children. It mainly causes generalized and pulmonary symptoms but may involve other parts of the body. Neck masses, parotid masses and facial nerve paresis may result. Cervical nodes are typically bilateral and nontender. The diagnosis is often suspected from the chest radiograph and can be confirmed by biopsy which shows typical noncaseating granulomas (see also Chapter 99, Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma). Treatment is usually conservative although steroids, and in some cases more potent antineoplastic agents, can be used.



Conditions which simulate lymphoma This heterogenous group of benign lymphoproliferative disorders is characterized by often prolonged unexplained cervical adenopathy which can give rise to concerns regarding lymphoma. Some authors refer to these conditions as ‘pseudolymphomata’. They include Rosai Dorfman disease, Castleman’s disease, Kawasaki syndrome and Kikuchi-Fujimoto disease.15



KAWASAKI SYNDROME



This is an acute multisystem vasculitis of unknown aetiology. It tends to affect children under five years of age and the clinical presentation is similar to many childhood infectious diseases. The diagnosis is clinical and children should have four of the five following criteria: 1. acute nonpurulent lymphadenopathy – usually unilateral; 2. erythema, oedema and desquamation of the hands and feet; 3. polymorphous exanthema; 4. painless bilateral conjunctival infection; 5. erythema and injection of the lips and oral cavity. There may be a thrombocytosis and pericardial effusion. In the subacute stage, coronary artery aneurysms develop in 15–20 percent of cases. The goal of management is to reduce inflammatory responses with antiinflammatory or gamma globulin therapy.16 The vasculitis is self-limiting but unfortunately causes permanent cardiac damage in around 20 percent of untreated patients. All patients should have an initial echocardiogram and cardiac follow-up. A mortality of 1–2 percent is associated with this disease due to the cardiac sequelae.



SINUS HISTIOCYTOSIS (ROSAI DORFMAN DISEASE)



Children present with massive cervical lymphadenopathy which is similar to infectious mononucleosis or



] 1217



lymphoma. This disease is thought to represent an abnormal histiocytic response to some precipitating cause, possibly a herpes virus or EBV.17 Fever and skin nodules may be present. Treatment is expectant but biopsy is usually required to rule out malignancy. Histopathologic examination reveals dilated sinuses, many plasma cells and marked proliferation of histiocytes.



KIKUCHI-FUJIMOTO DISEASE



This is an idiopathic disorder, first described in Japan. It is characterized by lymph gland enlargement which may occur anywhere in the body but is typically cervical. Fever chills and weight loss are common. Women and young adults are more commonly affected. The disease is self-limiting but biopsy is often performed to rule out malignancy. Histology shows a characteristic necrotizing lymphadenitis.



KEY POINTS  Lymphadenopathy is common in children and rarely requires investigation.  Viral adenitis is the most common cause of neck swelling.  Much anxiety can be caused by lesions that simulate lymphoma but careful clinical assessment supplemented by noninvasive investigations, and very rarely biopsy, will provide a diagnosis.  Immunocompromised patients pose a particular challenge and liaison with a paediatrician is essential in this group of children.



Best clinical practice [ A blood count is a useful screening investigation for haematological malignancy. [Grade B]



[ Ultrasound will help determine if a lesion is cystic or solid. An experienced ultrasonographer can comment on the internal architecture of lymph nodes and may raise suspicion of malignancy. [Grade B] [ In NTM, once a sinus has formed there is a relative indication to perform surgery as discharge is extremely disruptive to the life of an otherwise well child. [Grade C] [ Open biopsy may be the only way to exclude a lymphoma in the noninfective inflammatory lymphoproliferative disorders. [Grade C]



1218 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Deficiencies in current knowledge and areas for future research



$ $ $ $



The increasing survival of children with malignant disease, often with highly immunosuppressive chemotherapy, will produce therapeutic challenges in the management of infectious disease in the head and neck. Pooling of data on relatively uncommon conditions such as NTM may produce better treatment protocols. Imaging continues to be refined and as spiral CT scanners permit higher resolution images at lower radiation doses, imaging the architecture of cervical structures will improve. Aspiration biopsy cytology has proved disappointing in children and increasing specialization in pathology may expand the role of this technique.



REFERENCES











1. Kubba H. ‘Paediatric Neck Lumps – when does an enlarged lymph node need excising?’ in Summary of the proceedings of the V11 Annual Meeting of the EvidenceBased Management of Head and Neck Cancer. Clinical Otolaryngology. 2005; 30: 79–85. A succinct account of current indications for biopsy of neck nodes in children. 2. Stone ME, Walner DL, Koch BL, Egelhoff JC, Myer CM. Correlation between computed tomography and surgical findings in retropharyngeal inflammatory processes in children. International Journal of Pediatric Otorhinolaryngology. 1999; 49: 121–5. 3. Papesch M, Watkins R. Epstein Barr virus infectious mononucleosis. Clinical Otolaryngology and Allied Sciences. 2001; 26: 3–8. 4. Van der Poel LA, Faust SN, Tudor-Williams G. HIV-1 infection in children: current practice and future predictions. Advances in Experimental Medicine and Biology. 2004; 549: 135–48. Review of current stautus of HIV infection in children. 5. Craig FW, Schunk JE. Retropharyngeal abscess in children: clinical presentation, utility of imaging and current management. Paediatrics. 2003; 111: 1394–8. 6. Clarke MG, Kennedy NJ, Kennedy K. Serious consequences of a sore throat. Annals of the Royal College of Surgeons of England. 2003; 85: 242–4.







7. Brodsky L, Belles W, Brody A, Squire R, Stanievich J, Volk M. Needle aspiration of neck abscesses in children. Clinical Pediatrics (Philadelphia). 1992; 31: 71–6. 8. Spark RP, Fried ML, Bean CK, Figuero JM, Crowe Jr. CP, Campbell DP. Non tuberculous mycobacterial infections of the face and neck – practical considerations. American Journal of Diseases in Children. 1988; 142: 106–8. 9. Coulter JB, Lloyd DA, Jones M, Cooper JC, McCormick MS, Clarke RW et al. Nontuberculous mycobacterial adenitis: effectiveness of chemotherapy following incomplete excision. Acta Paediatrica. 2006; 95: 182–8. Large series treated by a combination of surgery and chemotherapy. 10. Rapp RP, Mc Craney SA, Goodman NL, Shaddick DJ. New macrolide antibiotics: usefulness in infections caused by mycobacterium other than mycobacterium tuberculosis. Annals of Pharmacotherapy. 1994; 28: 1255–66. 11. Hawkins, Clark J. ‘Management of the atypical tuberculous node in the neck; surgery or antibiotics? in Summary of the proceedings of the V11 Annual Meeting of the Evidence-Based Management of Head and Neck Cancer. Clinical Otolaryngology and Allied Sciences 2005. 1993; 30: 82–83. 12. Jackson LA, Perkins BA, Wenger JD. Cat scratch disease in the United States: an analysis of three national databases. American Journal of Public Health. 1993; 83: 1707–11. 13. Ridder GJ, Boedeker CI, Technau-Ihling K, Sander A. Catscratch disease: otolaryngological manifestations and management. Otolaryngology – Head and Neck Surgery. 2005; 132: 353–8. 14. McCabe RE, Brooks RG, Dorfman RF, Remington JS. Clinical spectrum in 107 cases of toxoplasmic lymphadenopathy. Reviews of Infectious Diseases. 1987; 9: 754. 15. Brown JR, Skarin AT. Clinical mimics of lymphoma. Oncologist. 2004; 9: 406–16. 16. Gersony WM. Diagnosis and management of Kawasaki disease. Journal of the American Medical Association. 1991; 265: 2699–703. 17. Levine PH, Jahan N, Murari P, Manak M, Jaffe ES. Detection of human herpesvirus 6 in tissues involved in sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman Disease). Journal of Infectious Diseases. 1992; 166: 291–5.



95 Diseases of the tonsil WILLIAM S MCKERROW



Structure and function Inflammatory disorders of the tonsil Noninflammatory disease Key points



1219 1220 1225 1226



Best clinical practice Deficiencies in current knowledge and areas for future research References



1226 1226 1227



SEARCH STRATEGY Searches were conducted of the Cochrane Library, Medline, Pubmed and Ovid databases using the key words tonsils, tonsillitis, pharyngitis, sore throat, and carrying out further subsearches for anatomy, microbiology, immunology, complications and therapy. These were complemented by hand searches of current English language texts and the following journals: Journal of Laryngology and Otology, Clinical Otolaryngology, Laryngoscope, Archives of Otolaryngology, Annals of Otolaryngology, International Journal of Pediatric Otolaryngology. The search was complemented by a hand search of the reference list in the MD thesis of Miss Ruth Capper, Doncaster, to whom I pay due acknowledgement.



STRUCTURE AND FUNCTION Structure The palatine tonsils consist of paired aggregates of lymphoid tissue. They are located in the pocket formed between the palatoglossus and palatopharyngeus muscles and the overlying folds of mucosa, which make up the anterior and posterior tonsillar pillars. With the lingual tonsils, the adenoids (see Chapter 84, The adenoid and adenoidectomy) and the diffuse aggregates of pharyngeal submucosal lymphoid tissue they make up Waldeyer’s ring, a complete circle of lymphoid tissue surrounding the entrance to the gastrointestinal and respiratory tracts. The tonsils share a common structure with lymphoid tissue elsewhere in the gastrointestinal and respiratory tracts including the adenoids and Peyer’s patches in the small intestine and within the appendix. Histologically they consist of lymphoid tissue with aggregates of lymphocytes



arranged in a follicular manner and embedded in a stroma of connective tissue. The stratified squamous mucosal covering of the tonsils extends irregular convoluted invaginations into the parenchyma forming pits or crypts. Microorganisms, desquamated epithelium and food debris are frequently present within the crypts and may be implicated in the development of acute and recurring inflammation.



Normal flora The range of organisms cultured from the tonsils both in health and disease is extremely variable, with recognized differences in bacterial flora retrieved from surface and from core samples.1, 2 The organism most commonly identified from the surface of the tonsil in disease is the group A beta haemolytic streptococcus (GABHS). Up to 40 percent of asymptomatic individuals will also have a



1220 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY culture positive for this organism.3, 4 Other surface organisms include Haemophilus, Staphylococcus aureus, alpha haemolytic streptococci, Branhamella sp., Mycoplasma, Chlamydia, various anaerobes and a variety of respiratory viruses.5, 6 [****] In a study of core samples obtained by fine needle aspiration in health and disease, core samples of normal tonsils usually failed to grow pathogenic organisms.7 In recurrent tonsillitis the samples grew a range of pathogens but the predominant organisms were Haemophilus influenzae and S. aureus. A mixed flora was also common. Beta haemolytic streptococci were less common. [****]



of the extensive ‘backup’ in the immune system.8 The precise role of bacteria and viruses in this process is also controversial. The evidence available suggests that they may act synergistically, with the presence of latent viruses (particularly Epstein–Barr virus, adenoviruses and herpes simplex) sensitizing the pathogenic bacteria frequently present on the tonsils of asymptomatic individuals.9 [**]



Function: role of the tonsils within the immune system



PRESENTATION AND FLORA



The tonsils have no afferent lymphatics. The lymphoid germinal centres are located immediately submucosally. They contain both B and T lymphocytes. B cells predominate, implying that both cell-mediated and humoral immune function is present in the tonsil and that B lymphocytes are actively produced in the tonsil. These cells have the capability to synthesize specific antibodies. They appear to be responsible for the final differentiation, induced by exposure to antigen, of B cells to principally immunoglobulin (Ig)G and IgA plasma cells. They allow positive selection of B cells according to receptor affinity for antigen and provide B cells specific for mucosal effector sites. It seems likely that they generate B cells which express polymeric IgA and which migrate to the upper respiratory tract mucosa and associated ‘front line’ mucosal surfaces. Contact with allergens in the upper respiratory tract therefore enhances local immunity and also contributes to the development of systemic immunity. Inflammatory change within the tonsils may be a manifestation of this function – hence the old adage that the function of the tonsils is ‘to develop tonsillitis’. Whether or not tonsillitis represents a swamping of defence mechanisms at local level or an exaggeration of the normal response is not known. The fact that tonsillitis is often recurrent suggests that there may be some intrinsic defect either in the tonsils themselves or in the patient’s immunity. The observation that tonsillitis tends to become less frequent with time suggests speculatively that the problem lies within the immune system rather than within the end organ itself. However, it has been shown that polymeric IgA production in tonsillar B cells is markedly reduced in children with recurrent tonsillitis. This suggests that conservation of tonsillar tissue is desirable, a point frequently emphasized by authors outwith the UK. There is no evidence that tonsillectomy per se results in impaired immunity, indeed, the converse appears more likely. In spite of a sizeable literature discussing the relationship of immune function to disease of the tonsils there is no tenable evidence of significant immune compromise, presumably as a result



INFLAMMATORY DISORDERS OF THE TONSIL Acute tonsillitis



Acute inflammatory episodes affecting the tonsils may occur as an isolated episode, or in association with a viral upper respiratory illness including generalized pharyngitis (see Figures 95.1 and 95.2). Tonsillitis may also present as part of a systemic infection such as infectious mononucleosis (see under Infectious mononucleosis below) when severe tonsil inflammation may be a prominent part of the clinical presentation. Classically, the causative organism in acute suppurative tonsillitis is



Figure 95.1 Tonsillopharyngitis (appearances could be either viral or bacterial).



Figure 95.2 Severe tonsillitis (GABHS were isolated in this case but the appearances are nonspecific).



Chapter 95 Diseases of the tonsil



GABHS. A wide variety of other organisms including anaerobes and viruses may be implicated in a clinically indistinguishable illness although whether they are primarily causative is undetermined.3, 4 [**] Epidemiological data about the true incidence of tonsillitis as opposed to pharyngitis or sore throat is confounded by the fact that there is no robust definition of tonsillitis, making assessment of the outcomes of management difficult. Studies of sore throat management frequently use bacteriological criteria for diagnosis but this does not reflect the situation in clinical practice nor is it a reliable ‘gold standard’. EPIDEMIOLOGY



Sore throat is a common reason for presentation to a primary care physician. The frequency in the UK is 0.1 consultations per capita per annum.10 [**] Not all these sore throats will be due to true tonsillitis; however, there are no case control or population studies of the epidemiology of sore throat in the recent literature, let alone any specifically looking at tonsillitis.11 Sore throat affects both sexes and all age groups but is much more common in children and during the autumn and winter months.12, 13 [**] CLINICAL EVALUATION



The diagnosis of acute tonsillitis is clinical. It is based on a history of a pyrexial illness, sore throat with a painful swallow, and the finding of pharyngeal erythema with or without tonsillar exudates and painful cervical adenopathy. Evaluating aggregated symptom complexes including tonsillar exudate, anterior cervical lymphadenopathy, the absence or presence of cough, pharyngeal erythema and the level of pyrexia has been attempted but the results of these studies are conflicting and the conclusions unconvincing. There is good evidence that the sensitivity and specificity of clinical diagnosis alone is between 20 and 50 percent.14 [***] It should not be relied upon in diagnosis of bacterial or viral aetiology. AETIOLOGY OF INFLAMMATORY DISEASE



The precise aetiology of inflammatory disease of the tonsils remains shrouded in controversy. It is widely accepted that both bacteria and viruses play a part in acute inflammation either separately or together. It is also probable that factors within the immune system of the individual patient render them susceptible to episodes of infection, which may be isolated or recurrent, but these remain unquantified and studies of immunoglobulin levels in such patients have been inconclusive. The point at which the tonsils become the cause of the disease rather than simply the site is undetermined, if indeed this ever becomes the case.



] 1221



With the exception of tonsillitis associated with infectious mononucleosis, there is no evidence that viral tonsillitis is more or less severe than bacterial tonsillitis or that the duration of the illness varies significantly in either case. DIAGNOSIS OF CAUSATIVE AGENT



Although the precise diagnosis may be of interest academically, from the practical point of view it is probably of little relevance in management. Both bacterial and viral tonsillitis tend to resolve quickly without treatment in most cases.15, 16 [****] Precise diagnosis of the causation may be of more relevance in protracted illness but is difficult and imprecise in practice. Possible options for diagnosis in clinical practice include clinical assessment, bacteriological culture or rapid antigen testing (RAT) of a throat swab for GABHS. RAT is in common use in North America but is seldom used in the UK. MICROBIOLOGICAL INVESTIGATION



Bacteriological culture of a throat swab may yield a positive culture for GABHS but this does not conclusively prove that the organism is causative. The incidence of a positive culture may be as high as 40 percent in asymptomatic carriers.3, 4 The organism has also been isolated from tonsillitis cases with no serological evidence of infection. In addition, the organisms cultured from the surface of the tonsil may vary greatly from the bacterial flora deep within the tonsillar crypts. It is by no means certain which is likely to be the more relevant to clinical symptoms.1, 2, 3, 4, 17 [***] Inevitably, there is a delay of 24–48 hours before results are available, rendering its value limited in treating a short-lived self-limiting illness. RAPID ANTIGEN TESTING



The use of RAT as an office procedure has superficial attractions, principally speed in reporting results (ten minutes), but the sensitivity measured against throat swab culture (itself a doubtful gold standard) was variable across trials at between 61 and 95 percent with specificity from 88 to100 percent.18 There may be variation between different laboratories.19 [**] The costs of both throat swab culture and RAT are not insignificant. PRIMARY MANAGEMENT



The management of acute tonsillitis is principally supportive, with the use of analgesics and adequate hydration until the symptoms subside. The choice of analgesia is governed by the severity of symptoms but the majority will find paracetamol in full dosage adequate. Nonsteroidal antiinflammatory drugs may be used in more severe cases but the higher side-effect profile makes them a second-line treatment.



1222 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY SPECIFIC TREATMENT



The relatively complex interplay between bacteria and viruses in acute tonsillitis is not fully understood and if no bacteria are cultured a viral aetiology is assumed. Even if bacterial pathogens are identified, their significance is uncertain. The average duration of an episode of acute tonsillitis seen in general practice is two to three days, with evidence that although the prescription of antibiotics will shorten the illness and may reduce the risk of sequelae this is not clinically significant.15, 20, 21 A Cochrane review of antibiotics for sore throat concluded that there was evidence of benefit in shortening the illness by a mean of one day around day three of the illness, by which time approximately 50 percent of patients will have settled spontaneously (OR 0.16; 95 percent CI 0.09, 0.26 with a positive throat swab; OR 0.65; 95 percent CI 0.38, 1.12 with a negative swab). They also found a reduction in the frequency of suppurative complications including quinsy (OR 0.16; 95 percent CI 0.07–0.35), but concluded that ‘blanket’ or routine prescription of antimicrobials was not justified.20 There is evidence that patients prescribed antibiotics for sore throat are more likely to reattend for antibiotic prescription on subsequent occasions,22 and the risks of widespread indiscriminate antibiotic prescription include genesis of resistant organisms, allergy and anaphylaxis. There are numerous studies of the efficacy of various alternative antibiotics (predominately cephalosporins) over penicillin in general and specifically in cases of proven GABHS infection. These show only marginal evidence of increased benefit insufficient to justify their use. In those patients in whom the illness shows no sign of improvement within 48–72 hours or in whom there is clinical concern because of the severity of symptoms, antibiotic therapy is appropriate and the drug of choice remains benzylpenicillin. The literature is unclear about the optimal duration of therapy but the recommendations in the British National Formulary, supported by recent evidence,23 are for a seven-day course. In cases where there is clinical concern about the severity of the illness, antibiotics should not be withheld. Recent evidence from several randomized controlled trials24 suggests that a single dose of dexamethasone as adjuvant therapy was of significant benefit in reducing pain in acute pharyngotonsillitis with no evidence of predisposition to abscess formation with the use of steroids. [****]



of GABHS – rheumatic fever and glomerulonephritis – are considered below under Immune complex disorders.



PERITONSILLAR ABSCESS



The principal complication of tonsillitis is peritonsillar abscess (quinsy) in which a collection of pus forms in the potential space between the tonsil and its bed. Prior to the formation of pus there is frequently a period of peritonsillar cellulitis without abscess formation. Clinically, the patient presents with a severe pharyngitis lateralized to one or other side, often with marked associated lymphadenopathy. There may be severe trismus limiting access for examination and treatment. This results in a severe local and systemic illness which may, if untreated, result in the spontaneous discharge of pus when the abscess points. More commonly, the course of the disease is modified by antibiotic therapy, needle aspiration or incision and drainage in hospital. The organism most commonly cultured from this pus is beta haemolytic streptococcus. Infection with other organisms such as Streptococcus viridans, S. aureus, H. influenzae and various anaerobes may also occur with anaerobes being commonly isolated if appropriate techniques are used.25, 26 [****] In severe cases airway compromise and dehydration due to the inability to swallow result, which may necessitate hospital admission for intravenous fluid therapy. This may occasionally happen in severe tonsillitis without abscess formation. In such cases immediate hospital admission and experienced clinical assessment of the airway is essential. Unskilled attempts to examine the throat may precipitate complete airway obstruction and must be avoided (Figure 95.3). Peritonsillar abscess formerly was regarded as an absolute indication for ‘interval tonsillectomy’ but most clinicians now prefer to take other factors



Complications of acute tonsillitis Acute tonsillitis may be complicated by systemic sepsis, including septicaemia and septic arthritis. GABHS may cause an acute exanthematous reaction with a macular rash – scarlet fever. The noninfective sequelae



Figure 95.3 Peritonsillar abscess (swelling is predominantly unilateral).



Chapter 95 Diseases of the tonsil



into account, for example a history of frequent tonsillitis. The majority of otolaryngologists are agreed that a second quinsy is a reasonable indication for operation. Tonsillectomy during the acute phase of the abscess has been established as a safe procedure but opinions differ about whether or not this is the most appropriate management for the condition. Protagonists emphasize the virtues of the avoidance of a second admission, convalescence from only a single episode, avoidance of loss to follow-up for interval tonsillectomy and the rapid relief of symptoms. The procedure requires emergency scheduling and the services of an anaesthetist experienced in the management of the difficult airway. The release of a large amount of pus into the posterior aspect of the oral cavity, either spontaneously or therapeutically, carries with it the risk of aspiration, especially in severely ill patients. The use of local anaesthesia prior to incision exacerbates this risk, which can be avoided, and satisfactory resolution hastened by aspiration of the abscess using a wide bore needle and syringe together with antibiotic therapy. This is now the management of choice. The antibiotics usually employed are high-dose penicillin intravenously or a cephalosporin but will be guided by bacteriology of the aspirated pus. Some clinicians also use metronidazole and this is logical as a significant proportion of peritonsillar abscesses grow anaerobes.26, 27



RETROPHARYNGEAL ABSCESS



A rare but serious complication of acute tonsillitis is retropharyngeal abscess, presenting mainly in infants and young children aged less than five years. It presents when infection has tracked into the lymphoid tissue between the posterior pharyngeal wall and the prevertebral fascia. The child is systemically ill and there may be evidence of airway compromise or an associated neck abscess. A plain x-ray of the neck may be helpful and the diagnosis can be confirmed by ultrasound or computed tomography (CT) scanning provided the child is well enough for investigation. In one study the sensitivity and specificity of CT in detecting pus was 81 and 57 percent, respectively. CT findings should, therefore, be interpreted with caution and correlated with the clinical picture in planning management. Treatment is initially high-dose antibiotic therapy with urgent incision and drainage when pus formation is suspected, under general anaesthesia with the airway protected by intubation by a skilled and experienced anaesthetist.28 [**] The abscess is usually drained perorally but occasionally external drainage via the neck may be appropriate. Rarely, tracheotomy is necessary. Retropharyngeal abscess due to tuberculosis, at one time relatively common, still occasionally occurs, more commonly in the developing world, and requires specific antibiotic therapy (see Chapter 94, Cervicofacial infections in children).



] 1223



PARAPHARYNGEAL ABSCESS



Peritonsillar and retropharyngeal abscess may occasionally be complicated by spread of infection to the parapharyngeal space with formation of a large abscess, which may require external drainage. The patient is usually severely systemically unwell, with severe trismus and possibly airway compromise. Management is urgent and if this diagnosis is suspected, imaging by ultrasound and CT scanning may be helpful in both diagnosis and in planning intervention, as noted above. The use of broadspectrum antibiotics intravenously is essential and initial therapy should include cover against streptococci and anaerobic organisms. Deep neck space sepsis may be complicated by progressive, life-threatening spread of infection including mediastinitis and even retroperitoneal sepsis. Marked neck swelling, airway compromise and female gender appear to be predictive of higher risk cases. Vigilance and a proactive management policy are thus essential in all cases (Figure 95.4).29 [**]



LEMIERRE’S SYNDROME



This is a rare and potentially fatal complication of oropharyngeal infection characterized by septic thrombophlebitis in the internal jugular vein, sometimes in association with metastatic abscesses. The organism is typically a fusiform bacillus. The condition should be considered when there is severe neck pain, septicaemia or a prolonged fulminant course in a patient with infection in the upper aerodigestive tract.30 It has also been described secondary to tympanomastoid infection. Skilled imaging will help to show the presence of a thrombus in



Figure 95.4 CT scan of parapharyngeal abscess (gas is visible in the extensive abscess cavity extending from the deep parapharyngeal space to lateral to mandible superficially).



1224 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY the neck veins (Figure 95.5). Treatment is with prolonged (six weeks) antibiotics, for example a beta-lactam with metronidazole or amoxycillin-clavulanate. Anticoagulation may be considered if there is evidence of spreading thrombophlebitis. There is a significant mortality.31, 32



IMMUNE COMPLEX DISORDERS



Acute tonsillitis and pharyngitis caused by GABHS are occasionally complicated by disease related to immune complex formation generated as part of the immune response to the infection. The two important diseases resulting from this phenomenon are acute rheumatic fever and acute glomerulonephritis. The incidence of these disorders is declining in the developed world but they are still common in certain ethnic groups, in particular, Australian Aborigines and New Zealand Maoris. The reasons for this are not clear but may be related to social as well as genetic factors. In the UK there is no evidence that treating acute tonsillitis aggressively with antibiotics has resulted in prevention or reduction in incidence of these complications. In communities where rheumatic fever is common, antibiotic therapy for sore throat may have a role in reducing the incidence of this complication.20 [**]



TONSILLITIS AND PSORIASIS



There is a probable association between acute tonsillitis due to GABHS and exacerbations of psoriasis, particularly of the guttate variety where numerous small psoriatic



lesions appear to be generated by each episode of acute tonsillitis. This appears to be an immune phenomenon. Some dermatologists and otolaryngologists advocate tonsillectomy but there is no good evidence that this relieves the condition.33 The most current (2003) review of the efficacy of antibiotic therapy and of tonsillectomy in children with psoriasis identified only one controlled trial, which found no significant effect for antibiotics in psoriasis. No controlled trials evaluating the effect of tonsillectomy were identified. Uncontrolled data suggest benefit in one-third to half of patients with tonsillectomy but 7 percent reported worsening of their condition after surgery.34 [**] RECURRENT TONSILLITIS



In a significant but unknown proportion of patients, acute episodes appear to follow a pattern of recurring infection every few weeks or months. This sequence of episodes may gradually abate but in some individuals runs a course amounting to several years. There is currently no way of predicting those individuals in whom the condition may run a protracted course, and epidemiological data are lacking to allow accurate generalizable prognostication about how long the susceptibility to recurrent infection will last. Supportive measures continue to be appropriate and management of more severe symptoms may, again, warrant antibiotic therapy. Some clinicians have used low-dose penicillin if episodes are happening close together but there is no convincing evidence of benefit from this or that any specific antibiotic is effective in aborting recurrent attacks of tonsillitis. The medical treatment of recurrent tonsillitis remains an elusive goal. SUBACUTE TONSILLITIS



Some patients follow what might be termed a subacute course where they are never free of low-grade discomfort in the throat associated with enlarged inflamed looking tonsils punctuated with acute episodes which may be either mild or severe. CHRONIC TONSILLITIS



Figure 95.5 Ultrasound scan of neck veins in Lemierre syndrome showing presence of thrombus (arrowed).



A further category of patients complain of chronic lowgrade symptoms affecting their quality of life because of throat discomfort, and the production of unpleasant smelly white or yellow debris from the tonsillar crypts. Rarely, this debris may become inspissated, calcify and form a tonsillolith, which may itself be complicated by acute sepsis. Sufferers may also complain of a feeling of low-grade ill-health, which they and their medical attendants may attribute to chronic tonsil sepsis. There is no scientific evidence addressing natural resolution and general morbidity, or the relationship to identifiable tonsillar pathology in these patients.



Chapter 95 Diseases of the tonsil



In children, historically, a very wide range of ailments including recurrent abdominal pain, general ill-health, failure to thrive and low body weight have been attributed to infection of the tonsils but there is no real scientific evidence of this. Conversely, although removal of the tonsils has been claimed to result in increased growth rate and improvement in general health, scientific evidence for this is not robust.



] 1225



swallowing results, a short course of high-dose corticosteroids is frequently employed as an adjunct, with evidence of benefit from several uncontrolled studies.35, 36 [**] Steroids should not be employed routinely in view of the risks in this disease, which is associated with lymphoproliferative disorders. Steroids should only be employed in combination with antibiotics. There is no evidence at present to support the use of antiviral medication such as acyclovir in this disease.37



Infectious mononucleosis Specific infections Acute pharyngotonsillitis is a frequent manifestation of infectious mononucleosis. This disease, commonly seen in young adults, is caused by the Epstein–Barr virus (EBV), one of the B lymphotropic human herpes viruses. In addition to its throat manifestations the disease causes severe systemic upset, haematological and liver function disturbance and splenomegaly, making the spleen vulnerable to abdominal trauma for a period of a month after cessation of symptoms. This is important in young people involved in contact sports. Diagnosis is by the Monospot blood test but this must be interpreted with caution, as the test sensitivity is o50 percent in children and 70–90 percent in adults. Confirmation is by specific EBV antibody tests. Among this group of patients symptoms are often serious, with dysphagia and occasionally dehydration from poor oral intake (Figure 95.6). Although the disease is viral, secondary infection of the tonsils happens in up to 30 percent of cases, commonly with beta haemolytic streptococci; antibiotics are routinely prescribed, usually penicillin in high dosage intravenously in those patients admitted to hospital. Occasionally, metronidazole is added to this regimen although some clinicians favour second- and thirdgeneration cephalosporins as an alternative. Ampicillin must be avoided in this condition as patients may suffer a severe allergic rash in consequence. If there is extreme swelling of the tonsils such that compromise of the airway and severe difficulty with



These include syphilis and tuberculosis. Owing to their relative rarity in the developed world they may potentially be difficult to diagnose. The lesion in syphilis classically takes the form of a ‘punched out’ ulcer but this is not invariable and, likewise, the appearances in tuberculosis may vary. The main differential diagnosis is with neoplasm and diagnosis is usually made by biopsy of the lesion.



NONINFLAMMATORY DISEASE Asymmetry A recent study of tonsil size after tonsillectomy revealed no significant difference in cases of apparent asymmetry.38 [**] Tonsil asymmetry is not an absolute indication for tonsillectomy but in such cases, particularly in adults, the clinician should be alert for the possibility of neoplasia, notably lymphoma. In clinical practice the apparent size of the tonsils per se is not well correlated with disease within them. The tonsils often swell when acutely inflamed, but there is wide variation in the degree to which the tonsils are buried in the lateral pharyngeal wall giving a false impression of the size of the structures themselves. The tonsils tend to involute during late childhood and early adult life, but in the presence of disease may remain prominent into adulthood. Indeed, some asymptomatic adults have impressively large tonsils. The rate of involution varies between individuals and sometimes this process varies between the two tonsils, giving an asymmetric appearance.



Spontaneous tonsillar haemorrhage



Figure 95.6 The throat in infectious mononucleosis (yellow slough is typical).



Occasionally, spontaneous bleeding from inflamed tonsils may take place but this is rarely serious. This may also happen in response to minor trauma in uninflamed tonsils. It may respond to cautery under local anaesthesia and, occasionally, if persistently troublesome, tonsillectomy may be indicated.



1226 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Neoplasia Asymmetry of the tonsils may give rise to the suspicion of malignancy as may an irregular or ulcerated appearance of one of the tonsils. In childhood it is not unusual for the tonsils to involute asymmetrically and, accordingly, a disparity in size of the tonsils is not always an indication for biopsy in childhood.38 However, any unusual appearances should be treated with caution and further investigated. This will normally include a full ENT examination with special attention to the neck, CT or magnetic resonance scanning of the neck, chest and abdomen if appropriate, followed by panendoscopy and excision biopsy, particularly in adults.



Lymphoma In keeping with a structure primarily of lymphoid origin, lymphoma may occur within the tonsils and although this is most likely in the adult age group, lymphoma is not unknown in childhood. The management of this disease includes detailed staging of the disease and treatment modalities may encompass surgery, radiotherapy and/or chemotherapy. The details are covered in Chapter 98, Tumours of the head and neck in childhood.



 Antimicrobial therapy has a small but measurable effect on outcome.  The tonsil rarely may be the site of presentation of lymphoma or malignant disease.



Best clinical practice [ In acute tonsillitis clinical diagnosis alone should not [ [



[



[



Squamous carcinoma In common with the rest of the upper aerodigestive tract and oral cavity, in particular, squamous carcinoma is the most common malignancy encountered. It is generally a disease of adult life and is covered in Chapter 193, Oropharyngeal tumours.



[



[



Obstructive sleep apnoea syndrome of childhood The entity of obstructive sleep apnoea of childhood is an increasingly recognized disorder resulting in impaired functioning due to exhaustion, failure to thrive and possibly leading to cardiovascular strain and the later development of pulmonary hypertension. The tonsils and adenoids are very significantly involved in the genesis of this disorder and adenotonsillectomy frequently indicated in its management. The condition and its management are covered in detail in Chapter 85, Obstructive sleep apnoea in childhood.



KEY POINTS  Acute tonsillitis is common and self-limiting.  Complications are rare.  Treatment is largely symptomatic with an emphasis on analgesia and rehydration.



[ [



be relied upon in distinguishing between a bacterial or viral aetiology. [Grade B] Both throat swab culture and RAT are of questionable value in guiding prescription of antibiotics for sore throat. [Grade B] Widespread indiscriminate antibiotic prescription promotes the genesis of resistant organisms, allergy and anaphylaxis. There is no justification for routine use of antibiotics in children with sore throat. [Grade A] In those patients in whom the illness shows no sign of improvement within 48–72 hours or in whom there is clinical concern because of the severity of symptoms, antibiotic therapy is appropriate and the drug of choice remains benzylpenicillin. A seven-day course is usually adequate. [Grade C] A single dose of dexamethasone as adjuvant therapy reduces pain in acute pharyngotonsillitis. [Grade B] Aspiration using a wide bore needle and syringe, together with antibiotic therapy, is now the management of choice for quinsy. As a significant proportion of peritonsillar abscesses grow anaerobes, metronidazole should be considered. [Grade C] Treatment of both parapharyngeal and retropharyngeal abscess is initially high-dose antibiotic therapy. When pus formation is suspected, incision and drainage under general anaesthesia, with the airway protected by intubation by a skilled and experienced anaesthetist is recommended. [Grade C] Ampicillin must be avoided in infectious mononucleosis as patients may suffer a severe allergic rash in consequence. [Grade C/D] Systemic glucocorticoids are of value in infectious mononucleosis where there is extreme swelling of the pharyngeal mucosa with impending airway compromise. [Grade C/D]



Deficiencies in current knowledge and areas for future research



$ $



The role of the tonsil in the development of immunity needs to be more fully understood. Data to shed light on the natural history of recurrent tonsillitis and define any subgroups in which natural



Chapter 95 Diseases of the tonsil



$



resolution of recurrent symptoms is likely to happen would be helpful. The precise place and value of antibiotic therapy is still not fully defined. Robust randomized controlled trials of antibiotic therapy versus placebo in sore throat/tonsillitis, both in the acute single episode situation and in recurrent tonsillitis, are required. The results should be applicable to everyday practice.







REFERENCES











1. Surow JB, Handler SD, Telian SA, Fleisher GR, Baranak CC. Bacteriology of tonsil surface and core in children. Laryngoscope. 1989; 99: 261–6. 2. Brook I, Yocum P, Shah K. Surface vs core-tonsillar aerobic and anaerobic flora in recurrent tonsillitis. Journal of the American Medical Association. 1980; 244: 1696–8. 3. Caplan C. Case against the use of throat culture in the management of streptococcal pharyngitis. Journal of Family Practice. 1979; 8: 485–90. 4. Feery BJ, Forsell P, Gulasekharam M. Streptococcal sore throat in general practice-a controlled study. Medical Journal of Australia. 1976; 1: 989–91. 5. Del Mar C. Managing sore throat: a literature review. 1. Making the diagnosis. Medical Journal of Australia. 1992; 156: 572–5. 6. Klein JO. Microbiology of diseases of the tonsils and adenoids. Workshop on tonsillectomy and adenoidectomy. Annals of Otology, Rhinology, and Laryngology. 1975; 84: 30–3. 7. Gaffney RJ, Cafferkey MT. Bacteriology of normal and diseased tonsils assessed by fine-needle aspiration: Haemophilus influenzae and the pathogenesis of recurrent tonsillitis. Clinical Otolaryngology and Allied Sciences. 1998; 23: 181–5. 8. Brandtzaeg P. Immunology of the tonsils and adenoids: everything that the ENT surgeon needs to know. International Journal of Pediatric Otorhinolaryngology. 2003; 67: S69–76. 9. Sprinkle PM, Veltri RW. The tonsils and adenoids. Clinical Otolaryngology and Allied Sciences. 1977; 2: 153–67. 10. Little P, Williamson I. Sore throat management in general practice. Family Practice. 1996; 13: 317–21. 11. Paradise JL, Bluestone CD, Bachman RZ, Karantonis G, Smith IH, Saez CA et al. History of recurrent sore throats as an indication for tonsillectomy. New England Journal of Medicine. 1978; 298: 409–13. 12. Moloney JR, John DG, Jagger C. Age, sex, ethnic origin and tonsillectomy. Journal of Laryngology and Otology. 1988; 102: 649–51. 13. Kljakovic M. Sore throat presentation and management in general practice. New Zealand Medical Journal. 1993; 106: 381–3.















] 1227



14. McIsaac W, Goel V, Slaughter PM, Parsons GW, Woolnough KV, Weir PT et al. Reconsidering sore throats. Part 1: Problems with current clinical practice. Canadian Family Physician. 1997; 43: 485–93. 15. Little P, Williamson I, Warner S, Gould C, Gantley M, Kinmonth AL. Open randomised trial of prescribing strategies in managing sore throat. British Medical Journal. 1997; 314: 722–7. 16. Del Mar C. Managing sore throat: a literature review. II Do antibiotics confer benefit? Medical Journal of Australia. 1992; 156: 644–9. 17. Capper R, Canter RJ. Is the incidence of tonsillectomy influenced by the family medical or social history. Clinical Otolaryngology and Allied Sciences. 2001; 26: 484–7. 18. White CB, Bass JW, Yamada SM. Rapid latex agglutination compared with the throat culture for the detection of group A streptococcal infection. Pediatric Infections Disease. 1986; 5: 208–12. 19. Joslyn SA, Hoekstra GL, Sutherland JE. Rapid antigen testing in diagnosing group A beta haemolytic streptococcal pharyngitis. Journal of the American Board of Family Practice. 1995; 8: 177–82. 20. Del Mar CB, Glasziou PP, Spinks AB. Antibiotics for sore throat (Cochrane review). In: The Cochrane Library, Issue 4, Oxford: Update Software, 2002. 21. Zwart S, Rovers MM, de Melker RA, Hoes AW. Penicillin for acute sore throat in children: randomised, double blind trial. British Medical Journal. 2003; 327: 1324. 22. Little P, Gould I, Williamson I, Warner G, Gantley M, Kinmonth AL. Reattendance and complications in a randomised trial of prescribing strategies for sore throat: the medicalising effect of prescribing antibiotics. British Medical Journal. 1997; 315: 350–2. 23. Zwart S, Sachs APE, Ruijs GJHM, Gubbels JW, Hoes AW, de Melker RA. Penicillin for acute sore throat: randomised double blind trial of seven days versus three days treatment or placebo in adults. British Medical Journal. 2000; 320: 150–4. 24. Wei JL, Kasperbauer JL, Weaver AL, Boggust AJ. Efficacy of single dose dexamethasone as adjuvant therapy for acute pharyngitis. Laryngoscope. 2002; 112: 87–93. 25. Lilja M, Raisanen S, Jokinen K, Stenfors L-E. Direct microscopy of effusions obtained from peritonsillar abscesses as a complement to bacterial culturing. Journal of Laryngology and Otology. 1997; 111: 392–5. 26. Brook I. Microbiology and management of peritonsillar, retropharyngeal and parapharyngeal abscesses. Journal of Oral and Maxillofacial Surgery. 2004; 62: 1545–50. 27. Brook I. The role of anaerobic bacteria in tonsillitis. International Journal of Pediatric Otorhinolaryngology. 2005; 69: 9–19. 28. Daya H, Lo S, Papsin BC, Zachariasova A, Murray H, Pirie J et al. Retropharyngeal and parapharyngeal infection in children: the Toronto experience. International Journal of Pediatric Otorhinolaryngology. 2005; 69: 81–6. 29. Wang LF, Kuo WR, Tsai SM, Huang KJ. Characterisation of life threatening deep cervical space infections: a review of



1228 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



30.



31.



32.



33.



one hundred ninety six cases. American Journal of Otolaryngology. 2003; 24: 111–7. Ajulo P, Qayyum A, Brewis C, Innes A. Lemierre’s syndrome: the link between a simple sore throat, sore neck and pleuritic chest pain. Annals of the Royal College of Surgeons of England. 2005; 87: 303–5. Jones JW, Riordan T, Morgan MS. Investigation of postanginal sepsis and Lemierre’s syndrome in the South West Peninsula. Communicable disease and public health/ PHLS. 2001; 4: 278–81. Pulcini C, Vandenbos F, Roth S, Mondain-Miton V, Bernard E, Roger PM et al. [Lemierre’s syndrome: a report of six cases] [Article in French]. La Revue de me´decine interne. 2003; 24: 17–23. Owen CM, Chalmers RJG, O’Sullivan T, Griffiths CEM. Antistreptococcal interventions for guttate and chronic plaque psoriasis (Cochrane review). In: The Cochrane Library, Issue 4, Oxford: Update Software, 2002.



34. Wilson JK, Al-Suwaidan SN, Krowchuk D, Feldman SR. Treatment of psoriasis in children: is there a role for antibiotic therapy and tonsillectomy? Pediatric Dermatology. 2003; 20: 11–5. 35. Chan SCS, Dawes PJD. The management of severe infectious mononucleosis tonsillitis and upper airway obstruction. Journal of Laryngology and Otology. 2001; 115: 973–7. 36. Tynell E, Aurelius E, Brandell A. Acyclovir and prednisolone treatment of acute infectious mononucleosis: a multicentre, double blind placebo controlled trial. Journal of Infectious Diseases. 1996; 174: 324–31. 37. Papesch M, Watkins R. Epstein-Barr virus infectious mononucleosis. Clinical Otolaryngology and Allied Sciences. 2001; 26: 3–8. 38. Spinou E, Kubba H, Konstantinidis I, Johnston A. Tonsillectomy for biopsy in children with unilateral tonsillar enlargement. International Journal of Pediatric Otorhinolaryngology. 2002; 63: 15–7.



96 Tonsillectomy WILLIAM S McKERROW AND RAY CLARKE



Introduction History of tonsillectomy The evidence-base Tonsillectomy technique Morbidity of tonsillectomy Perioperative management Alternatives to surgery



1229 1229 1229 1232 1235 1236 1237



Tonsils and variant Creutzfeld–Jakob disease Acknowledgements Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



1238 1238 1239 1239 1239 1239



SEARCH STRATEGY The opinions in this chapter are supported by a Medline search using the terms tonsil, tonsillectomy and child. Reference was also made to the Cochrane Library. This was supplemented by a hand search of current journals and by reference to the published results of the National Prospective Tonsillectomy Audit available on the Royal College of Surgeons of England website (www.rcseng.ac.uk). The UK Department of Health (www.dh.gov.uk) and the National Institute for Health and Clinical Excellence (NICE) (www.nice.org.uk) websites were also consulted.



INTRODUCTION Tonsillectomy is one of the most frequently undertaken procedures in otolaryngology. The indications remain controversial. Children are offered surgery primarily to reduce the frequency and severity of recurrent sore throats. A smaller number will have tonsillectomy – often with adenoidectomy – for relief of airway obstruction. This is discussed in Chapter 85, Obstructive sleep apnoea in childhood. This chapter focuses on the role of tonsil surgery in recurrent sore throats in children.



popularization of the operation in Victorian England. Sir Felix Semon (1849–1921) removed the tonsils from several of Queen Victoria’s grandchildren and the procedure became fashionable in the drawing rooms of the aristocracy.3 It was said to cure a variety of childhood ailments.4 Throughout much of the twentieth century, the practice – particularly in the UK and the USA – of more or less indiscriminate tonsillectomy (Figure 96.1), with a small but measurable mortality, led to hostility among paediatricians and public health physicians.5 The criteria for offering tonsillectomy have changed significantly over the years and are now much more stringent.



HISTORY OF TONSILLECTOMY Celsus in ‘De Medicina’ (14–37 AD) described ‘induration’ of the tonsils, which he advised could be removed by dissection with the fingernail. If this was not possible they could be grasped with a hook and pulled out with a ‘bistoury’.1 Improved instrumentation – particularly the snares and ‘guillotines’ used by Morrel McKenzie2 – led to



THE EVIDENCE-BASE Efficacy Despite the popularity of tonsillectomy and the enthusiasm with which it is offered and sought, high quality



1230 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Figure 96.1 Advertisement for tonsillectomy instruments, 1920s. Reprinted with permission of the John Q Adams Center for the History of Otolaryngology – Head and Neck Surgery Foundation, r 2007. All rights reserved.



evidence of efficacy is sparse.6 A systematic review concluded that there was little evidence on the use of tonsillectomy for recurrent throat infection7 and a Cochrane review concludes that there is no evidence from randomized controlled trials (RCT) to guide clinicians in formulating guidelines for surgery in children or adults.8 There is no good evidence that any benefit from tonsillectomy for recurrent sore throat in children is sustained for more than two years after surgery. The most celebrated trials make up the ‘Paradise study’ reported in 1984.9 These randomized and nonrandomized controlled trials prospectively compared groups of children who met strict entry criteria and who received surgery (tonsillectomy with, in some circumstances, adenoidectomy) or nonsurgical intervention. Participation was dependent on children having a history of seven episodes of sore throat in the year prior to the study, five or more in the preceding two years or three or more in each of the preceding three years. Well-documented clinical features had to be demonstrated to ensure eligibility. These trials showed that in these ‘severely affected children’ tonsillectomy was efficacious for two years, and possibly for a third year. [****] Efficacy was measured by a reduction in the number and/or severity of throat infections. Many of the children treated nonsurgically improved spontaneously. Such benefits as there were from surgery were slight. Critics point out that the study was poorly randomized. Children randomized to the nonsurgical group proceeded to surgery at the subsequent request of the parents. Children having adenoidectomy, as well as tonsillectomy, were included, making it impossible to exclude this intervention as relevant to the outcome of treatment. Children in the control limb of the study had active therapy, which may have minimized the morbidity. A number of randomized controlled studies of varying but generally poor quality have subsequently contributed



little. The ‘Paradise’ group has now reported on less severely affected children. Two parallel randomized controlled trials compared surgery with nonsurgical management for children with a history of recurrent episodes of throat infection, but with less stringent entry criteria than for the original studies. These reduced standards were considered to better reflect accepted practice, but were slightly more rigorous than the widely accepted USA guidelines as suggested by the American Academy of Otolaryngology Head and Neck Surgery, i.e. ‘three or more infections of tonsils and/or adenoids per year despite adequate medical therapy’.10 The incidence of throat infection was significantly lower in the surgical groups than in corresponding control groups during each of the three follow-up years. Adenotonsillectomy was no better than tonsillectomy alone. Benefits were marginal, such that the authors concluded that the benefits were outweighed in this group of children by the morbidity and risks of the operation.11 [****] A good quality open randomized controlled trial in the Netherlands – where surgical intervention rates for recurrent sore throats are considerably higher than in the UK or the USA – looked at 300 children under the age of eight years. Children who fulfilled the criteria of the original Paradise study (i.e. severely affected children) were excluded, so this study looked only at those with mild symptoms. Adenotonsillectomy was compared with watchful waiting. A number of outcome measures – including health-related quality of life and the frequency of episodes of pyrexia – were considered. The authors concluded that adenotonsillectomy had no major clinical benefits over watchful waiting in children with mild upper respiratory symptoms.6, 12 [****] It must be stressed that both of these studies looked at less severely affected children. In each case the criteria for intervention were considerably less robust than suggested by the Scottish Intercollegiate Guidance Network (SIGN)



Chapter 96 Tonsillectomy



guidelines (www.sign.ac.uk) and in the Netherlands’s study adenoidectomy was routinely combined with tonsillectomy. Perhaps inadequate knowledge of the natural history and expected outcome of recurrent sore throats in children has bedevilled the quest for a sound evidential base for current practice. Preliminary data from a study conducted during the ‘tonsillectomy embargo’ in the UK in 2001 resulting from concerns about variant Creutzfeldt-Jakob disease (vCJD) transmission (see Tonsils and variant Creutzfeld–Jakob disease) suggests that spontaneous improvement is to be expected in adults.13 [**/*] Given that overall the frequency of recurrent sore throat appears to decrease with time, there is no evidence predicting the rate of such improvement or which looks at any difference in natural history in those presenting in childhood compared with those presenting in adult life. In a randomized controlled trial in adults, 36 participants underwent immediate tonsillectomy. The control group – 34 patients – remained on a waiting list. Follow-up was short but at 90 days, streptococcal pharyngitis had recurred in 24 percent of the control group and in 3 percent of the tonsillectomy group. The authors concluded that adults with a history of recurrent streptococcal pharyngitis were less likely to have further streptococcal or other throat infections or days with throat pain if they had their tonsils removed.14 The morbidity associated with the operation must be considered and may outweigh any benefits.15 This study did not include children. Studies with a less rigorous design than the RCT provide anecdotal evidence of improvement following surgery. The overwhelming view of clinicians is that in judiciously chosen patients tonsillectomy brings about significant benefit. The results of the Scottish Tonsillectomy Audit found a high level of satisfaction among those undergoing surgery (97 percent satisfaction rate at one year).16 [**/*] There are, however, severe limitations in using satisfaction rates as an outcome measure and the follow-up was relatively short.



Current practice Tonsillectomy in children is one of the most frequently performed operations in the developed world. Although numbers are declining from a peak of over 200,000 annual tonsillectomies in the 1950s (UK), some 52,000 tonsillectomies were performed in England under the National Health Service in the year 2003/04. Data from the Department of Health Statistics show that around 30,000 of these were in children with an approximately equal gender ratio (www.dh.gov.uk). The majority are for recurrent sore throats with most of the remainder for airway obstruction. Parents in the UK initially present to a general practitioner. Referral patterns are dictated by thinking in primary care and many parents have firm



] 1231



views by the time they see an otolaryngologist. Despite the availability of guidelines and protocols in practice, a decision to undertake tonsillectomy is made by negotiation between the parent/carer and the otolaryngologist. Some units have introduced nurse-led clinics where strict protocols are used to determine eligibility for surgery. It will be illuminating to see if these protocols have a material effect on surgery rates.



Sociocultural factors There is no published evidence that the pattern of disease varies by racial, ethnic, climatic or cultural factors across the globe, but there are wide variations in the rate of tonsillectomy by geographical region.16 [***] A quoted incidence of 6.5 per 10,000 children in the UK National Health Service contrasts with 11.5 per 10,000 in the Netherlands and 5 per 10,000 in the USA. The figures for the UK may be higher as an estimated one-sixth of otolaryngology interventions occur in private practice where figures are not as widely available. In the 2005 National Prospective Tonsillectomy Audit (NPTA) (www.rcseng.ac.uk), some 13 percent of procedures were performed in independent hospitals. In the Scottish Tonsillectomy Audit, sizeable variations in tonsillectomy rate were found between different regions in the same country, from four per 10,000 in Forth Valley to ten per 10,000 in Dumfries and Galloway. In the developing world, the incidence of surgery is much lower, related perhaps to access to health care and parental expectations and preferences. Tonsillectomy is more common in large conurbations suffering higher levels of deprivation.17 [**] Recent work suggests that social class and parental smoking does not influence the number of reported episodes of sore throat or tonsillitis, but that a history of parental tonsillectomy and a family history of atopy have a positive association with the frequency with which children present with sore throat and tonsillitis.18 Parental enthusiasm for surgery varies. Some of the variation may be due to differences in demand for health care. There is evidence from the North of England and Scotland Study on Tonsillectomy and Adenoidectomy in Children (NESSTAC) that girls now present much more frequently than boys.19



Guidelines The British Association of Otolaryngologists Head and Neck Surgeons (BAO-HNS or ENT-UK, www.ent-uk.org) has published a series of reviews under the title ‘Clinical Effectiveness’ to assist clinicians with decision-making for common clinical problems.20 For tonsillectomy in children, the recommendations are in line with those of the Scottish Intercollegiate Guidance Network (www.sign.ac.uk).21



1232 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY SIGN [**] suggest what are felt to be reasonable indications based on current level of knowledge, clinical observation in the field and the results of clinical audit of outcomes. Patients should meet all the following criteria:



tonsillectomy. The results of evaluation of quality of life before and after the operation suggest that tonsil disease has a marked adverse effect on quality of life and that there is significant benefit from surgery.23, 24 [**]



   



TONSILLECTOMY TECHNIQUE



sore throats due to tonsillitis; five or more episodes of sore throat per year; symptoms for at least a year; episodes of disabling sore throat which prevent normal functioning.



As with any surgical procedure, the risks of surgery must be balanced against the potential benefit. Cognizance should be taken of whether the frequency of episodes is increasing or decreasing. The American Academy of Otolaryngologists/Head and Neck Surgeons (AAO HNS) guidelines are widely accepted by US health care insurance companies. These guidelines10 suggest that tonsillectomy should be considered for children with ‘three or more infections of tonsils and/or adenoids per year despite adequate medical therapy’. The difficulties associated with implementing guidelines and the paucity of effect of guidelines on clinical practice are well known. There is evidence that the rate of ENT interventions is not significantly altered by the introduction of guidelines. This may not be the case with strictly imposed insurance company constraints such as are used in ‘managed health care’. Moreover, when guidelines are broken, they tend to be broken more often in favour of offering rather than withholding surgery.22



Health economics Decision-making about the precise indications for tonsillectomy is complicated by the fact that each episode of tonsillitis may be treated individually and that there is a natural tendency for recurrent tonsillitis to improve with time. A relatively low cost but high volume procedure to alleviate a troublesome but nonlife-threatening condition, the operation has attracted the attention of those with an interest in the economics of health care. This in turn has focused on issues such as the value to the patient’s health of removing the tonsils, the contribution of the operation to reduction of morbidity from illness and to the increase in productivity of the individual if a tendency to recurrent debilitating illness is abolished. In children, the focus is on quality of life for parent and child and the effect on the child’s education.19 Tonsillitis is disruptive to school and family life, but tonsillectomy itself is associated with a period of morbidity. In uncomplicated cases, this usually involves at least two weeks absence from school with attendant implications for the parent or carer. Healthrelated quality of life indices and the economic aspects of sore throats in children have now been considered. A number of validated instruments evaluate benefit from childhood or adult ENT interventions, including



Introduction Recent concerns about the risk of transmission of vCJD (see Tonsils and variant Creutzfeld–Jakob disease), the incidence of primary and secondary haemorrhage and the effect of surgical technique on complications posttonsillectomy have been the subject of intense debate. The National Prospective Tonsillectomy Audit was established in 2003 on behalf of the Royal College of Surgeons of England and the BAO-HNS to look at some of these issues. Data were collected on over 50,000 patients undergoing tonsillectomy in England. Just under half were children (i.e. less than 15 years old). The results represent a comprehensive and up-to-date resume´ of current practice and provide a firm basis for recommendations to clinicians particularly in relation to surgical techniques.25 The optimal technique for removal of the tonsils must be as straightforward as possible for patient and surgeon. Given the frequency with which tonsil surgery is performed, it should be easy to teach, but above all it must be safe.26 The traditional methods for removing the tonsils are the so-called ‘cold steel’ techniques using metal instruments. In recent years, perceived advantages in terms of reduced blood loss, less pain, more rapid healing and easier surgical technique have led to the introduction of several new methods.



‘Cold steel’ tonsillectomy The most common method of ‘cold steel’ tonsillectomy is the dissection technique (Figure 96.2). In this, the tonsil is retracted medially, the mucosa overlying the tonsil capsule incised and the plane of loose areolar tissue between the tonsil and the pharyngeal musculature dissected with steel dissectors, gauze or cotton wool until the tonsil is fully mobilized (Figure 96.3). Blood vessels traversing the plane of dissection are dealt with either by ligature or diathermy as required. An alternative method of ‘cold steel’ tonsillectomy is the guillotine technique, whereby the tonsil is amputated using a specially designed guillotine device and haemostasis, secured as necessary by one of the above methods. Of these two techniques, traditional dissection remains the most frequently used. Arguably the advantage of speed in the guillotine technique is outweighed by the risk of leaving tonsil



Chapter 96 Tonsillectomy



Figure 96.2



] 1233



Steps in dissection tonsillectomy.



Figure 96.3 Tonsillectomy set as used by one of the authors (RWC).



tissue behind. Blood loss during tonsillectomy can be considerable and may constitute over 10 percent of total circulating blood volume.27 Cold steel dissection may be combined with diathermy to aid haemostasis, but many surgeons prefer ties or swabs. In the NPTA, the use of ‘cold steel’ dissection without diathermy was associated with the lowest haemorrhage rate. ‘Haemorrhage’ was defined as a bleed that prolonged the patient’s hospital stay, required blood transfusion, a return to the operating theatre, or in the case of ‘secondary’ haemorrhage readmission to hospital. The haemorrhage rate for various techniques is shown in Table 96.1. The NPTA recommendation is that all trainee surgeons should become competent in cold steel dissection with ties before considering alternative techniques.



Diathermy tonsillectomy In recent years, the technique has evolved of using diathermy not only as an aid to haemostasis when the



tonsil has been delivered, but to dissect the tonsil from its bed. This has the obvious advantage from the point of view of both operator and patient, particularly if a child, of reducing intraoperative blood loss to a minimum. Various claims and counterclaims have been made regarding the advantages and disadvantages of this technique, the most common alternative to traditional cold steel tonsillectomy. Major concerns about an increase in secondary haemorrhage rate with diathermy, following the introduction of disposable instruments for tonsillectomy during 2001, led to detailed retrospective audit of haemorrhage rates throughout the UK. This confirmed an incidence of secondary haemorrhage as high as 16.8 percent with diathermy in one published audit.28 A Cochrane review of dissection versus diathermy for tonsillectomy found that only two of the studies that addressed this were sufficiently robust for analysis. These demonstrated reduced intraoperative bleeding but increased pain in the diathermy group, with no difference in secondary haemorrhage rate.29 [**] These concerns were to the fore when the NPTA was set up. Surgeons use



1234 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY a variety of diathermy techniques and specifications often based on individual preference and on what is available in a particular institution. Power settings vary and both bipolar and monopolar equipment is used. In the NPTA, postoperative bleeding was more frequent with diathermy



Table 96.1



than with cold steel alone, (Table 96.1) and was particularly common with monopolar diathermy such that the authors recommended that surgeons using this method should consider changing. Table 96.2 lists the recommendations of the NPTA in full.



Surgical techniques and primary and secondary postoperative haemorrhage rates.



Surgical technique (n = 33,921)



Cold steel and ties/packs Cold steel and monopolar diathermy Cold steel and bipolar diathermy Monopolar diathermy forceps Bipolar diathermy forceps Bipolar diathermy scissors Coblation Other



Primary tonsillar haemorrhage rate (%)



Secondary tonsillar haemorrhage rate (%)



0.8 0.5 0.5 1.1 0.4 0.6 1.0 0.7



1.0 2.4 2.3 5.5 4.3 4.6 3.6 3.6



Reprinted from the National Prospective Tonsillectomy Audit, available on the Royal College of Surgeons of England website (www.rcseng.ac.uk), Table 4.6, with permission.



Table 96.2



Recommendations of the National Prospective Tonsillectomy Audit (2005).



 When a patient is counselled for surgery, the risk of tonsillectomy complications, and in particular postoperative haemorrhage, should be carefully explained to the patients/parents.  This risk should be quantified, preferably using the surgeon’s own (or department’s) figures.  National figures can be used, but this should be made clear to patients.  Surgeons using monopolar diathermy should consider using an alternative technique. There are no advantages to using this instrument over other methods.  All trainee surgeons should become competent in cold steel dissection and haemostasis using ties, before learning other techniques in tonsillectomy.  Emphasis must be placed on teaching the correct use of, and the potential hazards of, diathermy and other ‘hot’ techniques. Checks should be made of the power settings before starting the operation.  Inexperienced trainees must be supervised by a more senior surgeon until competency has been achieved. This recommendation is in agreement with the College’s Standards on Good Surgical Practice issued in 2002.  Irrespective of seniority and experience, surgeons who wish to start using new techniques, such as coblation, should undergo appropriate training.  All ENT departments should have regular morbidity and mortality meetings to monitor adverse incidents affecting patient outcome. For tonsillectomy, data should be presented by the surgeon, indicating the technique used for dissection and haemostasis and power settings if applicable, type of instrument used, and any difficulties encountered.  It is the responsibility of the surgeon, and if appropriate his trainer, to follow up any identified problems appropriately.  Use of single-use instruments should also be recorded, especially for cold steel dissection.  There is an urgent need for new standards for diathermy machines so that the amount of power used is obvious to the user. Manufacturers of diathermy machines should be encouraged to produce machines with information on the total amount of energy delivered to patients.  Hospitals should encourage the use of machines that provide clear information on power settings.  Manufacturers of single-use instruments should be encouraged to improve the quality of the instruments.  There is a need for further laboratory and clinical research to investigate the influence of diathermy and other ‘hot’ techniques on an open wound, such as the tonsillar bed.  In particular, there is a need to investigate the dose–response relationship between power used and complications. Reprinted from the National Prospective Tonsillectomy Audit, available on the Royal College of Surgeons of England website (www.rcseng.ac.uk), page vii, with permission.



Chapter 96 Tonsillectomy



] 1235



Coblations tonsillectomy



‘Capsulotomy’ techniques



This relies on the use of a specially designed bipolar electrical probe, which both coagulates and cuts the tissues as it develops the dissection plane between tonsil and capsule. The probes or ‘wands’ are single use and there is a cost consideration. The technique involves the use of the operating microscope.30, 31 Coblation has gained widespread acceptance in the United States. Good results have been reported, but the technique seems to be highly operatordependent. Postoperative bleed rates were unacceptably high in the NPTA. Some of these bleeds may reflect lack of familiarity with the technique; enthusiasts have reported bleed rates that are comparable with those obtained using cold-steel dissection. It is suggested that postoperative pain is less than with conventional dissection but one RCT has cast doubt on this and shown that morbidity was less with cold steel dissection.32 There are no good controlled studies comparing coblation with cold steel dissection without the addition of diathermy; current evidence is inadequate to justify its introduction in preference to cold steel dissection with ties and/or packs. The National Institute for Health and Clinical Excellence (NICE) advise that current evidence is sufficient to support the use of electrosurgery (diathermy and coblation) for tonsillectomy, but advise surgeons to ensure that they are appropriately trained and to make patients and parents/ carers aware of the risk of haemorrhage after these techniques.33



The above techniques are designed to remove the entire palatine tonsil. A preference for ‘office-based’ surgery particularly in the USA has led to the popularization of techniques to ablate a part of the tonsil, usually leaving the capsule intact. These ‘tonsillotomy’ techniques include thermal tissue ablation using radiofrequency volumetric reduction (RFVR) using a customized probe and surface laser surgery. They are widely used but have not been subject to good quality randomized controlled trials. They may be considered when tonsillectomy is undertaken in the very young where it may be desirable to leave some functioning lymphoid tissue.37



Ultrasonic dissection Ultrasonic dissection uses an oscillating blade, which acts as both a cutting and coagulating device. Enthusiasts for the ‘harmonic scalpel’ have claimed advantages over conventional techniques,34, 35 in terms of reduced pain and general morbidity, but evidence remains unconvincing.



MORBIDITY OF TONSILLECTOMY Psychosocial morbidity and pain Morbidity from the operation is significant. It includes both the expected adverse consequences and the possible complications. For most children it is their first experience of parental separation. Some show behaviour changes, such as attention seeking, temper tantrums and night waking. Depression and severe anxiety were commonplace but most units are now aware of the adverse consequences of separation and permit parents to accompany the child to the operating theatre and remain with him/her throughout the in-patient stay.38, 39 Pain and dysphagia are normal in the early postoperative period. Most children require at least a week to resume normal functioning and an average return to school or work time is one to two weeks.16 This is mainly due to pain preventing a return to normal diet and occasionally vomiting in the early postoperative period due either to the aftereffects of the anaesthetic or to the effect of swallowed blood on the stomach. Much attention has focused in the literature on the management of these symptoms, one of the goals being early discharge from hospital.



Mortality Laser tonsillectomy With the advent of the laser as a surgical tool, the use of this method of dissecting out the tonsil has been advocated as having advantages in terms of reduction of bleeding, postoperative pain and more rapid healing. Several studies have failed to confirm these advantages. There is convincing evidence that the rate of secondary haemorrhage and late postoperative pain is significantly greater with laser.36 [****] The evidence would suggest that this technique cannot be recommended as an alternative to conventional tonsillectomy on the grounds of cost, morbidity and safety.



The complications of tonsillectomy may be divided into those associated with the anaesthetic and those directly associated with the operation itself. As the operation is normally performed on children and young otherwise fit adults, for the majority of patients the risk of the short anaesthetic required for tonsillectomy is small. There are risks inherent in anaesthesia in very young children and for this reason tonsillectomy is seldom performed in children before the age of two years, even in the unlikely event of them fulfilling the above criteria for consideration for surgery. In general, tonsillectomy is not frequently indicated in children under the age of four



1236 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY years, the exception being children with obstructive sleep apnoea syndrome. There are some children who will be at increased risk from anaesthetic complications, principally related to the airway. These will include those with first and second branchial arch syndromes, Pierre–Robin sequence and Goldenhar syndrome and also those with Down syndrome, where airway compromise may occur. In the latter group, the additional hazard of atlantoaxial instability presents a further risk. Since its instigation in 1995, the Royal College of Surgeons in England audit on surgical mortality40, 41 records no deaths occurring from tonsillectomy, but there were two deaths reported in the British lay press in 2001. Extrapolating from US data42 and from Department of Health statistics, the potential mortality from tonsillectomy has been calculated at one per 24,000 operations26 or one in 16,000 to one in 35,000.19



while 1.44 percent required return to theatre for arrest of haemorrhage. In other quoted series, the rate of ‘secondary’ haemorrhage has been as high as 16.8 percent. Secondary haemorrhage can occur any time until the tonsil bed has healed, which may take as long as two weeks. It is attributed – on sparse evidence – to infection in the granulating tonsil bed, often with streptococcal organisms. Antibiotic therapy with penicillin, pending bacteriological guidance is appropriate. The reasons for this wide variation in bleed rates is not clear; it may be related both to the technique used and to the experience of the operator, although in the NPTA there was no significant difference in rates between trainees and more experienced surgeons.



Perioperative complications



Postoperative ‘infection’ is sometimes diagnosed in primary care. The presence of severe halitosis is the most prominent feature, usually associated with fever. It is almost certainly overdiagnosed as the appearance of the normally healing tonsil bed is of yellow/grey slough, which may give the impression of pus. Patients are routinely given advice to ensure a good intake of fluids and solids postoperatively on the grounds that this will clean the tonsil beds and avoid infection, but the evidence for this is scanty.



Occasionally, patients may experience temporomandibular joint dysfunction due to the mouth being opened too widely with the tonsillectomy gag. Small tears at the commissures of the mouth and cracks of the lip vermillion may be avoided by careful placement of the gag and the use of an emollient jelly on the lips prior to the commencement of surgery. Dissection outside the pharyngeal musculature may traumatize adjacent structures, such as the glossopharygeal nerve, the pharyngeal venous plexus and the carotid sheath, but such events are extremely rare. Nontraumatic atlantoaxial subluxation (Grisel syndrome) may occur secondary to any inflammatory process in the upper neck. It is thought to be due to infection in the periodontoid vascular plexus that drains the region, bringing about paraspinal ligament laxity. It is described following both tonsillectomy and adenoidectomy. Treatment consists of cervical immobilization, analgesia and antibiotics to reduce the risk of neurological deficit.43



Haemorrhage The main early complication is haemorrhage. This is defined as primary (within the first 24 hours postoperatively) or secondary, i.e. occurring after 24 hours and during the phase of healing of the tonsil bed. The use of the term ‘reactionary’ is confusing. It is not uniformly defined in published work and its use is not advised. Data from the NPTA on haemorrhage have already been presented. In a recent audit of 5646 tonsillectomy complications in Scotland in the years 2001 and 2002 (McKerrow, unpublished data), the rate of primary haemorrhage was quoted at 0.56 percent, rather lower than in some other published work and lower than that quoted in the NPTA. The readmission rate overall was 4.57 percent,



Infection



Late complications Late complications are generally associated with the scarring that inevitably takes place as the tonsil bed heals. On rare occasions, this may result in impairment of palatal functioning with velopalatine insufficiency. Even more infrequent is the complication of nasopharyngeal stenosis, which very rarely occurs after adenotonsillectomy. Pharyngoplasty may be required to minimize the effects of these complications. Concern has been raised about a possible increase in incidence of Hodgkin’s lymphoma in adults who had tonsillectomy in childhood,44 but a subsequent epidemiological study has not confirmed this.45



PERIOPERATIVE MANAGEMENT The main arms of management to minimize morbidity are skilled anaesthesia, analgesia and antiemetic therapy. There has also been interest in the use of steroid therapy pre- or postoperatively.



Anaesthesia Strict protocols are frequently used in the United States, while in the UK the technique varies with the preference



Chapter 96 Tonsillectomy



of the individual anaesthetist. Total intravenous anaesthesia with propofol and remifentanil is associated with fast ‘wake up’ and little ‘hangover’. Propofol has the added merit of being an antiemetic agent. Deep inhalational intubation obviates the need for muscle relaxants, speeds reversal and avoids the use of potentially emetic agents, such as neostigmine. The total duration of anaesthesia should be as brief as is practicable, certainly less than 30 minutes.



] 1237



Steroid therapy Peroperative glucocorticoids are widely used in the US but have gained less acceptance in the UK. A Cochrane review showed that a single intravenous dose of dexamethasone was an effective, relatively safe and inexpensive treatment for reducing morbidity from paediatric tonsillectomy.49 [****] Dexamethasone has the added advantage of being an antiemetic. Many units use a single dose of 2–4 mg.



Analgesia Antibiotics Adequate analgesia is important in the immediate postoperative phase. Narcotics have a potent emetic effect and should be used with caution if at all. A single dose of narcotic may be administered in the recovery phase and codeine may be used in the early postoperative period, but subsequent to this, paracaetamol is the drug of choice in the UK on the grounds of safety and efficacy. For some children this may not be adequate and a nonsteroidal antiinflammatory drug (NSAID) may be needed. There were concerns that the effect of these drugs on platelet adhesion might increase bleeding from the tonsil bed, but a recent metaanalysis found no such risk and a significant reduction in postoperative nausea and vomiting when compared with other analgesics notably narcotics.46 [****] Aspirin should not be used in children because of the risk of Reye syndrome. LOCAL ANAESTHESIA INFILTRATION



There has been recurring enthusiasm for the use of local anaesthetics infiltrated into the tonsil beds to reduce postoperative pain. The effect of injecting long- and short-acting local anaesthetics pre-, per- and postoperatively into the tonsil beds have all been studied. The results, including a Cochrane review, suggest that there is no current evidence of significant benefit from the use of these techniques and that further study is necessary.47



Antiemetic drugs Some patients suffer severely from vomiting postoperatively. In many cases this is related to swallowed blood irritating the stomach. In these cases, the symptoms usually settle after a single vomit of the stomach contents, but in those with prolonged vomiting parenteral antiemetic therapy may be indicated. Traditional antiemetics such as cyclizine and prochlorperazine may be adequate, but several studies suggest that the newer antiemetic ondansetron may have a useful role.48 [****]



A number of studies purport to assess the benefit of prophylactic antibiotics in minimizing postoperative morbidity. The very small reduction in time to resumption of normal activities apparent in some studies must be balanced against the potential side effects of antibiotics. Current evidence does not support their routine use.50, 51, 52 [****]



Day-case surgery In recent years, there has been a trend towards encouraging day-case and in some areas ‘23-hour admission’ tonsillectomy. The two main reasons for detaining posttonsillectomy patients in hospital are safety with regard to haemorrhage and to manage morbidity from pain and vomiting. The risk of primary bleeding diminishes to close to zero six hours after surgery so in theory the patient can be safely discharged at that time.53 [**] In most centres this is done according to a strict protocol to ensure that the patient is adequately supervised in the home, has access to a telephone and reasonable proximity to hospital. Frequently a visit by an appropriately trained nurse on the first postoperative day is arranged. For this regime of management to prove satisfactory, adequate analgesia is essential. Although there is evidence that this mode of management is safe, there is considerable variation in its acceptability to parents and carers.54 Currently, only a small proportion of tonsillectomies in England and Wales are done on a day-case basis (www.dh.gov.uk). The development of a successful and safe day-case tonsillectomy service requires the provision of a dedicated team of surgeon, anaesthetist, ward nurses and community backup so that the issues of analgesia, antiemesis and safety can all be addressed adequately.55



ALTERNATIVES TO SURGERY The concept that tonsillectomy surgery might be avoided by some alternative management regime is attractive to



1238 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY patients and health economists alike. Unfortunately this goal has proved elusive so far, but there are no conclusive studies that have looked at this issue in detail. There is no evidence that antibiotics will prevent recurrent tonsillitis. The most commonly used approach is to manage each episode independently and to await natural remission of the recurrent problem. It is worth bearing in mind that lengthy and indiscriminate courses of antibiotics may not only contribute to the development of antibiotic-resistant organisms, but may also provoke allergic reactions to useful antibiotics and occasionally anaphylactic shock with a fatal outcome. Unpleasant side effects of broadspectrum antibiotics, including vomiting and diarrhoea, are also frequent and relevant to the discussion. These issues are considered in more detail in Chapter 95, Diseases of the tonsil.



TONSILS AND VARIANT CREUTZFELD–JAKOB DISEASE Variant Creutzfeld–Jakob disease is a rare, fatal neurodegenerative disease. Histologically, it is characterized by intracellular vacuolation in nervous tissue, leading to a progressive spongiform encephalopathy. Tonsil biopsy is one of the methods of diagnosing vCJD and is less hazardous than brain biopsy, the main alternative. All possible precautions must be taken during this procedure to avoid any risk of cross infection and all equipment in direct contact with the patient must be destroyed by incineration. The biopsy material must be regarded as biohazardous and treated with appropriate precautions. Some 150 cases have been identified since the disease was described in 1996. The causative agent is almost certainly the prion protein PrP which is also responsible for bovine spongiform encephalopathy (BSE). Involvement of the lymphoreticular system is a defining feature of vCJD and immunohistochemical accumulation of prion protein in the lymphoreticular system remains the only technique that has been shown to predict neurological disease reliably in animal prion disorders. During the overt clinical stages of the disease, tonsillar tissue is invariably infected with the prion. The infecting agent is thought to have entered the human food chain because beef was contaminated by feeding sheep offal to cows, a practice that became widespread in Britain in the 1970s. Cases of BSE in cattle were described from 1985 onwards and reached epidemic levels in 1992 (Bates, personal communication). Stringent controls in the beef industry now prevent contaminated meat from entering the food chain.56 Nevertheless, there were concerns that the PrP might be present in the tonsils of ostensibly healthy individuals incubating the disease, or could have been present in those who had tonsillectomy during the long incubation



period (up to 20 years) when the prion may have been in the lymphoreticular system before the clinical features were manifest.57 Poor quality preparatory techniques for tonsillectomy instruments may result in particulate matter remaining adherent to them. Normal sterilization methods do not destroy prions. Iatrogenic transmission of prion disease – but not vCJD – has been reported and this raised the possibility that prions could be passed from one patient to another on contaminated tonsillectomy instruments.58 Transmission of vCJD has not been shown to occur in this way but the fear that it might do so led to the introduction of single-use instruments for tonsillectomy throughout the UK in early 2001. This policy was introduced by the Department of Health and the Chief Medical Officer. An improvement in the standards of decontamination of surgical equipment was held to reduce the risk of vCJD to acceptable levels, not so much by destroying the prion but by minimizing the amount of particulate matter (i.e. tissue) adhering to surgical instruments after cleaning. The perceived theoretical risk of prion transmission was very small and much less than the real risk of bleeding, thought to be increased by poor quality disposable instruments. Although the Department of Health rescinded the embargo on reusable tonsillectomy instruments in late 2001, single-use instruments continue to be used in Scotland and the issue remains controversial.59 Uncertainities relate to the incubation period of vCJD, the size of the ‘inoculum’ which puts patients at risk of developing overt disease, whether transmission of prion is possible by direct contact with tonsillectomy instruments and whether prion introduced in this way causes systemic or central nervous system infection. There are continuing concerns about the adequacy of sterilization of equipment. The prevalence of prion infection in the tonsils of healthy individuals is unknown. The National Anonymous Tonsil Archive (NATA) study, a large-scale screening programme of fresh tonsil tissue (www.hpa. org.uk), may clarify this, but despite the investment of significant resources in this project there are doubts that the power of the study will be adequate to give useful results. Other unknowns are whether healthy tissue can be inoculated with prion by direct transmission, whether direct contamination if it did occur in this way would result in disease, and what proportion of healthy individuals who have evidence of the prion in the tonsil tissue will go on to develop disease.60



ACKNOWLEDGEMENTS The authors would like to thank Professor Janet Wilson, Newcastle, for kindly reading the manuscript and commenting on it, and Grant Bates, Oxford, for assisting with the section on vCJD.



Chapter 96 Tonsillectomy



KEY POINTS  Tonsillectomy is one of the most commonly performed surgical procedures in the developed world.  The evidence-base for current practice is poor.  Tonsillectomy rates vary considerably in different populations. These variations are not accounted for by variations in disease prevalence.  Improvements following surgery are particularly small in less severely affected children. The morbidity of surgery usually outweighs any potential benefit in this group.  There is no evidence that the benefits of tonsillectomy for recurrent sore throat are prolonged beyond two years.  The operation is associated with significant morbidity, which may be minimized with careful perioperative management.



$ $



] 1239



consider outcomes other than reduction in number of episodes of tonsillitis. The North of England and Scotland Study on Tonsillectomy and Adenoidectomy in Children should help provide a sound evidencebase to guide clinical practice. The morbidity of tonsillectomy remains significant. The optimal strategy for control of pain and emesis post-tonsillectomy still requires to be defined. The use of newer antiemetic agents and peroperative glucocorticoids seems set to increase. The risk, if any, of prion transmission by surgical instruments remains undefined and has important implications for all types of surgery. Work is necessary to quantify the risk and to take appropriate steps to minimize any hazard to patients. The National Anonymous Tonsil Archive currently being collated under the aegis of the Department of Health in England aims to collect data on the incidence of vCJD prion in normal tonsil tissue and may assist in quantifying the risks.



Best clinical practice [ ‘Cold steel’ dissection tonsillectomy is widely



[ [ [ [ [ [



available and associated with the lowest postoperative haemorrhage rates in the hands of most surgeons. [Grade B] Surgeons in training must master traditional ‘cold steel’ dissection before considering alternative techniques. [Grade B] Adequate analgesia is essential in the postoperative care of children following tonsillectomy. [Grade D] Antiemetic agents, such as ondansetron, should be considered to reduce postoperative morbidity. [Grade D] There is now a sufficient body of evidence supporting the use of perioperative glucocorticoids to justify considering their use as routine. [Grade A] In secondary haemorrhage, surgery is rarely needed. Bleeding usually settles with antibiotic therapy alone. [Grade C] Continuing audit is essential; surgeons should familiarize themselves with the findings and recommendations of the NPTA. [Grade D]



REFERENCES







 Deficiencies in current knowledge and areas for future research



$



High quality randomized controlled trials of tonsillectomy versus standardized conservative management are essential to define the optimal management of recurrent tonsillitis. The trials need to







1. Spencer W (trans.). De Medicina. London: Loeb Classical Library, 1935; ii: 12. 2. MacKenzie M. A manual of diseases of the throat and nose: including the pharynx, larynx, trachea, oesophagus, nasal cavities and neck. London: J & A Churchill, 1880. 3. Harrison D. Felix Semon. London: Royal Society of Medicine Press, 2000: 1849–921. 4. Wilson TG. Diseases of the ear nose and throat in children. London: William Heineman, 1955. 5. Bolande RP. Ritualistic surgery – circumcision and tonsillectomy. New England Journal of Medicine. 1969; 280: 591–6. 6. van Staaij BK, van den Akker EH, van der Heijden GJ, Schilder AG, Hoes AW. Adenotonsillectomy for upper respiratory infections: evidence based? Archives of Diseases in Childhood. 2005; 90: 19–25. Summarizes the findings of the Netherlands RCT and a comprehensive review of current evidence. 7. Marshall T. A review of tonsillectomy for recurrent throat infection. British Journal of General Practice. 1998; 48: 1331–5. 8. Burton MJ, Towler B, Glasziou P. Tonsillectomy versus nonsurgical treatment for chronic/recurrent acute tonsillitis. Cochrane Database System Reviews. 2007; 7: CD001802. 9. Paradise JL, Bluestone CD, Bachman RZ, Colborn DK, Bernard BS, Taylor FH et al. Efficacy of tonsillectomy for recurrent throat infection in severely affected children. New England Journal of Medicine. 1984; 310: 674–83. The earliest RCT on tonsillectomy and still widely quoted as a baseline for current practice.



1240 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY















10. American Academy of Otolaryngology Head and Neck Surgery. Clinical indicators compendirum. Alexandria, VA: American Academy of Otolaryngology Head and Neck Surgery, 2000: 19. 11. Paradise JL, Bluestone CD, Colborn DK, Bernard BS, Rockette HE, Kurs-Lasky M. Tonsillectomy and adenotonsillectomy for recurrent throat infection in moderately affected children. Pediatrics. 2002; 110: 7–15. The most recent findings from the Paradise group. 12. van Staaij BK, Van den Akker EH, Rovers MM, Hordijk GJ, Hoes AW, Schilder AGM. Effectiveness of adenotonsillectomy in children with mild symptoms of throat infections or adenotonsillar hypertrophy: open randomised controlled trial. British Medical Journal. 2004; 329: 651–4. 13. McKerrow et al 2002. Unpublished data. 14. Alho O-P, Koivunen P, Penna T, Teppo H, Koskela M, Luotonen J. Tonsillectomy versus watchful waiting in recurrent streptococcal pharyngitis in adults: randomised controlled trial. British Medical Journal. 2007; 334: 939. 15. Little P. Recurrent pharyngo-tonsillitis. British Medical Journal. 2007; 334: 909. 16. Blair RL, McKerrow WS, Carter NW, Fenton A. The Scottish tonsillectomy audit. The Audit Sub-Committee of the Scottish Otolaryngological Society. Journal of Laryngology and Otology. Supplement. 1996; 20: 1–25. 17. Bisset AF, Russell D. Grommets, tonsillectomies and deprivation in Scotland. British Medical Journal. 1994; 308: 1129–32. 18. Capper R, Canter RJ. Is the incidence of tonsillectomy influenced by the family medical or social history. Clinical Otolaryngology. 2001; 26: 484–7. 19. Bond J, Wilson J, Eccles M, Vanoli A, Steen N, Clarke R et al. Protocol for north of England and Scotland study of tonsillectomy and adeno-tonsillectomy in children (NESSTAC). A pragmatic randomised controlled trial comparing surgical intervention with conventional medical treatment in children with recurrent sore throats. BMC Ear, Nose and Throat Disorders. 2006; 6: 13. An upto-date account of the current dilemmas and an outline of the NESSTAC study. 20. British Association of Otorhinolaryngologists Head and Neck Surgeons. Statements of clinical effectiveness. Otolaryngology. 1998. 21. Scottish Intercollegiate Guidelines Network. Management of sore throat and indications for tonsillectomy. SIGN publication No. 34. Available from: http://www.sign.ac.uk. 22. Donaldson L, Hayes JH, Barton AG, Howel D. The development and evaluation of best practice guidelines: tonsillectomy with or without adenoidectomy. Report to the Department of Health. University of Newcastle upon Tyne: Department of Epidemiology and Public Health, 1994. 23. Goldstein NA, Fatima M, Campbell TF, Rosenfeld RM. child behavior and quality of life before and after tonsillectomy and adenoidectomy. Archives of Otolaryngology, Head and Neck Surgery. 2002; 128: 770–5.







24. Kubba H, Swan IRC, Gatehouse S. The Glasgow Children’s Benefit Inventory: a new instrument for assessing healthrelated benefit after an intervention. Annals of Otology, Rhinology, and Laryngology. 2004; 113: 980–6. 25. The Royal College of Surgeons of England. National Prospective Tonsillectomy Audit 2005. Available from: www.rcseng.ac.uk/rcseng/content/publications/docs/ national_prospective 26. Brown P. How safe is paediatric tonsillectomy? International Journal of Paediatric Otolaryngology. 2006; 70: 575–7. A summary of the main findings of the NPTA and a succinct account of the vCJD saga. 27. Alatas N, San I, Cengiz M, Iynen I, Yetkin A, Korkmaz B et al. A mean red blood cell volume loss in tonsillectomy, adenoidectomy and adenotonsillectomy. International Journal of Pediatric Otorhinolaryngology. 2006; 70: 835–41. 28. Maini S, Waine E, Evans K. Increased post-tonsillectomy secondary haemorrhage with disposable instruments: an audit cycle. Clinical Otolaryngology. 2002; 27: 175–8. 29. Pinder D, Hilton M. Dissection versus diathermy for tonsillectomy. The Cochrane Library. Oxford: Update Software, 2002. 30. Timms MS, Temple RH. Coblation tonsillectomy: a double blind randomised controlled study. Journal of Laryngology and Otology. 2002; 116: 450–2. 31. Belloso A, Chidambaram A, Morar P, Timms MS. Coablation tonsillectomy versus dissection tonsillectomy: postoperative hemorrhage. Laryngoscope. 2003; 113: 2010–3. 32. Philpott CM, Wild DC, Mehta D, Banerjee AR. A doubleblinded randomized controlled trial of coblation versus conventional dissection tonsillectomy on post-operative symptoms. Clinical Otolaryngology. 2005; 30: 143–8. 33. National Institute for Health and Clinical Excellence. Procedure for Interventional Procedure Guidance Credentialing No. 150, 2005. Available from: www.nice.org.uk. 34. Sood S, Corbridge R, Powles J, Bates G, Newbegin CJ. Effectiveness of the ultrasonic harmonic scalpel for tonsillectomy. Ear Nose and Throat Journal. 2001; 80: 514–6, 518. 35. Willging JP, Wiatrak BJ. Harmonic scalpel tonsillectomy in children: a randomized prospective study. Otolaryngology – Head and Neck Surgery. 2003; 128: 318–25. 36. Auf I, Osborne JE, Sparkes C, Khalil H. Is the KTP laser effective in tonsillectomy? Clinical Otolaryngology. 1997; 22: 145–6. 37. Nelson LM. Radiofrequency treatment for obstructive tonsillar hypertrophy. Archives of Otolaryngology – Head and Neck Surgery. 2000; 126: 736–40. 38. Klausner RD, Tom LWC, Schindler PD, Potsic WP. Depression in children after tonsillectomy. Archives of Otolaryngology, Head and Neck Surgery. 1995; 121: 105–8. 39. Kotiniemi LH, Ryha¨nen PT, Moilanen IK. Behavioural changes following routine ENT operations in two- to



Chapter 96 Tonsillectomy



40.



41.



42.



43.



44.



45.



46.



47.



48.



49.



ten-year-old children. Paediatric Anaesthesia. 1996; 6: 45–9. Brown P, Ryan R, Yung M, Browne J, Copley L, Cromwell D et al. National prospective tonsillectomy audit, final report. London: Royal College of Surgeons in England, Clinical Effectiveness Unit, 2005. Royal College of Surgeons in England. Confidential enquiry into patient outcome and death (CEPOD). Available from: www.rcseng.ac.uk, 2005. Randall DA, Hoffer ME. Complications of tonsillectomy and adenoidectomy. Otolaryngology – Head and Neck Surgery. 1998; 118: 61–8. Yu KK, White DR, Weissler MC, Pillsbury HC. Nontraumatic atlantoaxial sublucation (Grisel syndrome): a rare complication of otolaryngological procedures. Laryngoscope. 2003; 113: 1047–9. Gledovic Z, Radovanovic Z. History of tonsillectomy and appendectomy in Hodgkin’s disease. European Journal of Epidemiology. 1991; 7: 612–5. Liaw KL, Adami J, Grindley G, Nyren O, Linet MS. Risk of Hodgkin’s disease subsequent to tonsillectomy: a population-based cohort study in Sweden. International Journal of Cancer. 1997; 72: 711–3. Cardwell M, Siviter G, Smith A. Non-steroidal antiinflammatory drugs and perioperative bleeding in paediatric tonsillectomy. Cochrane Database of Systematic Reviews. 2005; 18: CD003591. Hollis L, Burton MJ, Millar JM. Perioperative local anaesthesia for reducing pain following tonsillectomy. The Cochrane Library. Oxford: Update Software, 2002. Bolton CM, Myles PS, Nolan P, Sterne JA. Prophylaxis of postoperative vomiting in children undergoing tonsillectomy: a systematic review and meta-analysis. British Journal of Anaesthesia. 2006; 97: 593–604. Steward DL, Welge JA, Myer CM. Steroids for improving recovery following tonsillectomy in children. The Cochrane Library. Oxford: Update Software, 2003.



] 1241



50. O’Reilly BJ, Black S, Fernandes J, Parnesar J. Is the routine use of antibiotics justified in adult tonsillectomy? Journal of Laryngology and Otology. 2003; 117: 382–5. 51. Iver S, DeFoor W, Grocela J, Kamholz K, Varughese A, Kenna M. The use of perioperative antibiotics in tonsillectomy: does it decrease morbidity? International Journal of Pediatric Otorhinolaryngology. 2006; 70: 853–61. 52. Dhiwakar M, Eng CY, Selvaraj S, McKerrow WS. Antibiotics to improve recovery following tonsillectomy: a systematic review. Otolaryngology – Head and Neck Surgery. 2006; 134: 357–64. 53. Paranese A, Clarke RW, Yardley MP. Early post-operative morbidity following tonsillectomy in children: implications for day surgery. Journal of Laryngology and Otology. 1999; 113: 1089–91. 54. Kanerva M, Tarkkila P, Pitkaranta A. Day-case tonsillectomy: parental attitudes and conclusion rates. International Journal of Pediatric Otorhinolaryngology. 2003; 67: 777–84. 55. Brigger MT, Brietzke SE. Outpatient tonsillectomy in children: a systematic review. Otolaryngology and Head and Neck Surgery. 2006; 135: 1–7. 56. Colee JG, Bradley R, Liberski PP. Variant CJD (vCJD) and bovine spongiform encephalopathy (SE) 10 and 20 years on: part 1. Folia Neuropathologica. 2004; 44: 93–101. 57. Frosh A. Prions and the ENT surgeon. Journal of Laryngology and Otology. 1999; 113: 1064–7. 58. Frosh A, Joyce R, Johnson A. Iatrogenic vCJD from surgical instruments. British Medical Journal. 2001; 322: 1558–9. 59. Montague ML, Lee MS, Hussain SS. Post-tonsillectomy haemorrhage: reusable and disposable instruments compared. European Archives of Otorhinolaryngology. 2004; 261: 225–8. 60. Colee JG, Bradley R, Liberski PP. Variant CJD (vCJD) and bovine spongiform encephalopathy (SE) 10 and 20 years on: part 2. Folia Neuropathologica. 2006; 44: 102–10.



97 Salivary gland disorders in childhood PETER D BULL



Introduction Congenital disorders Inflammatory diseases Granulomatous diseases of the salivary glands Salivary gland tumours



1242 1242 1243 1244 1247



Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



1249 1249 1250 1250



SEARCH STRATEGY Data for this chapter were obtained from the author’s personal experience and bibliography, supported by a Medline search using the key words salivary gland, salivary disorders, sialolithiasis, congenital and child.



INTRODUCTION Many of the salivary gland conditions that occur in children mirror those found in adults but relative incidences vary. In addition, there are some conditions that are unique to childhood. Salivary gland disease in children is uncommon. Clinical expertise in this area tends to be concentrated in the hands of a few; as a result the correct diagnosis may be overlooked and the primary management may be inappropriate.



CONGENITAL DISORDERS



resonance imaging (MRI).1 Treatment of xerostomia is supportive and is focussed on meticulous oral hygiene and the use of artificial saliva.2



Congenital salivary cysts These are found most often in the parotid gland. They include cysts that occur in association with branchial cleft and branchial pouch anomalies (see Chapter 99, Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma). They may present as intraglandular masses or with secondary sialadenitis due to ductal compression.3



Absence of salivary glands Salivary gland agenesis or aplasia is extremely rare. If it involves all of the major salivary glands, the result is severe xerostomia with extensive dental caries. Partial agenesis may be missed as the remaining glands produce enough saliva to maintain the dentition. If in doubt, salivary tissue can be demonstrated by technetium 99m pertechnetate scanning, ideally augmented with magnetic



Ductal anomalies These are rare but include duplication and imperforate ducts. They may be demonstrated by contrast sialography but better still by MRI. If associated with secondary infection, they may require treatment. Marsupialization is usually adequate.3



Chapter 97 Salivary gland disorders in childhood



Ectopic salivary tissue Salivary acinar cell rests can be found as a focus of draining sinuses in the neck, the mandible and the cervical lymph nodes.3



INFLAMMATORY DISEASES Viral parotitis (mumps) A number of viruses, most typically mumps, may be implicated in acute sialadenitis. ‘Mumps’ is an acute infectious disease caused by the mumps virus, a paramyxovirus. Classically, children aged four to five are affected and 85 percent of cases occur below the age of 15 years. The incubation period is two to three weeks. Infection is spread by droplets from an infected person’s respiratory tract. Mumps causes a mild pyrexial illness in up to 40 percent of infected patients with enlargement of one or more of the salivary glands, nearly always the parotid. Parotid involvement is typically bilateral. The diagnosis is confirmed by measuring viral titres for the S and V antigen. The S antigen is positive at presentation and persists for 12 months. The V antigen appears at one month but persists for up to two years. No specific treatment is available and resolution is to be expected. Mumps may rarely be associated with severe complications, notably sensorineural deafness, orchitis, encephalitis and pancreatitis. The use of mumps, measles and rubella (MMR) vaccine in 92 percent of developed countries has resulted in mumps becoming a rare disease in Western communities, where complications are now almost never seen. [***] There has been a recent increase in the incidence of mumps in Britain, presumably due to a reduction in the number of children given the MMR vaccine.4, 5 Mumps deafness and encephalitis are still a significant global public health problem.6



] 1243



are S. aureus and Streptococcus pyogenes, together with anaerobic bacteria. The affected gland becomes swollen and progresses to suppuration. The baby will be ill, pyrexial and reluctant to feed. Ultrasonography may help determine when frank pus is present. In the early stages, adequate doses of intravenous antibiotics may halt progression. Surgical drainage should be considered with extreme care due to the superficial position of the facial nerve in the newborn. Needle aspiration is safer than incision (Figure 97.1). [*]



Relapsing acute parotitis This is the most common inflammatory disorder of the salivary glands in childhood. It is usually self-limiting and requires no surgery. [**] PRESENTATION



Though recurrent parotitis may occur at any age, it most commonly presents at five to six years. The first episode is often diagnosed as mumps. Usually one, but occasionally both, parotid gland(s) will become acutely swollen and painful. The salivary flow from the duct is reduced and may be turbid. Episodes typically last a few days and may



Acute suppurative sialadenitis Acute pyogenic infection of the salivary glands is characterized by painful swelling with erythema of the overlying skin. Organisms include Staphylococcus aureus and Streptococcus viridans. Dehydration and salivary calculi may be predisposing factors. Treatment is with antibiotics analgesia and adequate hydration. [*]



Acute parotitis of infancy Acute pyogenic infection of the parotid gland in infancy is a distinct clinical entity. It occurs in the newborn and is more common in prematurity. The most likely organisms



Figure 97.1 Acute suppurative parotitis in a baby aged three weeks.



1244 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY be months apart. There may be some erythema of the skin overlying the gland. The child is usually apyrexial. Although culture of the saliva is often sterile, the most common organisms are Streptococcus pneumonia and Haemophilus influenzae.7 INVESTIGATION



Ultrasound examination of the affected gland will demonstrate any calculi or intraparotid abnormalities and is well tolerated by children.1 Sialography is helpful in showing the changes of sialectasis that occur in most cases, but is difficult to perform in children (Figure 97.2). In any acute episode, a swab should be taken of the saliva from the duct of the affected gland, and if organisms are cultured appropriate antibiotic therapy should be given. In any case of recurring infection it is important to consider the child’s immune status, and to exclude diabetes mellitus, immunoglobulin deficiency and cystic fibrosis. The advice of a paediatrician should be sought if there is any doubt. MANAGEMENT



In almost all cases, recurrent parotitis of childhood resolves by the age of puberty. [***] Therefore, management



is conservative in the expectation of reduction in the frequency and severity of attacks. Individual episodes should be managed by adequate hydration, analgesics and gentle external massage of the gland in the direction of the parotid duct towards its papilla to encourage drainage. Antibiotics are not indicated unless there is generalized malaise and fever. [*] If the attacks do not resolve, lavage of the parotid duct with normal saline after cannulation may reduce the frequency of the attacks. This can be undertaken under general anaesthetic if necessary. There is little place for open parotid surgery. Even if irreversible structural changes in the gland have occurred, the condition is self-limiting and resolution of symptoms can be expected. If excision surgery is embarked upon, a subtotal parotidectomy will be required (Figure 97.3).8 [*]



GRANULOMATOUS DISEASES OF THE SALIVARY GLANDS The most common granulomatous infections of the salivary glands are mycobacterial but this group of disorders includes rare infective conditions such as actinomycosis and cat-scratch disease. These conditions may simulate malignancy.



Mycobacterial infection of the salivary glands This is probably not a true infection of the salivary glands but of the perisalivary lymph nodes in children. In Western communities, infection is more commonly by atypical or nontuberculous mycobacteria (NTM) than by Mycobacterium tuberculosis. [****] The most common organisms are M. avium-intracellulare complex, M. kansasii and M. malmoense. They can be identified with most certainty by culture in specialized reference laboratories.



Figure 97.2 sialectasis.



Parotid sialogram showing the changes of



Figure 97.3 Chronic irreversible parotitis with salivary fistulae.



Chapter 97 Salivary gland disorders in childhood



The condition of NTM salivary gland infection affects a young age group, typically three to four years old. The patients in the author’s series ranged from one year nine months to seven years ten months, with an average of three years ten months.9



PRESENTATION



Patients present with a mass in the neck or face, usually painless and often with a short history. The overlying skin usually becomes discoloured with a violet hue. Early necrosis and breakdown of the tissues may occur. There is no facial nerve weakness in uncomplicated cases (Figure 97.4). INVESTIGATION



The most important investigations are microscopy with Zeil Neilson staining, histology and culture of the affected tissue. Repeated culture of aspirated material may be negative. Fresh excised tissue is more likely to yield a diagnosis. It is important to inform the laboratory staff that the diagnosis is suspected before sending tissue. While specific antigens for skin testing to the various mycobacteria are available, they are not easily obtained. They may be used as further evidence if the diagnosis is in doubt. All cases of suspected mycobacterial infection should have a Mantoux test and chest x-ray to exclude M. tuberculosis. Imaging of the cervical mass is not particularly helpful. Ultrasound examination is often rendered difficult by the presence of skin necrosis.



] 1245



MANAGEMENT



Many cases of an enlarged cervical nodes infected by NTM will resolve spontaneously without necrosis. [***] Where there is skin discoloration and early abscess formation, long-term treatment with anti-tuberculosis chemotherapy is usually administered. [**] A typical regime is ciprofloxacin 15 mg/kg/day as two doses and azithromycin 10 mg/kg/day over some months. In cases of necrosis or abscess formation, surgical excision of the inflammatory mass with preservation of the facial nerve is curative.10 [**] The organism is more easily grown from the excised tissue than from the pus. Simple drainage of an abscess may result in a chronically discharging fistula which can take 18–24 months to heal fully, with considerable scarring. [**] Coulter et al.11 summarized the evidence base for current treatment regimes and concluded that there are no randomized controlled trials on which to base management strategies, but excision is curative and curettage is superior to incision and drainage. Single drug therapy rapidly induces resistance. Chemotherapy prior to surgery may help localize the lesion. Untreated the condition will spontaneously resolve over several years.



Sarcoidosis Sarcoid is a chronic granulomatous condition of unknown aetiology. The onset of sarcoidosis is usually in early adult life but it can present in the teenage years. Parotid gland involvement is present in about 10 percent of cases and it may present acutely with fever and uveitis – Heerford’s uveo-parotid fever. Facial palsy may occur in this condition. The changes in the gland are of noncaseating granulomata and may be demonstrated by biopsy of the minor salivary glands of the lower lip in 60 percent of cases.



Sjo¨gren’s syndrome



Figure 97.4 Infection with atypical (nontuberculous) mycobacteria.



Sjo¨gren’s syndrome is an autoimmune multisystem inflammatory disorder which almost always involves the salivary glands (see Chapter 147, Non-neoplastic salivary gland diseases). In long-standing cases it is characterized by dryness of the oral and conjunctival membranes (Sicca syndrome). It is rare for a diagnosis of Sjo¨gren’s to be made in childhood but, when it is, the commonest presentation is with parotitis. Girls are more commonly affected than boys. Sicca syndrome is uncommon until adulthood. If suspected, rheumatoid factor, antinuclear antibodies and SS-A and SS-B autoantibodies should be measured.12 Treatment should be multidisciplinary with involvement of a paediatrician and dental surgeon. Lifelong surveillance may be needed as patients are prone to the development of lymphomata.13 [***]



1246 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



HIV/AIDS



large enough, it may affect both breathing and swallowing (Figure 97.6).



Salivary gland involvement in juvenile cases of HIV infection is more common than in adult disease and is referred to as HIV-associated salivary gland disease (HIV-SGD).



HIV-SGD The main feature of HIV-SGD is the presence of lymphoepithelial cysts within the salivary glands, predominately the parotid. They are thought to be the result of hyperplasia of the intraglandular lymph nodes (Figure 97.5). Confirmation of the cystic nature can be made on ultrasound scanning and, unless neoplasia is suspected, no surgical intervention is required. No virus is present in the cyst fluid.14 The finding of multiple parotid cysts in childhood should alert the clinician to the possible diagnosis of HIV infection.



Ranula Ranula is the term to describe a cystic swelling arising in the floor of the mouth (ranula – L. frog). Ranula may be congenital or acquired and, on rare occasions, may be found in newborn infants.



Plunging ranula If extravasation of mucus occurs beyond the confines of the floor of the mouth through the mylohyoid muscle into the upper neck or submental region, a large cystic swelling develops. This can usually be defined on MR scanning but is difficult to clinically differentiate from cystic hygroma. Histologically, hygroma has a simple epithelial lining whereas a ranula is contained by loose connective tissue. TREATMENT



Simple aspiration or drainage of a ranula results in a high recurrence rate. Both simple and plunging ranulas should be excised, together with the cyst wall and the sublingual gland. [**] Care must be taken to identify and preserve the lingual nerve which lies in close proximity. A ranula presenting in the neck will usually need an external approach for excision. Marsupialization of a plunging ranula is inadequate with a recurrence rate of 80 percent.15 Excision of the sublingual gland results in a low rate of recurrence.



AETIOLOGY



Ranula is the result of obstruction of one of the sublingual salivary glands. It is essentially a retention cyst. This may be spontaneous or may result from surgery to the floor of the mouth, especially submandibular duct relocation.



Salivary gland stones Salivary gland stones or calculi are rare in children. Stones are formed by crystal formation from salivary solutes, mainly hydroxyapatite. These salts can be precipitated out



PRESENTATION



The ranula presents as a smooth cystic swelling under the tongue, usually to one side. It is often transparent or bluish in appearance with overlying small blood vessels. If



Figure 97.5 infection.



Lymphoepithelial cyst of the parotid gland in HIV Figure 97.6 Sublingual ranula.



Chapter 97 Salivary gland disorders in childhood



of solution by a change in pH in the saliva or by dehydration. They are much more common in the submandibular gland than the parotid because of the mucoid nature of the submandibular saliva. PRESENTATION



The stone will normally declare its presence by sudden enlargement of the affected salivary gland while the child is eating, caused by occlusion of the duct in the presence of salivary flow. The gland becomes swollen and painful and the swelling usually subsides over an hour or two. The stone may be visible or palpable within the duct. Parotid stones tend to be less calcified and therefore softer and less easy to feel. Sometimes, a large stone will be found coincidentally in the submandibular gland on routine radiology and may cause no symptoms. INVESTIGATION



If the stone is clinically apparent, no further investigation is necessary. Plain x-rays may show a stone but intraoral views of the duct are preferable. If doubt persists, MR scanning is extremely sensitive in demonstrating calcium salts. TREATMENT



If the stone is causing no symptoms, then it can be left alone. If the stone is accessible within the submandibular duct, it can be removed intraorally by making a linear incision in the floor of the mouth, extricating the stone and marsupializing the mucosal edges. Great care must be taken not to injure the lingual nerve which lies immediately deep to the duct. A stone within the substance of the submandibular gland which is causing persistent symptoms will necessitate excision of the gland. Endoscopy of the salivary ducts has been used to remove calculi but there is no account of this in children. Some radiologists report that a contrast sialogram may bring about improvement as the contrast flushes the debris from the gland.16 [*]



] 1247



salivary gland anatomy and salivary gland surgery. It is important to distinguish between haemangioma and vascular malformation. Haemangiomas are not present at birth but grow rapidly after birth. They are of endothelial origin and highly cellular. They initially grow rapidly for nine to twelve months and involution usually occurs by the age of three to five years. Most haemangiomas occur in the head and neck and many will involve the skin and the underlying salivary gland in the parotid region. Occasionally, the tumour may be deeply placed and lie within the parotid gland (Figure 97.7). TREATMENT



Almost all haemangiomas will involute and surgery should be confined to management of bleeding or ulceration, or to cases where the diagnosis is in doubt. [*] Diagnostic uncertainty in the case of an enlarging intraparotid mass may occasionally necessitate exploration and excision, with preservation of the facial nerve, even if haemangioma is suspected. High quality imaging has made this an extremely rare indication for surgery (Figure 97.8). If the haemangioma is of such a size or position as to be life-threatening, treatment with steroids or a-interferon has been shown to be of benefit in hastening resolution. [**] Each is associated with significant potential morbidity in babies and should only be considered with extreme care. Steroids may be administered systemically or directly into the lesion. Pulse-dye laser therapy may be considered for ulcerating lesions; interferon is used systemically for its inhibitory effect on angiogenesis.17 Vascular anomalies are developmental malformations that are present at birth and enlarge slowly as the child grows. They may be venous, lymphatic or arteriovenous in origin and include cystic hygroma and lymphangioma (see Chapter 99, Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma).



SALIVARY GLAND TUMOURS Salivary gland tumours are extremely rare in children. They comprise vasoformative swellings, true neoplasms of salivary tissue origin and proliferative disorders of the perisalivary lymphoid tissue (lymphoma). For salivary gland tumours in adults see Chapter 189, Benign salivary gland tumours and Chapter 190, Malignant tumours of the salivary glands.



Vasoformative lesions The commonest salivary gland swellings in childhood are vasoformative. While not truly salivary in origin, their management may require an extensive knowledge of



Figure 97.7 Parotid haemangioma histology showing highly cellular tumour within normal parotid tissue.



1248 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Figure 97.9 Pleomorphic adenoma of the submandibular salivary gland.



Figure 97.8 Capillary haemangioma involving the parotid gland and skin.



also occur in the submandibular gland or in extra glandular sites (Figure 97.9). In a large series from the Mayo clinic of salivary neoplasms in children, the only benign salivary neoplasm encountered was PSA,19 but all types found in adults do occasionally present. TREATMENT



Salivary gland neoplasms Neoplasms of the salivary gland account for 1 percent of all tumours in children. Fewer than 5 percent of all salivary tumours occur in patients under the age of 16 years. In the author’s series of salivary gland tumours in childhood, a total of 14 neoplasms has now been treated, plus three lymphomas arising within the major salivary glands. Of these 14 cases, seven have been malignant and seven benign. Of these benign cases, only five were truly of salivary cell origin. The rarity of these tumours is such that inappropriate management frequently occurs before referral.18



The treatment of benign salivary tumours is as it is in adults, by wide excision and facial nerve preservation. [**] The relatively superficial position of the facial nerve in children must be considered when performing parotidectomy. Particular care must be taken to achieve good surgical margins and avoid rupture of the tumour, since adjuvant radiotherapy cannot be safely used in children to salvage an unsatisfactory surgical procedure. Scar hypertrophy is more common in children than in adults, but in the author’s practice the incidence of Frey’s syndrome after parotidectomy is less.



Malignant tumours PRESENTATION



Salivary gland tumours in children present as an enlarging mass. Growth is usually slow, and rapid growth is more suggestive of lymphoma arising in an intraglandular lymph node. They are usually painless. Facial nerve palsy at presentation is indicative of malignancy. Eighty percent of all salivary neoplasms occur within the parotid gland.



Benign tumours The commonest benign tumour encountered is pleomorphic salivary adenoma (PSA) accounting for approximately 30 percent of all paediatric salivary neoplasms. The majority occur within the parotid gland but they may



The relative proportion of salivary tumours in children which prove to be malignant is higher in children than in adults, though of course overall, salivary neoplasms in children are extremely rare. In most series approximately half are malignant. Most malignant salivary gland neoplasms in childhood present simply as an enlarging mass. Growth may be slow and there is usually no other evidence of malignancy. Facial weakness and increasing pain are strong indications of malignant disease. PATHOLOGY



Most series show that the most common salivary malignancy in a paediatric age group is mucoepidermoid



Chapter 97 Salivary gland disorders in childhood



carcinoma (approximately 50 percent) followed by acinic cell carcinoma (20 percent). More aggressive tumours such as adenoid cystic carcinoma and adenocarcinoma may also occur (Figure 97.10). INVESTIGATION



All cases of salivary gland neoplasm in children should be treated as potentially malignant. [*] Aspiration (needle) biopsy and cytology are of limited value. In children this will usually require general anaesthetic and assessment of salivary gland histology is notoriously difficult. Ultrasonic imaging of the salivary mass is easy to perform, noninvasive and well tolerated by children. It gives information about the location and size of the lesion and will indicate whether it is cystic or solid. CT and MR may be needed, including imaging of the chest. MANAGEMENT AND TREATMENT OF MALIGNANT TUMOURS



Management will involve a multidisciplinary approach with input from a paediatric oncologist and his/her team. These children and their families are best treated in a



] 1249



designated centre where the whole range of support services for children with cancer is available. As with benign neoplasms, wide surgical excision is required for malignant tumours. Usually, the diagnosis cannot be made preoperatively unless there is associated facial palsy. It is essential to counsel the child and parents carefully prior to operation regarding the possible need for facial nerve sacrifice. The surgical margins on the specimen must be marked and evaluated by the pathologist and adequacy of excision confirmed. Provided the excision margins are satisfactory, no adjuvant treatment is indicated. [**] The author’s series has shown only one recurrence, that of an acinic cell tumour after four years, with subsequent freedom from relapse after ten years. The Mayo clinic series shows only one recurrence out of 14 malignant cases who had their primary treatment at the Mayo. The recurrence rate was higher in cases referred after initial treatment in a nonspecialist centre.19 [**] Adjunctive radiotherapy and chemotherapy are reserved either for recurrent disease or very aggressive disease at the time of the initial surgery and should be determined with the collaboration of a paediatric oncologist. Long-term follow-up is essential in these rare cases.



KEY POINTS  The relatively superficial position of the facial nerve should be constantly borne in mind when contemplating surgery in the parotid region in children.  Consider NTM in a child with a discharging lesion in the parotid region.  Relapsing parotitis is almost always selflimiting and resolves at puberty.  The rarity of salivary tumours in children is such that treatment should only be in specialist centres.



Best clinical practice [ Treatment of acute suppurative parotitis is with



Figure 97.10 gland.



Mucoepidermoid carcinoma of the parotid



analgesics hydration and antibiotics. Surgery is rarely indicated. [Grade C] [ Relapsing parotitis is almost always self-limiting and resolves at puberty. [Grade C] [ Consider NTM in a child with a discharging lesion in the parotid region. [Grade C] [ The traditional advice that a ranula can be adequately treated by simple marsupialization should be revised. These swellings are best excised. [Grade C]



1250 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY [ The finding of multiple parotid cysts in childhood should alert the clinician to the possible diagnosis of HIV infection. [Grade B] [ Almost all haemangiomas will involute and surgery is rarely needed. [Grade B] [ Salivary tumours in children are rare but a higher proportion are malignant than in adults. [Grade C]







Deficiencies in current knowledge and areas for future research



$ $ $ $



The implementation of vaccination programmes throughout the developing world would greatly reduce the morbidity of mumps encephalitis and deafness. The epidemiology of NTM and the reasons for the recent increased incidence in Western communities may become clear. The respective roles of surgery and chemotherapy in the management of NTM may become better established. Salivary gland neoplasms are so rare that randomized controlled studies may not be possible; a national register of these tumours and treatment only in centres where adequate expertise is available would improve outcome.



REFERENCES







1. King SJ. Salivary glands. In: King SJ, Boothroyd AE (eds). Pediatric ENT radiology. New York: Springer-Verlag, Berlin Heidelberg, 2002: 335–44. A summary of current practice in salivary gland imaging in children. 2. Hodgson TA, Shah R, Porter SR. The investigation of major salivary gland agenesis: a case report. Pediatric Dentistry. 2001; 23: 131–4. 3. Ibrahim HZ, Handler SD. Diseases of the salivary glands. In: Wetmore RF, Muntz HR (eds). Pediatric otolaryngology: principles and practice pathways. New York, Stuttgart: McGill TJ Thieme, 2000: 647–58. 4. Dobson R. Mumps cases on the rise in England and Wales. British Medical Journal. 2005; 330: 3241.







5. Bellaby P. Has the UK government lost the battle over MMR? Editorial. British Medical Journal. 2005; 330: 552–3. 6. Galazka AM, Robertson SE, Kraigher A. Mumps and mumps vaccine: a global review. Bulletin of the World Health Organization. 1999; 77: 3–14. A review of mumps as a global health problem. 7. Cohen HA, Gross S, Nussinovitch M, Frydman M, Varsano I. Recurrent parotitis. Archives of Disease in Childhood. 1992; 67: 1036–7. 8. O’Brien CJ, Murrant NJ. Surgical management of chronic parotitis. Head and Neck. 1993; 15: 445–9. 9. Jervis PN, Lee JA, Bull PD. Management of non-tuberculous mycobacterial peri-sialadenitis in children. Clinical Otolaryngology and Allied Sciences. 2001; 26: 243–8. 10. Dhooge I, Dhooge C, De Baets F, Van Caauwenberge P. Diagnostic and therapeutic management of atypical mycobacterial infections in children. European Archives of Oto-Rhino-Laryngology. 1993; 250: 387–91. 11. Coulter JB, Lloyd DA, Jones M, Cooper JC, McCormick MS, Clarke RW et al. Nontuberculous mycobacterial adenitis: effectiveness of chemotherapy following incomplete excision. Acta Paediatrica. 2006; 95: 182-8. 12. Anaya JM, Ogawa N, Talal N. Sjogren’s syndrome in childhood. Journal of Rheumatology. 1995; 22: 1152–8. 13. Bartunkova J, Sediva A, Vencovsky J, Tesar V. Primary Sjogren’s syndrome in children and adolescents: proposal for diagnostic criteria. Clinical and Experimental Rheumatology. 1999; 17: 381–6. 14. Schiott M. HIV associated salivary gland disease: a review. Oral Surgery, Oral Medicine, and Oral Pathology. 1992; 73: 164–7. 15. Dhaif G, Ahmed Y, Ramaraj R. Ranula and the sublingual salivary glands. Review of 32 cases. Bahrain Medical Bulletin. 1998; 20: 3–4. 16. Bull PD. Salivary stones. Hospital Medicine. 2001; 62.7: 396–9. 17. David LR, Malek MM, Argenta LC. Efficacy of pulse dye laser therapy for the treatment of ulcerated haemangiomas: a review of 78 patients. British Journal of Plastic Surgery. 2003; 56: 317–27. 18. Bull PD. Salivary gland neoplasia in childhood. International Journal of Pediatric Otolaryngology. 1999; 49: S235–8. 19. Orvidas LJ, Kasperbauer JL, Lewis JE, Olsen KD, Lesnick TG. Pediatric parotid masses. Archives of Otolaryngology, Head and Neck Surgery. 2000; 126: 177–84. Comprehensive review of salivary gland neoplasms in children.



98 Tumours of the head and neck in childhood FIONA B MacGREGOR



Introduction Epidemiology Presentation and assessment The child with cancer Oncology services The more common childhood cancers Long-term sequelae of treatment



1251 1251 1252 1253 1253 1254 1260



The dying child Key points Best clinical practice Deficiencies in current knowledge and areas for future research Acknowledgements References



1261 1261 1261 1262 1262 1262



SEARCH STRATEGY Systematic reviews were identified using the key words: childhood cancer, head and neck, thyroid cancer, rhabdomyosarcoma, nasopharyngeal carcinoma, neuroblastoma, lymphadenopathy, Hodgkin’s lymphoma, non-Hodgkin’s lymphoma, sarcoma, palliative care; and subject headings in the Cochrane database of systematic reviews. Individual articles were then identified using the same search terms in Ovid Medline. A hand-search of the ensuing bibliographies completed the search.



INTRODUCTION This chapter will focus on the epidemiology of childhood cancers of the head and neck region and the general principles of assessing and managing malignancy in paediatric patients. It will then discuss the more common histological types individually. The long-term sequelae of treatment in survivors, palliative care and future developments will be addressed.



EPIDEMIOLOGY After trauma, cancer is the most common cause of death in childhood. Approximately one-third of childhood malignancies are leukaemias, 25 percent are brain and spinal tumours, 15 percent are embryonal (neuroblastoma, retinoblastoma, Wilm’s tumour and hepatoblastoma) and 11 percent are lymphomas. The remainder are bone and soft tissue sarcomas and miscellaneous



tumours (see below under Miscellaneous tumours). Up to 12 percent of primary childhood malignancies originate in the head and neck area and so may present to the otolaryngologist. Lymphoma is the most common diagnosis in all series, followed by thyroid and neural tumours. Sarcomas and salivary gland tumours are less common and squamous cell carcinomas are rare.1 The distribution of histological types varies greatly depending on the age and sex of the child with neuroblastoma the most common in infants and thyroid carcinoma the most common malignancy in adolescent females. There is a bimodal age distribution of malignancy in children. The most common age group affected is 15 to 18 year olds, closely followed by the under fours.1, 2 Cancer data from many parts of the world have suggested a small but steady recent increase in cancer rates in the under 19s. This increase would appear to be more marked in head and neck malignancies.1 This elevation in numbers may in part be due to improved data collection. In parallel, the consistent use of effective



1252 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY multimodality treatments, including combination chemotherapy, surgery and radiotherapy, has resulted in a significant improvement in prognosis. A study from the UK observed five-year survival rates improving from 42 percent (diagnosis made between 1968 and 1977) to 71 percent (diagnosis made between 1988 and 1995).3 Early recognition is therefore important as treatment is becoming increasingly effective in reducing the mortality in childhood malignancy.4, 5 It is thought that only around 5 percent of malignancies in childhood are inherited. The underlying aetiology of the noninherited types remains unclear but certain chemotherapeutic agents, viruses (e.g. Epstein–Barr) and irradiation are known to play a role. Other potential carcinogens are suspected but not proven (pollution, parental exposure to toxins, electromagnetic fields).1, 6



PRESENTATION AND ASSESSMENT The most common presentation of a malignancy in the head and neck region in childhood is the asymptomatic mass. Otalgia, rhinorrhoea, otorrhoea and nasal obstruction may be present in both benign and malignant



disease. More worrying symptoms would include stridor, dysphagia and haemoptysis.7 Reactive lymphadenopathy in children is extremely common and gives rise to frequent therapeutic dilemmas. Studies have suggested that the best predictors of malignancy in cervical lymphadenopathy are the size of the node, the number of sites involved and the age of the patient. Other concerning features would be lymphadenopathy in the supraclavicular region, an associated abnormal chest x-ray or fixed lymph nodes.8, 9 [**/*] Examination would involve a complete assessment of the head and neck region and systemic evaluation with particular regard to the presence of lymphadenopathy elsewhere and the presence or absence of abdominal masses. Flexible nasendoscopy is possible in many children (Figure 98.1). Imaging with the help of a paediatric radiologist can often obviate the need for biopsy. However, full assessment of a mass may require examination under anaesthetic with biopsy.



Biopsy Fine needle aspirate (FNA), sometimes referred to as aspiration biopsy cytology (ABC), has a more limited role



Figure 98.1 Flexible nasendoscopy under local anaesthetic in the clinic setting.



Chapter 98 Tumours of the head and neck in childhood



to play in the child than the adult. Young patients may simply not tolerate this method of biopsy. With particular regard to cervical lymphadenopathy, the most likely neoplastic diagnosis is lymphoma and at the present time it is recognized that excision biopsy is the best method of confirming and typing this histologically.10, 11 FNA, often under ultrasound control, can be useful in the assessment of thyroid and salivary gland lesions and these tumours often present in older children who may tolerate such intervention. [**/*]



Imaging Ultrasonography, CT and MR scanning can all have their part to play in further assessment. Imaging of the neck is more accurate than clinical examination in detecting lymphadenopathy. CT and MR scanning may require a general anaesthetic in a younger child and it is helpful to discuss the available imaging options with a paediatric radiologist prior to proceeding, to avoid unnecessary and additional anaesthetics. Positron emission tomography (PET) scanning is also becoming increasingly used in the assessment of neoplastic lesions in the head and neck in children, but is not readily available at the present time in most departments.7



THE CHILD WITH CANCER The history and subsequent investigations of a child presenting with a suspicious lesion of the head and neck region must be tailored to the age and maturity of the child. In most situations, when the child is old enough, he or she should be included in any discussion about investigations and treatment and they should also be involved in consent for any procedures required. The diagnosis of cancer in a child has a tremendous impact not only on the patient, but also on parents, siblings, other family members and friends. Immediate involvement of a specialist paediatric multidisciplinary oncology team is mandatory. An open and realistic approach should include an explanation of what to expect from the investigations and treatment, the side effects and some idea of prognosis. This is vital in maintaining trusting relationships with the child and his or her family, in reducing uncertainty, preventing inappropriate hope and allowing proportionate adjustment. Parents may feel guilty that they waited too long before seeking medical advice. They may be concerned that they have, in some way, caused their child’s cancer. They can also feel guilty about making their child go through a series of invasive investigations and radical treatments. This can put an immense strain on the parents’ relationship with each other and with other family members and friends, not least the child involved. Also, siblings of the affected child may resent the additional time and



] 1253



attention that their sick brother or sister receives. Support services are therefore vital in reassuring and supporting all the individuals concerned. The amount of information that any child will require regarding their illness depends to some extent on the age and maturity of that individual. Most children aged six or more (and some more mature younger children) need to know their illness has a name and what that name is. All children require an explanation of the procedures to be performed and reassurance that any intervention is not a ‘punishment’. In children between the ages of six and eleven, these procedures and side effects of treatment may provoke much more anxiety than the illness itself. For example, loss of hair or a limb seem much more real and distressing than the prospect of death. Alterations in physical appearance can cause great insecurity in a child or adolescent, resulting in isolation and poor self-esteem. Children aged around 11 and over will have fears surrounding the diagnosis and its prognostic implications in addition to the above. Children should be encouraged to talk about their feelings or, if they are too young, they can express themselves in drawings or play.11



ONCOLOGY SERVICES In the UK, as in many other parts of the world, the importance of centralization of paediatric cancer services is recognized.4 This enables care to be provided by highly trained and specialist paediatric oncologists and allied staff, and facilitates the progress of research. In Britain, there are 22 United Kingdom Children with Cancer Study Group (UK-CCSG) centres. Working in conjunction with the medical staff are social workers, nurses, dieticians, psychologists and other health care professionals in order to provide comprehensive support for children and their families. The emphasis on centralization of paediatric oncology services, the sharing of data and the establishment of international working groups has resulted in the publication of a number of treatment protocols that are widely used in the management of children with cancer. The practicalities of investigating and treating children with cancer provide some particular challenges. Venous access for blood sampling and to administer chemotherapeutic agents can be difficult, and indwelling venous catheters are usually inserted at an early stage. Radiotherapy may require general anaesthesia in younger patients to ensure that the child remains still during irradiation. Other interventions (e.g. lumbar puncture and intrathecal injection) may also require general anaesthesia. Children tolerate the immediate side effects of chemotherapy and radiotherapy much better than adults but the long-term sequelae of such interventions can have a very significant effect on the health of childhood cancer survivors (see below under Long-term sequelae of treatment).



1254 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



THE MORE COMMON CHILDHOOD CANCERS Lymphoma Cervical lymphadenopathy in childhood is common and although the huge majority of cases will be due to reactive hyperplasia, it may be necessary to exclude malignancy. Lymph nodes in the neck larger than 2 cm are unusual in childhood and systemic symptoms such as weight loss, fever and organomegaly are usually indicators of serious pathology.8 FNA has a limited role and excision biopsy will provide the diagnosis.12 [**] Lymphomas are malignant neoplasms of the lymphoreticular system. Most lymphomas of the head and neck region in the paediatric age group present with enlarged cervical lymph nodes.



Hodgkin’s lymphoma This is distinguished morphologically by the presence of Reed Sternberg cells which are large and multinucleated with abundant cytoplasm. There are several classifications but the universally accepted one is the Rye classification. Within this there are four subtypes: lymphocyte predominant (LP), mixed cellularity (ML), lymphocytic depleted (LD) and nodular sclerosing (NS). NS is the most common type found in children and young adults. It has the propensity to involve the lower cervical, supraclavicular and mediastinal lymph node groups. No definitive causal factors have been identified but there is an association with previous infection with Epstein–Barr virus.13 Hodgkin’s lymphoma most commonly presents with lymphadenopathy in the neck and two-thirds of all children will have mediastinal lymphadenopathy at presentation. It rarely occurs under the age of five and there is a male predominance. ‘Constitutional’ upsets such as fever, night sweats and weight loss are present in 25 to 30 percent and this is associated with a poorer prognosis.7 Following careful examination and biopsy confirmation, the diagnostic workup of a child with Hodgkin’s disease (following thorough examination and biopsy confirmation) would include a chest x-ray, routine blood tests (although abnormal results are usually nonspecific) and staging scans (CT chest and MRI abdomen). Recently, PET scanning has become a routine part of staging and of assessing effectiveness of treatment. There is no longer a place for staging laparotomies. Bone marrow biopsy and bone scan are only indicated in children with more advanced disease.14 Currently, the Ann Arbor staging classification is used for Hodgkin’s disease (see Chapter 22, Haemato-oncology). Treatment depends on the age and physical maturity of the patient, the disease stage and bulk and the potential



treatment sequelae. In the paediatric population, the trend is to treat in multimodality fashion so as to reduce the morbidity and mortality associated with high doses of chemotherapy or radiation therapy needed for single modality treatment. Disease-free survival is over 90 percent in many series.15 Those children with intermediate risk of disease (constitutional symptoms, bulky disease or spleen involvement) may require an increased number of cycles of chemotherapy and an increased dose or volume of radiation therapy. In advanced disease, combined chemoradiotherapy is the treatment of choice. LP Hodgkin’s disease is associated with a better prognosis and separate shorter and less intense treatment protocols are implemented. The regimens used in the treatment of Hodgkin’s disease contain substantial doses of alkylating agents which are associated with the potential for significant morbidity.13 [***/**] Haematopoietic stem cell transplantation can be implemented in those children who relapse, but the risk of transplant-associated morbidity and mortality is not insignificant.13



Non-Hodgkin’s lymphoma Approximately 60 percent of paediatric lymphomas are non-Hodgkin’s lymphomas (NHL). There is a male predominance. The low grade, relatively indolent NHLs seen in adults are exceedingly rare in children. Paediatric NHLs tend to be aggressive with a propensity for widespread dissemination and half of these are small cell lymphomas (Burkitt’s and Burkitt’s-like). The classification of NHL is confusing and controversial but paediatric NHLs are usually divided into three main histological categories. These are lymphoblastic lymphoma (predominantly of T cell origin), small noncleaved cell lymphoma (Burkitt’s and non-Burkitt’s subtype of B cell origin) and large cell lymphoma (B or T cell origin).16 The Revised European American Lymphoma (REAL) classification is used in paediatric NHL and is a useful guide to management and prognosis of NHL. The relative frequency and incidence of NHL varies quite markedly from country to country. In equatorial Africa, Burkitt’s lymphoma accounts for approximately 50 percent of childhood cancers. In Europe and the US, approximately one-third are lymphoblastic lymphomas, one-third are Burkitt’s and Burkitt’s-like lymphoma and the rest are predominantly large cell lymphoma. In some parts of the world, an extremely high number of these tumours are positive for Epstein–Barr virus (EBV), e.g. parts of Africa. In contrast, the percentage of tumour positive for EBV in the US is much smaller. There is an increased incidence of NHL in association with immunosuppression and congenital and acquired immunodeficiency.16 All childhood NHLs are rapidly growing neoplasms and a significant number of children will have widespread



Chapter 98 Tumours of the head and neck in childhood



disease at the time of diagnosis, which may involve the bone marrow, central nervous system or both. Involvement of extranodal sites, especially Waldeyer’s ring, is particularly common in children. Lymphadenopathy occurs in 50–80 percent of all patients and 45 percent have cervical lymphadenopathy at the time of presentation. Bone marrow involvement is not infrequent and the replacement of more than 25 percent of the bone marrow by tumour cells is usually assigned a diagnosis of acute lymphoblastic leukaemia. Children with endemic Burkitt’s lymphoma frequently present with involvement of the jaw and this is particularly common in the younger age group.16



DIAGNOSIS AND STAGING



Because NHLs in children progress more rapidly, a speedy diagnosis is extremely important. A biopsy will provide tissue for histological confirmation and surgery then has little further role to play. Staging investigations will follow and will include relevant blood tests (full blood count, urea and electrolytes and liver function tests – lactic dehydrogenase (LDH) is a useful marker of disease – bone marrow biopsy and cerebrospinal fluid (CSF) examination. Staging laparotomy is not advocated in patients with NHL and ultrasonography, CT (chest) and MR (abdomen) scanning are used in assessing spread.16 The Ann Arbor staging classification can be applied to NHL (see Chapter 22, Haematooncology).



] 1255



because of their biological heterogeneity and long-term EFS ranges from 50 to 70 percent.14, 16 [***/**] Patients who relapse after receiving dose-intensive multiagent chemotherapy have an extremely poor prognosis. These patients are treated with high-dose chemotherapy regimes with or without bone marrow transplantation or would be candidates for drugs in phase II trials.



Rhabdomyosarcoma Rhabdomyosarcomas account for up to 60 percent of all sarcomas in the paediatric population and 40 percent occur in the head and neck region. Nearly half of these tumours occur in children under the age of five.4 The prognosis for this tumour used to be extremely poor but over the last 30 years survival rates have increased dramatically, particularly with the introduction of multimodality therapy in which surgery, multiagent chemotherapy and radiotherapy have been combined.17 Histologically, rhabdomyosarcomas resemble normal foetal skeletal muscle before innervation. Two types are identified and these are embryonal (good prognosis) and alveolar (poor prognosis). The alveolar type is found in older children and is often associated with metastatic spread. It is less common than the embryonal type.18, 19



PRESENTATION



The treatment of choice in childhood NHL is multiagent chemotherapy. The rapid doubling time of high grade NHL makes it particularly chemosensitive. Chemotherapeutic regimens vary depending on the histological classification of the disease. Radiation therapy has a limited role in NHL and is generally reserved for selected anatomical sites such as the cranium where a child has overt central nervous system (CNS) disease. CNS prophylaxis in children with high grade NHL can be achieved with intrathecally administered chemotherapy.14, 16 [***/**]



Rhabdomyosarcomas of the head and neck occur most frequently in the orbit or parameningeal sites and these include the paranasal sinuses, nose, nasopharynx and middle ear. The most common presenting symptoms are pain and swelling. Paranasal rhabdomyosarcoma may present with a gradual onset of nasal obstruction and bloody nasal discharge. Tumours within the ear may present with symptoms of bloody discharge and persistent otalgia, despite treatment. A polypoid mass may be visible in the ear canal or nasal cavity.19 The reported incidence of lymph node involvement varies between 3 and 36 percent.18 Metastases occur by both haematogenous and lymphatic spread.



PROGNOSIS



ASSESSMENT



Long-term event-free survival (EFS) is excellent in lymphoblastic lymphomas. It ranges from 80 to 90 percent in patients with limited disease and from 65 to 80 percent in patients with advanced disease. In Burkitt’s, where chemotherapeutic regimes are more intense and shorter, the long-term EFS ranges from 90 to 100 percent in patients with limited disease to 75 to 85 percent with patients with extensive disease. Even in patients with extensive bone marrow disease, the EFS is improving. Large cell lymphomas are more of a challenge to treat



Assessment should include a thorough examination of the upper respiratory tract and head and neck region including the cranial nerves. Flexible nasendoscopy may be employed in the clinic. An MR scan should be performed to evaluate the primary lesion and to rule out metastatic disease (Figure 98.2). A CT scan may be a useful complementary tool, particularly in the paranasal sinuses and skull base, to determine bony erosion and it is the best method to assess the chest. Bone marrow examination should also be performed. The Inter Group



TREATMENT



1256 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY TREATMENT AND PROGNOSIS



Since the establishment of the IRS Study Committee in 1972, protocols have been developed and are now widely adhered to. A multimodality approach has been adopted. In general, the role of surgery today is to simply evaluate the extent of the lesion and biopsy the tumour. Occasionally, when the rhabdomyosarcoma is an easily accessible polypoid lesion, then wide surgical removal may be appropriate. A debulking procedure may also be useful. Sophisticated skull base surgery can now be implemented in areas that were previously thought to be inaccessible.19 Multiagent chemotherapy and radiotherapy are then implemented as appropriate. [**/*] Prior to 1960, approximately 10 percent of patients survived five years. Now the prognosis is excellent in patients with early tumours (over 80 percent survival). With more advanced tumours the prognosis is still relatively poor and in those with meningeal involvement the five-year survival is less than 10 percent. Although nodal mestastases at initial presentation are not correlated with an unfavourable prognosis, development of nodes during follow-up does imply a poor outlook.19, 20, 21 [**/*] Figure 98.2 Paramedian T1-weighted MR scan through the left orbit of a three-month-old boy. There is a huge mass of rhabdomyosarcoma in the anterior orbit (short arrow) displacing the globe and submandibular lymphadenopathy (long arrow).



Table 98.1 IRS.



Staging of rhabdomyosarcoma according to the



Staging of rhabdomyosarcoma Group I



Group II



Group III Group IV



Localized disease completely resected. No regional nodes Confined to muscle or organ of origin Contiguous infiltration outside muscle or organ of origin Localized disease with microscopic residual disease or regional disease with no residual or with microscopic residual disease Grossly resected with microscopic residual disease (nodes negative) Regional tumour completely resected Regional nodes grossly resected but evidence of microscopic residual disease Incomplete resection or biopsy with gross residual disease Metastatic disease present at onset



Thyroid carcinoma Thyroid carcinoma in the paediatric population is uncommon. In the US there are five new cases per million per year. It is much more common in adolescents than in younger children and is also much more common in females with a ratio of 4:1. Approximately 45 percent of these lesions in children will be differentiated papillary carcinomas with a further 45 percent of mixed papillary/ follicular types with only 10 percent being follicular lesions.22 Medullary thyroid carcinoma (MTC) is rare (only 10 percent of thyroid malignancies in children) and must be suspected in children with multiple endocrine neoplasia (MEN) types IIa and IIb (see under Medullary thyroid carcinoma). Anaplastic and undifferentiated tumours are extremely rare in children and adolescents. AETIOLOGY



A clear relationship has been established between the development of thyroid carcinoma and previous irradiation. In one study, as many as 17 percent of patients had previously received irradiation to the neck.23 PRESENTATION



Rhabdomyosarcoma (IRS) study recommends the clinical staging shown in Table 98.1. The majority of children are stage II at the time of assessment20 as it is rarely possible to completely excise these tumours.



Patients usually present with an asymptomatic solitary mass in the anterior or lateral neck. At presentation, there is often regional lymph node involvement (74 percent) and distant parenchymal metastases (25 percent).22



Chapter 98 Tumours of the head and neck in childhood



INVESTIGATIONS



Investigations will include an ultrasound scan, usually in conjunction with an ultrasound-guided FNA. Regional and distant metastases can be assessed with a chest x-ray and CT scan. Thyroid function tests and plasma thyroglobulin levels should be obtained and also plasma calcitonin where a diagnosis of medullary carcinoma is suspected.



TREATMENT



Controversy remains over the optimum treatment in differentiated thyroid carcinoma in children because the long-term mortality in these patients is low and serious operative and postoperative complications can occur following radical surgery. These include recurrent laryngeal nerve damage, hypocalcaemia and airway obstruction requiring tracheostomy.22 It has become clear that these tumours in children are slow-growing and associated with prolonged survival rates, even in the presence of extensive disease. Some authors maintain that an aggressive approach is mandatory24, 25 while others have adopted a more conservative approach with the use of lobectomy and subtotal thyroidectomy for small and isolated lesions.22, 23, 26 Ideally, treatment should include complete surgical excision if possible. Total or subtotal thyroidectomy should be performed if adjuvent radioiodidine treatment is planned. Following surgery, a whole-body radioiodine scan is performed and ablative radioiodide treatment given if necessary. Plasma thyroglobulin can then be used as a tumour marker and suppressive levothyroxine should be given. Radiotherapy is rarely indicated in differentiated thyroid carcinoma in childhood.7



] 1257



on chromosome 10. If positive, the child should be considered for prophylactic total thyroidectomy.7 See Chapter 197, Thyroid cancer.



Nasopharyngeal carcinoma In the US and Europe, nasopharyngeal carcinoma (NPC) is an uncommon tumour comprising only 1–2 percent of paediatric malignancies, but in other geographical locations, such as parts of Africa, 10–20 percent of childhood malignancies are due to NPC. There is a bimodal age distribution of this disease with an early peak of 10 to 20 years and a second peak between 40 and 60 years. NPC is one of few malignant tumours in childhood that emerges from the epithelium and there is an association with EBV. Males are twice as likely as females to develop NPC. Children with NPC almost invariably have the undifferentiated variant that is associated with higher rates of advanced locoregional disease and distant metastases.27 Despite this, the five year disease-free survival is not significantly different to that of adults at 30–60 percent.28



PRESENTATION



Children may present with a cervical mass secondary to lymph node metastases. Other presenting symptoms and signs may include nasal congestion, epistaxis, otitis media with effusion, otalgia and cranial nerve palsy. As many as two-thirds of children with NPC have metastatic disease in the neck at presentation. Delay in the diagnosis occurs frequently because many of its symptoms mimic those of an upper respiratory tract infection.



ASSESSMENT PROGNOSIS



A long-term study of 329 patients under the age of 21 confirmed only eight deaths over a long period of time and of these only two were disease related. The risk of progression of disease was more common in younger patients and those with residual cervical disease after definitive thyroidectomy. The majority of recurrences are in cervical lymph nodes or thyroid bed (54 percent), or lungs (16 percent).26 The majority of relapses occur within the first seven years but have been seen as long as 25 years after treatment.22



Medullary thyroid carcinoma The detection of MTC in younger children is usually made following screening in ‘high risk’ individuals who have a family history of MEN 2.7 This is confirmed by elevated baseline levels of calcitonin or screening for the Ret (rearranged during transfection) protooncogene



Nasopharyngeal examination and biopsy is required for tissue diagnosis. CT or MR scanning allows precise evaluation of the primary tumour and confirmation of the presence or absence of metastases, most particularly in the neck. The American Joint Committee in Cancer staging of NPC (see Chapter 188, Nasopharyngeal carcinoma) is used by most.



TREATMENT



Undifferentiated NPC is a radiosensitive tumour and as such is usually treated with external beam radiotherapy. Such therapy is limited to the primary tumour and its regional metastatic spread. Chemotherapy is required in patients with disseminated systemic disease. There is now evidence that combined chemotherapy and radiotherapy provides better disease-free survival as compared with radiotherapy alone and combined chemoradiotherapy is therefore becoming routine practice in the UK.28, 29



1258 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Because of the tumour’s close proximity to major structures, radiotherapy in this region is difficult to give and is associated with significant post-treatment morbidity. This includes mucositis, xerostomia, neck fibrosis and panhypopituitarism (see below under Long-term sequelae of treatment). Newer radiotherapy techniques, such as intensity-modulated radiation therapy (IMRT) may offer reduced morbidity following treatment as it becomes more readily available (see Chapter 200, Developments in radiotherapy for head and neck cancer). As craniofacial surgical techniques improve, surgical resection is now becoming appropriate in certain situations, particularly in primary recurrence following treatment.



Neuroblastoma Neuroblastoma is a common malignancy of early childhood and is the most common malignancy in infants younger than one year. These tumours arise from undifferentiated sympathetic nervous system precursor cells of neural crest origin. The adrenal gland is the most common site of origin and additional sites include the sympathetic chain, posterior mediastinum and cervical regions. Neuroblastoma has a high propensity for lymphatic spread and regional lymph nodes are involved in up to 35 percent of cases. Metastases to the head and neck region are common, but primary neuroblastoma in the head and neck region is uncommon making up only 2–9 percent of the total.30



PRESENTATION



Symptoms and signs will be dictated by the size and position of both the primary lesion and any metastases. Children with primary cervical neuroblastoma may present with a firm mass in the lateral neck, occasionally associated with a Horner’s syndrome (due to cervical sympathetic chain involvement). Classical opthalmological manifestations include proptosis and periorbital ecchymosis (usually secondary to intraorbital metastatic deposits). Bilateral eye haematomas are a classical sign (racoon eyes). Interestingly, chronic destruction of the opthalmic sympathetic fibres in some children can lead to heterochromia of the irides. Involvement of the paranasal sinuses is also described.30



Figure 98.3 Metastatic neuroblastoma from an adrenal primary is extending out of both sides of the lateral orbital walls (more marked on left – arrows). Patient has incidental scaphocephaly.



and the extent of any metastatic spread (Figure 98.3). An iodine-123-metaiodobenzylguanidine (MIBG) scan is a useful method of assessing metastases. A bone marrow aspirate and trephine should be performed, with a bone scan reserved for those with a negative MIBG scan. Urinary catecholamine levels should be measured (24 hour urine collection) as they are raised in over 90 percent of cases.



TREATMENT



The choice of single modality or multimodality therapy depends on the individual patient. Localized cervical neuroblastoma may be treated by curative surgery. Multi-agent chemotherapy is usually indicated in patients if resection is incomplete or where there is evidence of metastases. The role of radiation therapy is not well defined and is usually reserved for unresectable tumours.



PROGNOSIS ASSESSMENT



A full assessment of the disease will include obtaining a histological diagnosis by means appropriate to the location of the presenting lesions. Examination should include a visual assessment, intraoral inspection and a thorough neurological examination. MR and CT scanning are usually employed to assess the primary



In young patients with resectable disease, complete excision offers the best chance of cure and at least a 90 percent survival rate. Primary neuroblastoma of the head and neck has a better prognosis than that of other sites. This is most likely due to the fact that it presents at an early stage. In disease in other sites the prognosis is poorer and it is worse in children over the age of one year at presentation.30 [**/*]



Chapter 98 Tumours of the head and neck in childhood



] 1259



Miscellaneous tumours



SQUAMOUS CELL CARCINOMA



A large variety of tumours can rarely affect the head and neck region. The small numbers involved can make meaningful analysis of data difficult and it is not really feasible to address therapeutic questions in the context of randomized controlled trials. However, the UK Children’s Cancer Study Group has formed a Rare Tumours Group which is developing guidelines for some of the rarer childhood cancers.



Squamous cell carcinoma (SCC) outside the nasopharynx is extremely rare in childhood. There is some evidence of an association between SCC and previous exposure to immunosuppressive medication, and also previous irradiation for laryngeal papillomatosis. Pre-existing xeroderma pigmentosum makes any patient extremely susceptible to the development of cutaneous SCC. The principles of management of SCC in the head and neck in children are the same as in adults.



SOFT TISSUE SARCOMAS



A variety of sarcomas rarely affect the head and neck region in children and these are listed in Table 98.2. There is a bimodal age distribution with incidence peaking in the under fives and in adolescents. Those in the younger age group tend to have lesions situated in the head and neck region while older children present with lesions in the extremities. In general, these tumours have a tendency to recur locally and to metastasize. The treatment of choice is neoadjuvant chemotherapy followed by surgery and radiotherapy, as deemed appropriate (Figure 98.4). The prognosis remains relatively poor despite multimodality treatment.19



Table 98.2



MALIGNANT TERATOMA



Malignant teratomas in the head and neck region are rare (most are benign) and usually present at birth, often as an airway emergency. Large tumours are usually diagnosed in utero, resulting in a multidisciplinary team being present at delivery to secure a safe airway. Tracheostomy may be necessary. Assessment should include nasendoscopy, alpha-fetoprotein and beta human chorionic gonadotrophin (hCG) levels, and imaging. Metastases are rare. Treatment is surgical excision although salvage chemotherapy and radiotherapy may be required. The prognosis is good in the absence of metastases.



Soft tissue sarcomas of the head and neck region other than rhabdomyosarcoma.



Histological type



Site



Presentation



Treatment and prognosis



Fibrosarcoma



Peripheral nerves



Slowly enlarging painless mass



Synovial sarcoma



Neck, larynx and oropharynx (see Figure 98.4) Foci of calcification are typical Peripheral nerves



Slowly enlarging mass often with compressive symptoms



Metastatic disease is infrequent Complete resection is treatment of choice Prognosis generally good Very rare in head and neck



Neurofibrosarcoma or maligant schwannoma



Haemangiopericytoma



Kaposi’s sarcoma Ewing’s sarcoma (extraskeletal)



Develops in 3–16 percent of children with Von Recklinghausen neurofibromatosis (NF1) Arise from pericytes of Zimmerman (capillaries) Skin lesions 1/ lymphadenopathy In soft tissue adjacent to cervical spine



Mass in neck 1/ pain and weakness



Nasal cavity and paranasal sinuses Nasal masses and epistaxis Now mainly seen in children who are HIV positive Neck mass Pain Spinal cord compression



Multimodality treatment Five-year survival approximately 50% Wide surgical excision and postoperative cranial nerve deficits common. Local recurrence and lung metastases common Poor prognosis if metastatic disease (especially in NF1)



Usually benign in infants but 20–35% malignant in older children Surgery 1/ chemotherapy and XRT Overall survival 50–70% Surgery and XRT Prognosis poor Surgery 1/ XRT and chemotherapy in similar fashion to rhabdomyosarcoma Prognosis poor



1260 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY to the maxilla in a child can result in asymmetrical growth of the midface.



Reproductive function



Figure 98.4 Cross section of a laryngectomy specimen from a 12-year-old boy with residual synovial cell sarcoma despite chemoradiotherapy.



CHORDOMA



This is a rare slow-growing bony tumour which is locally aggressive and which arises from embryonic remnants of the notochord. The presentation depends on the site of origin which, in the head and neck region, is most commonly found in the nasopharynx and adjacent skull base. Patients present with headache and diplopia and compression of the lower cranial nerves can result in a number of neurological signs (see Chapter 188, Nasopharyngeal carcinoma). Biopsy is necessary for diagnosis and MR and CT scanning will delineate the tumour extent. Complete surgical excision is rarely possible because of the anatomical location and adjacent structures and so postoperative radiotherapy is usually employed. The prognosis remains fairly poor.



An important issue for survivors of childhood cancer is the impact of the disease and its treatment on reproduction and the implications for the health of any offspring. In males there is evidence for impaired spermatogenesis after treatment but it appears that any sperm produced carries as much healthy DNA as produced by the population in general. However, it is not possible to predict fertility outcome in boys who receive treatment prior to puberty. Cryopreservation of sperm in young males (14–17 years) is effective but depends on the ability of the young patient to produce a specimen and, in the UK, consent for storage requires him to be ‘Gillick’ competent (see Chapter 63, The paediatric consultation). Radiotherapy to the hypothalamus or pituitary can result in precocious puberty in females and patients should therefore have their pubertal status checked regularly. Chemotherapy in general is less harmful to gonadal function in females but pelvic irradiation can affect ovarian function. Spontaneously conceived offspring of patients treated for cancer in childhood have no excess of congenital abnormalities or other diseases.32



Cardiac problems Anthracyclines have a significant cardiotoxic effect and can cause cardiac failure in later life. Long-term echocardiogram surveillance is recommended. Mediastinal radiotherapy can also result in impaired cardiac function (and an increased risk of breast carcinoma) and this should be monitored as appropriate.



LONG-TERM SEQUELAE OF TREATMENT The aim of cancer treatment in children is to maximize the chance of long-term surival and at the same time minimize the side effects, particularly in the longer term. With overall five-year survival rates in children now in the order of 60–70 percent, it is particularly important to examine the effects of the more recently employed multimodality treatments. Children tolerate the acute side effects of radiotherapy and chemotherapy reasonably well but other sequelae may not become apparent for several years.31



Thyroid disorders Thyroid dysfunction can result from radiotherapy to the neck or to the hypothalmic/pituitary axis. Chemotherapy is an independent risk factor. Survivors require regular thyroid function evaluation. Radiotherapy to the neck is a recognized aetiological factor in the development of thyroid carcinoma in later years.



Growth



Other otolaryngological manifestations of treatment



Cranial radiotherapy can result in growth hormone deficiency and growth retardation and chemotherapy can also have significant effects on growth. Localized tumour treatments can also affect growth and function of specific organs or tumour sites. For example, radiotherapy



Radiotherapy to the head and neck region results in a large variety of long-term sequelae that can affect function and cosmesis. Damage to the major and minor salivary glands can cause xerostomia and subsequent tooth and gum disease. Post-radiotherapy scarring within



Chapter 98 Tumours of the head and neck in childhood



the nasopharynx can result in middle ear effusions and subsequent deafness, and in the region of the midface can result in recurrent sinusitis, temperomandibular joint dysfunction and trismus. Certain chemotherapeutic agents – notably cisplatin – are known to cause deafness and tinnitus. Immunodeficiency with a propensity to develop opportunistic infections of the head and neck may complicate some chemotherapy regimens.31



Cognitive and psychological problems Surgery and/or radiotherapy to the brain or adjacent structures can result in neurocognitive defects such as low IQ, learning difficulties and an increased risk of seizures. Survivors of childhood cancer are at an increased risk for a wide range of disabling psychological problems such as low mood, low self-esteem and anxiety.



] 1261



able, mixing with friends and getting a break from the home enviroment. Others may benefit from a daily visit at home from a play specialist.33, 34 Symptomatic control in the dying child has not been formally assessed and is often based on clinical experience and adapted from general paediatric practice and palliative care of adults. The World Health Organization’s ‘three stepladder’ of analgesia can be applied in children with paracetamol, dihydrocodeine and morphine sulphate forming the standard steps. Laxatives and antiemetics may be required and appropriate sedatives may be required in the final stages of life.34 Regular respite should be offered to parents and it is important to explore the financial assistance, benefits and grants that families may be entitled to. Practical help with funeral arrangements should be given and it is important to continue to support the grieving family after their child’s death and to withdraw support slowly and appropriately.11, 33, 34



Follow-up It is important to follow-up these children in the long term and to educate them and their parents about the possible late consequences of their treatment. It is also important to promote a healthy lifestyle and discourage cancer promoting behaviours such as smoking and excessive sun exposure.32



THE DYING CHILD The death of a child is one of the greatest tragedies that can befall a family. It is important to recognize that palliative care of the dying child should address not only the control of pain but the social, psychological and spiritual needs of children and their families. Planned terminal care for children has become increasingly community-based and therefore requires involvement of the primary health care team, including the patient’s general practitioner at an early stage. An experienced and multiskilled team is required with a named key worker and access to advice and support 24 hours a day. The child should participate in the planning and provision of this care as much as is realistically possible.33 To ensure optimum palliative care, the family and carers need to work with identical aims and a mutual acceptance of the inevitability of death. An appropriate person should discuss the fact that active curative treatment is no longer in the child’s best interests and should start to negotiate a palliative care plan with the child and close family. Once the focus of treatment shifts from curative to palliative care, quality of life becomes of prime importance. Most children and their families will wish to maintain some semblance of normality while optimizing symptom control and minimizing medical intervention. For instance, many children will opt to remain at school for as long as they are



KEY POINTS  Head and neck tumours are uncommon in children.  Early diagnosis is essential.  Survival rates are improving.  Multimodality treatment has improved the prognosis while reducing both early and long-term complications.  Treatment should be provided by a specialized, centralized multidisciplinary paediatric oncology team.



Best clinical practice [ The vast majority of enlarged cervical lymph nodes in



[ [ [ [



children are harmless. Imaging should usually be considered before biopsy. Excision biopsy may be appropriate for: – node 42 cm; – supraclavicular area and/or fixed nodes; – weight loss and/or unexplained fever; – abnormal chest x-ray. [Grade C/D] FNA has a more limited role to play in the child than the adult. [Grade B] All children diagnosed with malignancy should be referred to a specialist centre. [Grade B] Children receiving chemotherapy should have central venous catheters placed for venous access and chemotherapy treatment. [Grade C/D] Disease-free survival is over 90 percent in many series of children with Hodgkin’s lymphoma. [Grade B]



1262 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



ACKNOWLEDGEMENTS Deficiencies in current knowledge and areas for future research



$ $ $ $ $ $ $



$



$



The long-term survival of children with cancer is improving, and a better appreciation of the long-term sequelae of treatment should result in an improved quality of life for children surviving cancer. We are likely to encounter more physical and psychological problems in adults related to cancer treatment in childhood. Basic science research should bring better understanding of the underlying aetiology of childhood cancer. This may help in introducing preventative measures such as vaccines and environmental interventions. Agreed protocols on the management of paediatric thyroid malignancy would be helpful. Centralization of services, the establishment of better databases and the implementation of multicentre trials will provide us with a better evidence base for the future management of childhood cancer. Better imaging techniques with wider access to sophisticated equipment will aid in more accurately assessing the initial disease and in detecting recurrence at an earlier stage. Greater knowledge of molecular genetics has resulted in new approaches to diagnosis and treatment. The advent of polymerase chain reaction (PCR) techniques has, for example, permitted the detection of cells bearing chromosomal translocations in the peripheral blood and bone marrow, thereby detecting minimal residual disease in lymphoma and even defining patients at high risk of relapse. Other novel strategies include the use of immunotherapy (immunotoxins or vaccines) to target specific cells or cell proteins. Radiotherapy techniques such as IMRT are allowing much greater flexibility in targeting tumours in awkward anatomical sites whilst avoiding some of the post-treatment complications. IMRT is an advanced mode of high precision radiotherapy where combinations of several intensity-modulated fields coming from different beam directions deliver a maximal dose to the tumour while minimizing the dose to surrounding tissues. Surgical advances, such as navigation systems with the use of endoscopes, mean that areas once inaccessible can now be approached surgically whilst maintaining good function and cosmesis.



In conclusion, we have made major advances in the investigation and management of childhood malignancy in the last 30 years and a number of promising diagnostic and treatment developments promise much in the next 30.



I wish to thank Dr Dermot Murphy, Consultant Paediatric Oncologist, Royal Hospital for Sick Children, Glasgow, for his comments on the manuscript and Dr Greg Irwin, Consultant Paediatric Radiologist, Royal Hospital for Sick Children, Glasgow, for providing the MR images.



REFERENCES







1. Albright JT, Topham AK, Reilly JS. Pediatric head and neck malignancies: US incidence and trends over 2 decades. Archives of Otolaryngology – Head and Neck Surgery. 2002; 128: 655–9. 2. Gurney JG, Severson RK, Davis S, Robison LL. Incidence of cancer in children in the United States. Sex, race, and 1-year age-specific rates by histologic type. Cancer. 1995; 75: 2186–95. 3. Cotterill SJ, Parker L, Malcolm AJ, Reid M, More L, Craft AW. Incidence and survival for cancer in children and young adults in the North of England, 1968-1995: a report from the Northern Region Young Persons’ Malignant Disease Registry. British Journal of Cancer. 2000; 83: 397–403. 4. Josephson GD, Wohl D. Malignant tumors of the head and neck in children. Current Opinion in Otolaryngology, Head and Neck Surgery. 1999; 7: 61. 5. Cunningham MJ, Myers EN, Bluestone CD. Malignant tumors of the head and neck in children: a twenty-year review. International Journal of Pediatric Otorhinolaryngology. 1987; 13: 279–92. 6. Narod SA, Stiller C, Lenoir GM. An estimate of the heritable fraction of childhood cancer. British Journal of Cancer. 1991; 63: 993–9. 7. Whittemore KR, Cunningham MJ. Malignant tumors of the head and neck. In: Bluestone CD, Stool SE (eds). Pediatric otolaryngology, 4th edn. Philadelphia: Saunders, 2003: 1703–36. 8. Soldes OS, Younger JG, Hirschl RB. Predictors of malignancy in childhood peripheral lymphadenopathy. Journal of Pediatric Surgery. 1999; 34: 1447–52. A useful study providing guidance on when to biopsy lymph nodes in children 9. Karadeniz C, Oguz A, Ezer U, Ozturk G, Dursun A. The etiology of peripheral lymphadenopathy in children. Pediatric Hematology and Oncology. 1999; 16: 525–31. 10. Kubba H. Summary of the proceedings of the VII annual meeting of the evidence-based management of head and neck cancer. Clinical Otolaryngology and Allied Sciences. 2005; 30: 79–85. 11. Langton H (ed.). The child with cancer. London: Balliere Tindall, 2000. 12. Tarantino DR, McHenry CR, Strickland T, Khiyami A. The role of fine-needle aspiration biopsy and flow cytometry in



Chapter 98 Tumours of the head and neck in childhood



 13. 14. 15.



 16.



17.



18.



19.



 20.



21.



 22.



the evaluation of persistent neck adenopathy. American Journal of Surgery. 1998; 176: 413–7. Hudson MM, Donaldson SS. Hodgkin’s disease. Pediatric Clinics of North America. 1997; 44: 891–906. An excellent review of Hodgkin’s disease. Patte C. Non-Hodgkin’s lymphoma. European Journal of Cancer. 1998; 34: 359–62. Louw G, Pinkerton CR. Interventions for early stage Hodgkin’s disease in children. The Cochrane Database of Systemic Reviews. 2005; 2. Shad A, Magrath I. Non-Hodgkin’s lymphoma. Pediatric Clinics of North America. 1997; 44: 863–90. A helpful review on management and treatment of non-Hodgkin’s lymphoma. McGill T. Rhabdomyosarcoma of the head and neck: an update. Otolaryngologic Clinics of North America. 1989; 22: 631–6. Coene IJ, Schouwenburg PF, Voute PA, Marion J, Burgers V, Hilgers FJ. Rhabdomyosarcoma of the head and neck in children. Clinical Otolaryngology and Allied Sciences. 1992; 17: 291–6. Lyos AT, Goepfert H, Luna MA, Jaffe , Malpica A. Soft tissue sarcoma of the head and neck in children and adolescents. Cancer. 1996; 77: 193–200. Pappo AS, Shapiro DN, Crist WM. Rhabdomyosarcoma. Biology and treatment. Pediatric Clinics of North America. 1997; 44: 953–72. Provides an overview of the staging, management and prognosis of children with rhabdomyosarcoma. Koscielniak E, Jurgens H, Winkler K, Burger D, Herbst M, Keim M et al. Treatment of soft tissue sarcoma in childhood and adolescence. A report of the German Cooperative Soft Tissue Sarcoma Study. Cancer. 1992; 70: 2557–67. Newman KD, Black T, Heller G, Azizkhan RG, Holcomb 3rd GW, Sklar C et al. Differentiated thyroid cancer: determinants of disease progression in patients o21 years of age at diagnosis: a report from the Surgical Discipline Committee of the Children’s Cancer Group. Annals of Surgery. 1998; 227: 533–41. A study looking at a very large cohort of children and adolescents with differentiated thyroid carcinoma.



] 1263



23. Viswanathan K, Gierlowski TC, Schneider AB. Childhood thyroid cancer. Characteristics and long-term outcome in children irradiated for benign conditions of the head and neck. Archives of Pediatrics and Adolescent Medicine. 1994; 148: 260–5. 24. Segal K, Shvero J, Stern Y, Mechlis S, Feinmesser R. Surgery of thyroid cancer in children and adolescents. Head and Neck. 1998; 20: 293–7. 25. Jocham A, Joppich I, Hecker W, Knorr D, Schwarz HP. Thyroid carcinoma in childhood: management and follow up of 11 cases. European Journal of Pediatrics. 1994; 153: 17–22. 26. La Quaglia MP, Corbally MT, Heller G, Exelby PR, Brennan MF. Recurrence and morbidity in differentiated thyroid carcinoma in children. Surgery. 1988; 104: 1149–56. 27. Ayan I, Altun M. Nasopharyngeal carcinoma in children: retrospective review of 50 patients. International Journal of Radiation Oncology, Biology, Physics. 1996; 135: 485–92. 28. Gasparini M, Lombardi F, Rottoli L, Ballerini E, Morandi F. Combined radiotherapy and chemotherapy in stage T3 and T4 nasopharyngeal carcinoma in children. Journal of Clinical Oncology. 1988; 6: 491–4. 29. Mertens R, Granzen B, Lassay L, Gademann G, Hess CF, Heimann G. Nasopharyngeal carcinoma in childhood and adolescence. Cancer. 1997; 80: 951–9. 30. Haase GM. Head and neck neuroblastoma. Seminars in Pediatric Surgery. 1994; 3: 194–202. 31. Fromm M, Littman P, Raney RB, Nelson L, Handler S, Diamond G et al. Late effects after treatment of twenty children with soft tissue sarcomas of the head and neck. Experience at a single institution with a review of the literature. Cancer. 1986; 57: 2070–6. 32. Long term follow up of survivors of childhood cancer. A national clinical guideline. Scottish Intercollegiate Guidelines Network. 2004: 76. 33. Anonymous. American Academy of Pediatrics. Committee on Bioethics and Committee on Hospital Care. Palliative care for children. Pediatrics. 2000; 106: 351–7. 34. Goldman A. ABC of palliative care. Special problems of children. British Medical Journal. 1998; 316: 49–52.



99 Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma PETER D BULL



The branchial arches First branchial arch (cervico-aural) fistula Second branchial arch fistula Third and fourth branchial arch abnormalities Disorders of the thyroglossal duct Lingual thyroid



1264 1264 1266 1267 1268 1268



Lymphangioma Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



1269 1271 1271 1271 1271



SEARCH STRATEGY The author has a personal bibliography which was supplemented by a Medline search using the key words branchial apparatus, child, congenital, thyroglossal duct, thyroglossal cyst, cystic hygroma and lymphangioma.



THE BRANCHIAL ARCHES Embryology The structures of the lateral face and neck are formed embryologically from the paired branchial arches, pouches and associated clefts (see Chapter 65, Head and neck embryology). The branchial apparatus appears between the fourth and fifth weeks of foetal development. It consists of six paired branchial arches separated by branchial clefts externally and branchial pouches internally. During normal development, the ventral parts of the first and second arches fuse to obliterate the ventral part of the first cleft. The remainder of the first cleft forms the cavum conchae and the external auditory meatus. The first branchial pouch forms the Eustachian tube and tympanic cavity. The fifth and sixth arches disappear early in foetal life. [****]



Terminology Persistence of remnants of the branchial apparatus gives rise to a number of well-recognized congenital anomalies in the head and neck. A persistent cleft will give rise to an external sinus, a blind ending opening onto the skin. A persistent pouch will cause an internal sinus typically opening into the pharynx whereas persistence of both cleft and pouch will cause a fistula with an internal and external (or upper and lower) opening. Anomalies that cause a single opening or sinus or the persistence of a tissue remnant are more common than fistulae and include pre-auricular sinuses, cysts and skin tags.



FIRST BRANCHIAL ARCH (CERVICO-AURAL) FISTULA This is rare and accounts for less than 5 percent of branchial cleft anomalies. The external fistulous opening



Chapter 99 Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma



in the neck varies in position but lies on a line from the tragus to the hyoid bone. The opening is often fairly inconspicuous and may look like a small skin-lined pit. The upper end of the fistula is also variable. There may be an opening anterior to the tragus; the track of the fistula may run under the floor of the external auditory canal and occasionally opens at the osseo-cartilaginous junction in the external auditory meatus. In some cases, a skin-covered band runs from the floor of the meatus to the umbo. If seen, it makes the diagnosis certain.



Histology The cervico-aural fistula is usually a sizeable track lined with squamous epithelium. Work1 has divided them into two types according to the presence or absence of mesothelial elements within the wall. Type 1 lesions are of ectodermal origin and are present medial to the concha. Type 2 lesions are both ectodermal and mesodermal in origin and contain mesodermal structures such as cartilage and hair follicles (Figure 99.1). [****] The lower opening of the type 2 fistula is usually below the angle of the mandible.



] 1265



Investigation If there is doubt about the nature of the fistula, a radioopaque sinogram using water-soluble contrast will demonstrate the extent of the track and confirm the position of the upper end (Figure 99.3). MR scanning may show the track but will not demonstrate its relationship to the facial nerve.



Treatment Because the fistulous track is lined with squamous epithelium, it will contain desquamated material, which may become infected. Surgical excision of these lesions is usually advocated. Surgery is easier before either acute infection or abscess has supervened. [**/*] Surgical excision necessitates the dissection of the track in its entirety. [*]



Presentation The opening at either end of the track is present at birth and may become more obvious by the discharge of epithelial or sebaceous debris (Figure 99.2). The fistula may become acutely infected and may rarely progress to abscess formation. Being lined with skin, cervico-aural fistulae do not discharge mucus.



Figure 99.2 The external opening of a cervico-aural fistula.



Figure 99.1



Figure 99.3 Sinogram showing the course of a cervico-aural fistula.



Histology of a first arch cervico-aural fistula.



1266 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY A parotidectomy type incision is modified to include the lower opening of the fistula. The fistula at this point can be followed for a small distance upwards and when its nature and direction are confirmed, the facial nerve trunk must be identified as it leaves the skull base and enters the parotid gland. The mandibular branch in particular should be followed so that it can be preserved while freeing the fistulous track. Sharp dissection will often be required to free the nerve from the track. In most instances the fistula will run deep to the branches of the facial nerve, which may be adherent if there has been recurrent infection, though the relationship is unpredictable (Figure 99.4). The upper part of the first arch fistula also has an unpredictable relationship to the cartilages of the external auditory canal (EAC). It may terminate at the canal or may form a duplicated canal parallel and inferior to the EAC proper. The fistula does not extend medial to the cartilaginous canal but a strand may cross the lumen of the canal from the osseo-cartilaginous junction to be attached to the umbo. Facial nerve exposure and protection is required in nearly every case of first branchial cleft fistula. If this is done, then the risk to the facial nerve is minimized. Peroperative facial nerve monitoring is advisable. [*] The very superficial position of the facial nerve in small children must be considered when making the approach to the nerve. The upper end of the fistula must be followed and excised to ensure that no squamous epithelium remains, and this may require opening of the external auditory canal. Closure of the wound with suction drainage and, if necessary, packing of the ear canal completes the operation.



branchial arch forms the epidermis of the upper neck and dorsal pinna. The mesoderm forms the facial muscles and the body of the hyoid, and the endodermal elements form the root of the tongue, the foramen caecum, the thyroid stalk and the tonsil. Malfusion of these structures may lead to the formation of the second arch branchial fistula. Second branchial arch fistula may form part of branchiooto-renal syndrome.2, 3



Presentation The second branchial arch fistula presents as a congenital opening on the lower neck, anterior to the sternomastoid muscle. The track of the second arch fistula is directed upwards and medially to pass between the internal and external carotid arteries. The upper end communicates with the pharynx through the tonsil (Figure 99.5). These fistulae nearly always leak clear or mucoid fluid from the lower opening and may become infected. On occasion, the infected track may form an abscess requiring surgical drainage.



Investigation As a rule, the diagnosis is evident and no further investigation is required. If there is doubt, a contrast sinogram can be performed to define the track.



Treatment



Second arch fistulae and sinuses are much the commonest of the branchial arch anomalies (95 percent). The second



Because of the risk of infection, excision of second arch branchial fistulae is advisable. Surgery can be performed at any age and aims to excise the track completely. An elliptical incision is made around the external opening and the track can be followed upwards. If the distance between the lower end and the pharynx is too great to



Figure 99.4 The surgical exposure of the fistula, crossed by the facial nerve.



Figure 99.5 A branchial fistula removed in continuity with the tonsil.



SECOND BRANCHIAL ARCH FISTULA



Chapter 99 Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma



allow good access, a second skin crease incision is made further up the neck to enable the upper part of the track to be visualized and excised. The track will be seen to enter the pharyngeal muscle at the level of the tonsil and can be ligated and excised at this level. Sometimes the track can be followed into the tonsil, which can be removed intraorally in continuity (Figure 99.6). Care must be taken to identify the lingual nerve in the upper part of the dissection. Complete excision of the fistula should prevent any recurrence.



] 1267



the piriform fossa (Figure 99.7). The appearance may be that of an acute suppurative thyroiditis. Typically, there is diagnostic delay and the child may have had several surgical interventions. Treatment requires endoscopy of the piriform fossa and complete excision of the track and the inflammatory mass.4, 5 [*]



Pre-auricular sinus EMBRYOLOGY



THIRD AND FOURTH BRANCHIAL ARCH ABNORMALITIES The third and fourth branchial arches form part of the hyoid bone and the muscles of the larynx and pharynx. The pouches also form the thymus gland, the parathyroids and part of the thyroid. Agenesis gives rise to the ‘DiGeorge’ anomaly. This is characterized by partial or complete aplasia of the thyroid and parathyroid glands, often with associated craniofacial and cardiac anomalies. Sinuses and fistulae of these arches are rare. Presentation is usually as an abscess in the neck, most commonly on the left, communicating with the apex of



The outer ear is formed from cartilagenous tubercles of first arch origin which fuse to form the pinna. A blindended sinus results from incomplete fusion and the inclusion of epithelial tissue forms a skin lining to the sinus. There may be a family history of such anomaly. The branchio-oto-renal syndrome is determined by an autosomal dominant gene and includes external ear abnormality, pre-auricular sinus and renal disorder.2, 3



PRESENTATION



The opening of the sinus is apparent at birth and is often bilateral. There may be some sebaceous discharge from the punctum. As a rule, these lesions give little trouble and can be safely left alone. In some patients there is recurrent episodic infection which may progress to abscess formation. Because the sinus is lined with squamous epithelium, spontaneous resolution does not take place.



TREATMENT



If the sinus is free of infection, it can be left alone, but many of the cases will be subject to repeated episodes of



Figure 99.6 A case of branchio-oto-renal syndrome, showing the external opening of a second arch branchial fistula and deformity of the outer ear.



Figure 99.7 Sinus opening at apex of left piriform fossa in a case of fourth branchial arch sinus. The patient had presented with recurrent neck abscesses.



1268 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY infection and should be excised. An ellipse of skin around the punctum is marked out and the track followed carefully. It can be helpful to instil some methylene blue dye into the track to make it more visible during the dissection. The aim of the operation is to remove all vestiges of squamous epithelium, so the track must be dissected out meticulously and in one piece.



DISORDERS OF THE THYROGLOSSAL DUCT Considered under this heading are:  thyroglossal duct cyst;  thyroglossal duct fistula;  lingual thyroid.



Embryology The thyroid primordium develops in the floor of the primitive pharynx at the site of the foramen caecum, which lies in the midline at the junction of the posterior one-third and anterior two-thirds of the tongue. The descending thyroid gland migrates through the tongue tissue, passing anterior to the hyoid bone to lie in the anterior neck. The line of descent from the foramen caecum is marked by the thyroglossal duct. If involution of the duct is incomplete, a cyst may develop at any point along the track.



Presentation Surprisingly for a congenital lesion, presentation of a thyroglossal cyst is often delayed well beyond infancy and into early childhood. The cyst is usually midline but in 10 percent of cases may present laterally, usually on the left.6 Protrusion of the tongue, by elevating the hyoid bone, results in the cyst (or fistulous opening) being pulled upwards (Figures 99.8 and 99.9).



Figure 99.8



A thyroglossal duct fistula.



Investigation If there is doubt about the diagnosis, ultrasound examination will clarify the nature and position of the cyst. Aspiration is unnecessary and may lead to infection and fistula formation. [*] The thyroglossal duct cyst, while derived from the descending thyroglossal track, does not contain the only functioning thyroid gland tissue. It may contain thyroid tissue in the wall, but unless there is a lingual thyroid, the thyroid gland will be in the normal position. It is not necessary to do isotope thyroid scans or thyroid function tests before surgery for thyroglossal cyst.



Treatment Treatment of thyroglossal duct cyst is by excision. The optimum operation is as described by Sistrunk.7 [**] A horizontal skin crease incision is made over the convexity of the cyst. If there is a fistula, an ellipse of skin is taken to include the opening. The cyst or track is then dissected out and the middle segment of the hyoid exposed and skeletonized. The thyroglossal duct goes through or deep to the body of the hyoid bone, so the body is excised in continuity. The track is then followed through the tongue in the midline. If no track is visible, a core of tissue is removed up to the foramen caecum. Removal of a thyroglossal cyst results in a recognized recurrence rate of between 4 and 10 percent. [**] Revision surgery is more unpredictable and requires an en bloc removal of inflamed tissue. Thyroid tissue has been observed histologically in 46 percent of cases of excised thyroglossal duct cysts.8



LINGUAL THYROID If the thyroid primordium fails to descend into the neck, thyroid tissue will remain at the foramen caecum. There may therefore be no thyroid tissue at the usual pretracheal



Figure 99.9 A thyroglossal duct fistula with the tongue protruded, showing the pull on the fistulous track.



Chapter 99 Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma



site and all the thyroid gland is represented by that at the foramen caecum.9



Presentation Lingual thyroid is extremely rare. In a review of thyroid abnormality in 29,000 autopsies, only four lingual thyroids were identified.10 [****] It usually presents as a mass on the base of the tongue in infancy or childhood, leading to dysphagia or airway compromise (Figure 99.10). It may be found coincidentally on isotope scanning in the investigation of hypothyroidism. Rarely, a lingual thyroid may present acutely as a result of bleeding into the ectopic thyroid tissue.



Investigation  MR or CT scanning will demonstrate the lesion in the sagittal or coronal plane (Figure 99.11).  Radioactive iodine scanning with I123 or I131 will demonstrate active thyroid tissue.  Thyroid function tests will usually demonstrate the presence of hypothyroidism.



Figure 99.10



A lingual thyroid.



] 1269



Treatment If the lingual thyroid is small and causing no symptoms, suppression of thyroid stimulating hormone (TSH) with oral thyroxine may result in reduction in size. [**] The advice of an endocrinologist, particularly in children, is essential. If the lesion is larger and causing dysphagia or airway obstruction, surgical excision is required. Since thyroid replacement therapy is in most cases necessary in any case, there need be no concern at removing all extant thyroid tissue. There is no hazard to the parathyroid glands, which are embryologically from a different origin and will not be within the lingual thyroid tissue. Peroral access to the lingual thyroid is usually possible, with midline splitting of the dorsal tongue to gain access to the substance of the lingual thyroid. If the ectopic tissue is more deeply buried, a suprahyoid external approach may be necessary. Postoperatively, surveillance of thyroid function is essential. [**]



LYMPHANGIOMA Embryology Lymphangioma (LA) refers to a group of developmental anomalies in which the lymphatics fail to connect in the normal way to the venous channels. As a result, there are large fluid-filled spaces occupying the tissues and expanding tissue planes. Histologically, these lesions consist of widely dilated lymphatic spaces with walls of varying thickness. The lesions may be macrocystic or microcystic. The macrocystic lesions are often large and compressible, though frequently multilocular, whereas the microcystic type is often smaller and firmer. The macrocystic lesion, with its well-defined margin, is commonly referred to as ‘cystic hygroma.’ Because the failure of proper development of the venous lymphatic relationship manifests most frequently in the developing paired jugular lymph sac systems, cystic hygroma is most common in the face, neck and mediastinum.



Presentation



Figure 99.11 CT scan showing the lingual thyroid.



The hygroma is usually evident at birth as a smooth, soft swelling, which may be fluctuant. Being filled with fluid, it can often be transilluminated. The hygroma may be enormous and cause immediate and fatal airway obstruction. It can be diagnosed antenatally by ultrasonography; steps to secure the airway can then be planned prior to delivery. The swelling may enlarge during childhood, sometimes triggered by intercurrent upper respiratory infections (Figure 99.12). This may lead in turn to acute infection within the hygroma, which can become tense and hard. If this happens to the



1270 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Figure 99.12 (a) and (b) Cystic hygroma of the left neck.



Figure 99.13 Histology of cystic hygroma infiltrating the parotid tissue.



intraoral or intralingual part, severe airway obstruction will result. If the lesion is extensive in the head and neck, there may be intraoral lesions forming small clear cysts in the mucosa. These are sometimes vascular and may bleed spontaneously but can be removed easily with laser or diathermy.



Treatment If the cystic hygroma is small and is cosmetically and functionally acceptable, no treatment other than observation will be required. [**]



Sclerotherapy Natural history Cystic hygromas of the head and neck do not involute spontaneously. Some progression may take place with enlargement, and infiltrative growth into previously unaffected areas sometimes happens. Recurrent episodes of acute infection may arise and the lesion will become larger, painful and inflamed. As a result of this infection, related lymph nodes can become enlarged. The activity in cystic hygroma may subside with increasing age. There is no potential for malignant change. Rarely, sequestration of fluid within a cystic hygroma, especially in neonates, may result in hypovolaemic shock and widespread intravascular thrombosis.



Macrocystic disease is amenable to aspiration and the instillation of sclerosant. Most experience is with OK 432, although other agents such as alcohol, bleomycin, ethanolamine oleate and trichloroacetic acid have been used.12 OK 432 is a lyophilized incubation material derived from group A pyogenic streptococcal culture, which induces an intense inflammatory reaction within the hygroma and obliteration of the cystic space. In a multicentre trial of OK 432,13 a group of patients received four doses of the material at six to eight week intervals. There was a successful outcome in 86 percent of the cases, predominantly for macrocystic disease. [**/*]



Surgery Investigation Although as a rule the diagnosis is apparent and no particular investigations are needed, ultrasonography will reveal the cystic nature of the lesion and its relationship to the surrounding tissues. It will also distinguish between macro- and microcystic disease. Cross-sectional imaging will be required if surgery is contemplated. [*] MR scanning is preferable as it provides better soft tissue delineation and avoids exposure to ionizing radiation.11 It will normally need to be carried out under general anaesthesia. There is little or no value in fine-needle aspiration for cytology unless underlying neoplasia is suspected.



Microcystic disease or macrocystic disease not amenable to sclerotherapy presents a formidable challenge (Figure 99.13). It will usually require surgical resection and complete extirpation may be impossible. A long-term review from Copenhagen indicated that 50 percent of such patients had residual disease, 44 percent had impairment of speech, swallowing or breathing and 36 percent had cosmetic deformity that caused them concern.14 [**/*] Surgery requires great attention to anatomical structures, which are enveloped in the hygroma tissue. Residual or recurrent hygroma is more likely in suprahyoid disease because of the more complex anatomy. Neurovascular structures should not be damaged in an



Chapter 99 Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma



attempt to remove every last remnant, since this will have long-term adverse consequences for the patient.15



REFERENCES



KEY POINTS  A knowledge and understanding of the embryology and often complex anatomy of congenital anomalies of the head and neck is essential before treating these disorders.  Investigation is not always required but careful imaging in consultation with a paediatric radiologist may be helpful in planning treatment.  The term ‘cystic hygroma’, though commonly used, refers to a group of disorders which are better considered as ‘lymphangiomas’.







Best clinical practice [ In first arch (cervico-aural) fistula, surgical excision



[ [



[ [



necessitates the dissection of the track in its entirety. Facial nerve exposure and protection is required in almost every case. [Grade D] Because of the risk of infection, excision of second arch branchial fistulae is advisable. [Grade D] Simple excision of a thyroglossal duct cyst is inadequate. The body of the hyoid must be excised and the track followed to the foramen caecum. [Grade C] Suppression of TSH with oral thyroxine may result in regression of a lingual thyroid. [Grade C] In lymphangioma surgery, neurovascular structures should not be damaged in an attempt to remove every last remnant. [Grade D]



Deficiencies in current knowledge and areas for future research



$ $ $



Thyroglossal duct abnormalities are increasingly dealt with by otolaryngologists with training and expertise in head and neck surgery; recurrence due to inadequate primary surgery should become less common. Continuing improvements in medical imaging should help treatment planning for complex head and neck abnormalities. The results of continuing studies on sclerotherapy for lymphangioma will help define the place of this method of treatment of these challenging disorders.



] 1271







1. Work WP. Newer concepts of first branchial cleft defects. Laryngoscope. 1972; 82: 1581–93. 2. Chen A, Francis M, Ni L, Cremers CW, Kimberling WJ, Sato Y et al. Phenotypic manifestations of branchio-oto-renal syndrome. American Journal of Medical Genetics. 1995; 58: 365–70. 3. Kemperman MH, Stinckens C, Kumar S, Joosten FB, Huygen PL, Cremers CW. The branchio-oto-renal syndrome. Advances in Oto-Rhino-Laryngology. 2002; 61: 192–200. 4. Nicollas R, Ducroz V, Garabedian EN, Triglia JM. Fourth branchial pouch anomalies: a study of six cases and review of the literature. International Journal of Pediatric Otorhinolaryngology. 1998; 44: 5–10. 5. Rea Pa, Hartley BE, Bailey CM. Third and fourth branchial arch anomalies. Journal of Laryngology and Otology. 2004; 118: 19–24. Good review of classification and treatment of these disorders. 6. Brewis C, Mahadadevax M, Bailey CM, Drake DP. Investigation and treatment of thyroglossal duct cysts in children. Journal of the Royal Society of Medicine. 2000; 93: 18–21. 7. Sistrunk WE. Technique of removal of cysts and sinuses of the thyroglossal duct. Annals of Surgery. 1920; 71: 121–4. 8. Chandra RK, Maddalozzo J, Kovarik P. Histological characterization of the thyroglossal tract: implications for surgical management. Laryngoscope. 2001; 111: 1002–5. 9. Vazquez E, Enriquez G, Castellote A, Lucaya J, Creixell S, Aso C et al. Ultrasound CT and MRI imaging of the neck lesions in children. Radiographic. 1995; 15: 105–22. 10. Williams ED, Toyn CE, Harach HR. The ultimobranchial gland and congenital thyroid abnormalities in man. Journal of Pathology. 1989; 159: 135–41. 11. Duncan AW. Congenital neck masses. In: King SJ, Boothroyd AE (eds). Paediatric ENT radiology. Berlin Heidelberg, New York: Springer Verlag, 2002: 175–98. Good overview of the role of imaging in evaluating neck masses in children. 12. Kim KH, Sung MW, Roh JC, Han MH. Sclerotherapy for congenital lesions in the head and neck. Otolaryngology Head and Neck Surgery. 2004; 131: 307–16. 13. Giguere CM, Bauman NM, Sato Y, Burke DK, Greinwald JH, Pransky S et al. Treatment of lymphangiomas with OK-432 (Picibanil) sclerotherapy: a prospective multi-institutional trial. Archives of Otolaryngology – Head and Neck Surgery. 2002; 128: 1137–44. 14. Charabi B, Bretlau P, Bille M, Holmelund M. Cystic hygroma of the head and neck – a long-term follow-up of 44 cases. Acta Oto-Laryngologica – Supplement. 2000; 543: 248–50. 15. Kennedy TL, Whitaker M, Pellitteri P, Wood WE. Cystic hygroma/lymphangioma: a rational approach to management. Laryngoscope. 2001; 111: 1929–37.



100 Gastro-oesophageal reflux and aspiration HAYTHAM KUBBA



Gastro-oesophageal reflux Key points Best clinical practice Deficiencies in current knowledge and areas for future research Aspiration



1272 1276 1276 1277 1277



Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



1278 1278 1278 1279



SEARCH STRATEGY Systematic reviews were identified first using the key words: reflux (gastric, gastro-oesophageal, gastropharyngeal, gastronasopharyngeal, gastrolaryngopharyngeal); and subject headings in the Cochrane Database of Systematic Reviews, the Database of Abstracts of Reviews of Effectiveness, and the ACP Journal Club. Individual articles were then identified using the same search terms in Ovid Medline. Articles were also identified from bibliographies and from the author’s own collection.



GASTRO-OESOPHAGEAL REFLUX



Natural history



Definitions



Physiologic reflux is almost universal in neonates. These reflux episodes are usually postprandial, brief and asymptomatic. One study showed healthy preterm infants with no symptoms or signs of reflux having between 40 and 145 (median 72) reflux events per 24 hour period, and these were more common immediately after feeds.1 Reflux becomes less frequent at certain predictable developmental milestones, such as sitting and weaning onto solids. At the age of four months, 67 percent of infants have at least one episode of regurgitation per day as reported by their parents: by the age of seven months this figure has fallen to 21 percent, and at a year regurgitation will only be present in 5 percent.2 Heartburn and regurgitation are reported by around 2 percent of three to nine year olds and 5–8 percent of 10–17 year olds, although few seek or receive treatment.3 It is likely that a significant proportion of these children will go on to have gastro-oesophageal reflux as adults.4, 5



Gastro-oesophageal reflux is the retrograde passage of stomach contents into the oesophagus. The refluxate need not be acidic, and is often neutral or even alkaline. Some degree of reflux is extremely common in neonates and usually has few, if any, symptoms: this is termed physiologic reflux. Pathologic reflux, on the other hand, is synonymous with gastro-oesophageal reflux disease and implies that the child is coming to harm in some way. This may be in the form of symptoms, or as tissue damage in the absence of symptoms (silent reflux). Secondary reflux occurs in a child predisposed by the presence of other abnormalities of the oesophagus, such as tracheooesophageal fistula or neuromuscular disease. Gastropharyngeal reflux ascends above the level of the upper oesophageal sphincter and has been implicated in a number of respiratory and otolaryngological diseases.



Chapter 100 Gastro-oesophageal reflux and aspiration



Pathophysiology The short length of the intra-abdominal portion of the oesophagus in neonates may predispose them to reflux to some degree, but anatomical factors are not thought to be a major factor in paediatric reflux. Hiatus hernia in particular is only present in a small minority of children with reflux.6 Nor is there evidence of an abnormality of resting muscle tone in the lower oesophageal sphincter. The evidence suggests that reflux in children of all ages is primarily caused by transient lower oesophageal sphincter relaxations.7, 8, 9, 10 The lower oesophageal sphincter is not a defined ring of muscle, but rather a physiological barrier consisting of an area of increased muscle tone in the lower oesophagus. Transient relaxations can occur for many reasons, including gastric distension, and are prolonged (more than ten seconds) compared with the normal relaxation that occurs during swallowing.11 Children with neuromuscular disorders may have upper gastrointestinal dysmotility with gastroparesis and retroperistalsis. Reflux is more likely in the presence of a nasogastric tube.1 It is also more likely in infants fed formula milk rather than breast milk, as the former passes from the stomach into the duodenum more slowly.12 Intolerance to cow’s milk may also play a role for some infants with reflux.13 A recent study has suggested the possibility of a gene on chromosome 13q14 predisposing infants to severe reflux,14 although further studies with more rigorous case definition are required to confirm this.



Symptoms and signs The most common manifestation of reflux is regurgitation of feeds, also known as posseting. Mild, infrequent regurgitation is common and is not usually regarded as a reason to intervene medically. More affected infants may show evidence of colicky pains and irritability, perhaps with arching of the back. With severe reflux (and now clearly in the realms of gastro-oeophageal reflux disease) infants may have severe vomiting, aversion to feeds and failure to thrive. Abnormal posturing of the head and neck (Sandifer syndrome) is an uncommon presentation of reflux disease.



Otolaryngological manifestations APNOEAS AND APPARENT LIFE-THREATENING EVENTS



Apnoeas are not uncommon in infants, particularly those born prematurely. Gastro-oesophageal reflux causing laryngospasm may be a factor in some, but other causes need to be considered, including tracheomalacia, seizures



] 1273



and central apnoea.15 Reflux events are common in neonates and are not always associated temporally with apnoeas,16 but if symptoms, signs and investigations suggest that reflux may be a factor, anti-reflux treatment may be effective.17 [**] Apparent life-threatening events (ALTEs) occur when a child becomes limp, bradycardic, cyanosed and clearly requires resuscitation. Infants with ALTEs are at risk of sudden infant death syndrome (SIDS). Approximately half of ALTEs are thought to be related to reflux-related laryngospasm or aspiration.18, 19 Structural airway anomalies are also common, but central apnoeas and nonaccidental injury should be considered in the differential diagnosis. AIRWAY DISORDERS



Gastro-oesophageal reflux may cause a variety of upper airway symptoms in children and has been found to be associated with hoarseness,20 recurrent croup21, 22 and ‘laryngitis’.23 Controlled studies of the effect of anti-reflux treatment on these symptoms are lacking. Gastro-oesophageal reflux is extremely common in children with laryngomalacia,24, 25, 26, 27 but it is difficult to separate cause from effect. The highly negative intrathoracic pressures generated by a child with upper airway obstruction will encourage reflux of gastric contents. Reflux may, in turn, worsen airway obstruction by causing laryngeal oedema and laryngospasm. Treatment of the laryngomalacia by aryepiglottoplasty produces significant improvement in reflux on pH testing.28 Equally, anti-reflux treatment can be very useful in improving the stridor of laryngomalacia, and may obviate the need for surgery in a proportion of those most affected. [*] Gastro-oesophageal reflux is common in children with subglottic stenosis and has been postulated as a contributory factor in the development of the stenosis.29, 30 As with laryngomalacia, however, it is difficult to know whether the reflux is a cause or an effect of the airway obstruction. Many surgeons feel that reflux is potentially very deleterious at the time of reconstructive surgery31 and prophylactic anti-reflux therapy is often prescribed,32 although evidence that this makes a difference is lacking.33 [*] Laryngeal and laryngo-tracheo-oesophageal clefts are associated with oesophageal dysmotility problems, and reflux is almost universal. Most surgeons would advocate anti-reflux treatment for these children, particularly around the time of cleft repair, either with medication34 or fundoplication.35 [*] Although it has been suggested that reflux may exacerbate the clinical course of children with recurrent respiratory papillomatosis36, 37 and choanal atresia,38 supporting evidence is lacking. Few surgeons use anti-reflux medications routinely in these situations at present.



1274 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY RHINOSINUSITIS



The reflux of stomach contents into the nasopharynx or even the nose has been postulated to cause inflammation and oedema, and thereby lead to chronic rhinosinusitis. Reflux might also cause nasal disease through colonization of the nose with Helicobacter pylori and autonomic dysfunction.39 Evidence in support of such an association in children is limited. A diagnosis of sinusitis and a history of sinus surgery are both more common in children with a diagnosis of gastrooesophageal reflux than in children without reflux.23 Previously diagnosed reflux is also common in children with chronic rhinosinusitis, and this has been demonstrated in uncontrolled studies, both prospective40 and retrospective.41 In most children this reflux is gastrooesophageal, but a substantial minority have demonstrable reflux into the nasopharynx.40 A single small controlled study has shown gastro-nasopharyngeal reflux on pH testing to be more common in children with chronic rhinitis than controls.42 Children undergoing adenoidectomy have been reported as having gastro-oesophageal reflux more often than children undergoing ventilation tube insertion, but the diagnostic criteria for reflux were variable and not standardized.43 Anti-reflux medication has been reported to produce a marked improvement in sinus symptoms in studies with small numbers of highly selected children with sinus disease and proven reflux.40, 41, 44 Two children with refractory sinusitis and otitis media associated with reflux have been reported as having complete resolution of their symptoms after fundoplication.45 The case for an association between reflux and rhinosinusitis is intriguing but remains unproven. Controlled studies with adequate numbers are required to establish the nature of any association and the response to antireflux treatment.



OTITIS MEDIA



A recent study by Tasker and co-workers46 showed extremely high levels of pepsin in middle ear effusions from children with otitis media (OME). The pepsin was not produced in the middle ear and was present in far higher concentrations than could be found in blood, so it could only have got there from gastric reflux. Gastric acid and pepsin could cause inflammation of middle ear mucosa and dysfunction of the Eustachian tube, and reflux is therefore a plausible cause of OME. However, supporting evidence for such an association is lacking. Although small studies have shown reflux on pH testing in more than half of children with OME or recurrent acute otitis media,47, 48 a large case-control study (9900 children) showed otitis media to be marginally less common in children with reflux than controls.23



It has been suggested that reflux may cause referred otalgia which, in a fretful child, may lead to an erroneous diagnosis of acute otitis media.49 This is an interesting suggestion that merits further study.



Diagnostic tests In clinical practice, probably the most common ‘diagnostic’ intervention for a child with suspected reflux is an empirical therapeutic trial of anti-reflux treatment. This reflects the fact that the available investigations, such as pH testing, are far from ideal. Diagnostic testing remains essential, however, when a child fails to respond to treatment, has atypical symptoms or is being considered for anti-reflux surgery. It is likely that, over the next few years, intraluminal impedance will prove to be the investigation of choice. Children with suspected intolerance of cow’s milk should be referred to a paediatric gastroenterologist for further investigation. Radiological contrast swallow studies are reasonably sensitive but not specific – if they show reflux, it is likely to be genuine, but a normal study certainly does not exclude significant reflux. The main value of a contrast study is in excluding rare anatomical anomalies such as malrotation, pyloric stenosis, hiatus hernia, strictures and webs.50 [***] Radionuclide scintigraphy (performed after the oral ingestion of technetium-labelled milk) has some advantages over pH monitoring in being able to show delayed gastric emptying, aspiration and nonacid reflux events.50 Like the contrast study, however, it lacks sensitivity and gives only a brief time for measurement, and is therefore not routinely used in clinical practice. [***] Currently, the ‘gold standard’ investigation for gastrooesophageal reflux is 24-hour ambulatory pH probe monitoring. The probe is passed transnasally into the distal oesophagus. Measurements are taken every few seconds and calculations are made of the frequency and duration of oesophageal acid exposure. The results are reasonably reproducible, and are especially valuable if reflux events can be temporally related to symptoms (Figure 100.1 and Table 100.1). Some degree of reflux is found in normal children at all ages, but is more common in infants. The results must be interpreted, therefore, with reference to the child’s age. The reflux index is the percentage of time spent with a pH less than 4: studies of normal children have suggested an upper limit of normal of 12 percent in the first year of life and 6 percent thereafter.50 [***] The correlation between otolaryngological symptoms and the presence of gastro-oesophageal reflux on conventional pH monitoring is, unfortunately, poor. The correlation is improved when dual-probe pH monitoring is performed, using a probe in the pharynx as well as one in the lower oesophagus.51, 52, 53, 54 The upper probe is placed above the cricopharyngeus



Chapter 100 Gastro-oesophageal reflux and aspiration



] 1275



pH Channel Compressed 24-hour pH Graph User def. Sleep Meal Heartburn Arbfact pH 9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 16:44 18:00 20:00 ch 1 (pH)



Figure 100.1



Table 100.1



22:00



0:00



2:00



4:00



6:00



8:00



10:00



12:00



14:00



16:00



17:44



A typical pH study result showing severe reflux.



Period table.



Item Duration of period (HH:MM) Number of acid refluxes (#) Number of long acid refluxes (#) Longest acid reflux (min) Total time pH below 4.00 (min) Fraction time pH below 4.00 (%)



Total



Upright



Supine



Meal



PostP



None



24:00 152 15 47 374 26.0



15:06 120 13 33 296 32.7



08:54 32 2 47 77 14.5



3:01 52 4 33 107 58.9



04:36 42 8 21 127 46.0



00:04 0 0 0 0 0.0



posterior to the larynx, ideally at a similar level to the vocal cords. An x-ray can be used to confirm position. The lack of control data make it difficult to establish what level of gastro-pharyngeal reflux, if any, can be considered normal, but a reflux index of 2 percent in infants and 1 percent in children has been used by some.54 [***] Unfortunately, pH probe results do not always seem to correlate with symptoms in clinical practice. It must be borne in mind that milk is alkaline and an infant’s stomach contents are likely to be of neutral pH (or only weakly acidic) for approximately two hours after each milk feed. Reflux events with a pH of greater than 4 will not be recorded by the pH probe. For neonates fed continuously or every one or two hours, this makes pH monitoring effectively useless. These events may still be injurious by causing laryngospasm or aspiration, for example. In addition, pepsin is active up to a pH of 6, and it may be pepsin rather than acid that causes the majority of tissue injury. Brief reflux events (less than 15 seconds) are also not recorded. Intraluminal impedance testing uses a probe with a series of electrodes along its length. It is passed into the oesophageal lumen in the same way as a pH probe. When



a bolus (solid or liquid) passes by, the electrical resistance between adjacent electrodes drops and the bolus can therefore be detected. Furthermore, the direction of travel of the bolus allows swallows to be distinguished from reflux events, and the height of a reflux event can also be detected. A pH probe at the distal end of the device completes the picture. With this technique it has been possible to show that 73 percent of reflux events in infants occur postprandially and are of neutral pH.55 In most of these events, the refluxate reaches the pharynx. Equipment suitable for ambulatory use has recently been developed, but normal values for children have yet to be established. [**] The findings on endoscopy can be highly suggestive of reflux, particularly the presence of hypopharyngeal cobblestone mucosa, oedema of the arytenoids, obliteration of the ventricle, vocal fold oedema and oedema of the posterior commissure (Figure 100.2). Other suggestive features include lingual tonsil hypertrophy, blunting of the carina and increased bronchial secretions. Some studies have shown a strong correlation between these findings and a diagnosis of gastro-oesophageal reflux,56, 57 while others have not.53 Oesophagitis is diagnostic, but is



1276 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Figure 100.2 Endoscopic view of the larynx in a one-year-old girl with Down syndrome and a repaired tracheo-oesophageal fistula. There is severe cobblestoning of the mucosa on the laryngeal surface of the epiglottis, and also a subglottic stenosis, both attributed to reflux.



present in only a minority of children with reflux. In fact, the appearance of the oesophageal mucosa is often normal despite the presence of reflux, and biopsy is required to establish the diagnosis with certainty. [***]



Treatment For many children with physiological reflux or very mild symptoms, expectant treatment is reasonable. If more significant reflux is suspected, a therapeutic trial of feed thickeners and advice on positioning, possibly with concomitant acid suppression, should be the first step. There have been a number of randomized controlled trials that support the effectiveness of feed thickeners in reducing regurgitation, although none met the quality criteria for inclusion in a recent Cochrane systematic review.58 Feed thickeners have the advantage of increasing the calorific value of the milk and thus promoting weight gain, but a modified teat may be required to allow the infant to feed effectively. [****] The ideal position to reduce reflux in children over a year old is the left lateral position with the head of the bed elevated.50 In infants, the prone position is best for reducing reflux, but has also been shown to increase the risk of SIDS. It is therefore not recommended. The left lateral position may be just as good as prone59 if it can be maintained safely, but the right lateral position makes reflux worse.60 Head elevation shows no benefit in infants and the use of an infant seat to achieve a semi-erect position actually makes reflux worse.61 [****] Antacids run the risk of causing toxicity in infants due to their high aluminium content and prolonged treatment is not recommended.50 Alginates (such as Gaviscon) are widely used in children but there is no published



paediatric evidence to support their use. Prokinetic agents were once popular but the only consistently effective agent, cisapride, has been withdrawn due to cardiac side effects. Domperidone has its advocates, particularly for use in children with neurological disorders. Acid suppression therapy is the mainstay of treatment for children with significant symptoms of gastro-oesophageal reflux disease. There are no trials, however, to demonstrate the effectiveness of such treatment for extraoesophageal symptoms of reflux in children. Placebocontrolled trials support the use of H2 receptor antagonists for children with oesophagitis.62, 63 Numerous trials have shown proton pump inhibitors (PPIs) to be more effective than H2 antagonists in adults, but studies of PPIs in children are limited to case series. It is difficult, therefore, to make clear recommendations about which drug to choose except to say that H2 antagonists are widely used and seem to be effective, while PPIs may be more effective but the evidence to support their use in children is lacking. Data on the effects of long-term acid suppression in children are also lacking. [**] Selected children with reflux resistant to medical therapy may be selected for surgery in the form of a fundoplication. The procedure (now often performed laparoscopically) involves wrapping the fundus of the stomach around the intraabdominal portion of the oesophagus to produce a valve-like effect. Case series suggest that the procedure can be extremely effective for selected children with extraoesophageal symptoms.45, 64 [**]



KEY POINTS  A degree of reflux is physiological in neonates.  The prevalence of symptomatic reflux declines with age after the neonatal period.  Some relationship between reflux and airway disorders has been established.  A relationship between reflux and disorders of the ear and nose has been postulated but remains unproven.



Best clinical practice [ Reflux is often physiological and can be managed expectantly if symptoms are mild. [Grade C]



[ A therapeutic trial of positioning, feed thickeners and acid suppression is an appropriate first step when significant reflux is suspected. [Grade C] [ The left lateral position is effective and in children above one year old a degree of head elevation should also be used. [Grade C]



Chapter 100 Gastro-oesophageal reflux and aspiration



[ Acid suppression with H2 receptor antagonists or PPIs appears to be safe and effective but supportive published evidence is lacking. [Grade C] [ Diagnostic tests have limitations and should be reserved for cases where medical treatments have failed. [Grade C] [ Fundoplication surgery should be considered where symptoms are significant and medical treatments have failed. [Grade C]



Deficiencies in current knowledge and areas for future research The relationship between reflux and otolaryngological disorders has been difficult to establish. In part, this may reflect the limitations of the currently available tests for the presence of reflux. Rigorous case definition is a prerequisite for any study, and particularly for studies on the effect of anti-reflux therapy on otolaryngological disorders. The introduction of intraluminal impedance testing may provide us with a tool to diagnose reflux with a degree of certainty that has been lacking until now.



ASPIRATION Definition Aspiration is the passage of foreign material beyond the vocal cords. This chapter will not cover acute airway obstruction due to aspiration of a foreign body, but rather the (usually chronic) aspiration of saliva, feeds and/or refluxed gastric contents.



Causes The air and food passages cross to a much greater extent in man than in other mammals as a result of adaptations for speech, making us particularly predisposed to aspirate. A host of mechanisms exist to compensate for this and a deficiency in any one can lead to aspiration. Such deficiencies include impaired or incoordinate swallowing, impaired laryngeal sensation, impaired laryngeal elevation on swallowing and impaired true vocal fold adduction. Most commonly, aspiration presents in children with some combination of oesophageal dysmotility, poorly coordinated oral and pharyngeal phases of swallowing and impaired laryngeal reflexes resulting from neuromuscular disorders or cerebral palsy. Occasionally, oesophageal dysmotility can be an isolated phenomenon



] 1277



in otherwise healthy infants.65 Swallowing dysfunction may also be due to inadequate maturation of neurological control in premature infants. Many children who aspirate have reflux, but only a small minority of children with reflux will aspirate.66 Laryngeal elevation is impaired by the presence of a tracheostomy tube and by endotracheal intubation, both of which predispose to aspiration, although aspiration can be reduced by applying positive pressure (positive-end expiratory pressure (PEEP) or continuous positive airway pressure (CPAP)).67, 68 Nasogastric tubes do not seem to be a risk factor.69



Symptoms and signs The typical clinical features in infants are apnoeas, bradycardias and choking attacks with feeds. There may be apparent life-threatening events (see above) and aspiration is not uncommon as a cause of sudden infant death.70 Chronic aspiration may present as recurrent pneumonia. Aspiration may also occur without overt signs (silent aspiration), manifesting as progressive deterioration in respiratory function.



Diagnostic tests The best available investigations at present are videofluoroscopy and fibreoptic endoscopic evaluation of swallow (FEES). The videofluoroscopic modified barium swallow (often referred to simply as videofluoroscopy) is usually carried out in conjunction with a speech and language therapist. Various consistencies of radio-opaque material, ranging from thin liquid to solid, are swallowed and followed on x-ray fluoroscopy. The investigation provides excellent information about the coordination and completeness of the oral and pharyngeal stages of the swallow, as well as the presence of aspiration. Videofluoroscopy is a detailed and dynamic investigation that provides considerably more information than a standard contrast swallow, whose only use in this regard is to exclude tracheo-oesophageal fistula and oesophageal stricture. [*] FEES has the advantage over videofluoroscopy of being performed at the bedside or in the outpatient clinic. It may also be more sensitive for aspiration.71 The swallow is observed using a fibreoptic endoscope positioned just behind the soft palate via the nose. Coloured liquid is easiest to see. The investigation gives information about palatal elevation, leakage into the pharynx during the oral phase, pharyngeal residue and aspiration. Fibreoptic endoscopic evaluation of swallowing with sensory testing (FEESST) is an extension of the technique that also provides information on laryngeal sensation, although its use in children is less widespread. With experience and a cooperative child, FEES is possible even in very young children.72 [*]



1278 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY It is possible to identify aspiration using oral administration of a radioactive substance which can be identified in the lungs on scintigraphy.73, 74 The procedure is extremely operator-dependent and the quality of the images obtained is poor. [*] Endoscopy is useful in a number of ways. Microlaryngoscopy-bronchoscopy under general anaesthetic is the only way to reliably exclude a laryngeal cleft, although a fibreoptic laryngoscopy should identify vocal cord palsy. Bronchoscopy (fibreoptic or rigid) allows bronchoalveolar lavage samples to be obtained for the identification of lipid-laden (‘foamy’) macrophages, although the sensitivity and specificity of this for aspiration has been reported to be poor.75, 76, 77 [***]



Management Minor degrees of aspiration may be managed by altering the consistency of feeds according to the results of the videofluoroscopy. A speech and language therapist should supervise this, and will also be able to give advice on head positioning during feeds. The decision to stop oral feeds in infants in favour of tube feeding should not be taken lightly as it may be extremely difficult for oral feeds to be re-established in a child who has not developed the necessary oral skills early in life. Tube feeding (either by nasogastric tube or gastrostomy), however, is often unavoidable. [*] Where aspiration of refluxed gastric contents is a major problem, control of the reflux by fundoplication may be helpful. Equally, if aspiration of saliva is a major issue despite tube feeding, a procedure to reduce saliva production may be beneficial. This may take the form of excision of the submandibular glands with bilateral parotid duct ligation,78 or alternatively simple ligation of all four major salivary gland ducts.79 Injection medialization of a paralyzed vocal cord may reduce aspiration80 but carries a risk of impairing the airway. [**] Tracheostomy will often make aspiration worse by preventing laryngeal elevation on swallowing. It does, however, allow easy access to the chest for suctioning and may, of course, be indicated for other reasons in these children with multiple problems. Even a cuffed tube does not prevent aspiration as secretions pool above the cuff and the seal is never perfect. Tubes are available with a low-pressure cuff and a suction port above it if this is a problem. [*] Ultimately, when all else has been tried, it may be necessary to consider a procedure to separate the air and food passages completely. Laryngotracheal separation is the procedure of choice and has the advantage over both laryngeal closure (suturing the false cords together above a tracheostomy81) and total laryngectomy, being reversible, at least in theory. The procedure involves transecting the cervical trachea and bringing out the lower end as



a permanent end-stoma. The upper end of the trachea can be closed off as a blind pouch or anastomosed end-to-side to the oesophagus. The procedure produces good results and may allow the resumption of oral feeds.82, 83 However, this is at the expense of voice. Most children considered for this procedure will be nonverbal but the decision to sacrifice speech is clearly one that cannot be taken lightly. [**]



KEY POINTS  The following are risk factors for aspiration: – cerebral palsy; – neuromuscular disorders; – impaired oesophageal motility; – prematurity; – gastro-oesophageal reflux; – tracheostomy.  The following otolaryngologic diagnoses should be considered: – laryngeal cleft; – tracheo-oesophageal fistula; – vocal cord palsy.



Best clinical practice [ A speech and language therapist should be involved from the outset. [Grade C/D]



[ Videofluoroscopy and/or FEES are the investigations of choice. [Grade C/D]



[ Endoscopy has a role in certain situations. [Grade C/D]



[ Alterations in feed consistency may be all that is required in mild cases. [Grade C/D]



[ Tube feeding is often necessary. [Grade C/D] [ Fundoplication and salivary surgery have a small but important role. [Grade C/D]



[ Laryngotracheal separation surgery is effective but sacrifices voice. [Grade C/D]



Deficiencies in current knowledge and areas for future research Few centres treat large numbers of children, so most of the published literature consists of relatively small retrospective case series. Any large, prospective evaluation of diagnostic tests or treatment options would be a useful addition to current knowledge.



Chapter 100 Gastro-oesophageal reflux and aspiration



REFERENCES 1. Peter C, Wiechers C, Bohnhorst B, Silny J, Poets C. Influence of nasogastric tubes in gastroesophageal reflux in preterm infants: a multiple intraluminal impedence study. Journal of Pediatrics. 2002; 141: 277–9. 2. Nelson SP, Chen EH, Syniar GM, Christoffel KK. Prevalence of symptoms of gastroesophageal reflux during infancy: a pediatric practice-based survey. Archives of Pediatrics and Adolescent Medicine. 1997; 151: 569–72. 3. Nelson SP, Chen EH, Syniar GM, Christoffel KK. Prevalence of symptoms of gastroesophageal reflux during childhood: a pediatric practice-based survey. Archives of Pediatrics and Adolescent Medicine. 2000; 154: 150–4. 4. Gold BD. Outcomes of pediatric gastroesophageal reflux disease: in the first year of life, in childhood, and in adultsyoh, and should we really leave Helicobacter pylori alone? Journal of Pediatric Gastroenterology and Nutrition. 2003; 37: s33–9. 5. el-Serag HB, Gilger M, Carter J, Genta RM, Rabeneck L. Childhood GERD is a risk factor for GERD in adolescents and young adults. American Journal of Gastroenterology. 2004; 99: 806–12. 6. Gorenstein A, Cohen A, Cordova Z, Witzling M, Krutman B, Serour F. Hiatal hernia in pediatric gastroesophageal reflux. Journal of Pediatric Gastroenterology and Nutrition. 2001; 33: 554–7. 7. Omari T, Barnett C, Snel A, Davidson G, Haslam R, Bakewell M et al. Mechanism of gastroesophageal reflux in premature infants with chronic lung disease. Journal of Pediatric Surgery. 1999; 34: 1795–8. 8. Omari T, Barnett C, Benninga M, Lontis R, Goodchild L, Haslam R et al. Mechanisms of gastro-oesophageal reflux in preterm and term infants with reflux disease. Gut. 2002; 51: 475–9. 9. Kawahara H, Dent J, Davidson G. Mechanisms responsible for gastroesophageal reflux in children. Gastroenterology. 1997; 113: 399–408. 10. Omari TI, Barnett C, Snel A, Goldsworthy W, Haslam R, Davidson G et al. Mechanisms of gastroesophageal reflux in healthy premature infants. Journal of Pediatrics. 1998; 133: 650–4. 11. Mittal RK, Holloway RH, Penagini R, Blackshaw LA, Dent J. Transient lower esophageal sphincter relaxation. Gastroenterology. 1995; 109: 601–10. 12. Heacock HJ, Jeffery HE, Baker JL, Page M. Influence of breast versus formula milk on physiological gastroesophageal reflux in healthy newborn infants. Journal of Pediatric Gastroenterology and Nutrition. 1992; 14: 41–6. 13. Cavataio F, Iacono G, Monalto G, Soresi M, Tumminello M, Carroccio A. Clinical and pH-metric characteristics of gastro-oesophageal reflux secondary to cows’ milk protein allergy. Archives of Disease in Childhood. 1996; 75: 51–6. 14. Hu FZ, Donfack J, Ahmed A, Dopico R, Johnson S, Post JC et al. Fine-mapping a gene for pediatric gastroesophageal



15.



16.



17.



18.



19.



20.



21.



22.



23.



24.



25.



26.



27.



28.



29.



] 1279



reflux on human chromosome 13q14. Human Genetics. 2004; 114: 562–72. Jeffery H, Rahilly P, Read D. Multiple causes of asphyxia in infants at high risk for sudden infant death. Archives of Disease in Childhood. 1983; 58: 92–100. Harris P, Munoz C, Mobarec S, Brockmann P, Mesa T, Sanchez I. Relevance of the pH probe in sleep study analysis in infants. Child: Care, Health and Development. 2004; 30: 337–44. Herbst J, Minton S, Book L. Gastroesophageal reflux causing respiratory distress and apnea in newborn infants. Journal of Pediatrics. 1979; 95: 763–8. MacFadyen UM, Hendry GMA, Simpson H. Gastrooesophageal reflux in near-miss sudden infant death syndrome or suspected recurrent aspiration. Archives of Disease in Childhood. 1983; 58: 87–91. McMurray JS, Holinger LD. Otolaryngic manifestations in children presenting with apparent life-threatening events. Otolaryngology-Head and Neck Surgery. 1997; 116: 575–9. Gumpert L, Kalach N, Dupont C, Contencin P. Hoarseness and gastroesophageal reflux in children. Journal of Laryngology and Otology. 1998; 112: 49–54. Contencin P, Narcy P. Gastropharyngeal reflux in infants and children: a pharyngeal pH monitoring study. Archives of Otolaryngology – Head and Neck Surgery. 1992; 118: 1028–30. Waki E, Madgy DN, Belenky WM, Gower VC. The incidence of gastroesophageal reflux in recurrent croup. International Journal of Pediatric Otorhinolaryngology. 1995; 32: 223–32. el-Serag H, Gilger M, Kuebeler M, Rabeneck L. Extraesophageal associations of gastroesophageal reflux disease in chidren without neurological defects. Gastroenterology. 2001; 121: 1294–9. Roger G, Denoyelle F, Triglia J, Garabedian E-N. Severe laryngomalacia: surgical indications and results in 115 patients. Laryngoscope. 1995; 105: 1111–7. Giannoni C, Sulek M, Friedman EM, Duncan NO. Gastroesophageal reflux association with laryngomalacia: a prospective study. International Journal of Pediatric Otorhinolaryngology. 1998; 43: 11–20. Matthews BL, Little JP, McGuirt WF, Koufman JA. Reflux in infants with laryngomalacia: results of 24-hour dualprobe pH monitoring. Otolaryngology-Head and Neck Surgery. 1999; 120: 860–4. Bibi H, Khvolis E, Shoseyov D, Ohaly M, Dor D, London D et al. The prevalence of gastroesophageal reflux in children with tracheomalacia and laryngomalacia. Chest. 2001; 119: 409–13. Hadfield PJ, Albert DM, Bailey CM, Lindley K, Pierro A. The effect of aryepiglottoplasty for laryngomalacia on gastrooesophageal reflux. International Journal of Pediatric Otorhinolaryngology. 2003; 67: 11–4. Halstead LA. Gastroesophageal reflux: a critical factor in pediatric subglottic stenosis. Otolaryngology-Head and Neck Surgery. 1999; 120: 683–8.



1280 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY 30. Walner D, Stern Y, Gerber M, Rudolph C, Baldwin C, Cotton R. Gastroesophageal reflux in patients with subglottic stenosis. Archives of Otolaryngology – Head and Neck Surgery. 1998; 124: 551–5. 31. Berkowitz R. Failed paediatric laryngotracheoplasty. Australian and New Zealand Journal of Surgery. 2001; 71: 292–6. 32. Ludemann J, Hughes C, Noah Z, Holinger L. Complications of pediatric laryngotracheal reconstruction: prevention strategies. Annals of Otology, Rhinology and Laryngology. 1999; 108: 1019–26. 33. Zalzal G, Choi S, Petal K. The effect of gastroesophageal reflux on laryngotracheal reconstruction. Archives of Otolaryngology – Head and Neck Surgery. 1996; 122: 297–300. 34. Kubba H, Gibson D, Bailey C, Hartley B. Techniques and outcomes of laryngeal cleft repair: an update to the Great Ormond Street Hospital series. Annals of Otology, Rhinology and Laryngology. 2005; 114: 309–13. 35. Hof E, Hirsig J, Giedion A, Pochon J-P. Deleterious consequences of gastroesophageal reflux in cleft larynx surgery. Journal of Pediatric Surgery. 1987; 22: 197–9. 36. Borkowski G, Sommer P, Stark T, Sudhoff H, Luckhaupt H. Recurrent respiratory papillomatosis associated with gastroesophageal reflux disease in children. European Archives of Oto-Rhino-Laryngology. 1999; 256: 370–2. 37. Holland B, Koufman J, Postma G, McGuirt WJ. Laryngopharyngeal reflux and laryngeal web formation in patients with pediatric recurrent respiratory papillomas. Laryngoscope. 2002; 112: 1926–9. 38. Beste DJ, Conley SF, Brown CW. Gastroesophageal reflux complicating choanal atresia repair. International Journal of Pediatric Otorhinolaryngology. 1994; 29: 51–8. 39. Loehrl T, Smith T. Chronic sinusitis and gastroesophageal reflux: are they related? Current Opinion in Otolaryngology Head and Neck Surgery. 2004; 12: 18–20. 40. Phipps C, Wood W, Gibson W, Cochran W. Gastroesophageal reflux contributing to chronic sinus disease in children: a prospective analysis. Archives of Otolaryngology-Head and Neck Surgery. 2000; 126: 831–6. 41. Bothwell MR, Parsons DS, Talbot A, Barbero GJ, Wilder B. Outcome of reflux therapy on pediatric chronic sinusitis. Otolaryngology – Head and Neck Surgery. 1999; 121: 255–62. 42. Contencin P, Narcy P. Nasopharyngeal pH monitoring in infants and children with chronic rhinopharyngitis. International Journal of Pediatric Otorhinolaryngology. 1991; 22: 249–56. 43. Carr M, Poje C, Ehrig D, Brodsky L. Incidence of reflux in children undergoing adenoidectomy. Laryngoscope. 2001; 111: 2170–2. 44. van den Abeele T, Couloigner V, Faure C, Narcy P. The role of 24h pH recording in pediatric otolaryngologic gastrooesophageal reflux disease. International Journal of Pediatric Otorhinolaryngology. 2003; 67: s95–100.











45. Suskind D, Zeringue G, Kluka E, Udall J, Liu D. Gastroesophageal reflux and pediatric otolaryngologic disease: the role of antireflux surgery. Archives of Otolaryngology-Head and Neck Surgery. 2001; 127: 511–4. 46. Tasker A, Dettmar P, Pearson J. Reflux of gastric juice in glue ear. Lancet. 2002; 359: 493. 47. Velepic M, Rozmanic V, Velepic M, Bonifacic D. Gastroesophageal reflux, allergy and chronic tubotympanal disorders in children. International Journal of Pediatric Otorhinolaryngology. 2000; 55: 187–90. 48. Rozmanic V, Velepic M, Ahel V, Bonifacic D, Velepic M. Prolonged esophageal pH monitoring in the evaluation of gastroesophageal reflux in children with chronic tubotympanal disorders. Journal of Pediatric Gastroenterology and Nutrition. 2002; 34: 278–80. 49. Gibson WS, Cochran W. Otalgia in infants and children – a manifestation of gastroesophageal reflux. International Journal of Pediatric Otorhinolaryngology. 1994; 28: 213–8. 50. Rudolph CD, Mazur LJ, Liptak GS, Baker RD, Boyle JT, Colletti RB et al. Guidelines for evaluation and treatment of gastroesophageal reflux in infants and children: recommendations of the North American Society for Pediatric Gastroenterology and Nutrition. Journal of Pediatric Gastroenterology and Nutrition. 2001; 32: s1–31. 51. Little JP, Matthews BL, Glock MS, Koufman JA, Reboussin DM, Loughlin CJ et al. Extraesophageal pediatric reflux: 24-hour double-probe pH monitoring of 222 children. Annals of Otology, Rhinology and Laryngology. 1997; 106: 1–16. 52. Bauman NM, Bishop WP, Sandler AD, Smith RJH. Value of pH probe testing in pediatric patients with extraesophageal manifestations of gastroesophageal reflux disease: a retrospective review. Annals of Otology, Rhinology and Laryngology. 2000; 109: 18–24. 53. McMurray JS, Gerber M, Stern Y, Walner D, Rudolph C, Willging JP et al. Role of laryngoscopy, dual pH probe monitoring, and laryngeal mucosal biopsy in the diagnosis of pharyngoesophageal reflux. Annals of Otology, Rhinology and Laryngology. 2001; 110: 299–304. 54. Rabinowitz SS, Piecuch S, Jibaly R, Goldsmith A, Schwarz SM. Optimizing the diagnosis of gastroesophageal reflux in children with otolaryngological symptoms. International Journal of Pediatric Otorhinolaryngology. 2003; 67: 621–6. 55. Skopnik H, Silny J, Heiber O, Schulz J, Rau G, Heiman G. Gastroesophageal reflux in infants: evaluation of a new intraluminal impedance technique. Journal of Pediatric Gastroenterology and Nutrition. 1996; 23: 591–8. 56. Carr MM, Nagy ML, Pizzuto MP, Poje CP, Brodsky LS. Correlation of findings at direct laryngoscopy and bronchcosopy with gastroesophageal reflux disease in children: a prospective study. Archives of Otolaryngology – Head and Neck Surgery. 2001; 127: 369–74.



Chapter 100 Gastro-oesophageal reflux and aspiration











57. Carr MM, Nguyen A, Poje C, Pizzuto M, Nagy M, Brodsky L. Correlation of findings on direct laryngoscopy and bronchoscopy with presence of extraesophageal reflux disease. Laryngoscope. 2000; 110: 1560–2. 58. Huang R-C, Forbes DA, Davies MW. Feed thickener for newborn infants with gastro-oesophageal reflux (systematic review). Cochrane Library. 2004: 3. 59. Ewer AK, James ME, Tobin JM. Prone and left lateral positioning reduce gastro-oesophageal reflux in preterm infants. Archives of Disease in Childhood Fetal and Neonatal Edition. 1999; 81: f201–05. 60. Omari TI, Rommel N, Staunton E, Lontis R, Goodchild L, Haslam R et al. Paradoxical impact of body positioning on gastroesophageal reflux and gastric emptying in the premature neonate. Journal of Pediatrics. 2004; 145: 194–200. 61. Carroll AE, Garrison MM, Christakis DA. A systematic review of nonpharmacological and nonsurgical therapies for gastroesophageal reflux in infants. Archives of Pediatrics and Adolescent Medicine. 2002; 156: 109–13. 62. Cucchiara S, Gobio-Casali L, Balli F, Magazzu G, Staiano A et al. Cimetidine treatment of reflux esophagitis in children: an Italian multicentric study. Journal of Pediatric Gastroenterology and Nutrition. 1989; 8: 150–6. 63. Simeone D, Caria MC, Miele E, Staiano A. Treatment of childhood peptic esophagitis: a double-blind placebocontrolled trial of nizatidine. Journal of Pediatric Gastroenterology and Nutrition. 1997; 25: 51–5. 64. Mattioli G, Sacco O, Repetto P, Pini-Prato A, Castagnetti M, Carlini C et al. Necessity for surgery in children with gastrooesophageal reflux and supraoesophageal symptoms. European Journal of Pediatric Surgery. 2004; 14: 7–13. 65. Sheikh S, Allen E, Shell R, Hruschak J, Iram D, Castile R et al. Chronic aspiration without gastroesophageal reflux as a cause of respiratory symptoms in neurologically normal infants. Chest. 2001; 120: 1190–5. 66. McVeagh P, Howman-Giles R, Kemp A. Pulmonary aspiration studied by radionuclide milk scanning and barium swallow roentgenography. American Journal of Diseases of Children. 1987; 141: 917–21. 67. Finder JD, Yellon R, Charron M. Successful management of tracheotomized patients with chronic saliva aspiration by use of constant positive airway pressure. Pediatrics. 2001; 107: 1343–5. 68. Janson BA, Poulton TJ. Does PEEP reduce the incidence of aspiration around endotracheal tubes? Canadian Anaesthetists Society Journal. 1986; 33: 157–61. 69. Bar-Maor JA, Lam M. Does nasogastric tube cause pulmonary aspiration in children? Pediatrics. 1991; 87: 113–4. 70. Iwadate K, Doy M, Ito Y. Screening of milk aspiration in 105 infant death cases by immunostaining with antihuman alpha-lactalbumin antibody. Forensic Science International. 2001; 122: 95–100.







] 1281



71. Hiss SG, Postma GN. Fibreoptic endoscopic evaluation of swallow. Laryngoscope. 2003; 113: 1386–93. 72. Hartnick CJ, Hartley BE, Miller C, Willging JP. Pediatric fiberoptic endoscopic evaluation of swallowing. Annals of Otology, Rhinology and Laryngology. 2000; 109: 996–9. 73. Boonyaprapa S, Alderson PO, Garfinkel DJ, Chipps BE, Wagner HJ. Detection of pulmonary aspiration in infants and children with respiratory disease: concise communication. Journal of Nuclear Medicine. 1980; 21: 314–8. 74. Cook SP, Lawless ST, Mandell GA, Reilly JS. The use of the salivagram in the evaluation of severe and chronic aspiration. International Journal of Pediatric Otorhinolaryngology. 1997; 41: 353–61. 75. Staugas R, Martin AJ, Binns G, Steven IM. The significance of fat-filled macrophages in the diagnosis of aspiration associated with gastro-oesophageal reflux. Australian Paediatric Journal. 1985; 21: 275–7. 76. Knauer-Fischer S, Ratjen F. Lipid-laden macrophages in bronchoalveolar lavage fluid as a marker for pulmonary aspiration. Pediatric Pulmonology. 1999; 27: 419–22. 77. Krishnan U, Mitchell JD, Tobias V, Day AS, Bohane TD. Fat laden macrophages in tracheal aspirates as a marker of reflux aspiration: a negative report. Journal of Pediatric Gastroenterology and Nutrition. 2002; 35: 309–13. 78. Gerber M, Gaugler MD, Myer CM, Cotton RT. Chronic aspiration in children: when are bilateral submandibular gland excision and parotid duct ligation indicated? Archives of Otolaryngology – Head and Neck Surgery. 1996; 122: 1368–71. 79. Klem C, Mair EA. Four-duct ligation: a simple and effective treatment for chronic aspiration from sialorrhea. Archives of Otolaryngology – Head and Neck Surgery. 1999; 125: 796–800. 80. Patel NJ, Kerschner JE, Merati AL. The use of injectable collagen in the management of pediatric vocal unilateral fold paralysis. International Journal of Pediatric Otorhinolaryngology. 2003; 67: 1355–60. 81. Pototschnig CA, Schneider I, Eckel HE, Thumfart WF. Repeatedly successful closure of the larynx for the treatment of chronic aspiration with the use of botulinum toxin A. Annals of Otology, Rhinology and Laryngology. 1996; 105: 521–4. 82. Cook SP, Lawless ST, Kettrick R. Patient selection for primary laryngotracheal separation as treatment of chronic aspiration in the impaired child. International Journal of Pediatric Otorhinolaryngology. 1996; 38: 103–13. 83. Takamizawa S, Tsugawa C, Nishijima E, Muraji T, Satoh S. Laryngotracheal separation for intractable aspiration pneumonia in neurologically impaired children: experience with 11 cases. Journal of Pediatric Surgery. 2003; 38: 975–7.



101 Diseases of the oesophagus, swallowing disorders and caustic ingestion LEWIS SPITZ



Embryology Oesophageal atresia Congenital oesophageal stenosis Achalasia Corrosive injury to the oesophagus



1282 1282 1287 1288 1290



Key points Best clinical practice Deficiencies in current knowledge and areas for future research References



1291 1292 1292 1292



SEARCH STRATEGY The author’s extensive bibliography on congenital and acquired diseases of the oesophagus in children was supplemented with a Medline search using the key words oesophagus, child, congenital, caustic, achalasia, oesophageal stenosis, tracheooesophageal fistula and tracheomalacia. The Cochrane library was also consulted. Mr Colin Bailie FRCS, Consultant Paediatric Surgeon at the Royal Liverpool Children’s Hospital advised the section editor before completion of the final draft.



EMBRYOLOGY The oesophagus and trachea first become identifiable as separate structures when the embryo is 22–23 days old. The lung bud appears as a median ventral diverticulum in the developing foregut. Shortly thereafter, the primitive stomach appears as a fusiform enlargement immediately caudal to the diverticulum. The oesophagus develops from the short area between the tracheal diverticulum and the stomach. As the trachea and oesophagus elongate, ridges appear in the lateral walls. Fusion in the midline of these ridges separates the trachea from the oesophagus. The separation process commences caudally, proceeds cranially and is complete between days 34 and 36 of gestation. Elongation of the oesophagus relative to the rest of the developing foetus begins in the distal portion and is complete by seven weeks. Kluth cast doubt on the lateral ridge theory of oesophagotracheal separation and proposed an alternative dorsal and lateral ridge theory.1



The circular musculature of the oesophagus appears in the sixth week and by the end of that week innervation by the vagus nerve has commenced. During the seventh and eighth weeks, the epithelium of the oesophagus proliferates to such an extent that the lumen is virtually, but not completely, occluded. Initially, the epithelium is ciliated but it is gradually replaced by stratified squamous epithelium (see Chapter 65, Head and neck embryology).



OESOPHAGEAL ATRESIA In oesophageal atresia, part of the wall of the oesophagus fails to develop. In most cases this is associated with a failure of complete separation of the developing trachea – hence the persistence of a tracheo-oesophageal fistula. This was a uniformly fatal congenital abnormality until 1939 when Levin and Ladd independently reported the



Chapter 101 Diseases of the oesophagus, swallowing disorders and caustic ingestion



first two survivors. Both infants required multiple procedures – cervical oesophagostomy, feeding gastrostomy and subsequently oesophageal substitution. The first successful primary repair of the defect, which paved the way for future developments, was achieved by Haight, in 1941.2, 3 It is now rare for an infant with oesophageal atresia to succumb from oesophageal atresia alone, unless it is associated with extreme prematurity or a major congenital cardiac defect.4, 5



Incidence Abnormalities in oesophageal development are present in 1:3000–4000 live births. There is no standard genetic pattern of inheritance, although the condition has been documented in siblings, in one and very occasionally in both twins and in two generations.6



Pathogenesis The anomaly is thought to arise between the third and sixth weeks of intrauterine development. The precise cause and mechanism are unknown. Failure of complete separation of the foregut from the respiratory tract would appear to be the basis for the development of the various types of defects.7, 8, 9 Oesophageal atresia may occur as an isolated anomaly but at least 50 percent of cases have additional malformations.



Types of anomaly The variety and incidence of the different types of tracheo-oesophageal abnormalities are shown in Table 101.1. It is especially important for the otolaryngologist to be alert to the possibility of a tracheo-oesophageal fistula without oesophageal atresia (‘H’ type) as this may present in later childhood with recurrent respiratory infections. It may be discovered as an incidental finding at bronchoscopy, when a suction canula in the tracheal lumen is found to enter the fistula. Often the tracheal orifice can be concealed in a mucosal fold. It may Table 101.1 Different types of tracheo-oesophageal anomalies. Type of anomaly Oesophageal atresia with distal tracheo-oesophageal fistula Oesophageal atresia without tracheo-oesophageal fistula Oesophageal atresia with proximal tracheo-oesophageal fistula Oesophageal atresia with proximal and distal tracheooesophageal fistula Tracheo-oesophageal fistula without oesophageal atresia



] 1283



sometimes be recognized at flexible bronchoscopy in a newborn baby.10



Associated anomalies Additional congenital malformations are found in approximately one-half of infants with tracheo-oesophageal anomalies. The various systems affected are as shown in Table 101.2 with multiple defects occurring in many patients.11 The mortality in oesophageal atresia is directly related to the severity of associated congenital anomalies particularly cardiac defects, and the degree of prematurity (Table 101.3).4, 5



The VACTERL complex The VATER complex of associated anomalies was described by Quan and Smith in 1973.12 (The acronym stands for V = vertebral, A = anorectal, T-E = tracheooesophageal fistula and (o)esophageal atresia, R = radial and renal dysplasia.) This was subsequently extended to the VACTERL association to include cardiac and limb abnormalities. Ventricular septal defects are the single most common cardiac malformation.13 Of the gastrointestinal anomalies, duodenal atresia, Meckel’s diverticulum and malrotation are most commonly encountered. A variety of genitourinary anomalies may be present in association Table 101.2 Congenital anomalies associated with oesophageal atresia/tracheo-oesophageal fistula. Type of anomaly



%



Cardiovascular defects Vertebrae anomalies Gastrointestinal (excluding anorectal) anomalies Genitourinary anomalies Anorectal malformations Skeletal defects Respiratory anomalies Genetic/chromosomal defects Miscellaneous anomalies



34 17 14 12 11 11 6 2 10



% 87 6–7 2 1 3–4



Table 101.3 Risk classification for infants with oesophageal atresia. Group Birth weight (g) I II III



41500 o1500 o1500



Major cardiac anomaly No Or Present PLUS Present



% Survival 495 60 25



1284 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY with oesophageal atresia, the most serious being bilateral renal agenesis (Potter’s syndrome), which is fatal. The anorectal anomalies are equally divided between the supralevator (high) and translevator (low) defects.



Diagnosis Polyhydramnios is nonspecific but is present in approximately 90 percent of mothers of infants with oesophageal atresia. Antenatal ultrasound scan may be indicative of oesophageal atresia without fistula when there is failure to demonstrate the presence of intragastric fluid.14 To confirm the diagnosis of oesophageal atresia suspected on prenatal ultrasound scan, Langer, and later Levine, have proposed magnetic resonance imaging.15, 16, 17 The overall prognosis for these foetuses is poor, due to the high incidence of chromosomal abnormalities.18 Of 16 cases of prenatally diagnosed oesophageal atresia, only 4 (25 percent) survived the neonatal period. The infant, at birth, is ‘excessively mucusy’ and requires repeated suction as it is unable to swallow saliva. Failure to recognize the anomaly at this stage will expose the infant to choking episodes and aspiration pneumonitis with the first feed. The diagnosis is confirmed by passing a large calibre (No. 10 French) firm catheter through the mouth and into the oesophagus. The position of arrest of the tube is confirmed on a chest x-ray (Figure 101.1). If the tube enters the stomach, there is no oesophageal atresia. In the majority of cases with this condition the tube cannot be advanced more than 10 cm beyond the lower gum margin. It is important to include the abdomen on the original x-ray in order to assess the presence of intestinal gas shadows (Figure 101.2). Gas within the gastrointestinal tract implies the presence of a distal tracheo-oesophageal fistula while the distribution of the gas may indicate an additional intestinal anomaly, e.g. duodenal atresia. The chest radiograph should be assessed for pulmonary pathology and the configuration of the heart shadow may be indicative of cardiac defects, e.g. Fallot’s tetralogy.



Figure 101.1 Plain x-ray of the chest and upper abdomen in an infant with oesophageal atresia. The radiopaque tube can be seen in the upper oesophagus. Gas in the intestines indicates the presence of a distal tracheo-oesophageal fistula.



aspiration of saliva. Definitive repair may have to be postponed in the presence of aspiration pneumonitis, which generally responds very rapidly to broad-spectrum antibiotics and physiotherapy. [***] In the infant with severe respiratory distress requiring mechanical ventilatory support, consideration should be given to emergency ligation of the distal tracheo-oesophageal fistula to facilitate respiratory support and to prevent overdistension of the stomach and intestine.19



Management Treatment of infants with oesophageal atresia should be concentrated in centres where the surgical expertise, supportive services (anaesthesia, paediatric intensive care, radiology, pathology) and specialized nursing care are available. Transfer to such centres should be prompt to avoid exposing the infant to the risks of aspiration pneumonitis. The infant is transported in a portable incubator either in a lateral position or in the prone position in order to discourage reflux of gastric juice into the distal tracheo-oesophageal fistula, while continuous suction is applied to the upper pouch to prevent



SURGICAL REPAIR



The operative procedure is carried out under general endotracheal anaesthesia. Preliminary bronchoscopy to document the level of entry of the tracheo-oesophageal fistula, to assess tracheomalacia and to exclude an upper pouch fistula is recommended. Access is achieved via a right posterolateral extrapleural thoracotomy through the fourth or fifth intercostal space. After dividing the azygos vein, the distal oesophagus is identified and traced proximally to its site of entry into the trachea. The fistula is divided and the tracheal defect closed with fine



Chapter 101 Diseases of the oesophagus, swallowing disorders and caustic ingestion



] 1285



the fashioning of a gastrostomy exposes the infant to an increased incidence of gastro-oesophageal reflux which predisposes the anastomosis to stricture formation.24 [**/*] Oral feeding has been introduced as early as the second postoperative day but is generally postponed until three to four days after surgery. A contrast oesophagogram may be performed on the fifth postoperative day to check the anastomosis. Patients with a long gap between the proximal and distal segment (particularly those with an isolated oesophageal atresia) require special attention. The alternative approaches available are to delay the repair pending differential growth of the oesophageal segments towards each other (three months), or to perform a cervical oesophagostomy and carry out an oesophageal substitution at a later stage – colonic interposition,25 gastric tube26 or gastric transposition.27



Results and prognosis The survival rate for infants born with oesophageal atresia has improved significantly due to improved operative technique, better anaesthesia and greatly improved postoperative intensive care. The main determinants of survival are birth weight above 1500 g and the presence or absence of a significant cardiac anomaly.



Figure 101.2 Plain x-ray of the chest and abdomen in an infant with oesophageal atresia. The tip of the catheter can be seen at the thoracic inlet. The gas pattern in the abdomen indicates the presence of a duodenal atresia (‘double bubble’).



interrupted nonabsorbable sutures. The proximal blind end of the oesophagus is identified in the apex of the chest and mobilized sufficiently to effect an anastomosis with as little tension as possible. An end-to-end anastomosis is performed using a single layer of sutures (Figure 101.3). The operative correction of oesophageal atresia has recently been carried out thoracoscopically, but this requires considerable expertise and can only be safely carried out in specific centres.20, 21 [**/*] An oesophagomyotomy,22 oesophagoplasty23 or the use of elective paralysis and mechanical ventilation for a few days postoperatively in an extremely tight anastomosis may be effective in reducing the incidence of anastomotic complications. The passage of a fine transanastomotic nasogastric tube through the nose into the stomach will allow enteral feeding to commence on the second or third postoperative day. A gastrostomy tube is no longer indicated in the routine repair of an oesophageal atresia. In fact,



Complications Complications of oesophageal atresia repair may be divided into early and late.  Early complications include anastomotic leak, strictures, recurrent tracheo-oesophageal fistula and vocal cord palsy.  Late complications comprise tracheomalacia, gastrooesophageal reflux and disordered oesophageal peristalsis.



ANASTOMOTIC LEAKS



The incidence varies from 4 to 36 percent, depending on the vigour with which the diagnosis is pursued. [***] The majority of leaks are insignificant and are mainly detected on routine contrast studies. Minor leaks seal spontaneously but may lead to stricture formation. Major leaks present within 48–72 hours postoperatively and cause respiratory distress due to tension pneumothorax. They may be amenable to direct repair if promptly diagnosed, or to conservative treatment by intercostal drainage. Cervical oesophagostomy and gastrostomy may be the safest approach if there is a major disruption of the anastomosis.



1286 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



(a)



(b)



(c)



Figure 101.3 The technique of repair of oesophageal atresia with distal tracheo-oesophageal fistula. (a) Division of distal tracheo-oesophageal fistula; (b) closure of tracheal defect with interrupted sutures; (c) end-to-end oesophago–oesophageal anastomosis with a single layer of interrupted sutures.



ANASTOMOTIC STRICTURE



The stricture produces symptoms such as dysphagia or recurrent respiratory infections due to aspiration. The incidence varies from 10 to 50 percent. Strictures are



caused by poor anastomotic technique, ischaemia, anastomotic leakage or gastro-oesophageal reflux. Most strictures respond to one or two dilatations but intractable strictures generally fail to respond to dilatations



Chapter 101 Diseases of the oesophagus, swallowing disorders and caustic ingestion



] 1287



alone until the associated gastro-oesophageal reflux is corrected.11 [***]



cases.31, 32 Gastro-oesophageal reflux is covered in detail in Chapter 100, Gastro-oesophageal reflux and aspiration.



RECURRENT FISTULA



DISORDERED PERISTALSIS



Recurrent fistulae develop in approximately 8 percent of patients.28 The diagnosis should be suspected in a child who develops respiratory symptoms during feeding or who suffers repeated respiratory infections. The investigation of choice is a tube oesophagogram with confirmation at bronchoscopy. Surgical division of the fistula is the only reliable method of treatment.



Swallowing difficulties may persist for many years as a result of the inherent oesophageal dysmotility, affecting the distal segment in particular.33 The infant gradually learns to cope with this problem unless there is an associated anatomical defect (anastomotic stricture, gastro-oesophageal reflux or distal oesophageal stenosis). RESPIRATORY INFECTION



TRACHEOMALACIA



This is due to a weakness of the tracheal cartilaginous ring structure and is characterized by partial or complete occlusion of the tracheal lumen, primarily during expiration.29 It occurs in 10–20 percent of patients and is responsible for ‘near miss’ apnoeic and/or cyanotic attacks or recurrent respiratory infections. The condition affects the distal trachea and is diagnosed on bronchoscopic examination, which reveals the slit-like aperture of the trachea during expiration. Treatment is controversial. Many authors claim that the problem resolves spontaneously in time but the infant is at risk of sudden death. Aortopexy, in which the ascending segment and arch of the aorta are elevated and attached to the posterior surface of the sternum, should be considered.30 Tracheomalacia is covered in Chapter 88, Congenital disorders of the larynx, trachea and bronchi. VOCAL CORD PALSY



Vocal cord palsy may take place in association with a tracheo-oesophageal fistula or one or both recurrent laryngeal nerves may be traumatized during tracheooesophageal fistula surgery. The nerves are at most risk during the cervical repair of an H-type tracheo-oesophageal fistula. Vocal cord palsy is considered in Chapter 88, Congenital disorders of the larynx, trachea and bronchi. GASTRO-OESOPHAGEAL REFLUX



Careful investigation will reveal gastro-oesophageal reflux in up to 60 percent of children undergoing successful repair of oesophageal atresia. Various factors have been implicated in the pathogenesis of the reflux, particularly disordered oesophageal motility, displacement of the gastro-oesophageal junction and the use of a gastrostomy. The reflux causes recurrent vomiting which may lead to aspiration and respiratory infections or may produce a stricture at the anastomotic site. The diagnosis is established on barium oesophagogram and on pH monitoring. Medical treatment is successful in many cases but antireflux surgery may be necessary in selected



These infants are also prone to recurrent respiratory infections during the first few years of life. Children who have had a tracheo-oesophageal fistula repair often develop a troublesome recurrent cough (‘tracheooesophageal fistula (TOF) cough’), due to a combination of factors including a tendency to aspiration, mild tracheomalacia and gastro-oesophageal reflux. Many require ongoing care from a multidisciplinary team including an otolaryngologist.



CONGENITAL OESOPHAGEAL STENOSIS Congenital stenosis of the oesophagus is rare. Stenosis most commonly arises from acquired lesions, e.g. reflux oesophagitis, corrosive ingestion, foreign body impaction or secondary to surgical resection and anastomosis. The congenital form may be caused either by a membranous web or diaphragm, or may arise as a result of intramural deposits of tracheobronchial cartilaginous tissue. The latter pathology has been most frequently reported in association with oesophageal atresia and/or tracheooesophageal fistula.34



Clinical features In the presence of a complete web, presentation is similar to that of oesophageal atresia, with symptoms appearing on the first day of life. In other cases, symptoms may develop at any stage of life through to adulthood but generally arise in early infancy. The symptoms include dysphagia, vomiting with food ingestion, failure to thrive, recurrent respiratory infections and foreign body impaction.



Diagnosis Gastro-oesophageal reflux as a cause for the stenosis should always be excluded. The barium oesophagogram will identify the site of the lesion. Congenital webs are most commonly located in the middle third of the oesophagus and appear as shelf-like projections within



1288 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY the oesophageal lumen. Tracheobronchial remnants are generally located in the lower third of the oesophagus or at the gastro-oesophageal junction and they cause sharp narrowing at this point. The precise nature and anatomical location of the lesion should be confirmed by endoscopic examination.



Treatment Dilatation alone may be sufficient for many oesophageal webs. [**/*] Surgical resection of tracheobronchial cartilaginous tissue is generally recommended.35 [**/*]



ACHALASIA Achalasia is a motility disorder of the oesophagus characterized by an absence of peristalsis and failure of relaxation of the lower oesophageal sphincter.36 This causes obstruction at the level of the oesophago-gastric junction.



Incidence Achalasia is rare in children. The incidence is one per 100,000 and approximately 5 percent of all patients with achalasia are symptomatic before the age of 15 years. There are few reports of achalasia being present in siblings37, 38 and it has been reported in association with a number of syndromes, e.g. Riley-Day syndrome. Evidence for a familial incidence of the disease is lacking. Both sexes are equally affected.



Aetiology and pathogenesis The aetiology of achalasia is unknown. Oesophageal dysmotility is also found in Chaga’s disease, scleroderma, oesophageal atresia, diabetes and secondary to gastrooesophageal reflux, but the unique feature of achalasia is the constantly nonrelaxing lower oesophageal sphincter. There are numerous theories regarding pathogenesis, the primary defect being described variously as neurogenic, myogenic and hormonal. There is evidence that it is the result of an abnormality of parasympathetic innervation. Absence of ganglion cells in the myenteric plexus in the dilated portion of the oesophagus with normal ganglion cells in the distal nondilated segment has been described.39 This, however, is not a constant feature and is reflected in the variable reports of the histopathology of some of the specimens of oesophageal muscle. These range from the total absence of ganglia, to the presence of normally ganglionated muscle or abnormal ganglion cell morphology. [****] Histochemical staining for acetylcholinesterase may reveal the presence of ganglion cells and nerve trunks in the myenteric plexus,



although their numbers are slightly reduced. Reports using electron microscopy and intestinal polypeptide hormonal essay support the theory that this is a neurogenic disorder.40



Diagnosis SYMPTOMATOLOGY



The principal symptoms of achalasia in childhood consist of vomiting, dysphagia, chest pain and weight loss. Dysphagia with the sensation of food sticking in the lower oesophagus and postprandial vomiting are the most frequent presenting symptoms. Retrosternal or epigastric pain manifests in one-third of the patients and in a few cases it is the primary presenting symptom. Weight loss of varying extent occurs in one-half of the patients. Nocturnal regurgitation may give rise to respiratory symptoms resulting in recurrent respiratory infections. Diagnostic delay is the norm with an average duration of symptoms prior to diagnosis of 24 months. Many children are treated for long periods for ‘cyclic vomiting’ or for ‘anorexia nervosa’.



RADIOLOGICAL FEATURES



If achalasia is suspected, order a chest x-ray and a barium swallow in consultation with a colleague in the radiology department. The plain chest radiograph may show a dilated food-filled oesophagus with an air-fluid level in the distal third. In addition, there may be radiological signs of repeated aspiration pneumonitis. The chief characteristics on barium oesophagogram are a dilated oesophagus, the absence of a stripping wave, incoordinated oesophageal contractions and obstruction at the oesophagogastric junction with prolonged retention of barium in the oesophagus. Failure of relaxation of the lower oesophageal sphincter leads to the classical rat-tail deformity of funnelling and narrowing of the distal oesophagus (Figure 101.4).41 ENDOSCOPY



Oesophagoscopy contributes little to the diagnosis, but retained food may be found within the dilated oesophagus. The main value of endoscopy is to exclude organic causes of obstruction in the oesophagus.



OESOPHAGEAL MANOMETRY



The diagnosis of achalasia is best confirmed by oesophageal motility studies using a constantly perfused catheter technique. The criteria for the diagnosis are:  a high-pressure (430 mmHg) lower oesophageal sphincter zone;



Chapter 101 Diseases of the oesophagus, swallowing disorders and caustic ingestion



] 1289



dilatation and oesophageal myotomy with or without the addition of an antireflux procedure.



PHARMACOLOGICAL TREATMENT



The manipulation of oesophageal motility disorders using pharmacological and dietary measures has been disappointing. [*] Reports of the use of isosorbide dinitrate and nifedipine have been more encouraging.42 [**/*] Nifedipine is a calcium entry blocker and since calcium ions are directly responsible for the activity of myofibrils and consequently the tension generated, their use in reducing the pressure in the lower oesophageal sphincter in achalasia, or for the vigorous oesophageal spasms, seems logical. Prostaglandin E2 has also been employed with some success. Its value in the long-term treatment of achalasia remains to be proven. The injection of botulinum toxin has not been a success in children.43, 44 [**/*]



FORCEFUL DILATATION Figure 101.4 Barium swallow in a child with achalasia of the oesophagus with the classical ‘rat-tail’ deformity of the distal end.



 failure of the lower oesophageal sphincter to relax in response to swallowing;  absence of propulsive peristalsis;  incoordinated tertiary contractions in the body of the oesophagus (Figures 101.5 and 101.6).



Treatment Three treatment options are available for the management of achalasia: pharmacological manipulation, forceful Normal



Good palliation may be obtained by dilatation. The most commonly used dilator consists of a single bag of fixed diameter which is inflated with water (Plummer) or air (Browne-McHardy, Rider-Moeller). Dilatation has been advocated as the treatment of choice in adults. Fellows et al.41 showed that, following pneumatic dilatation in adults, only 10 percent of patients subsequently required cardiomyotomy. In children, success rates ranging from 40 to 60 percent have been reported.45, 46 [***] The aim of forceful dilatation is to disrupt the muscle fibres of the lower oesophageal sphincter. There is, however, no evidence that the muscle fibres tear rather than stretch. Vantrappen and Janssens44 were unable to distinguish histologically between sphincter segments in dogs and monkeys subjected to forceful dilatations and those from normal controls. In general,



20 mmHg



Proximal



Mid



20 mmHg



LOS



20 mmHg



100 mm|min DS



DS



Figure 101.5 Oesophageal motility study in a normal child showing progression of the peristaltic waves through the oesophagus. DS, dry swallow.



1290 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY Achalasia



Proximal



20 mmHg



Mid



20 mmHg



LOS



20 mmHg



100 mm|min



WS



WS



WS



Figure 101.6 Oesophageal motility study in a child with achalasia showing completely incoordinate oesophageal contractions. WS, wet swallow.



WS



older patients respond better to pneumatic dilatation.47 A report of 899 adult patients treated at the Mayo Clinic concluded that myotomy was more successful and safer than dilatation, poor results being obtained twice as frequently following dilatation as after myotomy.45 The incidence of perforation following pneumatic dilatation varies from 1 to 5 percent.



SURGICAL PROCEDURE



Cardiomyotomy, as originally described by Heller in 1914, is the basis of all surgical procedures. The technique involves splitting the muscle fibres along the whole length of the lower oesophageal sphincter. The controversies concern the length of the oesophageal myotomy, the distance which the myotomy extends onto the stomach and the necessity for including an additional antireflux procedure.48



CORROSIVE INJURY TO THE OESOPHAGUS Accidental ingestion of caustic substances by children has become relatively uncommon in the developed world. This is as a result of government legislation regulating the use of caustics in commercially available drain cleaners and the introduction of child-proof containers. Parental awareness of the dangers has also contributed. In the developing world, lye burns to the oesophagus are still an enormous public health problem and a significant cause of mortality and of much long-term morbidity.49



Pathophysiology Caustic soda (sodium hydroxide) ingestion can cause severe injury to the oropharynx, oesophagus or stomach. The strong alkali rapidly penetrates the body tissues



producing an intense acute inflammatory reaction and oedema. If the concentration of the solution is high, transmural penetration occurs with resulting destruction of the musculature of the oesophagus, penetration into the peri-oesophageal tissues with mediastinitis or frank oesophageal perforation. The acute phase is followed by sloughing of the necrotic tissue and replacement by granulation tissue. The final outcome varies from complete resolution in the mild case to extensive fibrosis of the entire oesophagus. The extent and severity of the injury are directly related to the concentration of the lye ingested (liquid lye is more damaging than the granular form), the quantity ingested and the duration of contact.50, 51



Clinical presentation There is extensive oedema and swelling of the mouth and lips and the child is unable to swallow. There may be chest pain if large quantities of lye reach the stomach. Haematemesis, dyspnoea, stridor and other respiratory symptoms develop as a consequence of the resulting oedema or from direct laryngeal injury. Fibrosis of the lips and temporomandibular joint may develop as a result of severe oropharyngeal burns (Figure 101.7).



Diagnosis It is important to ascertain whether the lye was actually ingested or whether it entered the oral cavity only. Early endoscopic examination, within 12–24 hours of the injury, should be undertaken to determine whether or not the oesophagus is affected. Assessment of the extent of oesophageal burn is not possible at this early stage and endoscopic examination is terminated, once evidence of oesophageal injury is encountered, to minimize the risk of perforation.



Chapter 101 Diseases of the oesophagus, swallowing disorders and caustic ingestion



] 1291



Figure 101.7 Fibrosis of the lips following severe oropharyngeal burns secondary to caustic soda ingestion.



An early contrast oesophagogram will reveal the extent of the injury, determine the presence of a perforation and act as a baseline for evaluating future stricture formation (Figure 101.8).



Treatment EMERGENCY



Admission to hospital of all suspected cases is imperative. Vomiting should not be induced. No neutralizing agents should be given as these produce heat when interacting with the corrosive substance and increase the severity of the injury.



CONTINUED MANAGEMENT



Initial treatment should consist of broad-spectrum antibiotics and intravenous fluids. [**] The use of steroids (prednisolone 2 mg/kg/day)52 as advocated by Haller and Bachman,53 has been questioned and, in a prospective controlled trial, Anderson found steroids to be of no benefit.53, 54 [**/*] Ulman55 similarly found no objective evidence to support the use of steroids; nevertheless many clinicians feel a three-week course of prednisolone is helpful and it is widely used. If, on endoscopy, no oesophageal injury is found, active treatment is stopped and the patient is discharged. If a burn is found, antibiotics are continued for ten days. Oral feeds are commenced as soon as the child is able to tolerate fluids. A gastrostomy for feeding purposes may be necessary in cases with severe burns. In these patients, the opportunity may be taken to pass a string through the oesophagus to act as a guide for future dilatations. Another option is to stent the oesophagus early after ingestion of the caustic substance.56, 57 [**/*] A repeat oesophagogram is carried out three weeks after ingestion of the lye and oesophageal dilatations are commenced if a stricture is found. The dilatations



Figure 101.8 Barium oesophagogram in a child following ingestion of caustic soda. There is extensive stricture formation of the distal half of the oesophagus.



are repeated at regular intervals until the stricture is eliminated. Ninety percent of oesophageal strictures will respond to dilatation. The remainder will require oesophageal replacement.58



KEY POINTS  Polyhydramnios is nonspecific but is present in approximately ninety percent of mothers of infants with oesophageal atresia.  Fifty percent of babies with oesophageal atresia have additional malformations.  It is now rare for an infant to succumb from oesophageal atresia alone, unless it is associated with a major heart defect or extreme prematurity.



1292 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY  Ten to twenty percent of children with tracheo-oesohageal fistula have tracheomalacia which is responsible for ‘near miss’ apnoeic and/or cyanotic attacks or recurrent respiratory infections.  In the developing world, lye burns to the oesophagus are still a serious public health problem, a significant cause of mortality and the source of much long-term morbidity.



Best clinical practice [ Treatment of infants with oesophageal atresia should be concentrated in centres where the surgical expertise, supportive services (anaesthesia, paediatric intensive care, cardiology, radiology, pathology) and specialized nursing care are available. Transfer to such centres should be prompt. [Grade B] [ The diagnosis of oesophageal atresia is confirmed by passing a catheter through the mouth and into the oesophagus. If the tube enters the stomach there is no oesophageal atresia. [Grade C/D] [ If achalasia is suspected, order a chest x-ray and a barium swallow in consultation with a colleague in the radiology department. [Grade C/D]



Deficiencies in current knowledge and areas for future research



$ $ $ $



Knowledge of the embryology of the gastrointestinal tract and the genetic basis for congenital anomalies is growing. Better support services including paediatric intensive care facilities for infants with major congenital anomalies of the aerodigestive tract continue to improve survival. Multidisciplinary clinical input – including otolaryngologists – is likely to be increasingly accepted as routine for continuing care of these children. Greater awareness of the dangers of lye burns and legislation to improve containers for caustic agents would greatly reduce the misery of these injuries, particularly in the developing world.



REFERENCES 1. Kluth D, Fiegel H. The embryology of the foregut. Seminars in Pediatric Surgery. 2003; 12: 3–9.







2. Myers NA. The history of oesophageal atresia and tracheooesophageal fistula – 1670-1984. Progress in Pediatric Surgery. 1986; 20: 106–57. 3. Ashcraft KW, Holder TM. The story of esophageal atresia and tracheoesophageal fistula. Surgery. 1969; 65: 332–40. 4. Beasley SW, Myers NA. Trends in mortality in oesophageal atresia. Pediatric Surgery International. 1992; 7: 86–9. 5. Spitz L, Kiely EM, Morecroft JA, Drake DP. At risk groups in oesophageal atresia for the 1990’s. Journal of Pediatric Surgery. 1994; 29: 723–5. 6. Orford J, Glasson M, Beasley S, Shi E, Myers N, Cass D. Oesophageal atresia in twins. Pediatric Surgery International. 2000; 16: 541–5. 7. Ioannides AS, Chaudhry B, Henderson DJ, Spitz L, Copp AJ. Dorsoventral patterning in oesophageal atresia with tracheo-oesophageal fistula: evidence from a new mouse model. Journal of Pediatric Surgery. 2002; 37: 185–91. 8. Mere JM, Hutson JM. Embryogenesis of tracheo esophageal anomalies: a review. Pediatric Surgery International. 2002; 18: 319–26. 9. Berrocal T, Madrid C, Novo S, Gutierej J, Arionilla A, Gomez L. Congenital anomalies of the tracheobronchilal tree lung and mediastinum. Embryology, radiology and pathology. Radiographics. 2004; 24: 1527–33. Good review with an emphasis on medical imaging. 10. Goyal A, Potter F, Losty PD. Transillumination of H-type tracheoesophageal fistula using flexible miniature bronchoscopy: an innovative technique for operative localization. Journal of Pediatric Surgery. 2005; 40: e33–4. 11. Chittmitatrapap S, Spitz L, Brereton RJ, Kiely EM. Anastomotic stricture following repair of esophageal atresia. Journal of Pediatric Surgery. 1990; 25: 508–11. 12. Quan L, Smith DW. The VATER association. Journal of Pediatrics. 1973; 82: 104–7. 13. Greenwood RD, Rosenthal A. Cardiovascular malformation associated with tracheoesophageal fistula and esophageal atresia. Pediatrics. 1976; 57: 87–91. 14. Shulman A, Mazkereth R, Zalel Y, Kuint J, Lipitz S, Avigad I et al. Prenatal identification of esophageal atresia: the role of ultrasonography for evaluation of functional anatomy. Prenatal Diagnosis. 2002; 22: 669–74. 15. Langer JC, Hussain H, Khan A, Minkes RK, Gray D, Siegel M et al. Prenatal diagnosis of esophageal atresia using sonography and magnetic resonance imaging. Journal of Pediatric Surgery. 2001; 36: 804–7. 16. Stringer MD, McKenna KM, Goldstein RB, Filley RA, Scott Adzick N, Harrison MR. Prenatal diagnosis of esophageal atresia. Journal of Pediatric Surgery. 1995; 30: 1258–63. 17. Levine D, Barnewolt C, Mehta T, Trop I, Estrof J, Wong GF et al. thoracic abnormalities: MR imaging. Radiology. 2003; 228: 379–88. 18. Malone PS, Kiely EM, Brain AJ, Spitz L, Brereton RJ. Tracheo-oesophageal fistula and pre-operative mechanical ventilation: a dangerous combination. Australian and New Zealand Journal of Surgery. 1990; 60: 525–7.



Chapter 101 Diseases of the oesophagus, swallowing disorders and caustic ingestion 19. Rothenberg SS. Thoracoscopic repair of tracheoesophageal fistula in newborns. Journal of Pediatric Surgery. 2002; 37: 869–72. 20. Zee DC, Bax NM. Thoracoscopic repair of esophageal atresia with distal fistula. Surgical Endoscopy. 2003; 14. 21. Lividitis A. Esophageal atresia: a method of over-bridging large segmental gaps. Zeit Kinderchirugie. 1973; 13: 298–306. 22. Davenport M, Bianchi A. Early experience with oesophageal flap repair for oesophageal atresia. Pediatric Surgery International. 1990; 26: 89–91. 23. Kiely E, Spitz L. Is routine gastrostomy necessary in the management of oesophageal atresia. Pediatric Surgery International. 1987; 2: 6–9. 24. Waterston DJ. Reconstruction of the esophagus. In: Mustard WT, Ravitch MM, Snyder WH (eds). Pediatric surgery. Chicago: Year Book Medical, 1969: 400. 25. Anderson KD, Randolph JG. Gastric tube interposition. A satisfactory alternative to the colon for esophageal replacement in children. Annals of Thoracic Surgery. 1978; 25: 521–5. 26. Spitz L. Gastric transposition for esophageal substitution in children. Journal of Pediatric Surgery. 1992; 27: 252–9. 27. Ghandour KE, Spitz L, Brereton RJ, Kiely EM. Recurrent tracheo-oesophageal fistula: experience with 24 patients. Journal of Paediatric Child Health. 1990; 26: 89–91. 28. Wailoo MP, Emery JL. The trachea in children with tracheoesophageal fistula. Histopathology. 1979; 3: 329–38. 29. Kiely EM, Spitz L, Brereton R. Management of tracheomalacia by aortopexy. Pediatric Surgery International. 1987; 2: 13–5. 30. Ashcraft KW, Goodwin CD, Amoury RA, Holder TM. Early recognition and aggressive treatment of gastroesophageal reflux following repair of esophageal atresia. Journal of Pediatric Surgery. 1977; 12: 317–21. 31. Jolley SG, Johnson DG, Roberts CC, Herbst JJ, Matlak ME, McCombs A et al. Patterns of gastroesophageal reflux in children following repair of esophageal atresia and distal tracheoesophageal fistula. Journal of Pediatric Surgery. 1980; 15: 857–62. 32. Laks BH, Wilkinson RH, Schuster SR. Long-term results following correction of esophageal atresia and tracheoesophageal fistula. A clinical and cine-fluorographic study. Journal of Pediatric Surgery. 1972; 7: 591–7. 33. Payne WS, King RM. Treatment of achalasia of the esophagus. Surgical Clinics of North America. 1983; 63: 963–70. 34. London FA, Raab DE, Fuller J, Olsen AM. Achalasia in three siblings: a rare occurrence. Mayo Clinical Proceedings. 1977; 52: 97–100. 35. Stoddard CJ, Johnson AG. Achalasia in siblings. British Journal of Surgery. 1982; 69: 84–5. 36. Gallone L, Peri G, Galliera M. Proximal gastric vagotomy and anterior fundoplication as complementary procedures to Heller’s operation for achalasia. Surgery, Gynecology and Obstetrics. 1982; 155: 337–41.











] 1293



37. Aggestrup S, Uddman R, Sundler F, Fahrenkrug J, Hakanson R, Sorenson HR et al. Lack of vasoactive intestinal polypeptide nerves in esophageal achalasia. Gastroenterology. 1983; 84: 924–7. 38. Gelfond M, Rosen P, Gilat T. Isosorbide dinitrate and nifedipine treatment of achalasia: a clinical, manometric and radionuclide evaluation. Gastroenterology. 1982; 83: 963–9. 39. Hurwitz M, Bahar RJ, Ament ME, Tolia V, Molleston J, Reinstein LJ et al. Evaluation of the use of botulinum toxin in children with achalasia. Journal of Pediatric Gastroenterology and Nutrition. 2000; 30: 509–14. 40. Ip KS, Cameron DJ, Catto-Smith AG, Hardikar W. Botulinum toxin for achalasia in children. Journal of Gastroenterology and Hepatology. 2000; 15: 1100–4. 41. Fellows IW, Ogilvie AL, Atkinson MP. Pneumatic dilatation in achalasia. Gut. 1983; 24: 1020–3. 42. Vane DW, Cosby K, West K, Grosfeld JL. Late results following esophagomyotomy in children with achalasia. Journal of Pediatric Surgery. 1988; 23: 515–9. 43. Babu R, Grier D, Cusick E, Spicer RD. Penumatic dilatation for childhood achalasia. Pediatric Surgery International. 2001; 17: 505–7. 44. Vantrappen G, Janssens J. To dilate or operate? This is the question. Gut. 1983; 24: 1013–9. 45. Payne WS, King RM. Treatment of achalasia of the esophagus. Surgical Clinics of North America. 1983; 63: 963–70. 46. Buick RG, Spitz L. Achalasia of the cardia in children. British Journal of Surgery. 1985; 72: 341–3. 47. Patti MG, Albanese CT, Holcomb 3rd GW, Molena D, Fisichella PM, Perretta S et al. Laparoscopic Heller myotomy and Dor fundoplication for esophageal achalasia in children. Journal of Pediatric Surgery. 2001; 36: 1248–51. 48. Ellis FH, Gibb SP, Crozier RE. Esophagomyotomy for achalasia of the esophagus. Annals of Surgery. 1980; 192: 157–61. 49. Spitz L. Management of ingested foreign bodies in childhood. British Medical Journal. 1971; 4: 469–72. 50. Hamza A, Abdelhay S, Sherif H, Hasan T, Soliman H, Kabesh A et al. Caustic esophageal strictures in children: 30 years experience. Journal of Paediatric Surgery. 2003; 38: 828–33. Reports an extensive experience of caustic strictures in a tertiary referral centre, 850 cases, the majority requiring oesophageal replacement. 51. Spitz L, Kiely E, Pierro A. Gastric transposition in children a twenty-one year experience. Journal of Paediatric Surgery. 2004; 39: 276–81. Account of the author’s extensive experience including secondary repairs of oesophageal atresia and caustic strictures. 52. Spitz L, Hirsig J. Prolonged foreign body impaction in the oesophagus. Archives of Disease in Childhood. 1982; 57: 551–3. 53. Haller JA, Bachman K. The comparative effect of current therapy of experimental caustic burns of the esophagus. Journal of the American Medical Association. 1963; 186: 262.



1294 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY 54. Vaishnav A, Spitz L. Alkaline battery-induced tracheooesophageal fistula. British Journal of Surgery. 1989; 76: 1045. 55. Ulman I, Mutaf O. A critique of steroids in the management of caustic esophageal burns in children. European Journal of Paediaric Surgery. 1998; 8: 71–4. 56. Anderson KD, Rouse TM, Randolph JG. A controlled trial of corticosteroids in children with corrosive injury of the



oesophagus. New England Journal of Medicine. 1990; 323: 637–40. 57. de Jong AL, Macdonald R, Ein S, Forte V, Turner A. Corrosive esophagitis in children: a 30-year review. International Journal of Pediatric Otorhinolaryngology. 2001; 57: 203–11. 58. Viiala C, Collins B. use of multiple self-expanding metal stents to treat corrosive induced esophageal strictures. Endoscopy. 2001; 33: 291–2.



102 Imaging in paediatric ENT NEVILLE WRIGHT



Introduction Imaging modalities Radiation protection Patient preparation Key points



1295 1296 1300 1301 1302



Best clinical practice Deficiencies in current knowledge and areas for future research References



1302 1303 1303



SEARCH STRATEGY This chapter is supported by a PubMed search using the key words children, ENT, imaging, radiology, ultrasound, magnetic resonance, computed tomography, bronchography, play therapy, scintigraphy and sialography.



INTRODUCTION Paediatric ear, nose and throat (ENT) imaging is a challenging subject both for the radiologist and the otolaryngologist. In addition to the anatomical complexities of middle and inner ear structure, the convolutions of the turbinates, the pneumatization pattern of the sinuses, and the dynamics of the swallowing mechanism, both the radiologist and the surgeon need to be aware of the variations during normal growth of the child.



Many of the imaging techniques used in paediatric ENT radiology are similar to those used in adult practice (Table 102.1), but the diseases encountered may be entirely different. Consideration of the need for sedation or general anaesthesia is an important factor when deciding to image a child. The number of plain radiographs performed in children has significantly diminished over recent years with the widespread availability of computed tomography (CT), especially for imaging the ear and sinuses. It is



Table 102.1 Imaging modalities pertinent to paediatric ENT radiology. Imaging modalities Plain radiography Fluoroscopy Ultrasound Computed tomography Magnetic resonance imaging Nuclear medicine Interventional procedures and angiography



Mastoid views, lateral soft tissue neck, limited role for sinus views Videofluoroscopy, contrast swallows, bronchography, sialography, chest and diaphragm screening Including Doppler and colour flow studies Standard, helical, high resolution, contrast-enhanced and angiography Including angiographic sequences, perfusion and possibly functional studies Thyroid, parathyroid and salivary scintigraphy, perchlorate discharge test Percutaneous biopsy and aspiration, balloon dilatation, stent placement, embolization



1296 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY important to minimize radiation exposure in children and in this context magnetic resonance imaging (MRI), which avoids the use of ionizing radiation, has an expanding and important role. Dynamic contrast bronchography is also being used in some centres for ‘real-time’ assessment of the upper airway, particularly in ventilated children when it can be combined with simultaneous pressure measurements to give valuable information about opening airway pressures.



IMAGING MODALITIES Plain radiographs (x-rays) Much of the older radiological literature details an array of radiographic projections and plain x-ray tomographic techniques to assess the airway, petrous bone and mastoid region. These have now largely been replaced by crosssectional imaging (CT and MRI). There is still a limited role for mastoid views to define the extent of pneumatization of the air cells, but this should be restricted to requests from ENT specialists. The projections obtained are no different to those used in adult practice. Plain radiography of the airway is usually limited to a lateral film and in this instance air acts as a negative contrast medium outlining the anatomical details. For the most reliable results, the film should be obtained during inspiration with the neck slightly extended. This is because there is considerable variation in the appearances of the soft tissues of the nasopharynx in the young child dependent on posture. This is especially so when the neck is flexed and the child is breathing out, when apparent soft tissue swelling can mimic a pharyngeal mass. A high kilo-voltage radiographic technique with added tube filtration (‘Cincinnati view’) also used to be recommended for airway assessment, but this is no longer necessary following the development of digitally enhanced computed radiography, CT and MRI, all of which give much better detail. Sinus x-rays are difficult to interpret in the young child, where mucosal thickening can be a normal finding. They should be avoided in the under five year olds,1 [***] and should be restricted to requests by ENT specialists only. In the acute setting, they are indicated to exclude fractures and foreign bodies. Radiology for nasal bone fractures should be deferred until 10–14 days have elapsed since injury, and even then are rarely clinically indicated. Despite all the advances in imaging technology, it is important to remember that a chest x-ray can still be very informative, particularly to detect air-trapping and foreign body inhalation.



Fluoroscopy Fluoroscopy is the study of sequential dynamic images. Techniques pertinent to paediatric ENT radiology include



videofluoroscopy of the swallowing mechanism, videofluoroscopy during phonation, barium or contrast swallows, tracheo-oesophageal fistul-o-grams (TOFo-grams), chest and diaphragm screening, sialography and dynamic contrast enhanced bronchography. Radiographic assessment of the swallowing mechanism and phonation studies require fluoroscopy of the upper airway and palate. The assistance of a speech and language therapist is mandatory for best results. For the assessment of swallowing, sequential analysis of the swallowing cycle requires recording of data on videotape or other media storage device for later analysis. In general, screening in the lateral position will suffice. Most children require assessment with various types of food consistency. Yoghurt mixed with barium, thickened barium and liquid barium should generally be assessed, with other foodstuffs used as appropriate. The examination should be modified depending on the child’s age and feeding habits. For phonation studies, ideally a small quantity of barium is administered to each nostril and then sniffed to coat the soft palate. Clearly barium may have to be omitted in the very young child. Fluoroscopy is performed with the child in a modified Towne’s (chintuck) position to obtain a tangential view of the velopharyngeal portal and then subsequently in the lateral projection. When no barium has been administered, only the lateral projection is performed. Barium and water-soluble contrast studies are useful for assessing the oesophagus for strictures, vascular rings, gastro-oesophageal reflux and oesophagitis. Barium is the preferred contrast agent, unless aspiration is highly likely, or in the postoperative state where a suitable watersoluble agent is preferred. The procedure is similar to that performed in adults, although it will need alteration with young children who are usually fed supine rather than erect. A modified technique is required to demonstrate a tracheo-oesophageal fistula (TOF), the technique often referred to as the ‘TOF-o-gram’. In this procedure the child has a nasogastric tube placed into the stomach and lies in a prone position. The tube is incrementally withdrawn while injecting a suitable contrast agent and simultaneously screening the oesophagus in a lateral projection. The effect of gravity and the careful withdrawal of the tube under direct vision is felt to enhance the chances of identifying a fistula. [*] Diaphragm screening is a simple procedure involving visualization of diaphragmatic movement during normal respiration, coughing, sniffing and Valsalva manoeuvres. The degree of diaphragmatic excursion and presence or absence of paradoxical movement should be recorded. Fluoroscopy can also be used to assess air-trapping if inhaled foreign bodies are suspected. Some children require dynamic ‘real-time’ evaluation of the airway using bronchography. A small quantity of nonionic water-soluble contrast is injected into the trachea and bronchi and the airway is then screened to assess changes in airway calibre during the respiratory



Chapter 102 Imaging in paediatric ENT



cycle. This is limited to children requiring ventilatory support and should only be carried out with full resuscitation facilities available. Clearly its use is restricted to specialized paediatric centres. It is mainly used for confirmation of suspected anatomical abnormalities where there is doubt or to confirm the extent of tracheobronchomalacia and evaluate airway opening pressures.2, 3 [**] Sialography of the parotid or submandibular glands can be performed in children. The technique is similar to that performed in adults, although the procedure may need to be carried out under anaesthesia in the younger child. Sialectasis, stone formation and ductal abnormalities may be identified. Sialography may be combined with CT or MRI examinations. Sialography has also been performed successfully using a digital subtraction technique to obtain superior image quality.4 [**] MR sialography, using heavily T2-weighted sequences, is a noninvasive alternative to cannulation techniques, but is unable to provide the excellent spatial resolution (and therefore image quality) that can be obtained with a formal sialogram.5 [**]



Ultrasound Ultrasound (sonography) is a noninvasive, portable and easily accessible method of assessing the neck and thoracic inlet. Importantly, it uses no ionizing radiation. Ultrasound should be the first imaging investigation for a neck mass, although it has a limited role in assessing deeply placed pathologies (Figure 102.1). It is especially good at distinguishing solid from cystic lesions (e.g. thyroglossal cysts). Combined with Doppler and colour flow imaging, ultrasound provides vital information about the vascularity of masses, such as haemangiomas and cystic



Figure 102.1 Ultrasound is a readily available, noninvasive technique which can quickly distinguish solid from cystic lesions – this scan shows a lymph node.



] 1297



hygromas. Ultrasound has also been used successfully to assess the unossified, cartilaginous structures of the larynx.6 [**] It has the advantages of providing real-time images under physiological conditions and may be useful in assessing vocal cord movement,7 but it has not yet replaced direct inspection of the cords at endoscopy. [***] Ultrasound can also evaluate diaphragm movement, although in practical terms this is often restricted to the intensive care unit setting when the child cannot be moved easily for fluoroscopic assessment. It can also guide biopsy and aspiration of neck lesions, although this is carried out much less frequently in children than adults.



Computed tomography Plain x-ray tomography is now a technique of the past. CT provides good soft tissue characterization, but excels in the definition of fine bony detail. With the development of faster scan times, helical scanning and improved resolution, CT can now provide superb images of the middle and inner ears (Figure 102.2), although it is less effective at assessing the retrocochlear auditory pathway.8 [**] The major drawback for CT, however, is its use of ionizing radiation. Generally, high resolution 1 mm scans should be obtained in two planes, the transverse and the coronal, although with some of the more modern scanners transverse images can be obtained and then the coronal images reconstructed from the scan dataset, without loss of detail and negating the need to do two scans. This helps to reduce the radiation exposure to the child. The images should all be post-processed using a sharp filter to improve bony detail. CT of the petrous temporal bone complements MRI and is essential in preoperative assessment for cochlear implants. CT also has a role in the assessment of children’s sinuses and the nasal airway, although this is much less frequently required than that in adults. It is particularly pertinent in children with orbital cellulitis, where contrast-enhanced CT will define the presence and extent of retro-orbital disease much more clearly than clinical examination. [**] Evaluation of choanal atresia is also facilitated by defining the bony elements on CT, although this often requires mucosal decongestants and suction of nasal secretions to improve the sensitivity of the scan. Volume acquired CT information using helical or electron beam scanners can be used to assess the upper airway, including the larynx,9 [**] and may be used to generate ‘virtual’ bronchoscopic images (Figure 102.3). This is a significant step forward in imaging terms, but it still falls short of the fine detail which can be seen with direct vision. CT angiography is rapidly approaching the degree of definition previously provided only by formal invasive angiography, and this can be extremely helpful in conditions such as juvenile nasopharyngeal angiofibroma.



1298 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Figure 102.3 Single transverse image from CT virtual bronchogram dataset clearly shows the carina.



Figure 102.2 Transverse CT image through the petrous bone showing: (a) a dilated vestibule; and (b) a Mondini abnormality of the cochlea.



Magnetic resonance imaging MRI is a noninvasive, multiplanar imaging technique with excellent soft tissue discrimination. Importantly for children, it does not use ionizing radiation, but unfortunately may require sedation or anaesthesia. It is a rapidly advancing technology in which numerous acronyms are used to describe the multitude of imaging sequences used. In simple terms however, T1-weighted sequences provide anatomical definition and T2-weighted sequences demonstrate areas of high fluid content, in particular oedema. MRI excels in the demonstration of the brain and spinal cord. However, there is increasing use of MRI, not just to assess the posterior fossa and other intracranial structures, but also inner ear abnormalities. T2-weighted sequences



provide the optimum information about the inner ear, clarifying the presence of fluid filled structures, such as the cochlea, vestibule and semicircular canals, but occasionally T1-weighted and post-contrast images may be helpful. Vestibular schwannoma is an uncommon condition in children and most imaging in the cerebello-pontine angle is focussed on identifying other mass lesions, the brainstem and the seventh and eighth nerves. When intracranial extension of disease processes in the ear or paranasal sinuses occurs, there may be involvement of the venous sinuses. Magnetic resonance (MR) venography provides a noninvasive method of assessing involvement of these venous sinuses (Figure 102.4).10 Further advances in MR technology now also permit exquisitely detailed three-dimensional reconstructions of the inner ear.11 [**] Velopalatine movement can also be successfully assessed with MRI, using fast T1-weighted sequences during phonation of different sounds and replaying them in cine mode.12, 13 [*] In cooperative children this may be a satisfactory alternative to videofluoroscopy, although this does require MR scanners of high specification.14 The excellent soft tissue discrimination of MRI can also be effectively used in imaging the neck when, after ultrasound, MRI is the next most useful imaging modality (Figure 102.5). It is especially helpful in defining the extent and nature of cystic masses, such as cystic hygromas (Figure 102.6), when extension into the thoracic inlet can also be assessed. Parotid and submandibular gland assessment is also more accurate with MRI. Fat-suppressed MRI (e.g. STIR or ‘short tau inversion recovery’ sequences) are especially useful in identifying lymphadenopathy and adenoidal and tonsillar tissue.



Chapter 102 Imaging in paediatric ENT



] 1299



MR angiography is now a viable alternative to invasive formal angiography, especially for the assessment of vascular malformations. However, formal angiography may still be required for endovascular treatment of lesions.



Nuclear medicine Nuclear medicine has an important role in assessment of the thyroid and parathyroid glands. It can be used to



Figure 102.4 Coronal MR venogram, a noninvasive study, showing absent flow in the right transverse venous sinus consistent with venous thrombosis.



Figure 102.6 MR images include (a) coronal T1-weighted and (b) transverse T2-weighted scans of the neck showing an extensive cystic hygroma extending into the nasopharynx and posteriorly around the neck.



Figure 102.5 Coronal T1-weighted MR images of the neck show a discrete lesion in the right parotid gland later proven to be a pleomorphic adenoma: (a) pre-contrast; (b) post-contrast.



1300 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY assess the presence, size and distribution of thyroid and parathyroid tissue, and is most frequently used in young children with hypothyroidism. The usual technique for evaluating the thyroid gland involves the injection of 99m technetium (Tc) pertechnate and is similar to that performed in adults. Alternatively 123iodine can be used. In some children there may be abnormalities of organification of iodine into tyrosine and this can be assessed with the perchlorate discharge test (Figure 102.7).15 [**] Parathyroid assessment is now more easily performed using 99mTc-labelled sestamibi to identify increased parathyroid metabolism.16 [**] Salivary scintigraphy can also be performed in children to assess parenchymal function and excretion of the salivary glands.17 [*] Its main indication is in the assessment of connective tissue disease and occasionally to look for salivary duct obstruction when sialography cannot be performed.



insertion of metallic stents. Balloon dilatation is performed under fluoroscopic guidance, and can be repeated if necessary. Stent placement again requires fluoroscopic guidance, but the long-term effects of these placements have yet to be fully evaluated.18 [*] Formal angiography is an invasive technique requiring placement of an intravascular catheter usually by the femoral artery, and subsequent contrast injection to delineate the vessels (Figure 102.8). Obtaining vascular access in young children can be a challenge and selective vessel catheterization may also be difficult, although children do not usually suffer from atherosclerotic vessels encountered in adult practice! Diagnostic studies should be combined with interventional procedures if possible. Embolization for vascular head and neck lesions can be used for definitive treatment or preoperatively and involves selective catheterization and usually particulate embolization or injection of sclerosant.19 [**]



Interventional procedures and angiography



RADIATION PROTECTION



Most interventional techniques in children require the assistance of sedation or, more commonly, general anaesthesia to ensure a safe, controlled procedure. Percutaneous biopsy of head and neck lesions is much less commonly performed in children, but the technique is similar to that performed in adults. Aspiration of cysts under ultrasound or CT control is effective and when appropriate can be combined with sclerotherapy, for example for treatment of extensive cystic hygroma. The primary treatment for tracheobronchial stenosis remains surgery, but there are now radiological interventional alternatives, such as balloon dilatation and



Of paramount importance when considering whether a child needs a diagnostic test is balancing the need and quality of the information the test will provide with the risk associated with the test. A useful test is one in which the result, whether positive or negative, will affect patient management or improve confidence in the clinician’s diagnosis. The prime area of concern in radiological



270



Perchlorate given



250 230 210



Counts/sec



190 170 150 130 110 90 70 50 30 0



2



4



6



8



10 Min



12



14



16



18



20



Figure 102.7 Time-activity curve from a perchlorate discharge test demonstrates displacement of activity from the thyroid confirming an organification defect.



Figure 102.8 Formal angiography demonstrates abnormal vascularity within a lobular mass in the right side of the neck.



Chapter 102 Imaging in paediatric ENT



imaging is the risk from ionizing radiation. Radiological tests are an accepted part of medical practice in which clinical benefits should clearly exceed the small radiation risk. Statutory regulations in the UK, The Ionizing Radiation (Medical Exposure) Regulations 2000, require all involved to reduce unnecessary exposure of patients to radiation.20 This is especially important in children, in whom there is increased sensitivity to the effects of radiation. One important way of reducing the radiation burden to a patient is ensuring unnecessary tests are avoided, especially avoiding repeat examinations. It is useful to consider radiation dose in terms of the total radiation risk to the body, calculated from the sum of the doses to a number of body tissues. The weighting factor for each tissue varies upon its relative sensitivity to radiation-induced cancer or severe hereditary effects. This gives a single effective dose to the body. It is also useful to consider the effective dose in terms of chest x-ray equivalents. Table 102.2 gives typical effective doses for a number of examinations pertinent to paediatric ENT radiology.21 It is important to note that a CT examination administers a relatively high radiation dose and thus it is certainly worthwhile considering alternative imaging modalities such as ultrasound and MRI.



] 1301



hugely beneficial, and make the procedure both pain free and uneventful. Indeed, the importance of access to play therapy for children has recently been recognized at governmental level in the UK National Service Framework (NSF) document.22 When intravenous contrast agents will be required, it is important to notify the radiology department of any allergies or previous problems with contrast agents to ensure the procedure can run smoothly. The choice of contrast agent, the examination technique or even the particular modality to be used to image the child may be modified depending on the history provided. Where complex procedures (such as angiography) require sedation or general anaesthesia, the child will need admission to a ward and appropriate preparation, depending on the particular procedure to be performed. Informed consent is mandatory for invasive procedures.



Sedation and general anaesthesia For some imaging tests, particularly those requiring long periods of immobility such as MRI, young children will require sedation or general anaesthesia. Plain x-rays and ultrasound rarely require either, whereas interventional studies and angiography for head and neck lesions almost invariably require general anaesthesia. Particular care must be exercised in using sedation in children with airway problems, especially in the MR environment where the child may spend some time within the bore of the magnet. It is also mandatory to have appropriate MR compatible monitoring equipment in such circumstances (Figure 102.9). The choice of sedation or anaesthesia will depend on local practice and expertise. In general, children under the age of four to five years will require some form of assistance to maintain adequate immobility to ensure diagnostic images. Very young children, under three months, can often simply be fed and wrapped up well. Children older than five years or weighing more than 20 kg are often difficult to sedate adequately, and it may



PATIENT PREPARATION General approach It is important that children and their parents understand why any imaging test has been requested and what to expect from the examination. This information is best addressed initially in the outpatient environment during the consultation process, but is usually supplemented by information provided by the radiology department. Information, both verbal and written, needs to be appropriate for both child and adult. Some examinations require venous access and this can be distressing for both the child and their carer. The use of topical anaesthetics and play therapists using distraction techniques can be Table 102.2 Typical effective doses from diagnostic medical exposure. Diagnostic procedure



Chest (single PA film) Skull Barium swallow CT head Thyroid (Tc-99 m) radionuclide scan Ultrasound Magnetic resonance imaging a



Typical effective dose (mSv)a



Equivalent number of chest x-rays



Approximate equivalent period of natural background radiation



0.02 0.06 1.5 2.0 1



1 3 75 100 50



3 days 9 days 8 months 10 months 6 months



0 0



UK average background = 2.2 mSv per year, range 1.5–7.5 mSv.



0 0



0 0



1302 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



Figure 102.9 MR compatible monitoring equipment is mandatory for safe practice in the MR environment – the magnetic field is permanently on!



be prudent to proceed directly to general anaesthesia. Some of the faster CT scanners are reducing the tendency to require sedation or anaesthesia, and play therapists can be especially useful in the MR environment for fidgety children.23 [**/*] There are many different regimens for sedation and Tables 102.3 and 102.4 show only two examples. Sedation protocols will need to be developed depending on local equipment and expertise and with close cooperation between the departments of radiology, ENT and anaesthesia. It is important to reiterate that the MR environment requires specialized, MRI-compatible monitoring equipment, which must at all times be readily available.



Table 102.3 Possible sedation regimens, part 1. Age



oThree months 4Three months



Fasting time (hours) Clear fluids



Milk



Solids



2 2



4 6



6 6



Table 102.4 Possible sedation regimens, part 2. Weight o20 kg



420 kg



Suggested drug, 1 month to 18 years Oral chloral hydrate 100 mg/kg body weight, maximum dose 2000 mg May be supplemented with rectal paraldehyde, unless neutropenic 0.3 mL/kg body weight 1 0.3 mL/kg of olive oil, maximum dose 10 mL paraldehyde 1 10 mL olive oil Consider oral quinalbarbitone 7.5–10 mg/kg body weight, maximum dose 200 mg



KEY POINTS  Plain sinus radiography is of limited use in children.  Ultrasound is the first line imaging investigation for a neck mass.  Bony detail is best demonstrated by CT.  When cross-sectional imaging is considered necessary, ultrasound and magnetic resonance imaging should be considered as alternatives to CT to minimize the radiation burden to the child.  Magnetic resonance imaging provides excellent soft tissue imaging in multiple planes and can be used to provide noninvasive angiographic information.  Sedation or anaesthesia may be needed for examinations which are time-consuming and require prolonged immobility of the child.



Best clinical practice The following is a brief summary of imaging pathways for investigating some common paediatric ENT problems. The reader is referred to the relevant areas elsewhere in the book for more detail.



Craniofacial malformations [ CT with multiplanar reconstruction – for bone and soft tissue demonstration.



[ MRI – for midline defects with a possible intracranial extension.



Chapter 102 Imaging in paediatric ENT



$



Deafness and prior to cochlear implantation [ CT to assess the temporal and parietal bones and middle ear, cochlear morphology and cochlear patency.24 [ MRI to assess cochlear morphology, cochlear patency and the retrocochlear auditory pathway.25



$



] 1303



The development of imaging agents which act at a cellular or molecular level may provide therapeutic as well as diagnostic advances. The role of functional MR in evaluating the auditory pathway needs further investigation.26



Trauma to the head and face [ CT for complex fractures. Paranasal sinus disease Plain sinus radiographs should be limited to requests from ENT specialists. Indications include: [ single occipitomental (OM) or lateral view for foreign body; [ single OM view to confirm maxillary opacification prior to antral washout in suspected maxillary empyema; [ otherwise low-dose CT.



Neck masses including lymphadenopathy [ Ultrasound, with or without colour Doppler imaging – often the only imaging examination required to confirm lymphadenopathy. [ CT or MRI.



Airway obstruction [ [ [ [



Lateral soft tissue view of the neck. Chest x-ray. Barium swallow. Contrast enhanced CT or MRI – if a vascular ring is suspected.



Deficiencies in current knowledge and areas for future research



$



$ $



Many of the imaging techniques used in paediatric ENT radiology have not been evaluated in an evidence-based manner, but have crept into routine use as custom and practice allowed. As some techniques, such as helical CT, impart a considerable radiation burden on the child and are becoming more readily available, it is important that any advances are carefully assessed to ensure their impact on clinical care is beneficial. Although advances in imaging technology have provided the opportunity to display pathology as a three-dimensional entity, it remains unclear what quantifiable benefits this provides the surgeon. The long-term effect of interventional procedures requiring insertion of implants, such as tracheal or bronchial stents, is unknown.



REFERENCES







1. RCR Working Party. Making the best use of a department of clinical radiology: guidelines for doctors, 5th edn. London: The Royal College of Radiologists, 2003. Guidance on best practice. 2. MacIntyre P, Peacock C, Gordon I, Mok Q. Use of tracheobronchography as a diagnostic tool in ventilatordependent infants. Critical Care Medicine. 1998; 26: 755–9. 3. Little AF, Phelan EM, Boldt DW, Brown TC. Paediatric tracheobronchomalacia and its assessment by tracheobronchography. Australasian Radiology. 1996; 40: 398–403. 4. Ilgit ET, Cizmeli MO, Isik S, Arac M, Altin M, Koker E. Digital subtraction sialography: technique, advantages and results in 107 cases. European Journal of Radiology. 1992; 15: 244–7. 5. Kalinowski M, Heverhagen JT, Rehberg E, Klose KJ, Wagner HJ. Comparative study of MR sialography and digital subtraction sialography for benign salivary gland disorders. American Journal of Neuroradiology. 2002; 23: 1485–92. 6. Garel C, Contencin P, Polonovski JM, Hassan M, Narcy P. Laryngeal ultrasonography in infants and children: a new way of investigating. Normal and pathological findings. International Journal of Pediatric Otorhinolaryngology. 1992; 23: 107–15. 7. Vats A, Worley GA, deBruyn R, Ponter H, Albert DM, Bailey CM. Laryngeal ultrasound to assess vocal fold paralysis in children. Journal of Laryngology and Otology. 2004; 118: 429–31. 8. Nikolopoulos TP, O’Donoghue GM, Robinson KL, Holland IM, Ludman C, Gibbin KP. Preoperative radiologic evaluation in cochlear implantation. American Journal of Otology. 1997; 18: S73–4. 9. Korkmaz H, Cerezci NG, Akmansu H, Dursun E. A comparison of spiral and conventional computerized tomography methods in diagnosing various laryngeal lesions. European Archives of Oto-rhino-laryngology. 1998; 255: 149–54. 10. Connor SE, Jarosz JM. Magnetic resonance imaging of cerebral venous sinus thrombosis. Clinical Radiology. 2002; 57: 449–61. 11. Murugasu E, Hans P, Jackson A, Ramsden RT. The application of three-dimensional magnetic resonance imaging rendering of the inner ear in assessment for cochlear implantation. American Journal of Otology. 1999; 20: 752–7.



1304 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY 12. Akguner M. Velopharyngeal anthropometric analysis with MRI in normal subjects. Annals of Plastic Surgery. 1999; 43: 142–7. 13. Fitch WT, Giedd J. Morphology and development of the human vocal tract: a study using magnetic resonance imaging. Journal of the Acoustical Society of America. 1999; 106: 1511–22. 14. Vadodaria S, Goodacre TE, Anslow P. Does MRI contribute to the investigation of palatal function? British Journal of Plastic Surgery. 2000; 53: 191–9. 15. El-Desouki M, al-Jurayyan N, al-Nuaim A, al-Herbish A, Abo-Bakr A, al-Mazrou Y et al. Thyroid scintigraphy and perchlorate discharge test in the diagnosis of congenital hypothyroidism. European Journal of Nuclear Medicine. 1995; 22: 1005–8. 16. Rauth JD, Sessions RB, Shupe SC, Ziessman HA. Comparison of Tc-99m MIBI and TI-201/Tc-99m pertechnetate for diagnosis of primary hyperparathyroidism. Clinical Nuclear Medicine. 1996; 21: 602–8. 17. Klutmann S, Bohuslavizki KH, Kroger S, Bleckmann C, Brenner W, Mester J et al. Quantitative salivary gland scintigraphy. Journal of Nuclear Medicine Technology. 1999; 27: 20–6. 18. Sommer D, Forte V. Advances in the management of major airway collapse: the use of airway stents. Otolaryngologic Clinics of North America. 2000; 33: 163–77. 19. Robson CD. Vascular lesions of the head and neck in children. In: King SJ, Boothroyd AE (eds). Pediatric ENT radiology. Berlin: Springer, 2002: 267–88.











20. Health and Safety Executive. The ionising radiation (medical exposure) regulations 2000. London: HMSO, 2000. 21. National Radiological Protection Board. Radiation exposure of the UK population from medical and dental x-ray examinations, NRPB-W4. Didcot: NRPB, 2001. Regarding dose implications of radiological procedures. 22. Department of Health. Getting the right start: National service framework for children. Standard for hospital services. Department of Health, 2003. 23. Pressdee D, May L, Eastman E, Grier D. The use of play therapy in the preparation of children undergoing MR imaging. Clinical Radiology. 1997; 52: 945–7. Evaluating play therapy in facilitating MR scanning. 24. Antonelli PJ, Varela AE, Mancuso AA. Diagnostic yield of high-resolution computed tomography for pediatric sensorineural hearing loss. Laryngoscope. 1999; 109: 1642–7. 25. Westerhof JP, Rademaker J, Weber BP, Becker H. Congenital malformations of the inner ear and the vestibulocochlear nerve in children with sensorineural hearing loss: evaluation with CT and MRI. Journal of Computer Assisted Tomography. 2001; 25: 719–26. 26. Zur KB, Holland SK, Yuan W, Choo DI. Functional magnetic resonance imaging: contemporary and future use. Current Opinion in Otolaryngology and Head and Neck Surgery. 2004; 12: 374–7.



103 Medical negligence in paediatric otolaryngology MAURICE HAWTHORNE



Introduction Hearing disorders Cholesteatoma Congenital fixation of the stapes Atlanto-axial instability Tonsillectomy and adenoidectomy Acute epiglotitis Tracheostomy Foreign bodies



1305 1306 1306 1307 1308 1308 1309 1309 1309



Neck masses Epistaxis Nasal trauma Key points Best clinical practice Deficiencies in current knowledge and areas for future research References Further reading



1310 1310 1310 1311 1311 1311 1311 1311



SEARCH STRATEGY Much of the material for this chapter is based on anecdotal cases. This is supported by a manual search of Clinical risk (www.rsmpresss.co.uk/cr.htm) and a search of the site http://www.lawreports.co.uk. Standard textbooks of medical law were also consulted.



INTRODUCTION The reader is referred to Chapter 48, Medical jurisprudence and otorhinolaryngology, for an outline of the principles of medical jurisprudence. This chapter and the foregoing are based on the law in England and Wales and the definition of ‘negligence’ as applied by courts in this jurisdiction. In summary, ‘negligence’ requires that the aggrieved party establish that there was a duty of care, that there was a breach of that duty and that harm followed. The standard of care is that of the reasonably skilled and experienced doctor. The appropriate test, known as the Bolam (Bolam v Friern Hospital Management Committee)1 test, states that it is: the standard of the ordinary skilled man exercising and professing to have that special skill. A man need not possess the highest expert skill; it is well-established law that it is sufficient if he exercises the ordinary skill of an ordinary competent man exercising that particular art.



Despite differences in the legal framework within which otolaryngologists work, depending on their geographic area of practice, it is hoped that the principles discussed here are applicable internationally. The focus is on optimum care for children. Approximately two-fifths of the workload of a typical National Health Service department of otolaryngology is in the management and treatment of ENT disorders in children. The vast majority of children attending an ENT department have common conditions that are not especially challenging to diagnose; the operations involved are usually straightforward to perform, with low surgical risk. Tumours in general, and head and neck cancers in particular, are rare in children. Recognition of rarities, such as cavernous sinus thrombosis or tuberculosis of the middle ear in an infant, may be delayed and give rise to litigation, but patterns of error with more common pathologies occur and the aim of this chapter is to convey practical advice as to how best to avoid these patterns rather than dwell on the esoteric.



1306 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



HEARING DISORDERS



Negligent causation



Delayed diagnosis



Profound deafness at birth may arise from a negligent act in the antenatal period. For example, in one case a mother was established as lacking immunity to rubella following the birth of her first child. The hospital failed to arrange a rubella vaccination, as did the GP. A year later the mother attended at nine weeks gestation with her second pregnancy, having just contracted rubella. The general practitioner reassured her that, as it was coming to the end of her first trimester, there was no risk to the foetus of damage from the maternal rubella infection. As a result of this advice, the mother decided not to pursue a termination of the pregnancy, which had been her original intention. The child was born profoundly deaf, and after many years of protracted litigation, the case settled in 1994 in favour of the claimant, for a total of £300,000, including £50,000 general damages.



Although parents may be convinced that there has been a significant delay in the diagnosis of a child with profound hearing loss, it is often difficult for them to prove this. Often the child will have passed the health visitor screening tests early in life; the claimant may find it difficult to establish that these were improperly undertaken and led to the wrong conclusion. A child may be born with a mild to moderate hearing impairment, which gradually deteriorates. A delay in diagnosis of meningitis may also account for a claim for profound hearing impairment. In the case of pneumococcal or meningococcal meningitis this may amount to only a matter of hours, and there is often dispute as to whether the signs and symptoms of meningitis were evident at the time when the first doctor missed the diagnosis at the onset of the illness. If not, a case can be defended successfully as there is evidence that the onset of profound deafness is early in the course of the disease, especially in pneumococcal cases. In one case, tuberculous meningitis was misdiagnosed as a viral infection, despite the fact that at the same time the child’s mother had just been diagnosed with open tuberculosis. Nobody had linked the presence of the tubercular cough as being a likely source of infection and cause of the baby’s ill health.



Establishing damages Even if there is clear evidence from general practitioner (GP) records that a child’s mother attended on frequent occasions complaining about the child’s ability to hear, the mother may have grave difficulties in establishing that the child has been damaged by any delay that may have occurred. The problem becomes more difficult if the child has other developmental abnormalities, particularly if they affect language development. A child in whom the diagnosis is delayed until the age of, say, two and a half years may then be fitted with hearing aids and it is uncertain what deleterious effect the delay brought about. If the child is older and the optimum time for fitting a cochlear implant has been missed, then the prospect of the child developing the ability to understand spoken language and to communicate orally may have been lost. With a trend toward earlier implantation (Chapter 70, Paediatric cochlear implantation) it may become more difficult to argue that diagnostic delay did not bring about adverse consequences. This is a complex issue and there are those in the deaf community who would hold that not having the opportunity of a cochlear implant is no loss at all.



Deafness wrongly diagnosed Litigation may also arise due to decisions made on the false assumption that the child has a profound hearing impairment. In the young infant and toddler, a detected hearing loss assessed by brain stem evoked response audiometry, may give thresholds in the region of 80 dB. The child may be fitted with hearing aids, and educational decisions based on a hearing difficulty may lead the child to be placed in a school for children with special needs. Such a child may also have a fluctuating conductive hearing loss, due to otitis media with effusion, and the true sensorineural thresholds may only become established when the child reaches five or six. At this point, it may become clear that the hearing impairment is only in the order of 30 to 40 dB, and that placement in a mainstream school can be considered. Such parents often feel very aggrieved. They may feel that their child is being disadvantaged educationally and attempt to seek compensation. Although the diagnosis in such circumstances of the degree of hearing loss may have been incorrect, it is usually impossible for the claimant to establish that alternative management would firstly have been instituted, and secondly would have led to a materially different and improved outcome.



CHOLESTEATOMA Diagnostic or treatment delay Delay in diagnosis usually occurs due to a failure of the GP to refer a child with a chronic aural discharge. Occasionally, particularly in those cases of congenital cholesteatoma with an intact eardrum, a mother may



Chapter 103 Medical negligence in paediatric otolaryngology



have attended either the GP, and/or the hospital, complaining that their child has a unilateral hearing impairment. It is quite permissible in those children under the age of eight, particularly if the tympanogram is abnormal, to wait the usual three months for a middle ear effusion to resolve. If the condition has not resolved, then serious consideration should be given to an examination of the ear under anaesthesia, and if there is any doubt, a CT scan. Plain radiography often gives a clue in these cases, due to the loss of the septae in the air cell system, as well as the cloudiness due to the presence of the cholesteatoma. Although parents may be aggrieved by the delay in diagnosis, they are usually unable to establish that any additional damage has occurred as a result of the delay, unless there has been an intracranial complication. If children have been referred and diagnosed and placed on the waiting list for excision of a cholesteatoma, complications such as facial nerve paralysis, meningitis or other intracranial sepsis may occur whilst the child is on the waiting list. In these cases it is rarely possible to prove negligence, as it is difficult for the claimant to establish that the waiting list has been managed such as to amount to providing substandard care. However, whether increased delays by some patients cause by prioritizing ‘long waiting’ patients amounts to negligence has not been tried in the courts. Cholesteatomas are often large in children despite a relatively short history and consequently it is probably only a matter of time until there is a challenge to waiting list management arguing that a child with a cholesteatoma should be given priority over an adult who may have waited longer. This may be especially so today as waiting list management is increasingly being taken out of the hands of consultants who know the patients and is being controlled more by hospital managers. As such managers are not medically qualified, there is an argument that the Bolam principle should not apply.



Negligent management Sadly, surgical management of paediatric cholesteatoma is poor in many areas. In matters of consent, the consequences of open cavity surgery on future employment prospects are often not discussed. If a parent discovers that the open-cavity operation which his son received barred him from working offshore, when he was not told of the long-term effects of such surgery, he may seek redress through the courts. As there is a potential for a claim for life-long loss of earnings, such an error on consent could be expensive. Technical errors in performing mastoid surgery are still all too common. In one case, an aural polypectomy was performed which revealed cholesteatoma. Several months later, an atticotomy for cholesteatoma was undertaken on the same ear. Five days following surgery the child was admitted to another hospital with meningism and



] 1307



cerebritis. An exploration of the ear was undertaken by an experienced ear surgeon who found that extensive cholesteatoma still remained. This was despite the operation note for the atticotomy indicating that all cholesteatoma was removed. The case was settled for an interim payment of £15,000 as fortunately, the child had made a full recovery from the brain infection, but should epilepsy subsequently develop then further compensation would be paid. Facial nerve injury may complicate cholesteatoma surgery in children. Occasionally there is a defence. From time to time the nerve is in an abnormal position or more commonly is exposed by extensive disease. Provided the principle of identifying the nerve in an area of normal anatomy and an appropriate drill technique has been employed, a defence may be possible. Ideally, video or photographs of the operative findings make the matter much easier. A detailed operative note and a corroborating intraoperative note made by another surgeon are helpful. Nevertheless, most cases cannot be defended. The situation is often confounded by incorrect management of the palsy in the postoperative period due to a failure to explore the nerve or to provide good reasons with electroneuronographic evidence for not doing so.



CONGENITAL FIXATION OF THE STAPES Rarely, a child may present with a conductive hearing loss and no middle ear effusion or significant past history of middle ear suppuration. This may be misdiagnosed in the older child as otosclerosis and surgery recommended. In these circumstances a range of congenital abnormalities can be present and not be recognized by the inexperienced. When a stapedotomy is attempted, the result is frequently a dead ear secondary to a perilymph ‘gusher’. Problems may also arise from attempting to perforate a mobile thin footplate. This happens when the anterior crus is attached to the promontory rather than to the footplate. This abnormality, rather than a fixed footplate, accounts for the conductive loss. Correct management is to either divide the anterior crus alone or else close the ear. It is rare that poor outcomes from stapes surgery in children can be defended. Not only do surgical errors occur, but also often the consent has been inadequate, not taking into account the special problems of such surgery in children. Where a stapedectomy is planned in a child under ten, most experts in the field would expect that high definition CT scans of the cochlear and vestibular aqueducts had been obtained beforehand. Congenital abnormalities of the inner ear are a specific contraindication to such surgery. Matters become more complex in the case of the adult with congenital fixation of the stapes, as not surprisingly there are often no records from childhood. The past history of childhood hearing loss is then erroneously ascribed to otitis media with effusion.



1308 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY



ATLANTO-AXIAL INSTABILITY Children with trisomy 21 (Down syndrome) have a high risk of atlanto-axial instability. This can result in excessive movement of the odontoid process during anaesthesia, such as to cause pressure on the spinal cord and neurological dysfunction, including permanent tetraplegia (Figure 103.1). It is often the anaesthetist that is blamed, especially if the induction has been fraught and the parent was present. However, it is more likely that it is the surgeon and the theatre team that cause this event by over-vigorous movement of the head while the child is hypotonic under the anaesthetic, often with muscle relaxants. High-risk children should be assumed to have the disorder. The practice of ordering routine plain radiographs of the cervical spine prior to surgery on Down syndrome children is outdated and no longer appropriate. It is more important to ensure that all theatre personnel move the child’s head with extreme care. There is a salutary lesson to be learned with reference to this: a case was brought after a myringoplasty from which the child awoke with a partial tetraparesis. Postoperative radiology indicated that it was a rotational injury probably caused by the surgeon pushing the head away from himself to get a view of the anterior margin of the perforation whilst trying to position the graft. The mother dropped the case when she learned that it was more likely that the surgeon rather than the anaesthetist had caused the injury, as she liked the surgeon and thought he was a kind and caring doctor.



TONSILLECTOMY AND ADENOIDECTOMY Dental damage Most anaesthetists will advise that loose milk teeth are removed prior to surgery and consequently inhalation



of a milk tooth is unusual. Occasionally, secondary dentition can get chipped in the older child. Many hospitals now have speedy internal access to dental services to effect a repair, which often forestalls litigation.



Consent and bleeding One of the commonest sources of complaint revolves around postoperative haemorrhage. Many parents when they are giving consent do not grasp that a haemorrhage can be large, require a blood transfusion and even become life-threatening. Most parents will acknowledge that they were warned that bleeding could occur, but most do not have in their mind a collapsed, shocked child with a young doctor desperately trying to get venous access – or even worse medical and nursing staff showing signs of panic as they realize they are losing control of the situation. Parents who have come through this scenario with their child almost all say that they would not have consented had they known of this risk. Actual haemorrhage rates are between 5 and 10 percent with severe bleeds happening in approximately 1–2 percent (see Chapter 96, Tonsillectomy). A landmark case in October 2004 in the House of Lords (Chester v. Afshar)2 now means that it is no longer necessary for the claimant to establish that had they known of the risk they would not have had the surgery; it is just necessary to establish that they would have delayed surgery whilst they sought further advice. This effectively means that it is now essential that every parent be warned of the risk of major haemorrhage and blood transfusion. Occasionally, the haemorrhage can go unrecognized by nursing staff. This can happen when nursing staff do not follow postoperative instructions on monitoring, or worse carefully chart a rising pulse in a restless child up to the point that the child has a massive haematemesis or even loses consciousness. Errors in the resuscitation of the shocked child may be less common now with the introduction of paediatric life support and advanced paediatric life support training courses.



Poor surgical technique



Figure 103.1



Atlanto-axial subluxation.



Diathermy burns in the mouth and on the lips are still produced. Considerable care needs to be taken especially if using diathermy dissecting instruments or scissors as these can cause large burns, particularly on the tongue. It is generally impossible to mount a defence. Perforation of the palate – deemed to be due to negligent surgery – has been reported with diathermy dissection.



Chapter 103 Medical negligence in paediatric otolaryngology



ACUTE EPIGLOTITIS Since the introduction of haemophilus influenzae B (Hib) vaccination, epiglottitis is on the decrease. Only one child died from epiglottitis in England and Wales between 2000 and 2002, while there were six adult deaths. In the recent past it was failure to diagnose or suspect the condition that would lead to a patient being sent home only to return hours later with grave breathing problems or even to die at home. Not only may it be a failure to recognize clinical symptoms and signs, but also the classical radiographic appearance can go unrecognized thus leading to an indefensible claim (see Chapter 87, Acute laryngeal infections). Acute airway obstruction is more commonly due to acute laryngotracheobronchitis (ALTB or croup), especially in infants. If an artificial airway is needed this is now best managed by intubation (see Chapter 87, Acute laryngeal infections). The anaesthetist may wish to have an ENT surgeon standing by to undertake a tracheostomy should intubation be impossible and the airway become compromised. Emergency tracheotomy in this situation is nowadays extremely rare (see Chapter 86, Stridor). If the airway is not quickly secured, brain damage ensues. Whether substandard care is deemed to have occurred will depend on each individual case.



TRACHEOSTOMY Problems with the operation being undertaken as a planned procedure are unusual. Rarely, in very young children a pneumothorax can develop as can injury to the brachiocephalic vein. In the immediate postoperative period, should the tracheostomy tube become dislodged then there can be problems with maintaining the airway. This is why the tube may be sutured to the skin or stay sutures placed on either side of the tracheal incision to help find the opening should the tube being dislodged (see Chapter 93, Tracheostomy and home care). Most litigation relating to tracheotomy surgery can be attributed to complications arising from poor nursing care. Many patients who have tracheotomies have been long-stay patients on intensive care. These patients frequently have chest infections with production of thick secretions. The secretions leading to acute airway obstruction can block the tracheostomy tube. Speedy identification of the problem by close monitoring of the patient usually results in aggressive suction of the airway with saline lavage or bronchoscopy with lavage if necessary. If this does not cause a rapid relief of the problem, the tracheostomy tube can be removed altogether. Unfortunately problems arise on general wards where the nursing staff may have many patients to look after and are unfamiliar with tracheostomy care. Ward nurses on ENT and neurosurgery wards are often experienced in nursing tracheotomies patients but this



] 1309



cannot be said for the other wards. The timid nurse is often frightened to introduce a suction catheter deep into the lungs. In fact it is failing to pass the catheter right through the tube into the trachea that leads to the deep part of the tube obstructing. Also, these nurses may not have been trained to undertake bronchial lavage to loosen secretions and so bronchial plugging happens which reduces the patient’s oxygenation. When the patient’s airway obstructs, speedy removal of the ‘inner’ tube and aggressive suction is required. If there is no inner tube then rapid removal of the whole tracheostomy tube and insertion of a clean tube is required. The tracheostomy tube is then connected to an Ambu-bag for inflation of the lungs. Despite the obvious fact that the tracheostomy was undertaken for obstruction to the airway above the stoma, there have been cases where the nursing staff attempted ventilation via the mouth with a facemask and ignored the tracheostomy. It was only when the cardiac arrest team arrived that the blocked tracheostomy was dealt with, but by then severe brain damage had been sustained.



FOREIGN BODIES The ear The safest way to deal with a foreign body in the ear of a small child is under a general anaesthetic. Attempts to remove a foreign body from the ear without a general anaesthetic by syringing or the use of a strong electromagnet are often successful and carry a low risk of injury. However, instrumentation of the ear in a fractious child can lead to serious injury; such an action cannot usually be defended. Fortunately the injury is usually just limited to laceration of meatal skin, but perforation of the eardrum and even complete avulsion of the stapes has occurred.



The nose Failure to diagnose a foreign body is a common cause of complaint and even litigation. The presence of a unilateral or even bilateral foul nasal discharge, which does not settle, should alert the otolaryngologist to the possibility of a foreign body in the nose in a child. In most cases the mother has attended on several occasions at the GP, the accident and emergency department and the ear, nose and throat department where a junior doctor has failed to realize the significance of the problem. It is a common practice to wrap the small child in a blanket, then to remove the foreign body. However, if this method is employed, it is vital that all care is taken to avoid injury and the practice should not be used on those at risk of atlanto-axial subluxation. Injury occurring in



1310 ] PART 12 PAEDIATRIC OTORHINOLARYNGOLOGY these circumstances is difficult to defend when it could be argued that a general anaesthetic carries less risk.



NECK MASSES Adenopathy



The oesophagus Problems usually only arise with sharp foreign bodies (Figure 103.2). The main risk is of perforation but often this can be defended with a sharp foreign body. However, litigation is more likely to arise when there is a delay in recognizing the perforation or inappropriate management once recognized (see Chapter 92, Foreign bodies in the ear and the aerodigestive tract in children).



The airway Occasionally an inert foreign body can be present for some time before it is recognized. Complaint usually only arises when a significant symptom such as stridor is ignored and then subsequently it is found to have been due to a foreign body. An unjustifiable delay in undertaking an examination where a foreign body has been inhaled may lead to problems. This is especially so when the foreign body is of a vegetable material such as a peanut. The delay can lead to swelling of the peanut with subsequent obstruction to that lung segment and bronchiectasis.



Button batteries This modern foreign body can cause considerable damage if left in place for any length of time. If they are left in place in the nose or oesophagus they cause caustic burns with marked swelling and a risk of subsequent stenosis. Litigation may arise if a child is not treated as an emergency but is left until the next available routine list.



Surgery may be considered to establish a histological diagnosis in persistent cervical adenopathy. Trauma to the accessory nerve with resultant shoulder drop can arise as a result of dissection in the posterior triangle. Parents need to be explicitly warned of this risk prior to surgery. Ultrasound imaging may avoid the need for surgery in some neck masses.



Salivary gland tumours in children This is a complex group of disorders (see Chapter 99, Branchial arch fistulae, thyroglossal duct anomalies and lymphangioma). Claims usually arise because of delay in diagnosis or facial paralysis.



EPISTAXIS In children, the commonest site for repeated epistaxis is the nasal septum. Cautery with silver nitrate is still commonly practised. This rarely causes a problem, but excessive use on both sides of the nasal septum at the same time can lead to a septal perforation. This can also occur when hot wire cautery is used. In occasional circumstances, a severe vestibulitis following the cautery may be a significant contributory factor. Usually, such a complication cannot be defended. Trichloroacetic acid is still used in some centres to cauterize the nasal septum. This needs to be handled very carefully as if too much is applied it can run on to the upper lip and cause a permanent scar (see Chapter 81, Epistaxis in children).



NASAL TRAUMA Delayed treatment



Figure 103.2



Sharp oesophageal foreign body.



Nasal fractures with bony deformity should be reduced within at most ten days following the injury to avoid callus formation and rigid bony union. Delayed treatment may lead to permanent osseocartilaginous deformity which can only be corrected by septohinoplasty. This may lead to a successful claim for negligence. It is essential to have a referral mechanism between accident and emergency departments and otolaryngology departments so that early assessment can take place and arrangements can be made for manipulation of a nasal fracture under anaesthesia. The decision to manipulate a nasal fracture is clinical. Plain x-rays have little to offer.



Chapter 103 Medical negligence in paediatric otolaryngology



] 1311



[ Iatrogenic facial palsy in tympanomastoid surgery is [ [ [ [ Figure 103.3



Septal haematoma.



[



Septal haematoma [



Failure to recognize acute septal haematoma can lead to a permanent deformity if a septal abscess subsequently develops (Figure 103.3). In most instances of septal abscess, a doctor has not previously examined the child but occasionally a haematoma is missed or is diagnosed but not drained. The saddle nose deformity that results can be difficult to correct and as such the compensation can be relatively expensive to cover the pain and suffering of additional surgery and the cosmetic deformity.



Deficiencies in current knowledge and areas for future research The recent case of Chester v. Afshar will have a radical effect on consent in the future. It has moved practice toward the former American model of warning about everything, and virtually overturned the Bolam principle on consent. Furthermore, it applies even when the operation has been carried out in a competent manner. In the future there will be an increased use of detailed leaflets for patients and, with the exception of emergencies, a ‘cooling-off’ period during which patients can reflect on the issues and consider whether they wish to go ahead with the prescribed treatment being considered best practice for all elective surgery. There may even be two consent forms to sign, the first when the patient agrees to surgery and the second after the ‘cooling-off ’ period stating that they had considered the advice carefully.



KEY POINTS  Medical negligence cases involving children are relatively unusual given the number of children undergoing otolaryngological management.  Children are eligible to apply for ‘legal aid’ in their own right irrespective of the wealth of their parents.  Some common pitfalls which give rise to litigation may be avoided by the application of simple ground rules.  The information which must be conveyed to parents, carers and children prior to treatment to ensure that consent is properly informed is much greater than at any time in the past.



REFERENCES



 Best clinical practice [ Suspicion of hearing loss in children warrants prompt and thorough investigation. [ Early recognition and urgent treatment of meningitis reduces the risk of post-meningitic deafness. [ Cholesteatoma in children may be rapidly progressive. Always exclude the diagnosis in a discharging ear and intervene promptly if it is suspected.



almost always avoidable with good anatomical knowledge and attention to surgical technique. Should iatrogenic facial palsy occur, immediate reexploration must be considered. Ensure that parents and carers are aware that postoperative bleeding following adenotonsillectomy can be severe. It is prudent to warn that such bleeding may require blood transfusion Button batteries are potentially corrosive. When they present as foreign bodies they must be removed as soon as possible. Excision of neck nodes in the posterior triangle may be complicated by accessory nerve palsy; this should be explicitly discussed prior to surgery. Children with nasal trauma should be assessed by an otolaryngologist within days of the injury.



1. Bolam v. Friern Hospital Management Committee [1957] 1 Weekly Law Reports 582. 2. Chester v. Afshar [2004] (UKHL 41) Appeal Cases 41, House of Lords. Landmark judgement.



FURTHER READING



Cherry JR. Ear nose and throat for lawyers. London: Cavendish Publishing, 1997. Becoming a standard text for lawyers and expert witnesses. Powers M, Harris N (eds). Medical negligence. Scotland: Butterworth’s Law, 1994.



This page intentionally left blank



Scott-Brown’s Otorhinolaryngology, Head and Neck Surgery



Scott-Brown’s Otorhinolaryngology, Head and Neck Surgery 7th edition Lead editor: Michael Gleeson



Volume Part 1 Part 2 Part 3 Part 4 Part 5 Part 6 Part 7 Part 8 Part 9 Part 10 Part 11 Part 12



1 Cell biology, edited by Nicholas S Jones Wound healing, edited by Nicholas S Jones Immunology, edited by Nicholas S Jones Microbiology, edited by Nicholas S Jones Haematology, edited by Nicholas S Jones Endocrinology, edited by Nicholas S Jones Pharmacotherapeutics, edited by Martin J Burton Perioperative management, edited by Martin J Burton Safe and effective practice, edited by Martin J Burton Interpretation and management of data, edited by Martin J Burton Recent advances in technology, edited by Martin J Burton Paediatric otorhinolaryngology, edited by Ray Clarke



Volume 2 Part 13 The nose and paranasal sinuses, edited by Valerie J Lund Part 14 The neck, edited by John Hibbert Part 15 The upper digestive tract, edited by John Hibbert Part 16 The upper airway, edited by John Hibbert Part 17 Head and neck tumours, edited by John Hibbert Volume 3 Part 18 Plastic surgery of the head and neck, edited by John C Watkinson Part 19 The ear, hearing and balance, edited by George G Browning and Linda M Luxon Part 20 Skull base, edited by Michael Gleeson Index CD-ROM



George G Browning MD FRCS Professor of Otorhinolaryngology, MRC Institute of Hearing Research, Glasgow Royal Infirmary, Glasgow, UK Martin J Burton MA DM FRCS Senior Clinical Lecturer, University of Oxford; and Consultant Otolaryngologist, Oxford Radcliffe NHS Trust Oxford, UK Ray Clarke BSc DCH FRCS FRCS (ORL) Consultant Paediatric Otolaryngologist, Royal Liverpool University Children’s Hospital, Alder Hey, Liverpool, UK Michael Gleeson MD FRCS Professor of Otolaryngology and Skull Base Surgery, Institute of Neurology, University College London; and Consultant, Guy’s, Kings and St Thomas’ and the National Hospital for Neurology and Neurosurgery, London UK; and Honorary Consultant Skull Base Surgeon, Great Ormond Street Hospital for Sick Children, London, UK John Hibbert ChM FRCS Formerly Consultant Otolaryngologist, Department of Otolaryngology, Guy’s Hospital, London, UK Nicholas S Jones MD FRCS FRCS (ORL) Professor of Otorhinolaryngology, Queen’s Medical Centre, University of Nottingham, Nottingham UK Valerie J Lund MS FRCS FRCS (Ed) Professor of Rhinology, The Ear Institute, University College London, London, UK Linda M Luxon BSc MBBS FRCP Professor of Audiovestibular Medicine, University of London at University College London, Academic Unit of Audiovestibular Medicine; and Consultant Physician, National Hospital for Neurology and Neurosurgery; and Honorary Consultant Physician, Great Ormond Street Hospital for Children, London, UK John C Watkinson MSc MS FRCS (Ed, Glas, Land) DL0 Consultant Head and Neck and Thyroid Surgeon, Department of Otorhinolaryngology/Head and Neck Surgery, Queen Elizabeth Hospital, University of Birmingham NHS Trust, Birmingham, UK



Scott-Brown’s Otorhinolaryngology, Head and Neck Surgery 7th edition



Volume 2



Edited by



Michael Gleeson George G Browning, Martin J Burton, Ray Clarke, John Hibbert, Nicholas S Jones, Valerie J Lund, Linda M Luxon, John C Watkinson



Hodder Arnold www.hoddereducation.com



First published in Great Britain in 1952 by Butterworth & Co. Second edition 1965 Third edition 1971 Fourth edition 1979 Fifth edition 1987 Sixth edition 1997 This seventh edition published in Great Britain in 2008 by Hodder Arnold An imprint of Hodder Education, a part of Hachette Livre UK, 338 Euston Road, London NW1 3BH http://www.hoddereducation.com & 2008 Edward Arnold (Publishers) Ltd All rights reserved. Apart from any use permitted under UK copyright law, this publication may only be reproduced, stored or transmitted, in any form, or by any means with prior permission in writing of the publishers or in the case of reprographic production in accordance with the terms of licences issued by the Copyright Licensing Agency. In the United Kingdom such licences are issued by the Copyright Licensing Agency: Saffron House, 6-10 Kirby Street, London EC1N 8TS Whilst the advice and information in this book are believed to be true and accurate at the date of going to press, neither the author[s] nor the publisher can accept any legal responsibility or liability for any errors or omissions that may be made. In particular (but without limiting the generality of the preceding disclaimer) every effort has been made to check drug dosages; however it is still possible that errors have been missed. Furthermore, dosage schedules are constantly being revised and new side-effects recognized. For these reasons the reader is strongly urged to consult the drug companies’ printed instructions before administering any of the drugs recommended in this book. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN



978 0 340 808 931



1 2 3 4 5 6 7 8 9 10 Commissioning Editor: Joanna Koster Project Editor: Zelah Pengilley Production Controller: Lindsay Smith / Andre Sim Text and Cover Designer: Amina Dudhia Cover photograph & MEHAU KULYK/SCIENCE PHOTO LIBRARY Typeset in 10 pt Minion by Macmillan India Printed and bound in India. What do you think about this book? Or any other Hodder Arnold title? Please send your comments to www.hoddereducation.com



Contents



How to use this book



xi



PART 13 THE NOSE AND PARANASAL SINUSES – EDITED BY VALERIE J LUND



1313



104



Anatomy of the nose and paranasal sinuses H Stammberger and Valerie J Lund



1315



105



Nasal endoscopy Rodney J Schlosser and David W Kennedy



1344



106



Physiology of the nose and paranasal sinuses Adrian Drake-Lee



1355



107



Measurement of the nasal airway Ronald Eccles



1372



108



Classification and differential diagnosis of rhinosinusitis Ian S Mackay and Valerie J Lund



1380



109



Allergic rhinitis Glenis Scadding and Stephen Durham



1386



110



Nonallergic perennial rhinitis Claus Bachert



1408



111



Occupational rhinitis Hesham Saleh



1415



112



Food allergy and intolerance Carsten Bindslev-Jensen and Morten Osterballe



1424



113



Rhinosinusitis Michael S Benninger



1439



114



Fungal rhinosinusitis Jean Michel Klossek



1449



115



Specific chronic infections S Bahadur and A Thakar



1458



116



Medical management of chronic rhinosinusitis Glenis Scadding



1469



117



Surgical management of rhinosinusitis Valerie J Lund and Julian Rowe-Jones



1478



118



The frontal sinus Wolfgang Draf



1500



119



Mucocoeles Valerie J Lund



1531



120



Complications of rhinosinusitis Robert Slack and Richard Sim



1539



vi ] Contents 121



Nasal polyposis Niels Mygind and Valerie J Lund



1549



122



The relationship between the upper and lower respiratory tract Jean Bousquet and Antonio M Vignolay



1560



123



The septum Adriaan F van Olphen



1569



124



Nasal septal perforations Charles East and Santdeep Paun



1582



125



The management of enlarged turbinates Luisa F Grymer



1589



126



Epistaxis Gerald W McGarry



1596



127



Nasal fractures Brent A McMonagle and Michael Gleeson



1609



128



Fractures of the facial skeleton Simon Holmes and Michael Gleeson



1618



129



Cerebrospinal fluid rhinorrhoea Andrew H Marshall and Nicholas S Jones



1636



130



Granulomatous conditions of the nose David J Howard and Valerie J Lund



1645



131



Abnormalities of smell Richard L Doty and Steven M Bromley



1660



132



Orbital and optic nerve decompression Valerie J Lund and Geoffrey E Rose



1677



133



Dacryocystorhinostomy Neil Fergie and Nicholas S Jones



1689



134



Conditions of the external nose Michael O’Connell



1699



135



The diagnosis and management of facial pain Tim J Woolford and Nicholas S Jones



1718



136



Medical negligence in rhinology Maurice Hawthorne



1730



PART 14 THE NECK – EDITED BY JOHN HIBBERT



1737



137



Surgical anatomy of the neck Chris R Jennings



1739



138



Examination and imaging of the neck Sheila C Rankin and John Hibbert



1754



139



Neck trauma Johannes J Fagan and Andrew J Nicol



1766



140



Benign neck disease: infections and swellings Peter Clarke



1777



PART 15 THE UPPER DIGESTIVE TRACT – EDITED BY JOHN HIBBERT



1789



141



Anatomy of the mouth and dentition Barry KB Berkovitz



1791



142



Benign oral and dental disease Crispian Scully and Jose-V Sebastian Bagan



1816



Contents



] vii



143



Abnormalities of taste Steven M Bromley and Richard L Doty



1840



144



Salivary gland anatomy Graham J Cox



1852



145



Physiology of the salivary glands Roderick Cawsony and Michael Gleeson



1858



146



Imaging of the salivary glands Nicole JM Freling



1871



147



Non-neoplastic salivary gland diseases Stephen R Porter



1898



148



Cysts and tumours in and around the jaws, including sarcoma Mark MCGurk, Anna Cassoni and Lisa Pitkin



1921



149



Anatomy of the pharynx and oesophagus Nigel Beasley



1942



150



Physiology of swallowing Paula Leslie and Stephen McHanwell



1954



151



Functional investigations of the upper gastrointestinal tract Lisa J Hirst



1964



152



Acute and chronic pharyngeal infection Marcelle Macnamara



1981



153



Causes of dysphagia Elfy B Chevretton



2025



154



Globus pharyngeus Petros D Karkos and Janet A Wilson



2037



155



Pharyngeal pouch Grant Bates



2043



156



Oesophagal diseases Robert C Mason



2062



157



Neurological disease of the pharynx Robert J Sandersony, Paul Tierney and Richard Adamson



2074



158



Dysphagia: management and intervention Alison Perry



2084



159



Management and treatment of intractable aspiration Elfy B Chevretton



2094



PART 16 THE UPPER AIRWAY – EDITED BY JOHN HIBBERT



2105



160



Anatomy of the nasopharynx Alexander C Vlantis and C Andrew van Hasselt



2107



161



Benign conditions of the nasopharynx Victor J Abdullah and C Andrew van Hasselt



2116



162



Anatomy of the larynx and tracheobronchial tree Nigel Beasley



2130



163



Assessment and examination of the upper respiratory tract Jean-Pierre Jeannon and Marcelle Macnamara



2145



164



Physiology of the larynx Lesley Mathieson and Paul Carding



2155



165



Voice and speech production Paul Carding and Lesley Mathieson



2164



166



Objective evaluation of the voice Julian McGlashan and Adrian Fourcin



2170



viii ] Contents 167



Disorders of the voice Julian McGlashan



2192



168



The professional voice Meredydd Harries



2211



169



Speech therapy in ENT practice: scope, science and evidence for intervention Alison Perry



2216



170



Phonosurgery Meredydd Harries



2234



171



Acute infections of the larynx Andrew C Swift



2248



172



Chronic laryngitis Kenneth MacKenzie



2258



173



Laryngeal trauma and stenosis Lisa Pitkin



2271



174



Upper airway obstruction Paul Pracy



2286



175



Tracheostomy Paul Pracy



2292



176



Physiology of sleep and sleep disorders John Fleetham



2305



177



Obstructive sleep apnoea: medical management Dev Banerjee



2313



178



The surgical management of snoring Marcelle Macnamara



2325



PART 17 HEAD AND NECK TUMOURS – EDITED BY JOHN HIBBERT



2341



179



Epidemiology of head and neck cancer Randall P Morton and Mark E Izzard



2343



180



Aetiology of head and neck cancer Nicholas D Stafford



2351



181



Staging of head and neck cancer Nicholas J Roland



2359



182



Data collection in head and neck cancer Richard Wight



2372



183



Prognostic indicators and serum markers Andrew S Jones



2378



184



Skin cancer of the head and neck Garrick A Georgeu and Michael Gleeson



2395



185



Mucosal malignant melanoma Andrew S Jones



2406



186



Nasal cavity and paranasal sinus malignancy Brent A MCMonagle and Michael Gleeson



2417



187



Juvenile angiofibroma Michael Gleeson



2437



188



Nasopharyngeal carcinoma John Kong Sang Woo and C Andrew van Hasselt



2445



189



Benign salivary gland tumours Michael Gleeson and Roderick Cawsony



2475



190



Malignant tumours of the salivary glands Andrew S Jones



2493



] ix



Contents 191



Tumours of the parapharyngeal space Andrew S Jones



2522



192



Oral cavity tumours including the lip Keith Webster



2543



193



Oropharyngeal tumours Patrick J Bradley



2577



194



Tumours of the larynx Martin A Birchall and Laysan Pope



2598



195



Rehabilitation after laryngectomy Andrew J Parker



2623



196



Tumours of the hypopharynx and oesophagus Andrew S Jones



2633



197



Thyroid cancer James Ramsden and John C Watkinson



2663



198



Management of the patient presenting with neck lymphadenopathy and an unknown primary carcinoma Jawaher Ansari and John Glaholm



2702



199



Metastatic neck disease John C Watkinson



2711



200



Developments in radiotherapy for head and neck cancer John Glaholm



2753



201



Quality of life in head and neck cancer Randall P Morton and Hisham Mehanna



2766



202



Palliative care for head and neck cancer Charles Diamond and Claud Regnard



2781



203



Medical negligence in head and neck surgery Arnold GD Maran



2803



Please note: The table of contents for all three volumes can be found on the Scott-Brown website at: www.scottbrownENT.com. The list of contributors, preface and list of abbreviations are included in the prelims for Volume 1. The index for all three volumes is included in Volume 3.



This page intentionally left blank



How to use this book



This new edition of Scott-Brown’s Otorhinolaryngology, Head and Neck Surgery incorporates some special features to aid the readers’ understanding and navigation of the text. These are described below.



SEARCH STRATEGY The majority of the chapters feature a search strategy indicating the key words used by the author when conducting their literature review in order to prepare the chapter, so that the reader can repeat and develop the search.



EVIDENCE SCORING For the major sections in each chapter, the authors have used a hierarchical system to indicate the level of evidence supporting their statements. This is shown in the text in the form [***], with the number of stars indicating the level of evidence. The key to this system is shown in the table below. Level



Category of evidence



****



Systematic reviews, meta-analyses of randomized controlled trials and randomized controlled trials



***



Non-randomised studies



**



Observational or non-experimental studies



*



Expert opinion



Where no level is shown, the quality of supporting evidence, if any exists, is of low grade only (for example, case reports, clinical experience etc.). For more information on evidence scoring, please refer to Chapter 304, Evidence-based medicine; and 305 Critical appraisal skills.



CLINICAL RECOMMENDATIONS The authors have indicated the basis on which they have made clinical recommendations by grading them according to the level of the supporting evidence. This is shown in the text in the form [Grade A], with the grade indicating the level of evidence supporting the recommendation. The key to this system is shown in the table below.



xii ] How to use this book Grade



Nature of supporting evidence



A



Recommendation based on evidence from meta-analyses of randomized controlled trails



B



Recommendation based on evidence from high quality case-controlled or cohort studies



C



Recommendation based on evidence from low quality case-controlled or cohort studies



D



Recommendation based on evidence from clinical series or expert opinion



Recommendations are graded where the author is satisfied that the literature supports such a grading; otherwise a grading may not be given.



REFERENCE ANNOTATION The reference lists are annotated with an asterisk, where appropriate, to guide readers to key primary papers and major review articles. We hope that this feature will render the lists of references more useful to the reader and will encourage self-directed learning among both trainees and practicing physicians.



PART



13



THE NOSE AND PARANASAL SINUSES EDITED BY VALERIE J LUND



104 Anatomy of the nose and paranasal sinuses H Stammberger and Valerie J Lund



1315



105 Nasal endoscopy Rodney J Schlosser and David W Kennedy



1344



106 Physiology of the nose and paranasal sinuses Adrian Drake-Lee



1355



107 Measurement of the nasal airway Ronald Eccles



1372



108 Classification and differential diagnosis of rhinosinusitis Ian S Mackay and Valerie J Lund



1380



109 Allergic rhinitis Glenis Scadding and Stephen Durham



1386



110 Nonallergic perennial rhinitis Claus Bachert



1408



111 Occupational rhinitis Hesham Saleh



1415



112 Food allergy and intolerance Carsten Bindslev-Jensen and Morten Osterballe



1424



113 Rhinosinusitis Michael S Benninger



1439



114 Fungal rhinosinusitis Jean Michel Klossek



1449



115 Specific chronic infections S Bahadur and A Thakar



1458



116 Medical management of chronic rhinosinusitis Glenis Scadding



1469



117 Surgical management of rhinosinusitis Valerie J Lund and Julian Rowe-Jones



1478



118 The frontal sinus Wolfgang Draf



1500



119 Mucocoeles Valerie J Lund



1531



120 Complications of rhinosinusitis Robert Slack and Richard Sim



1539



121 Nasal polyposis Niels Mygind and Valerie J Lund



1549



122 The relationship between the upper and lower respiratory tract Jean Bousquet and Antonio M Vignolay



1560



123 The septum Adriaan F van Olphen



1569



124 Nasal septal perforations Charles East and Santdeep Paun



1582



125 The management of enlarged turbinates Luisa F Grymer



1589



126 Epistaxis Gerald W McGarry



1596



127 Nasal fractures Brent A McMonagle and Michael Gleeson



1609



128 Fractures of the facial skeleton Simon Holmes and Michael Gleeson



1618



129 Cerebrospinal fluid rhinorrhoea Andrew H Marshall and Nicholas S Jones



1636



130 Granulomatous conditions of the nose David J Howard and Valerie J Lund



1645



131 Abnormalities of smell Richard L Doty and Steven M Bromley



1660



132 Orbital and optic nerve decompression Valerie J Lund and Geoffrey E Rose



1677



133 Dacryocystorhinostomy Neil Fergie and Nicholas S Jones



1689



134 Conditions of the external nose Michael O’Connell



1699



135 The diagnosis and management of facial pain Tim J Woolford and Nicholas S Jones



1718



136 Medical negligence in rhinology Maurice Hawthorne



1730



104 Anatomy of the nose and paranasal sinuses H STAMMBERGER AND VALERIE J LUND



Development of nose and paranasal sinuses Comparative anatomy Anatomy of the nose and paranasal sinuses Key points



1315 1321 1322 1341



Deficiencies in current knowledge and areas for future research References



1342 1342



SEARCH STRATEGY The data in this chapter are supported by a literature search using the key words endoscopic anatomy, paranasal sinuses, nasal cavity and external nose.



DEVELOPMENT OF NOSE AND PARANASAL SINUSES The external nose and nasal cavity The nose develops from a number of mesenchymal processes around the primitive mouth (Figure 104.1).1, 2 The nasal cavity is first recognizable in the 5.6 mm (crown–rump distance) embryo in the fourth intrauterine week as the olfactory or nasal placode, a thickening of the ectoderm above the stomatodaeum.3 This placode sinks to form the olfactory pit lying between the proliferating mesoderm of the medial and lateral nasal folds of the frontonasal process. This deepens to form the nasal sac by the fifth week (Figure 104.2). In the 12.5 mm embryo, the maxillary process of the first branchial arch grows anteriorly and medially to fuse anteriorly with the medial nasal folds and the frontonasal process which closes the nasal pits off to form widely separated primitive, nasal cavities (Figure 104.3). The primitive nasal cavity and mouth are separated initially by a bucconasal membrane. This gradually thins as the nasal sacs extend posteriorly and eventually breaks down at the 14–15 mm stage to form the primitive choanae. These are



more anteriorly placed than the definitive choana due to continuous posterior growth of the palate.4 The floor anterior to the choana forms from mesenchymal extensions of the medial nasal folds to produce the premaxilla and ultimately the upper lip and medial crus of the lower lateral cartilages. The maxillary process also grows ventrally from the dorsal end of the mandibular process (first visceral arch) to join the lateral nasal fold around the nasomaxillary groove. Ectoderm in this region eventually canalizes to form the nasolacrimal duct. The lateral nasal folds also form the nasal bones, upper lateral cartilages and lateral crus of the lower lateral cartilages.



The palate and nasal septum The primitive palate begins to form anteriorly with fusion of the maxillary and frontonasal processes by the 13.5 mm embryo stage. A midline ridge develops from the posterior edge of the frontonasal process in the roof of the oral cavity and extends posteriorly to the opening of Rathke’s pouch (Figure 104.4). This becomes the nasal septum which is continuous anteriorly with the partition between



1316 ] PART 13 THE NOSE AND PARANASAL SINUSES Forehead



Nasal sac



Olfactory epithelium



Lateral nasal fold



Nasal fin



Medial nasal fold



Primitive palate



Frontonasal process



Eye Primitive posterior naris



Maxillary process



Rathke's pouch



Oronasal membrane Medial nasal fold Olfactory nerve



Figure 104.2 The roof of the stomatodaeum of a 12 mm human embryo illustrating the development of the primitive palate and posterior nares by approximation of the maxillary processes to the lateral and medial nasal folds. The previous site of attachment of the buccopharyngeal membrane is represented by a dotted line and part of the left maxillary process has been removed. After Hamilton and Mossman1.



Lateral nasal fold Medial nasal fold Nasolacrimal furrow



Anterior naris Posterior naris



Rudiment of nasal septum



Maxillary process Eye



Primitive palate Palatal process of maxillary mesoderm



Figure 104.1 Development of the nasal sac, nasal cavity and primitive palate. After Hamilton and Mossman1. Rathke's pouch



the primitive nasal cavities. As the nasal cavities enlarge, the palatal processes, derived from the lateral maxillary mesoderm, grow medially towards each other and the septum. Initially, they lie lateral to the tongue, but as this moves ventrally with further growth, the palatal processes swing medially and fuse horizontally (Figures 104.5 and 104.6). The fusion begins along the posterior margin of the primitive palate and is complete except for a midline dehiscence at the future site of the incisive canal. Fusion continues between the palatal processes and the septum from anterior to posterior, separating the nasal and oral cavities and most posteriorly the nasopharynx and oral cavity as the palatal processes complete the soft palate and uvula. On either side of the anterior septum, in relation to the paraseptal Jacobson’s cartilage, an invagination of ectoderm forms the vomeronasal organ, which largely disappears in man, leaving only a blind tubular pouch, 2–6 mm long.5 Longitudinal strips of cartilage 7–15 mm in length may be identified in the embryo, lying adjacent to the vomeronasal organ on either side of the septal cartilage. These may occasionally remain as



Figure 104.3 The roof of the stomatodaeum of a 12.5 mm embryo. After Hamilton and Mossman1.



Primitive Frontonasal palate Anterior process Medial naris nasal fold Lateral Rudiment nasal fold of nasal septum Forehead



Rathke’s pouch



Nasolacrimal furrow Maxillary process Eye Primitive posterior naris Palatal process of maxillary mesoderm



Figure 104.4 The roof of the stomatodaeum of a 13.5 mm embryo. After Hamilton and Mossman1.



discrete entities in the adult but more usually involute, leaving only a small cartilaginous bulge. The primitive septum is initially made entirely of cartilage. The superior



Chapter 104 Anatomy of the nose and paranasal sinuses



] 1317



Septal cartilage Developing inferior turbinate Vomeronasal cartilage Tongue Palatal process of maxilla Tooth bud Meckel's cartilage



Figure 104.5 Section through the developing palate of a 20 mm human foetus. After Hamilton and Mossman1.



Septal cartilage Developing turbinate Vomeronasal organ Fused palatal and septal processes



Tongue



Meckel's cartilage



part ossifies to form the perpendicular plate of the ethmoid (from crista galli downwards6, 7) and the vomer in the posteroinferior portion, leaving an anteroinferior quadrilateral cartilaginous plate. Two ossification centres appear for the vomer at the eighth foetal week on either side of the cartilage, uniting to form a deep bony groove in which the cartilage sits. As growth continues, part of the cartilage absorbs as the two bony lamellae fuse. By puberty, the lamellae are almost completely united with everted alae and an anterior groove as indications of the vomer’s bilaminar origin. The nasal bones arise during the tenth and eleventh weeks.



Figure 104.6 Section through the developing palate of a 48 mm human foetus. After Hamilton and Mossman1.



The paranasal bones and sinuses THE MAXILLA



The maxilla arises during the sixth and seventh weeks from five ossification centres. In the fourth foetal month these fuse to form the alveolar, palatine, zygomatic and frontal processes and the floor of the orbit. A further centre appears in the medial floor of the pyriform aperture, forming the premaxilla in which the upper incisor teeth develop. The premaxilla forms the anterior nasal spine and fuses with the vomeronasal cartilages laterally and septal cartilage superiorly.



1318 ] PART 13 THE NOSE AND PARANASAL SINUSES THE ETHMOID BONE



The ethmoid bone ossifies in the cartilaginous nasal capsule from three centres; one for each labyrinth and one for the perpendicular plate. The centres for the labyrinth are present from the fourth or fifth intrauterine month, so they are partially ossified at birth. The perpendicular plate and crista galli develop from one centre during the first year after birth and fuse with the labyrinths at the beginning of the second year. Both this centre and those for the labyrinth contribute to the cribiform plate.



THE FRONTAL BONE



The frontal bone ossifies in membrane from two centres, one in each superciliary ridge appearing in the eighth intrauterine week. At birth, the bone is composed of two halves separated by a frontal or metopic suture, which begin to fuse from the second year. This is usually complete by the eighth year though may persist in some races, such as the Japanese.8



THE SPHENOID BONE



The sphenoid is divided into two parts, a presphenoidal portion anterior to the tuberculum sellae continuous with the lesser wings made up of six separate ossification centres, and the post-sphenoidal part composed of the sella turcica and dorsum sellae associated with the greater wings and pterygoid processes derived from eight centres. The pre- and post-sphenoidal parts of the body fuse around the eighth intrauterine month. At birth, the bone consists of three pieces, a central portion consisting of the



body and lesser wings and two lateral parts, each consisting of the greater wing and pterygoid process which begin to fuse at one year after birth.



THE TURBINATE BONES



A series of elevations appear on the lateral wall of the nose from the sixth foetal week which will ultimately form the turbinates. The most inferior or maxilloturbinal forms the inferior turbinate. The middle, superior and supreme turbinates result from reduction of the complex ethmoturbinal system found in lower mammals. Similarly, the primitive nasoturbinal is represented by the agger nasi region and uncinate process of the ethmoid.



THE MAXILLARY SINUS



The maxillary sinus is the first sinus to appear (seven to ten weeks) as a shallow groove expanding from the primitive ethmoidal infundibulum into the mass of the maxilla (Figure 104.7). Absorption and expansion results in a small cavity at birth which measures 7  4  4 mm.9 It continues to grow during childhood at an estimated annual rate of 2 mm vertically and 3 mm anteroposteriorly10 and in particular with development of the middle third of the face as the dentition erupts. This process slows down around the seventh year of life, followed by a second growth phase thereafter. At the twelfth year of life, pneumatization may reach laterally just under the lateral orbital wall at the insertion of the zygomatic process, inferiorly to the level of the nasal floor and after the second dentition below the nasal floor. After dentition,



Figure 104.7 (a) A coronal section through the lateral nasal wall, orbit and maxilla of a newborn. Clearly the uncinate process can be identified and lateral to it, the ethmoidal infundibulum. The arrows indicate a small blister-like maxillary sinus, which has already developed from the infundibulum. 1, uncinate process; 2, maxillary sinus; 3, dental germ; cm, middle turbinate; ci, inferior turbinate; s, septum. (b) The development of the maxillary sinus from the ethmoidal infundibulum in the foetus and infant. This evolution explains why the ostium of the fully developed maxillary sinus is normally found at the floor of the ethmoidal infundibulum, and why drainage and ventilation of the maxillary sinus pass through the ethmoidal infundibulum. i, ethmoidal infundibulum; s, septum. (Dissection W. Anderhuber, Graz; drawing modified from Peter.)



Chapter 104 Anatomy of the nose and paranasal sinuses



the sinus only gradually enlarges, reaching its final size in the seventeenth to eighteenth year of life. A ‘standard’, maxillary sinus in an adult has a volume of around 15 cm2 and is roughly pyramid shaped. The base of this pyramid is formed by the medial wall of the maxillary sinus with the apex of the pyramid towards the zygomatic recess. The sinus may become relatively enlarged in old age as a result of resorption of the alveolus secondary to loss of teeth.



THE ETHMOID SINUS



During the ninth and tenth week of gestation, six to seven folds appear in the lateral wall of the nasal capsule of the foetus. These folds are separated from each other by corresponding grooves. Over the next weeks, the folds fuse into three to four remaining crests with an anterior ‘ascending’ and a posterior ‘descending’ portion (ramus ascendens and ramus descendens) (Figure 104.8). All permanent ethmoidal structures develop from these crests and the furrows in between. From the main folds, two to a maximum of three ethmoidal turbinates develop: the persisting middle, superior and – if developed – supreme nasal turbinate. The persisting inferior turbinate, the socalled maxillo-turbinal, develops from the most inferior of the folds and is considered a separate bone not connected to the ethmoid complex. Not all primary folds result in permanent nasal turbinates. The crest of the first fold usually regresses during development, its ascending portion resulting in the agger nasi (sometimes called the naso-turbinal), the



S1 NT



S3



S2



ET1



ET2



] 1319



descending portion develops into the uncinate process. The second crest develops into the bulla lamella which when pneumatized, usually results in the largest and most constant cell of the anterior ethmoid complex. Uncinate process, ethmoidal bulla, middle, superior and supreme turbinate are attached to the lateral nasal wall by their basal or ground lamella. This reflects their origin from the folds or crests which arose in the lateral wall of the nasal capsule. The attachment of the middle turbinate is considered the most important of these ground lamellae. Cells and clefts opening and draining antero–inferiorly to this lamella by definition are part of the anterior ethmoid, cells and clefts opening and draining posterior to it are considered cells and spaces of the posterior ethmoid complex.11 The uncinate process and ethmoidal bulla can already be identified during the first third of gestation. Between these structures develops an initially shallow, then deeper cleft orientated anteriorly and inferiorly, the ethmoidal infundibulum. At birth, these structures can clearly be identified and the initially very small maxillary sinus can be seen developing into the maxilla out of the ethmoidal infundibulum (Figures 104.7 and 104.9).



THE SPHENOID SINUS



The sphenoid sinus is recognizable at around the third intrauterine month as an evagination from the sphenoethmoidal recess and again a small cavity is found at birth



S4 S 5 ET3 ET S 4 6 ET5



MT To



Figure 104.8 A schematic drawing of the primary folds and the lateral wall of the nasal capsule of the human foetus in their original positional relationship. ET1 to ET5, primary ethmoturbinals, which fuse into two to three persisting ethmoidal turbinates. MT, maxilloturbinal; NT, nasoturbinal; S1 to S6, major furrows (grooves) between primary ethmoturbinal crests; To, pharyngeal opening of Eustachian tube. Note: ascending and descending portion of S and ET. Modified from Killian and Peter.



Figure 104.9 A view of the lateral nasal wall in a newborn. 1, inferior turbinate; 2, middle turbinate. This still shows the sharp crease at the transition from its descending part to the ascending part, which is almost at right angles. 3, longitudinal torus above a sagittal groove in the medial surface of the middle turbinate, which may occasionally persist in an adult. 4, superior turbinate; 5, supreme turbinate; 6, remnants of the last ethmoturbinal in sphenoethmoidal recess; 7, Bertini ossification centres of the nonpneumatized sphenoid; 8, spheno-occipital synchondrosis; 9, dental germ (upper incisor); an, agger nasi. By permission of W. Anderhuber, Graz.



1320 ] PART 13 THE NOSE AND PARANASAL SINUSES (2  2  1.5 mm). At the third year of life, pneumatization of the sphenoid bone progresses and at age seven has frequently reached the floor of the sella. In adults, degrees of pneumatization vary greatly, asymmetry being the rule rather than the exception. In extreme cases, internal carotid artery, optic nerve, V2 and the vidian nerve can be exposed in the walls of the sphenoid sinus with almost no protective bony walls (Figure 104.10). Pneumatization is said to progress at a rate of 0.25 mm each year from the age of four years but this is not constant.12



continue throughout adolescence. The configuration of the sinuses and the frontal recess is subject to considerable variation. Diverticula and septation are not uncommon. The frontal recess, although frequently a narrow hourglass segment, may have a longer, more tortuous path encroached upon by adjacent anterior ethmoidal cells. These variations are a direct consequence of the sinus’ embryological development as it is the last to complete its growth, in early adulthood. ANOMALIES



THE FRONTAL SINUS



The frontal sinus is the most variable in size and shape and may be regarded embryologically as an anterior ethmoidal cell. In the fourth foetal month, development of the frontal sinus out of the so-called frontal recess can be seen. From this most anterior and superior segment of the anterior ethmoid complex, the frontal bone is gradually pneumatized, resulting in frontal sinuses of variable size. At birth, the frontal sinuses are small and, on x-rays, cannot usually be differentiated from other anterior ethmoidal cells. In contrast to the maxillary sinus, the process of pneumatization is initially very slow. However, pneumatization is evident in computed tomography at the end of the first year of life. In the fifth year of life, pneumatization extends superiorly and at twelve years the sinus is largely developed. Pneumatization may



Figure 104.10 CT scan (coronal) through sphenoid sinus of an adult. The anterior clinoid processes are pneumatized, resulting in an infraoptic recess on both sides. Note the bulging of the internal carotid artery during its course through the cavernous sinus. Pneumatization extends far laterally exposing the round foramen with V2. At the floor of the sinus, the vidian canal can clearly be identified.



Fusion of the processes takes place from anterior to posterior. Partial or complete failure of fusion results in abnormalities which range from a bifid uvula to clefts of varying severity. Failure of fusion between a maxillary process and the corresponding premaxilla causes a cleft lip. Failure of the maxillary and lateral processes to fuse produces a facial cleft with an open nasolacrimal furrow. Nonfusion between the palatine processes and nasal septum results in a cleft palate. In its most severe form there is a completely deficient palate and bilateral clefts in the upper lip on either side of the philtrum. Clefts of the lip and palate are seen in just under 0.1 percent of neonates. Interestingly, more clefts develop on the left than the right.13 Occasionally, median clefts are encountered which can produce anomalies ranging from a widened columella and nasal tip to complete dehiscence of the nasal bridge and a bifid nose. Failure of the oronasal membrane to rupture results in choanal atresia, which may be either membranous or bony and may lie at a variable distance from the nasopharynx because of changes in the position of the posterior choana during development. Rarely, the olfactory placode fails to develop, with complete or partial absence of the nose. This is usually associated with bilateral choanal atresia and hypoplastic maxillae. Unilateral choanal atresia is much more common. Unilateral maldevelopment of the olfactory placode produces a proboscis lateralis where the mesenchyme grows out in a complete ring resembling a trunk and the maxilla and eye are often affected. Median facial anomalies result from abnormal first and second arch development and are generally fatal. Premature ossification or the sphenoid or fusion of the pre- and post-sphenoidal portions may produce an abnormal depression of the nasal bridge (often seen in achondroplasia) and anomalous fusion of the presphenoid results in hyperteleorism. Cysts arise from epithelial entrapment in the lines of fusion between the various processes. Nasolabial cysts develop in the furrow between the maxillary and median nasal elevation and globulomaxillary cysts at the junction of the primitive palate and palatine processes, in the alveolar process between lateral incisor and canine teeth. The dermoid cyst is a relatively common congenital abnormality. It is a median nasal lesion found on the glabella, dorsum or tip of the nose between the alar



Chapter 104 Anatomy of the nose and paranasal sinuses



cartilages or on the columella. It may be superficial or communicate with a deeper component through a tract between the nasal bones which may extend intracranially through the cribiform plate. Two theories have been proposed for its formation; the cranial theory postulates that as dura mater recedes, it pulls nasal ectoderm forming a sinus. Pinching off the sinus leads to cyst formation. Alternatively, ectoderm may become trapped between the two medial nasal folds.



COMPARATIVE ANATOMY There is a basic pattern to the mammalian nasal cavity which can be recognized in man and offers important teleological information (Figures 104.11, 104.12, 104.13 and 104.14).14, 15 In all mammals the nasal fossa is divided into two by a midline septum and its functions are respiratory (moistening, cleaning and warming inspired air) and olfactory. This is evident in the macroscopic and microscopic configuration of the nose. The lateral nasal wall bears three sets of turbinals; ethmoturbinals, nasoturbinals and maxilloturbinals. The ethmoturbinales are arranged in two rows, the ectoturbinals laterally and endoturbinals medially. They increase surface area for greater olfactory acuity. Ethmoturbinals are particularly well developed in macrosmatic animals and can be very numerous. The olfactory area is augmented by the ethmoturbinals by a factor of 5 in sheep, 2.5 the dog and 1.34 in man.16 The anthropoid and human nose are remarkably similar demonstrating the increasing importance of vision over olfaction.17, 18, 19, 20 There is a reduction in anteroposterior length, resulting in a high-roofed cavity between two large orbits which face forwards. The ectoturbinals disappear during foetal development and the endoturbinals are reduced to three; middle, superior



Figure 104.11 Coronal section through midfacial block from Mandrillus sphinx (mandrill) showing absence of maxillary sinus.



] 1321



and supreme. The nasoturbinal is represented by the uncinate process and agger nasi of the ethmoid and the maxilloturbinal is equivalent to the inferior turbinate. A secondary olfactory organ, the vomeronasal or Jacobson’s organ, is found lying either side of the anterior nasal septum, usually protected by small cartilages. This organ has an important influence on feeding and oestrus in many mammals, though there is considerable variation in its morphology.21 The paranasal sinuses are a characteristic feature of terrestrial, placental mammals and are absent in monotremes, marsupials and cetacean. The distribution and configuration of the sinuses is varied and, despite much consideration and some interesting suggestions,22 a convincing explanation of their origin and function has not been forthcoming. The maxillary sinus ranges from an extensive cavity, meeting in the midline of the hard palate, e.g. chimpanzee23 and pneumatizing the nasal, frontal, lacrimal, palatine and nasoturbinal bones, e.g. bears, horses and elephants, to a shallow lateral recess, e.g. the baboon. The size of the maxillary sinus increases with the size of the skull and orbit in monkeys and higher primates, though not with body size.24 However, the



Figure 104.12 Coronal section through head of Pan troglodytes (chimpanzee) showing well-developed maxillary sinus and reduced turbinal structure.



1322 ] PART 13 THE NOSE AND PARANASAL SINUSES



Figure 104.13 Skull of Miopithecus talapoin (dwarf gueron). The maxilla is encroached upon by the orbit and dentition. It therefore does not have a functional maxillary sinus.



maxillary sinus is not a constant feature and there does not appear to be any specific factor which correlates with its presence or absence among the anthropoids. In the great apes, the maxillary sinus and nasal configuration of the gorilla and chimpanzee resemble that of man with the natural ostium opening into the middle meatus. Most mammals have additional pneumatized chambers communicating with the olfactory region and opening between the ethmoturbinals into which the olfactory epithelium may extend. These may correspond to ethmoid, frontal or sphenoid sinuses but there is considerable interspecies and intraspecies variation. The big ungulates, elephants and carnivores, have considerable frontal pneumatization. In the majority of monkeys, gibbons and orangs, there is little frontal development. Humans and the African apes can be distinguished by the presence of an ethmoid complex in addition to a true frontal and sphenoid sinus, though the pattern of mammalian sinus pneumatization does not clarify the purpose of the paranasal sinuses.



ANATOMY OF THE NOSE AND PARANASAL SINUSES



Figure 104.14 Coronal section through nasal cavity of Castor canadensis (Canadian beaver) showing extensive ethmoturbinal system.



VESTIBULE AND SKIN



The vestibule is the dilated passageway leading from the external nares into the nasal fossae, demarcated by the limen nasi, at the superior margin of the lower lateral cartilage. It is lined by skin bearing coarse hairs or vibrissae (though without erector muscles), sebaceous glands and sweat glands. With increasing age, the overall area of the vestibule enlarges at the expense of the respiratory region of the nasal cavity. The thickness of the skin and soft tissues of the nasal bridge vary. Over the dorsum and sides of the nose, it is thin and loosely adherent. It becomes thicker and more adherent over the tip and alar cartilages where it contains numerous large sebaceous glands. The elasticity and mobility of the skin over the nose also varies, dependent upon the quality and anchorage of the collagen fibres running between the skin and underlying layers.



External nose and vestibule A considerable literature is available, devoted to descriptions of the external appearances of the nose as it pertains to cosmetic surgery (Figure 104.15).



MUSCLES OF THE EXTERNAL NOSE



The nose has a number of muscles which, in man, have assumed an almost vestigial importance (Figures 104.16



Chapter 104 Anatomy of the nose and paranasal sinuses



] 1323



Nasion



Dorsum



Lobule



Supratip area Lobule Tip



Supratip area Alar groove



Tip Columella



Nasolabial fold Nasolabial angle



Figure 104.15 External nose showing surface anatomy.



Procerus



Corrugator



Levator labii superioris Nasalis (alar part)



Nasalis (transverse part)



Depressor septi nasi Nasalis



Figure 104.16 External musculature of the nose.



and 104.17). As muscles of facial expression, they are all supplied by branches of the facial nerve. The depressor septi nasi muscle attaches between the alveolus and the medial crus of the lower lateral cartilage. Its function is to



depress the septum and tip, expanding the external nares during forced inspiration. The nasalis muscle is composed of an alar and a transverse part. The transverse fibres run from the pyriform aperture onto the dorsum of the nose



1324 ] PART 13 THE NOSE AND PARANASAL SINUSES



Nasal bones Frontal process at maxilla Upper lateral cartilages Septal angle Alar cartilages



1 1



Lateral cartilage



Nasal bones



2



2 Lateral cartilage



Supra-septal groove Septum



Figure 104.17 Bony and cartilaginous anatomy of the nose.



into a thin aponeurosis attached to the transverse muscle fibres of opposite side. Its action is to contract the nasal aperture. The alar component arises beneath the nasomaxillary suture and runs inferiorly, laterally and anteriorly, attaching by a short thin tendon to the skin of the nasal ala. Contraction produces shortening and dilatation of the nostril. A continuation of frontalis forms the procerus muscle. It shortens the nose when contracted but also produces facial movement of the area between the eyebrows, hence its alternative name, depressor glabellae.25 The development of this muscle and the associated fat pad is responsible for the shape of the root of the nose. Levator labii superioris alaequae nasi arises from the frontal process of the maxilla and blends with the perichondrium of the lateral crus of the lower lateral cartilage. It pulls this superiorly, dilating the nostril and in addition elevates the upper lip, hence its name. Electromyography shows that the nasal dilators are respiratory muscles, as their action correlates with ventilatory resistance.26 The supporting framework of the external nose is composed of a bony skeleton provided by the nasal bones, frontal processes of the maxillae and nasal part of the frontal bone and a cartilaginous framework consisting of



septum, upper and lower lateral cartilages and a variable number of minor accessory alar cartilages.



NASAL BONES



The nasal bones unite with each other in the midline, with the frontal bone superiorly at the nasofrontal suture and laterally with the frontal process of the maxilla at the nasolacrimal suture (Figure 104.18). They are supported by the nasal spine of the frontal bone and by the perpendicular plate of the ethmoid, both of which groove the bones. The nasal bone is wedge-shaped, usually convex and smooth on its outer surface and concave and roughened internally. The bones are grooved by adjacent neurovascular bundles. There is considerable ethnic and individual variation in the shape and size of the nasal bones.27



PYRIFORM APERTURE



The pyriform aperture is bounded below and laterally by the maxilla, and above by the nasal bones. The anterior nasal spine lies in the middle of the inferior border. It can



Chapter 104 Anatomy of the nose and paranasal sinuses



Figure 104.18 Human skull (anterior aspect) showing bones surrounding the pyriform aperture.



be up to 15 mm in length28 and is related superiorly to the anteroinferior free end of the septal cartilage.



CARTILAGES OF THE EXTERNAL NOSE AND COLUMELLA



The nasal cartilages are composed of hyaline cartilage which may be ossified. They prevent collapse of the vestibule on inspiration. The upper cartilages are triangular flat expansions lying inferior to the nasal bones and are overlapped by them, by the adjacent frontal processes of the maxillae and by the lower lateral cartilages in 72 percent of cases,27 to all of which they are attached by fibrous tissue. The groove between the upper and lower lateral cartilages is known as the limen nasi, which is the site of intercartilaginous incisions. The medial aspect of the upper lateral cartilages are continuous with the nasal septal cartilage which is bifid in this area. The lower lateral or alar cartilages form the lower third of the nose. They are each composed of a medial and lateral crus which meet at the dome of the tip, though the highest point can be on the lateral crus. The medial crura are loosely attached to each other in the midline and contribute to the columella, anterior to the quadrilateral cartilage. The lower margin of the lateral crus does not follow the margin of the nostril but ascends away from the margin laterally. Between one and four (average 2.3) minor sesamoid cartilages are found between the upper and lower lateral cartilages. The part of the septum running between the tip of the nose and philtrum is called the columella. It bounds the anterior nares medially and is thicker posteriorly because of the contribution made by the medial crura of the lower lateral cartilages.



BLOOD SUPPLY



Branches of the facial artery supply the alar region while the dorsum and lateral walls of the external nose are



] 1325



supplied by the dorsal branch of the ophthalmic artery and the infraorbital branch of the maxillary. There are significant anastomoses between these vessels on each side and between the right and left sides.27 The venous networks do not parallel the arterial supply but correspond to territories termed arteriovenous units.27 The frontomedian area drains to the facial vein and the orbitopalpebral area to the ophthalmic vein with interconnections to the anterior ethmoidal system and thence cavernous sinus which can be of clinical significance. The facial vein arises by the confluence of the supratrochlear and supraorbital veins at the inner canthus where it is termed the angular vein in its superior portion. Usually, a transverse venous anastomosis exists between the right and left supratrochlear veins.



NERVE SUPPLY



The skin of the external nose receives its sensory supply from the two upper divisions of the trigeminal nerve; ophthalmic and maxillary. The ophthalmic has an infratrochlear branch supplying the lateral surface of the root of the nose and an external nasal branch supplying the skin over the root and dorsum as far as the tip of the nose. The infraorbital branch of the maxillary nerve gives external and internal nasal branches which supply the nasal alae and skin of the nasal vestibule respectively, inferior palpebral and superior labial branches which form the pes anserinus minor with superior buccal branches of the facial nerve, and the anterior superior alveolar branch with its supply to the anterior lateral wall.



LYMPHATIC DRAINAGE



This drains from the external nose with the anterior face to the submandibular and submental nodes, with buccal nodes adjacent to the facial vein sometimes intervening. There may also be bilateral drainage and flow to the parotid region is possible.



Nasal cavity The nasal cavity extends from the external nares or nostrils to the posterior choanae, where it becomes continuous with the nasopharynx and is narrower anteriorly than posteriorly. Vertically, it extends from the palate to the cribiform plate, being broader at its base than superiorly where it narrows to the olfactory cleft. The nasal cavity is divided into two by a septum. The configuration and dimensions show considerable individual and ethnic variation. Each half has a floor, a roof, a lateral wall and a medial (septal) wall. The floor is concave from side to side, anteroposteriorly flat and almost horizontal. Its anterior three-quarters are composed of the palatine process of the maxilla, its posterior



1326 ] PART 13 THE NOSE AND PARANASAL SINUSES one-quarter by the horizontal process of the palatine bone. Approximately 12 mm behind the anterior end of the floor is a slight depression in the mucous membrane overlying the incisive canals. This contains the terminal branches of the nasopalatine nerve, the greater palatine artery and a short mucosal canal (Stenson’s organ). Occasionally, incisor and canine teeth can protrude into the floor of the nasal cavity. The roof is narrow from side to side, except posteriorly, and may be divided into frontonasal, ethmoidal and sphenoidal parts, related to the respective bones. As both the frontonasal and sphenoidal parts of the roof slope downwards, the highest part of the nasal cavity relates to the cribiform plate of the ethmoid which is horizontal. This area is covered by olfactory epithelium which spreads down a little distance onto the upper lateral and medial walls of the nasal cavity. The rest of the nasal cavity (with the exception of the nasal vestibule) is lined by respiratory mucous membrane which is intimately adherent to the underlying periosteum and perichondrium and is continuous with that of the paranasal sinuses, nasolacrimal duct and nasopharynx.



Nasal septum The nasal septum is composed of a small anterior membranous portion, cartilage and several bones: the perpendicular plate of the ethmoid, the vomer and two bony crests of the maxilla and palatine (Figure 104.19). The cartilaginous portion is composed of a quadrilateral cartilage with a contribution from the lower and upper lateral alar cartilages forming the anterior nasal septum. The quadrilateral cartilage is 3–4 mm thick in its centre but increases to 4–8 mm anteroinferiorly, an area which Nasal spine of frontal bone Nasal bone



has been termed the footplate. The upper margin of the cartilage also expands where it is connected to the upper lateral cartilages, forming the anterior septal angle, just cranial to the domes of the lower lateral cartilages. It is bound firmly by collagenous fibres to the nasal bones, and to the perpendicular plate of the ethmoid and vomer and, where it sits inferiorly in the nasal crest of the palatine process of the maxilla, the fascial attachment effects a pseudoarthrosis. It abuts the maxillary spine at the inferior septal angle. Anteriorly, it is attached by a thin membranous septum to the medial crura of the lower lateral cartilages. The perpendicular plate forms the superior and anterior bony septum, is continuous above with the cribriform plate and crista galli and abuts a variable amount of the nasal bones.27 The vomer forms the posterior and inferior nasal septum and articulates by its two alae with the rostrum of the sphenoid, thereby creating the vomerovaginal canals which transmit the pharyngeal branches of the maxillary artery. Occasionally, the sphenoid sinus may pneumatize the vomer. The inferior border of the vomer articulates with the nasal crest formed by the maxillae and palatine bones. The anterior border articulates with the perpendicular plate above and the quadrilateral cartilage inferiorly. The posterior edge of the vomer forms the posterior free edge of the septum. The nasal septum, and in particular the quadrilateral cartilage, is of crucial importance in the development of the middle third of the face. This has been the subject of considerable experimental and longitudinal clinical studies.29, 30, 31, 32, 33, 34, 35, 36, 37, 38 The surface area of the septum measures between 30 and 35 cm2 in adults.39 Deflections may develop at any of the septal articulations and spurs may also be found where the quadrilateral Frontal sinus Crista galli of ethmoid bone Cribriform plate



Medial edge of upper lateral nasal cartilage



Perpendicular plate of ethmoid



Septal cartilage Rostrum of sphenoid Membranous septum Medial crus of alar cartilage



Vomer Crest of palatine bone



Vomeronasal cartilage



Greater palatine foramen



Incisive canal



Crest of maxilla



Figure 104.19 The cartilaginous and bony septum of the nose.



Chapter 104 Anatomy of the nose and paranasal sinuses



cartilage sends small processes between the ethmoid and vomer. Theile40 found septal deviations were more often to the left than the right.27 As such deflections are far commoner in men than women, they are most likely to be acquired due to trauma than be congenital and this has been substantiated by work on identical twins in relation to deformities of the anterior septum.41



Axis



] 1327



Peripheral microtubule



Radial spoke



Spoke head Central microtubule



Ciliary membrane



HISTOLOGY Dynein arms



The mucoperichondrium and mucoperiosteum of the septum are separate from that overlying the maxillary crest, reflecting their embryological development. The mucous membrane is predominantly respiratory with a small area of olfactory epithelium superiorly adjacent to the cribriform plate. Respiratory epithelium is composed of ciliated and nonciliated pseudostratified columnar cells, basal pluripotential stem cells and goblet cells. The columnar cells are 25 mm in height and 7 mm wide, tapering to 2–4 mm at the basement membrane. Each cell bears 300–400 microvilli, irrespective of the presence of cilia. These are finger-like cytoplasmic extensions, 2 mm in length and 0.1 mm diameter. Their function is to increase surface area and thus prevent drying. Where cilia are present, there are 50–100 per cell though the number varies with their position in the nose and age. The cilia are composed of the classical axonema of nine peripheral doublet and two central single microtubules. Each peripheral pair (A and B) connects to the next doublet and to the central microtubule with hexin links. The A microtubule bears an outer and inner dynein arm, composed of ATPase which can attach to the B microtubule, leading to axonemal displacement and cilial beating (Figure 104.20). Seromucinous glands are found in the submucosa and are more important in mucus production in the nasal cavity than the goblet cells which are more numerous in the sinuses. On the septum, the number of goblet cells increases from anterior to posterior and from superior to inferior.42 By contrast, the glands decrease from anterior to posterior and from superior to inferior. In newborns, the septal mucosal surface is 450 mm2 with 17–18 glands/ mm2, compared with the adult septum of 1700 mm2 and 8.5 glands/mm2 (Figure 104.21). The olfactory epithelium spreads down from the cribriform plate onto the upper septum. It is composed of receptor cells, supporting cells with microvilli and basal stem cells conferring on olfactory epithelium the capacity for regeneration. Each receptor cell has approximately 17 cilia, but these differ from their respiratory counterparts in their radial arrangement, greater length and poorly developed ultrastructure. Dynein arms are not present, preventing linking between the microtubules and conventional beating. The sensory endings have a characteristic knob-like vesicular structure from which olfactory fibres join the axonal bundle (Figure 104.22). There is a



Bridge of hexin links



Axoneme Microtubule B Microtubule A



Figure 104.20 Diagrammatic cross section through cilium near the base, showing the ultrastructure the axonema and demonstrating the classical 9 1 2 pattern of microtubules.



Figure 104.21 Scanning electron micrograph of respiratory epithelium showing ciliated cells with globules of mucus (magnification 3300).



Figure 104.22 Scanning electron micrograph of olfactory epithelium showing a senosory bulb (magnification 24,000).



1328 ] PART 13 THE NOSE AND PARANASAL SINUSES sharp transition zone between the olfactory and respiratory epithelium though the relative area of each varies with age and reflects the decrease in olfactory acuity. Secretion for the olfactory epithelium is provided by Bowman’s glands.



Anterior ethmoidal Posterior ethmoidal



BLOOD SUPPLY



Superior labial



The external and internal carotid arteries are responsible for the rich blood supply to the nose. The sphenopalatine artery (branch of the maxillary artery and thus external carotid artery) supplies the posteroinferior septum and the antero–inferior face of the sphenoid sinus, as well as the posterior halves of inferior and middle turbinates. The greater palatine artery (also a branch of the maxillary) supplies the anteroinferior portion entering the nasal cavity via the incisive canal. The superior labial branch of the facial artery contributes anteriorly, in particular to Kiesselbach’s plexus, which is composed of unusually long capillary loops and is situated in Little’s area on the anterior septum – a common source of epistaxis. The internal carotid artery supplies the septum superiorly via the anterior and posterior ethmoidal arteries and also contributes to Kiesselbach’s plexus (Figure 104.23). There is a sinusoid system in the nasal submucosa under autonomic control which has been well described in relation to the turbinates but is also present on the septum adjacent to the inferior turbinate and on the most anterior septum (Figure 104.24). This anterior septal tubercle or intumescence was first described by Morgagni6 and may be related to control of airflow into the olfactory cleft. A similar structure is seen on the posterior septum in two-thirds of individuals. The cavernous venous system drains via the sphenopalatine vessels into the pterygoid plexus posteriorly and into the facial veins anteriorly. Superiorly, the ethmoidal veins communicate with the superior ophthalmic system and there may be direct intracranial connections through the foramen caecum into the superior sagittal sinus.



Greater palatine Sphenopalatine



Figure 104.23 Blood supply to the nasal septum. Reprinted from Ref. 43 by courtesy of Georg Thieme Verlag.



NERVE SUPPLY



The maxillary division of the trigeminal nerve provides the sensory supply to the majority of the nasal septum (Figure 104.25). The nasopalatine nerve supplies the bulk of the bony septum, entering the nasal cavity via the sphenopalatine foramen, passing medially across the roof of the upper septum and running down and forwards to the incisive canal to reach the hard palate. The anterosuperior part of the septum is supplied by the anterior ethmoidal branch of the nasociliary nerve and a smaller anteroinferior portion receives a branch from the anterior superior alveolar nerve. The posteroinferior septum also receives a small supply from the nerve to the pterygoid canal and a posterior inferior nasal branch of the anterior palatine nerve.



Figure 104.24 Coronal CT scan showing anterior septal tubercle of anterior nasal septum.



The sensory nerves are accompanied by postganglionic sympathetic fibres to blood vessels and postganglionic parasympathetic secretomotor fibres pass to glands with the branches from the pterygopalatine ganglion. The olfactory epithelium covers the inferior surface of the cribiform plate spreading down to cover a variable area on the upper septum and adjacent lateral wall, over



Chapter 104 Anatomy of the nose and paranasal sinuses



the medial surface of the superior concha. With increasing age, the area is encroached upon by respiratory epithelium and in the adult covers an area of approximately 2–5 cm2. The surface area is considerably increased by the cilia on the receptor cells. Nerve fibres arising from the olfactory receptors are slim (0.2 mm in diameter) and nonmyelinated. They join up into approximately 20 bundles which traverse the cribiform plate to reach the olfactory bulbs. Each bundle carries a tubular sheath of dura and pia-arachnoid, which may be sheared in head injuries, destroying olfaction and potentially producing cerebrospinal fluid leakage. The fibres synapse in the glomeruli of the olfactory bulbs which in turn connect with the olfactory tract and with other cells (mitral and tufted) within the bulb which contribute to feedback loops from higher cortical centres. The olfactory tract passes to the trigone from which diverging bundles, the medial and lateral olfactory striae, pass around the anterior perforating substance to connect with the hypothalamus, amygdala and hippocampus. These complex linkages subserve the interactions of smell, taste, feeding and reproductive behaviour, representing one of the most primitive areas of the brain.



] 1329



anterior and middle third. In adults, this ranges from 1.6 to 2.3 cm (mean 1.9 cm) at 1.6 cm along the bony lateral wall.45 The nasolacrimal duct opens into the inferior meatus usually just anterior to its highest point. There is no true valve, the opening being covered by small folds of mucosa. It can be endoscopically identified in life by gentle massage of the lacrimal sac at the medial canthus.



INFERIOR TURBINATE



This structure is composed of a separate bone, the inferior concha which has an irregular surface, perforated and grooved by vascular channels to which the mucoperiosteum is firmly attached (Figures 104.26 and 104.27). The bone has a maxillary process which articulates with the inferior margin of the maxillary hiatus. It also articulates with the ethmoid, palatine and lacrimal bones, completing the medial wall of the nasolacrimal duct. The inferior concha has its own ossification centre which appears around the fifth intrauterine month. The turbinate



F



LYMPHATIC DRAINAGE



The anterior septum drains with the external nose to the submandibular nodes while drainage is to the retropharyngeal and anterior deep cervical nodes posteriorly.



S



ST MT IT



The lateral nasal wall INFERIOR MEATUS



The inferior meatus44 is that part of the lateral wall of the nose lateral to the inferior turbinate. It is the largest meatus, extending almost the entire length of the nasal cavity. The meatus is highest at the junction of the



Figure 104.26 Sagittal section through the nasal cavity showing structures of lateral nasal wall. F, frontal sinus; IT, inferior turbinate; MT, middle turbinate; S, sphenoid sinus; ST, superior turbinate.



P A



Figure 104.25 Nerve supply to the nasal septum. Reprinted from Ref. 43 by courtesy of Georg Thieme Verlag.



Figure 104.27 Inferior turbinate, medial surface. A, anterior; P, posterior. Reprinted from Ref. 43 by courtesy of Georg Thieme Verlag.



1330 ] PART 13 THE NOSE AND PARANASAL SINUSES possesses an impressive submucosal cavernous plexus with large sinusoids under autonomic control which provides the major contribution to nasal resistance. The turbinate is covered by respiratory epithelium, with a high number of goblet cells (approximately 8/mm2) which decrease in density towards the posterior end.42



MIDDLE MEATUS



The middle meatus is that portion of the lateral nasal wall lying lateral to the middle turbinate (Figures 104.28 and 104.29). It receives drainage from the frontal, maxillary



and anterior ethmoidal sinuses. Considerable confusion has arisen with regard to terminology in this area, as many terms originally defined in the last century by the German and French have been used interchangeably. In the past, when radical sinus surgery was predominantly used for most pathology, this was of less significance. The advent of endoscopic surgery has led to an increased interest in the detailed anatomy of the region and a need for consensus in terminology. The configuration of the structures of the middle meatus are complex and variable, but can more readily be understood when the embryological development of the area is considered. If the topographical anatomy is considered in the sagittal plane, a number of structures are apparent, covered by the middle turbinate. In a disarticulated skull, the maxillary bone has a large opening in its medial wall, the maxillary hiatus. In the articulated skull this is filled in by adjacent bones:  inferior: the maxillary process of the inferior turbinate bone;  posterior: the perpendicular plate of the palatine bone;  anterosuperior: a small portion of the lacrimal bone;  superior: the uncinate process and bulla of the ethmoid. A portion of the maxillary hiatus is nevertheless left open by these osseous attachments, which in life is filled by the mucous membrane of the middle meatus, the mucous membrane of the maxillary sinus and the intervening connective tissue – the membranous portion of the lateral



F



HS UP



S



BE



AF



Figure 104.28 (a) Schematic drawing of relation of anterior ethmoid complex to frontal sinus. 1, frontal sinus with frontal sinus infundibulum; 2, frontal recess; 3, uncinate process; 4, hiatus semilunaris