of 353/353
The Society of Cardiothoracic Surgeons of Great Britain and Ireland Fifth National Adult Cardiac Surgical Database Report 2003 Improving outcomes for patients Compiled by Bruce E. Keogh KBE BSc MD FRCS FECS FETCS Society of Cardiothoracic Surgeons Robin Kinsman BSc PhD Dendrite Clinical Systems Ltd

The Society of Cardiothoracic Surgeons of Great Britain and Ireland · 2013-01-24 · The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical

  • View
    1

  • Download
    0

Embed Size (px)

Text of The Society of Cardiothoracic Surgeons of Great Britain and Ireland · 2013-01-24 · The Society...

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland

    Fifth National Adult Cardiac Surgical Database Report2003

    Improving outcomes for patients

    Compiled by

    Bruce E. KeoghKBE BSc MD FRCS FECS FETCSSociety of Cardiothoracic Surgeons

    Robin Kinsman BSc PhDDendrite Clinical Systems Ltd

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland

    Fifth National Adult Cardiac Surgical

    Database Report 2003

    Improving outcomes for patients

    Compiled by

    Prof. Bruce Keogh Dr Robin Kinsman

    KBE BSc MD FRCS FESC FETCS BSc PhD

    Society of Cardiothoracic Surgeons Dendrite Clinical Systems Ltd

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    2

    The Society of Cardiothoracic Surgeons of

    Great Britain and Ireland

    gratefully acknowledge the assistance of

    Dendrite Clinical Systems Ltd

    for

    data harvesting, data aggregation,

    data analysis, risk modelling

    and publishing this report.

    Dendrite Clinical Systems Ltd is registered

    under the Data Protection Act

    Data Protection Act Register Number

    Y33 54 136

    This document is proprietary information that is protected by copyright. All rights reserved. No part of this document may be photocopied, stored in a retrieval system, transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the permission of the publishers and without prior written consent from the Society of Cardiothoracic Surgeons of Great Britain and Ireland, the Royal College of Surgeons, 35/43 Lincoln’s Inn Fields, London, WC2A 3PN.

    Windows and Excel are registered trademarks of the Microsoft Corporation.

    Crystal Reports is a registered trademark of Crystal Decisions.

    The Patient Analysis & Tracking System is a registered trademark of Axis Clinical Software Inc.

    July 2004 Edition. A catalogue record for this book is available from the British Library. ISBN 1-903968-09-7

    © 2004 Dendrite Clinical Systems Ltd and the Society of Cardiothoracic Surgeons of Great Britain and Ireland. All rights reserved.

    Published by:

    Dendrite Clinical Systems Ltd, 59A Bell Street, Henley-on-Thames, Oxfordshire RG9 2BA, United Kingdom

    Contact details: Phone: +44 1491 411 288, Fax: +44 1491 411 377, E-mail: [email protected]

    £60.00 €90.00

    Printed and Bound by CPL Associates

    ABC

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    3

    Executive summary

    • This report documents the nature of contemporary cardiac surgery practice in the United Kingdom and Ireland.

    • Cardiac surgery in the UK and Ireland is safe by international standards • The operative mortality rate for all major operation groups continues to fall despite a worsening

    casemix.

    • Coronary surgery activity has settled at about 25,000 cases a year. Combined valve and coronary surgery continues to increase with over 3,000 combined operations performed in the financial year 2002-2003.

    • Analysis of contemporary risk profiles of cardiac surgery patients shows that surgeons are presented with patients of increasing complexity year-on-year; this trend is quite marked.

    • This report focuses on coronary surgery, which represents about 70% of cardiac surgical activity. Our results for coronary surgery are comparable with the best in the world and continue to improve:

    • We are better at treating the elderly. The proportion of patients over 75 years old has increased by more than 4½-fold over the last decade. During the last 5 years the mortality in this group has fallen by 35%.

    • We are better at treating diabetics. The proportion of diabetic patients has risen from 15% to 22% since 1997, but the operative mortality has fallen from 5.4% to 2.6% a reduction of 50%.

    • We are better at treating women with heart disease. Around the world women have a higher operative mortality than men, but this has fallen by over 30% in the UK and Ireland since 1997.

    • We are better at treating people with high blood pressure. The proportion of patients with high blood pressure has risen from 48% in 1998 to 65% in 2003. During this time the mortality for these patients has fallen from 2.8% to 2.1%. A reduction of 25%.

    • We are better at treating patients with very sick and weak hearts. About 10% of patients have very weak hearts. With improving techniques the operative mortality for this group has been reduced by 30% over the last 5 years.

    • Two-thirds of coronary surgery patients are overweight or frankly obese. • Only 70% of coronary bypass operations are elective. There is marked variation between hospitals

    in the proportion of elective and urgent operations. The operative mortality for non-elective coronary cases is twice as high as for elective patients. This is not changing.

    • Patients who have had a previous heart attack are almost twice as likely to die at the time of surgery compared with people who have not had a heart attack.

    • For the first time in the UK surgical results for individual surgeons have been analysed for a mainstream specialty in the UK. With the statistical methodology employed, all surgeons meet acceptable standards over the three financial years 2001-2003.

    • Data submission for basic activity and mortality for coronary surgery is voluntary but complete for all consultant surgeons in the UK.

    • Data capture of variables that allow for comparisons in casemix between institutions and surgeons is incomplete preventing reliable and meaningful risk adjusted comparisons.

    • Crude mortality, not adjusted for casemix, has therefore been used for comparative analyses.

    • A surgeon’s crude mortality naturally fluctuates from year to year.

    • The level of variation in crude mortality between surgeons is less than that seen in New York or New Jersey where similar data are published.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    4

    Preface Mr Patrick Magee President of the Society of Cardiothoracic Surgeons of Great Britain and Ireland

    The first National Adult Cardiac Surgical Database Report was produced in 1996. It consisted of a stapled pamphlet with data from 12 centres. Since then we have come a long way. This report of over 340 pages is an almost complete record of contemporary cardiac surgical practice in the UK today. It provides a mine of information, which will be of value to surgeons, patients and those interested in healthcare.

    The information is compiled from several data sources each of which brings a different perspective to our understanding of national practice. The UK Cardiac Surgical Register tells us exactly how many different operations were performed in 2002/2003 and what the results were in each unit. Data from The National Adult Cardiac Surgical Database on over 210,000 patients paints a picture of how patient populations have changed over the last decade and how they vary between different units in the UK. The UK Heart Valve Registry data provides long-term mortality data on patients undergoing heart valve replacement in the UK over the last 15 years.

    Information of this nature is interesting enough on its own. However, following events in Bristol we have also focussed on how to measure performance. Within the specialty we have developed ways of measuring and predicting the risk of surgery both within the UK and internationally. We have applied this knowledge to give better informed consent to our patients and to monitor our own performance and compare it with others. These techniques are clearly described in the report and are applicable to many other specialties.

    The last couple of years have engendered considerable unease as we have struggled over how best to present individual surgeon’s results. Although we know every surgeon’s results for coronary surgery our data is not yet good enough to allow us to cater for variations in casemix, which can be very different. An experienced surgeon is likely to operate on different patients to a less experienced surgeon. We have done the best we can with the available data. The methodology and its rationale are clearly described. All consultant surgeons have contributed voluntarily to this endeavour, which is a mark of our commitment to transparency and quality. We remain concerned, however, that unless this process is handled properly patients will suffer if surgeons feel compelled to avoid the higher risk patients.

    This report is a remarkable achievement, which would not have been possible without the encouragement, vision and support of many. Dr Peter Walton of Dendrite Clinical Systems has been instrumental in establishing this Blue Book and has fully supported this project over the years. Robin Kinsman has personally merged and checked all the submitted data and brought increasing intellectual independence and scrutiny to the process of data management and analysis.

    Bruce Keogh has provided leadership and encouragement as this process has evolved over the years, but in addition has put in endless hours analysing and preparing the information for presentation. This impressive document is evidence of this effort. Finally, of course, none of this would be possible without the enthusiastic participation and support of the members of the Society of Cardiothoracic Surgeons.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    5

    Foreword Mr James Roxburgh Secretary of the Society of Cardiothoracic Surgeons of Great Britain and Ireland

    A quarter of a century ago the Society of Cardiothoracic Surgeons of Great Britain and Ireland agreed to collect raw mortality data for a wide range of cardiac and thoracic surgical procedures. The idea was that all units would submit unit specific data to the Society who would then collate the data and publish annual mortality figures. It was expected that units would then compare these national figures with their own and look for variations and if there was cause for concern, subject their own results to a more detailed review. However as The Bristol Inquiry showed simple collection of the data was not enough, we needed to be more proactive and monitor performance centrally 1. There can have been few surgeons who did not feel that during The Bristol Inquiry we were hung by our own petard when it came to data collection! What had started out as an innovative and ground-breaking project was used as a stick to beat us with; but far from cowering under the pressure of the press coverage and the politicians we have slowly but surely built up a performance review mechanism that is second to none.

    It has been said that by publishing individual performance indicators for Consultant Cardiac Surgeons we have crossed the Rubicon; this is indeed true but we have done so flanked by our involvement in two major projects, the NCEPOD study into deaths following first-time coronary artery surgery, and the cardiac surgical component of the Central Cardiac Audit Database (CCAD). The NCEPOD study is a detailed examination of all deaths following surgery for first-time coronary artery surgery over a 3-year period and will involve some 1500 patients. The study will gather information about the patient’s treatment from their first presentation to hospital through up to the events surrounding the death. The outcome of cardiac surgery is multi-factorial and this project is, we believe, the first of its kind to be undertaken anywhere in the world; since it is a joint undertaking with an organisation with a proven track record in the analysis of peri-operative deaths. The CCAD study will allow near real-time submission of data to a central database that will then feed back to the units comparative figures of their performance against national values. This will involve the submission of around 150 data points for every single cardiac surgical procedure undertaken in the United Kingdom, this high volume of data will allow units to receive comparative data on a whole variety of performance indicators such as death, complication rates and length of stay. It will also provide information about the case-mix of unit workload, an important factor in understanding outcome measures. The information from these projects will of course be published in future editions of the Blue Book and will be a unique resource of high quality performance indicators that are available to the public and the profession alike.

    The sporting cliché There is no 'i' in team should be borne in mind when discussing the release of surgeon-specific data. The outcome of a cardiac surgical procedure is par excellence an example of teamwork. There are many factors; such as the patient’s general condition, the other medical staff (cardiologists, anaesthetists, intensivists and junior medical staff), the post-operative care (nurses, physiotherapists, pharmacists etc.), the hospital facilities (infection rates, physical plant) and so on, that can have a bearing on the surgical outcome that it is invidious to name an individual as responsible for the success or failure of a cardiac surgical operation. In this day and age many of these factors are simply outside the day-to-day control of the Consultant Surgeon. We have though not used this argument to refuse to publish individual performance figures but hope that the format in which it is published underlines the need to examine surgical outcomes on a unit basis. The result the major projects that are currently underway will, I believe, persuade most people that unit-specific data is of more use to the patients and the public as a whole. The Society has monitored the results of surgeons on an individual basis for over 5 years and will of course continue to do so. Many of the techniques used in performance monitoring and risk management have been developed in the aviation industry; and it has been argued that the airline pilot/passenger and the surgeon/patient relationships are synonymous in risk management terms. However I would argue that that the patient’s journey through hospital is analogous to a holiday abroad rather than a single airline flight. One would no more blame or praise the pilot for the hotel, weather or food than one should hold an individual surgeon accountable for the functioning of the whole hospital. We must take a system-based approach to these complex matters and not one directed towards an individual. The performance individuals within that system must of course be monitored and the Society of Cardiothoracic Surgeons has been responsible for the review of Consultant Cardiac Surgical performance for over 5 years, the results have not, until now, been published.

    Cardiac surgery in the United Kingdom can justly claim to be world leaders in the way it provides high quality meaningful performance data to the profession and public alike. It is now high time that other high volume interventional specialities demonstrated their commitment to performance review and the public presentation of this information.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    6

    Acknowledgements to key contributors

    SCTS Executive Committee 2002 -2003

    Member Position Tenure City

    Mr Pat Magee President 2004-2006 London

    Prof. Bruce Keogh President-elect 2004-2006 London

    Mr James Roxburgh Secretary 2004-2009 London

    Mr Babulal Sethia Treasurer 2004-2009 London

    Mr Graham Cooper Meeting Secretary 2000-2005 Sheffield

    Mr Peter Goldstraw Chairman, Specialist Advisory Committee on Higher Surgical Training

    1999-2005 London

    Mr Andrew Murday Chairman, Intercollegiate Examination Board 2004-2007 Glasgow

    Mr Leslie Hamilton Cardiothoracic Dean 1999-2004 Newcastle

    Mr Jonathan Hyde Young Consultants’ Representative 2002-2004 Brighton

    Mr Christopher Munsch Cardiothoracic Tutor, Royal College of Surgeons of England

    2000-2004 Leeds

    Mr Alfred Wood Representing the Republic of Ireland From 2003 Dublin

    Mr Michael Lewis Trainees’ Representative 2002- Birmingham

    Mr Steven Hunter Elected member 2002-2005 Middlesbrough

    Mr Alan Faichney Elected member 2002-2005 Glasgow

    Mr Simon Kendall Elected member 2003-2006 Middlesbrough

    Mr Richard Page Elected member 2003-2006 Liverpool

    Mr Stephen Livesey Elected member 2004-2007 Southampton

    Mr Graham Venn Elected member 2004-2007 London

    Ms Jacqueline Nicol Nursing Representative 2004-2007 Glasgow

    Recent Executive Committee members contributing to the development of this report

    Member Position Tenure City

    Mr Colin Hilton President 2002-2004 Newcastle

    Mr Jim Monro President 2000-2002 Southampton

    Mr Andrew Murday Chairman, Intercollegiate Examination Board 2001-2004 Glasgow

    Mr Terence Lewis Elected member 2000-2003 Plymouth

    Mr Robert Bonser Elected member 2001-2004 Birmingham

    Mr Mark Jones Elected member 2001-2004 Manchester

    Mr Sam Nashef Elected member 1999-2002 Papworth

    Mr David Luke Representing the Republic of Ireland Till 2003 Dublin

    All UK surgeons

    This report would not have been possible without the ongoing support of the members of the Society who have generously submitted their heart surgery data for analysis.

    Other contributors who deserve acknowledgement

    Dr Alan Bailey, who co-founded and managed the Cardiac Surgical Register for the last 25 years; Dr Roger Boyle, the National Director for Heart Disease in England; Dr Jill Pell, Chairman of the Scottish Cardiac Registers Steering Committees; Mr Martin Old, project manager of CCAD; Mrs Maria Shortis, founder of Constructive Dialogue for Clinical Accountability, for ensuring that the issues of clinical excellence and patient safety remain central to our endeavours.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    7

    Data managers

    Data management 2003 CITY HOSPITAL

    Surgical audit lead Data manager

    Aberdeen Royal Infirmary Mr El Shafei Ms E Anderson

    Belfast Royal Victoria Hospital Mr A Graham Ms B Scullion

    Birmingham Queen Elizabeth Hospital Mr B Keogh (to 2003) Mr D Pagano (2004)

    Ms V Barnett

    Blackpool Victoria Hospital Mr J Au Ms C Malpas

    Brighton Royal Sussex County Hospital Mr A Forsyth Mrs J Gayfer

    Bristol Royal Infirmary Mr A Bryan Mr D Finch

    Cardiff University Hospital of Wales Prof. U Von Oppell Ms C Bray

    Cork Cork University Hospital Mr A O’Donnell Ms M Delaney

    Coventry Walsgrave Hospital Mr R Patel Mrs A Nugent

    Dublin Mater Misericordiae Mr A E Wood Mr P Marsden

    Dublin St James’s Hospital Miss E McGovern Ms A Hughes

    Edinburgh Royal Infirmary Mr W Walker Mr A Deegan

    Glasgow Royal Infirmary Mr A Murday Ms A McOuat

    Glasgow Western Infirmary Mr G Berg Ms M Kinnaird

    Hull Castle Hill Hospital Mr L Guvendik Mr N Richards

    Leeds General Infirmary Mr P Kay Mr J Peckover

    Leicester Glenfield Hospital Mr T Spyt Mr P Stafford

    Liverpool Cardiothoracic Centre Mr B Fabri Mr A Ward

    London Guy's & St Thomas’ Hospitals Mr J Roxburgh None

    London Hammersmith Hospital Mr P Punjabi Ms S Krikler

    London Harefield Hospital Mr A Khaghani Mr P Kimberley

    London King's College Hospital Mr J Desai Mr J Omigie

    London London Bridge Hospital Mr G Venn Mrs L Katsumbe

    London The Heart Hospital Mr D Lawrence Mr M Curtis

    London Royal Brompton Hospital Mr N Moat Mr P Kimberley

    London Bart’s and the London Mr A Wood Mr N Silverside

    London St George's Hospital Mr J Smith Dr O Valencia

    London St Mary's Hospital Mr B Glenville (to 2003) Mr R Stanbridge (2004)

    None

    Manchester Royal Infirmary Mr D Keenan Ms S Chaisty

    Manchester Wythenshawe Hospital Mr B Bridgewater Mr S Bullough

    Middlesbrough South Cleveland Hospital Mr A Owens Ms T Smailes

    Newcastle Freeman Hospital Mr S Barnard Ms S Jamieson

    Nottingham City Hospital Mr I Mitchell Mr M Whitworth

    Oxford John Radcliffe Hospital Mr C Ratnatunga Mr C Evans

    Papworth Papworth Hospital Mr S Nashef Ms S Powell

    Plymouth Derriford Hospital Mr A Marchbank Mrs K Caley

    Sheffield Northern General Hospital Mr D Hopkinson Ms J Daly Ms L Kirk

    Southampton Southampton General Hospital Mr S Livesey Ms A Green

    Stoke-on-Trent N Staffordshire Royal Infirmary Mr C Smallpeice Mrs J O’Bree

    Swansea Morriston Hospital Mr A Youhana None

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    8

    Summary of contents

    This report is divided into the following sections:

    1. Acknowledgements to key contributors (page 6-7).

    2. Preface and foreword from the President and the Secretary of the Society (pages 4-5).

    3. Introductory section (pages 13-38).

    Regional perspectives on data issues, progress against the Bristol Royal Infirmary Inquiry Report and basic information for patients.

    4. The United Kingdom Cardiac Surgical Register (pages 40-59).

    The UK Cardiac Surgical Register was established in 1977 to collect activity and mortality data for all cardiac surgery conducted in NHS hospitals in the UK.

    The UK Cardiac Surgical Register was terminated in 2003 and has been replaced by the National Adult Cardiac Surgical Database.

    5. Report from the United Kingdom Heart Valve Registry (pages 60-71).

    The UK Heart Valve Registry was established in 1986 to collect data on all heart valve implants in the NHS and to track those patients for re-operation and death. Mortality tracking is conducted through the Office of National Statistics (England, Scotland & Wales) and the Central Services Agency (Northern Ireland).

    6. The National Adult Cardiac Surgical Database (pages 72-212).

    The National Adult Cardiac Surgical Database was established as a pilot project in 1994 and first reported in 1996. It aims to collect detailed information on all patients undergoing heart surgery in the UK. This section focuses on coronary surgery and describes the national patient population, how this population changes over time and how it varies between hospitals. It also describes how changes in surgical practice and surgical techniques have resulted in ever-improving outcomes for patients.

    7. Principles of risk measurement and stratification for coronary surgery (pages 213-243).

    This section describes ways of measuring surgical risk in a way that is of value to for individual patients and also for measuring performance.

    8. Missing data (pages 234-239).

    9. Performance indicators and the presentation of surgical results (pages 244-263).

    This is another new section. It provides information on performance for both units and surgeons.

    10. Institutional and individual surgeon performance (pages 264-309).

    11. Long-term follow-up for patients with different risk factors (pages 310-318).

    This is a new section made possible by the linkage between a test centre and the Central Cardiac Audit Database to provide long-term mortality tracking through the Office of National Statistics. This adds an entirely new dimension to the potential for analyses that are available on the national data, and will further aid the decision-making process for both patients and surgeons.

    12. Appendices (pages 320-340)

    A detailed table of contents can be found on the following pages.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    9

    Table of contents

    Executive summary 3 Preface 4 Foreword 5 Acknowledgements to key contributors 6

    SCTS Executive Committee 2002 -2003 6 Data managers 7

    Summary of contents 8 Introduction 13 Perspective from the English Department of Health 14 The value of national data collection: a Scottish perspective 15 A cautionary note on the interpretation of surgical results 16 Background to the Society’s data-collection initiatives 17

    Harmonising datasets for the CCAD 23 Difficulties in data collection 25 The National Service Framework for Coronary Heart Disease 28

    Audit requirements defined within the National Service Framework 28 Milestones relating to clinical audit 28

    Progress against the Bristol Royal Infirmary Inquiry Report 29 The need for data validation 31 Collaboration between the SCTS and the NCEPOD 33 Improving informed consent 34 Informing patients and the public 35 Basic principles of cardiac surgery 36

    Valvular heart disease 36 Coronary artery disease 36 Surgical correction of valvular and coronary artery disease 36 Types of bypass grafts 37

    The United Kingdom Cardiac Surgical Register (UKCSR) 40 Overall trends in cardiac surgery since 1977 40 Trends in coronary artery bypass surgery since 1977 41 Trends in valve surgery since 1977 42 UKCSR: Data for financial year 2003 43

    UKCSR: summary data financial year 2003 43 UKCSR: single open valve operations financial year 2003 44 UKCSR: multiple open valve operations financial year 2003 44 UKCSR: miscellaneous operations for IHD financial year 2003 45 UKCSR: miscellaneous operations for acquired heart disease financial year 2003 46 UKCSR: cardiac surgical activity by centre financial year 2003 47 UKCSR: results by centre financial year 2003; first-time isolated CABG 50 UKCSR: results by centre financial years 2001-2003; first-time isolated CABG 51 UKCSR: results by centre financial year 2003; first-time isolated AVR 52 UKCSR: results by centre financial years 2001-2003; first-time isolated AVR 53

    Don’t jump to conclusions 54 Comparison of different sources of data on isolated CABG procedures 56

    Final transition from the UKCSR to the NACSD 57

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    10

    The United Kingdom Heart Valve Registry (UKHVR) 60 Background 60 Data collection 60 Follow up data 61

    Re-operation 61 Notification of death 61

    Feedback information 61 Publication 62 Research output from the HVR 62

    The National Adult Cardiac Surgical Database (NACSD) 72 Introduction to the database 72 Contributing centres and trusts 72

    Contributing centres 73 Data harvest, import, merge and analysis methodology 74

    Data collection facilities by centre as at June 2004 76 General cardiac surgical outcomes 77

    Mortality rate by procedure 77 Post-operative stay 78

    Valve surgery 79 General 79 Aortic valve surgery 80 Mitral valve surgery 82 Mitral valve replacement and repair 84

    Coronary artery bypass surgery 87 Age 89 Gender 98 Size – body surface area and body mass index 104 Priority 110 Ejection fraction 116 Previous myocardial infarctions (MIs) 122 Percutaneous coronary intervention (PCI) 132 Left main stem (LMS) disease 134 Previous surgery 140 Diabetes 148 Hypertension 154 Peripheral vascular disease (PVD) 160 Renal disease 166 Angina 172 Dyspnoea (breathlessness) 178 Interactions between risk factors 184

    Re-operation for post-operative bleeding 186 Re-operation for post-operative bleeding by procedure 186 Re-operation for post-operative bleeding after isolated CABG 188

    Aspects of coronary surgery practice in the UK and Ireland 195 Number of bypass grafts 195 The use of arterial grafts 197 Protecting the heart during coronary surgery 201

    Influence of myocardial protection technique on mortality 203 On- and off-pump CABG 204

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    11

    International comparisons 205 United Kingdom, Ireland and the USA 210

    Evolution of risk factors in coronary bypass surgery 211 Experience from the Society of Thoracic Surgeons National Database (1980-2000) 211 Development of trends 211 Practical considerations 211 Conclusions 212

    Risk stratification for coronary surgery 214 What risks are attached to my operation? 214 The Parsonnet score 215

    Parsonnet score distributions through time 215 Mortality by Parsonnet score 217 Completeness of Parsonnet score 218

    The EuroSCORE 219 EuroSCORE distributions through time 220 Mortality by EuroSCORE 221 Completeness of EuroSCORE data 222

    The Logistic EuroSCORE 223 Logistic EuroSCORE distributions through time 223 Mortality by Logistic EuroSCORE 224 Completeness of logistic EuroSCORE data 225

    Bayesian modelling 226 What do Bayes tables do? 226 Two Bayesian risk models for isolated coronary artery bypass surgery 227 Simple (5-factor) Bayes model for isolated CABG 228 Complex (9-factor) Bayes model for isolated CABG 228 Adoption of the complex Bayes model 229

    The complex Bayes score 230 Complex Bayes score distributions through time 230 Mortality by complex Bayes score 232 Completeness of the complex Bayes score data 233

    Missing data 234 Missing simple Bayes variables 234 Missing complex Bayes variables 236 Missing EuroSCORE variables 238

    Testing a risk scoring system 240 Receiver Operating Characteristic (ROC) curve 240 Calibration plot 240

    Comparisons of risk scoring systems 241 ROC curve areas 241 Calibration 242

    Performance indicators and the presentation of surgical results 244 The application of risk stratification 244

    Why do we need statistical analysis at all? 244 Setting a benchmark threshold for comparison 245 The graphical presentation of results 245

    Average outcomes over a given timeframe 245 Ranking analyses 245 Funnel plots 247

    Sequential analyses 249 Standard CUSUM 249 Setting control limits around CUSUM curves 251

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    12

    Variable Life-Adjusted Display (VLAD) 253 Plotting cumulative risk 253 The VLAD plot 253 Risk Adjusted CUSUM (RA-CUSUM) 256 Sequential probability ratio testing (SPiRiT) 258 Beyond mortality 260

    Institution- and individual-surgeon performance 264 A general perspective on the publication of cardiac surgical results 264

    Different data-sources tell different stories 277 Hospital Episode Statistics 280 Paul Ayling, Clinical Senior Lecturer in Epidemiology and Public Health, Assistant Director 280 Dr Foster Unit at Imperial College 280 Background to the publication of individual surgeon’s performance against standards 281

    Why publish individual surgeon’s results? 281 The way forward 282 Could crude mortality be presented? 283 Examining institutional influences on outcomes 285 Key points in interpreting the hospital tables including surgeons’ performance 286

    Long-term follow up through the Central Cardiac Audit Database 310 Appendices 320

    Governance arrangements for adult cardiac surgical audit components 321 The Parsonnet score 324 The EuroSCORE 325 The simple (5-factor) CABG Bayes score 326 The complex (9-factor) CABG Bayes score 327 Facts of life the figures can hide 328 Guidelines for the audit of cardiothoracic surgical practice 330 Quality accreditation programme for adult cardiac surgery 332 The minimum dataset 335

    References 348

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    13

    Intro

    du

    ctory sectio

    n

    Introduction Prof. Bruce Keogh Co-ordinator, UK Cardiac Surgical Register and National Adult Cardiac Surgical Database

    This is the most authoritative and comprehensive public documentation of national contemporary cardiac surgical practice available anywhere in the world.

    It is authoritative because it has been primarily written by surgeons for surgeons, but in such a way that we believe will be of interest and value to non-medical readers. It is based on professionally collected data with surgically relevant analyses determined by constructive feedback, comment and support from the whole cardiothoracic surgical community in the UK and Ireland. We hope that this authority is strengthened by an open and transparent approach, which acknowledges the flaws and deficiencies in the data and the limitations this imposes on both the analyses and their interpretation.

    It is comprehensive because it includes:

    • Data from over 210,000 individual patient records.

    • Data capture from 100% of NHS centres from England, Scotland, Wales and Northern Ireland.

    • Increasing contributions from the Republic of Ireland and the private sector.

    This fifth edition builds on previous editions with some new innovations:

    • Long-term survival data with tracking through the Central Cardiac Audit Database (CCAD).

    • New data from the UK Heart Valve Registry.

    • 25-year trend data on activity and hospital mortality for different types of operations.

    • 10-year trend data on risk factors.

    • Presentation of volume and outcome graphs for both institutions and surgeons.

    • Individual named surgeons’ reports.

    We continue to publish unadjusted mortality data for coronary bypass and aortic valve replacement surgery for every NHS unit in the UK. This year the unit results are supplemented with an analysis of the results of coronary bypass surgery for every consultant surgeon within the UK. For this we have used crude mortality, unadjusted for casemix, because our data is not yet good enough to apply casemix adjustment to every surgeon’s practice. We also have some concerns around the accuracy of risk adjustment statistics when applied to small individual practices. Nevertheless, we are the first mainstream specialty to have achieved consensus on how to scrutinise individual surgeons’ results, and the first to publish how those results relate to transparent standards. This represents a level of professional maturity, which should be reassuring to our patients.

    However, our patients should also be reassured by other information presented in the book. Our national results compare very favourably with international standards and continue to improve. As our patients become older so our results for elderly patients improve (see pages 90-91), as we treat more people with diabetes so the results for diabetic patients improve (see pages 148-149) and as we increasingly understand how to protect the heart itself from the insult of major surgery so the results of surgery for patients with very weak hearts improve (see page 117).

    We are now working with the Healthcare Commission, the Central Cardiac Audit Database and the Heart Team in England to develop a robust mechanism for data validation (see pages 31-32). The data feels right, but tighter focus on data quality is essential if we are to engage in safe and sometimes sensitive comparative analyses (see pages 264-309).

    The importance of accurate data was highlighted by the Bristol Royal Infirmary Inquiry Report 2 and the Government’s response 3. The SCTS concurs with this important principle as reported in The Times 4, the BMJ 1, 5 6 Heart 7 and in a comprehensive report of a joint study with the Nuffield Trust, the RAND organisation and the California Office of Statewide Health Planning and Development 8, 9 in which the methodology was defined. Clearly the level of validation will determine how the data can be used.

    The publication of this book and previous reports reflects considerable effort, enthusiasm and commitment from a cohesive body of clinicians and managers in a successful collaboration with Dendrite Clinical Systems who have provided the data management expertise. This report demonstrates that where vision and determination exist it is possible to collect, merge, analyse and present national information for a whole specialty. Such information is the essential bedrock for ensuring safe and effective clinical governance and improving outcomes for our patients.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    14

    Intr

    od

    uct

    ory

    sec

    tio

    n

    Perspective from the English Department of Health Dr Roger Boyle National Director for Heart Disease, England

    The last report from the National Audit Cardiac Surgical Database showed quite clearly how the Society of Cardiothoracic Surgeons was leading the way in making clinical outcome data available to the general public. This report shows for the first time, tables that identify individual surgeons in a fashion that allows patients to be reassured that their surgeon meets the professional standards set by the Society while, at the same time, avoids the pitfalls of year-on-year fluctuations in outcome data. This is vitally important since, in the past, where concerns have been raised about individual centres and surgeons, the solutions have been found not in problems with individuals but in the management of the whole system in the complex settings that cardiac centres have become.

    The last 12 months has seen further major progress towards improving cardiac services in England. Major capital schemes are underway across the land to improve the clinical setting in which our patients are to be treated and waiting times have continued to improve. No patient waits more than 6 months now for coronary bypass surgery and we are on target to achieve a maximum wait of 3 months by April 2005. At the time of writing there are, on average, just 19 patients waiting for each of our consultant cardiac surgeons.

    Another point that is important to highlight is the extremely high standard of care that is being offered in this country. Despite ever-increasing complexity of case-mix, the mortality rates here compare favourably with those anywhere in the world. This is a tribute to the high standards of training being made available in Britain and to the commitment to quality that the Society of Cardiothoracic Surgeons promulgates.

    There are challenges ahead. More and more cardiologists are finding ways of treating cases by angioplasty leaving a much more complex series of patients for surgeons to tackle. There will be inevitable effects on outcomes as well as length of stay in ICU and in hospital overall. These changes will need to be factored in as we move to the collection of risk stratified data into the servers of the Central Cardiac Audit Database (CCAD). This developing system already contains some 500,000 episodes of care, mostly admissions with heart attack but increasingly admissions for bypass surgery and angioplasty. Soon it will be possible to track not just the mortality rates after an event but also event-free survival, probably a much better measure of quality.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    15

    Intro

    du

    ctory sectio

    n

    The value of national data collection: a Scottish perspective Dr Jill Pell Consultant in Public Health, Greater Glasgow NHS Board Chairman of the Scottish Cardiac Registers Steering Committees

    Since 1996, we have collected data prospectively on all adult cardiac surgery procedures performed in Scottish NHS hospitals. The following year, we set up a second register to collect data on all percutaneous coronary interventions performed in Scottish NHS hospitals. In 2000, we expanded the interventional cardiology register to include diagnostic catheterisation and, in the last year, we have started to receive data from private hospitals.

    We have separate steering committees for the cardiac surgery and cardiology registers. However, we have long recognised that the distinction is relatively meaningless in the clinical setting where patients may be considered for or undergo both surgery and percutaneous intervention. Therefore, since 1997, the cardiac surgeons have distributed their annual report to interventional cardiologists and vice versa. This year we took a further step towards full integration by producing our first annual report to cover cardiac surgery, percutaneous intervention and coronary angiography.

    In Scotland, we are fortunate that data on all admissions to hospitals have been collected routinely since 1981. Since the inception of our cardiac registers, we have linked our data to the national hospital admission database and death certificate data to provide follow-up information on events such as myocardial infarction, readmission to hospital, re-investigation, repeat procedures and survival. As a result, we have been able to monitor trends in outcome over time, by procedure and between different sub-groups of patients. We have been able to demonstrate, for example, that the well-recognised improvements in in-hospital mortality following coronary artery bypass grafting have translated into statistically significant improvements in longer-term survival (see figure below).

    There are a number of benefits to having a Scottish register, such as linkage to routine data sources, fostering local ownership and therefore improvements in data quality, and ensuring that decisions on service developments are evidence-based. However, Scottish clinicians are also aware of the importance of contributing data to the UK registers, such as the Society of Cardiothoracic Surgeons’ database, in order to provide better benchmarking for use in clinical audit and governance. Also, there is much to be gained from sharing ideas and experience of how best to analyse, present and disseminate data. Therefore, as in previous years, the Scottish cardiac registers are pleased to contribute to this important report.

    Isolated CABG: Actuarial survival curves for Scotland (n=8,181)

    1997-1998 1999-2000 2001-2002

    80%

    82%

    84%

    86%

    88%

    90%

    92%

    94%

    96%

    98%

    100%

    0 200 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000

    Time after operation / days

    Surv

    ival

    rat

    e

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    16

    Intr

    od

    uct

    ory

    sec

    tio

    n

    A cautionary note on the interpretation of surgical results Prof. Tom Treasure, Guy’s and St Thomas’ Hospital, London

    Simplistic data may disadvantage patients as well as surgeons

    UK cardiac surgeons are entitled to feel proud of the quantity and quality of data that are available and the innovative ways in which our data are presented. The UK Cardiac Surgical Register was a professional initiative started years before the buzzwords clinical audit, the quality agenda and clinical governance were coined 10 and a very long time before CHI, CHAI and NICE were invented. The objectives of the founding fathers of the UK Register were clear: to improve standards. Without the need to pillory anyone it was possible for us all to see the sort of results that were being achieved nationally and to judge whether results in our own units were about the same, better, or worse. It was not the way of things to make public exposés but it would be a mistake to think that the hospitals of Britain were populated by gung-ho surgeons who would carry on regardless, putting lives at risk. Not all the thoracic surgeons of the 1950s made the successful transition to open heart surgery. Not all of the cardiothoracic surgeons of the sixties and seventies took to coronary surgery. In the 1980s it became much less acceptable to do a few congenital operations in children amongst a largely adult practice. Many unpublicised decisions were made on the basis of reflective reading of the annual reports. Some surgeons made these decisions for themselves; others reached a policy in discussion with colleagues; and some had the decision made for them because the clinical referrals were deliberately steered in another direction. Local policies were informed by the data provided in the Register, which we surgeons would seize and scrutinise as it came through the letterbox each year. That unpublicised but quietly effective process is not well known and not remembered. Unfortunately it did not always work that way. What will be remembered are the instances where that process of introspection has let us down and surgeons have had to be publicly forced to stop.

    During what I like to refer to as the Bristol show trial some credit was given to cardiac surgeons for collecting data, but we were criticised for making them anonymous. Now the identity and reputation of the surgeon is on the line at the outset. The problem with legislation designed to stop the same thing happening again is that the rules may change practice in a way that was not anticipated. A well-known example is the legislation following the scandalous behaviour of the London landlord of the early 1960s, Rachman (who has given his name to the word Rachmanism, listed in the Oxford English Dictionary). New laws were drafted to protect poor tenants from an unscrupulous and, in that case, a brutal landlord. The law gave to the tenant such security that well intentioned private property owners willing to let a flat, a bed sitter, or the basement of a house now too large for them, would not take the chance. Instead of improving the situation for the tenants, rented property became scarce and more expensive.

    The rules under which we now must live and work are being made to stop Bristol happening again and there is no point beating about the bush for Bristol is named by everyone in politics, the press, and the profession. I fear we are already seeing the Rachman effect. As we use surgeon specific outcome data to try to improve things for the patients we actually make things worse. There is no doubt in the mind of any cardiologist that I have talked to that it is much harder to persuade a surgeon to take on the more risky coronary cases. There are patients who we predict have a slim chance of survival without surgery who willingly accept a 20% or 30% risk. Now let us look at that from the point of view of say a surgeon with a mixed cardiothoracic practice, other duties and three or four half day operating sessions a week and a share of the on call rota. Suppose the coronary operations number 50 in the year. Every death is 2%, three deaths in the year makes 6%, and that is the trigger for the cautionary word. Knowing the 95% confidence limits of 3/50 (1.25% to 16.6%) does not make the problem go away. The figures 1.25% to 16.6% span the best but also the very worst of contemporary outcomes. In journalists’ terms 6% is twice the national average mortality and that bald statement, without risk adjustment, is the one that the Medical Director and the Chief Executive may be given. It does not take a mathematician to tell us that it is a dumb use of statistics. These are the rules that were intended to protect patients from the imagined coterie of unskilled surgeons who are as arrogant as they are reckless. It is in the higher risk cases that we make the biggest impact on survival and yet the new rules may deprive these patients of care. We must count but we must use the numbers intelligently. Data sets should not supplant common sense.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    17

    Intro

    du

    ctory sectio

    n

    Background to the Society’s data-collection initiatives

    The Society of Cardiothoracic Surgeons has a 25-year history of voluntary data collection and analysis, which is described in detail below:

    Society of Cardiothoracic Surgeons Parallel Initiatives

    UK Cardiac Surgical Register See page 40 Established 1977 by Sir Terence English and Dr Alan Bailey

    Basic activity and outcome data of 250 categories of adult cardiac operations. Data received from 100% of UK NHS and all Republic of Ireland units. Aggregated national data fed back to each unit to allow comparisons of local results with national average. Discontinued in 2003 to encourage focus on the National Adult Cardiac Surgical Database. Since 1997 this has included individual surgeon’s results for coronary surgery. This will continue until the Central Cardiac Audit Database is fully functional and can take over this role. Funded by Society membership.

    UK Thoracic Surgical Register Established 1978 by Prof. Hugoe Matthews and Dr Alan Bailey. Currently administered by Prof. Tom Treasure

    Basic activity and outcome data on many categories of adult thoracic surgical procedures. Data from 100% of NHS cardiothoracic and thoracic surgical units. Funded by Society membership.

    UK Heart Valve Registry See page 60 Established 1986 by Professor Ken Taylor

    Initially funded by the English Department of Health but more recently by the Commission for Health Improvement and then the Healthcare Commission. Collects patient details on about 60-65% of all heart valve patients in the UK. Linkage to the Office of National Statistics (England, Scotland & Wales) and the Central Services Agency (Northern Ireland) provides long-term mortality tracking now approaching 20 years. Unique international resource. Funding withdrawn 2004.

    National Adult Cardiac Surgical Database See page 72 Established 1994 by Mr Bruce Keogh and Dr Peter Walton

    The basis of this report. 100% of NHS units contribute individual patient data. Compatible with American, European and other international initiatives.

    No external funding.

    Central Cardiac Audit Database See page 23 Pilot established 1996. Funded by the English Department of Health through the Healthcare Commission and administered by the NHS Information Authority Managed by Dr David Cunningham and Mr Martin Old

    Collects patient orientated data from 100% of UK paediatric cardiac surgical units and provides mortality tracking through the Office of National Statistics. Will become data collection route for all cardiac procedures and will incorporate the UK Heart Valve Registry and the National Adult Cardiac Surgical Database. Society will continue to merge and analyse Scottish, Welsh and Irish data for its members and the public.

    Funded by the Healthcare Commission in England.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    18

    Intr

    od

    uct

    ory

    sec

    tio

    n

    The history and principles of data collection by the Society

    Patients have a right to good surgery. They also have a right to informed consent so they can make the right decision. It is our responsibility to provide appropriate information so that all parties understand the basis on which a decision for surgery is made. Similarly, the purchasers with constrained budgets and difficult decisions will also benefit from quantitative information on the nature of the patient population being referred for and undergoing cardiac surgery. We cannot provide this information without robust local and national data.

    In principle the model for a successful national database is simple. There are two basic aspects. Firstly, each contributing unit should collect all variables in the agreed dataset for each and every patient. The data should be collected in a system allowing data export in a predefined format. Merging of the data centrally is then easily manageable. Secondly, the data should be robust and validated.

    Introduction to 1999-2000 Report

    The Society of Cardiothoracic Surgeons has a long history of national data collection. In 1977 Sir Terence English established the UK Cardiac Surgical Register, which collected activity and mortality data on all cardiac operations performed in NHS units in the UK. Although simple in concept, the process represented the first attempt in Britain by any speciality to collect national activity and outcome data. These data were unit specific, but were not in any way risk-stratified which limited its value for statistically valid outcome comparisons.

    In 1986 Professor Ken Taylor and Mr John Dark (senior) set up the UK Heart Valve Registry. This was a unique visionary step forward. It was different in several ways.

    • Firstly, data on heart valve implantation was to be collected not by surgeons, but by an independent nominated person working in the cardiac surgical operating theatres in each unit.

    • Secondly, the form requesting information, including patient identifiers was no longer than one side of A4 paper

    • Thirdly, using the patient identifiers the data was linked to the Office of National Statistics so the patients could be flagged up on the system when they died. This provided the basis for the first long-term follow up of valve surgery patients in the world

    • Finally, the system allowed identification of re-operations This provided us with a unique resource, which has provided insight into a number of aspects of valve surgery and patient survival.

    In 1994 the Society of Cardiothoracic Surgeons recognised the need to collect comprehensive data on patients undergoing cardiac and thoracic surgery and supported a proposal to develop a patient orientated National Adult Cardiac Surgical Database to replace the UK Cardiac Surgical Register over a ten year timeframe starting from 1996. At that time there were signs of a move away from simple, aggregated data to patient-by-patient data. This move had been driven by the fact that aggregated data could not be used to give the same detailed analysis options that patient-specific data would afford. The current dataset includes demographic data, procedural data and outcome data for each patient. The reasons for collecting more comprehensive data were becoming increasingly clear. Firstly, there was growing public and political interest on both sides of the Atlantic in cardiac surgical outcomes. In the United States there was clear evidence that misconception of operative risks was a leading cause of malpractice actions. Secondly, the Society of Thoracic Surgeons in the United States had shown that hospitals, cardiologists and in some cases surgeons, were ignorant of the changing patterns of patient populations undergoing coronary surgery and that this perpetuated the professional and public misconception that coronary surgery carried little or no risk. Finally, the release of raw mortality data on Medicare patients into the public domain in the late 1980s by the Healthcare Financing Administration without respect for any of the known patient-specific risk factors or co-morbidities created considerable concern within the cardiothoracic community. Then in 1990 Newsday successfully sued the New York Department of Health under the Freedom of Information Law to release risk adjusted surgeon specific outcome data starting the trend towards the public reporting of surgical outcomes.

    Contemporaneously in the UK the purchaser – provider split in healthcare provision provided some parallels with developments in the United States. The similarities indicated that if we were to succeed in the new

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    19

    Intro

    du

    ctory sectio

    n

    healthcare market we would need to understand the nature of our own patient population in the United Kingdom and have a grasp of surgical outcomes for patients with differing severities of illness. The Society of Thoracic Surgeons in the US had responded by establishing a National Cardiothoracic Surgical Database with an agreed comprehensive dataset in 1989-1990. It seemed sensible that we should not re-invent the wheel and we therefore agreed, through the Standards of Care Committee and the Executive Committee, a reduced version of the US dataset, which was broadly compatible with existing databases in the UK.

    Recognition of the requirement for an agreed dataset and national database has been slow in some quarters. However, momentum has grown recently, in part as a response to events in Bristol, but mainly as surgeons and healthcare administrators recognise the value of comprehensive data collection for understanding severity of illness, resource allocation and outcome analysis. This enthusiasm has been complemented by the government’s initiative on clinical governance, which further emphasises the surgeon’s duty to understand the quality of care he is providing.

    This report demonstrates significant progress in this endeavour. We have clearly shown that electronic downloading of anonymous patient specific data from centres that collect such data is feasible. However, there is much to do and enormous opportunities remain. The issues that we must focus our attention on are the same issues that were outlined in the 1999-2000 Report:

    • Standardising data collection. • Ensuring data collection from all units. • Developing and understanding the strengths and limitations of methods for measuring and

    comparing differences in casemix.

    • Recognising the difference between hospital mortality and quality. • Producing sensitive quality indicators to help improve the quality of hospital care. • Tracking long-term follow-up of patients. • Building a national resource for the Society and its members and patients and their families.

    Standardising data collection

    There are still considerable differences in the datasets collected by different units. These differences need to be ironed out if we are to realise the potential benefits of the National Adult Cardiac Surgical Database project. In England and Wales we have had considerable support in this endeavour through the National Service Framework for Coronary Heart Disease (NSF) launched in March 1999 and the CCAD project. The NSF document outlines the requirements for hospital audits, one of which is risk adjusted mortality following coronary artery bypass surgery. This is supported further by the inclusion in the NSF document (Chapter 5, pages 24–40) of the Society’s recommended dataset, which is designed to facilitate stratification for casemix. Furthermore, the NSF imposes a timeframe for collection of this data. Milestone 3 states that:

    … by April 2002, in every local network of cardiac care hospitals … should have clinical audit data no more than 12 months old that describes the relevant audit criteria. Where relevant these data are derived from participation in national audits.

    Unfortunately the NSF recommendations do not apply to the whole of the UK, so national differences in datasets may persist. This can be accommodated by a mapping process, termed correspondence, but a common dataset would ultimately bring more value to all participants.

    The recommended dataset was agreed by the Society Executive in 1996. It was modified slightly in September 2002 with a view to implementation of the changes in April 2003. It has been designed not only to allow stratification of patients according to severity of illness using most current algorithms, but also to track changes in surgical practice, explore specific aspects of surgical training and to examine outcomes. Although a smaller dataset would be easier to collect, it would be of less and less value in the longer term. Relatively simple analyses of the data in the National Adult Cardiac Surgical Database have demonstrated that there are interesting and significant trends in various risk factors over time. Not all of these trends could have been predicted when the project was first launched, and the exclusion of one or more risk factors might have meant that these trends would have remained permanently hidden (see changing risk factors, page 211).

    It would also mean that the UK data would be less compatible with parallel, evolving international initiatives, with the implication that useful international comparisons might not be possible. The current dataset is reasonably comparable to the new Society of Thoracic Surgeons (USA) dataset both in terms of size and detail. It is also compatible with the European and US cardiac surgical databases, and the well-established UK Heart Valve Registry. As such, the current dataset should serve to facilitate good clinical audit and governance within each unit and provide comparative national data for benchmarking, through this report.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    20

    Intr

    od

    uct

    ory

    sec

    tio

    n

    Any central database is only as reliable as its component parts. Therefore, data validation is a key issue for all of the people who use the data. This will include surgeons who wish to compare their local populations against a national standard and benchmark their results against their peers, but also the patients and relatives of patients who wish to glean information pertinent to their forthcoming operations. Data could be validated at any one of several points in the process, but there is no substitute for good, regular internal review and validation. Without an individual dedicated to those tasks it is unlikely, if not impossible, to accumulate reliable data. These internal review processes must be subject to external scrutiny, in the same way that all commercial financial accounts are independently audited. The value of this was highlighted in a study conducted with the Nuffield Trust, the Rand Organisation and the California Department of Health, which has been published elsewhere 5, 9.

    The long-term success of this database project is dependent on development and co-operation at several levels.

    • At a local level each unit should collect the Society’s recommended dataset on all patients who undergo cardiac surgery.

    • The data should be collected and maintained in a computer-based system that allows for an electronic export of the data, in a format that the Society has defined in conjunction with the CCAD. The definitions of the data fields and the export specifications are available on both the Society and CCAD websites.

    Separate databases kept by individual surgeons are not helpful. Such an approach makes downloading of data and validation more difficult and time consuming and suggests a lack of commitment to integrated audit within the unit.

    Ensuring data collection from all units

    This has now been achieved. What we have not achieved is consistent quality of data from all units.

    This will take a couple of years to resolve. It will be helped by the reminder of the obligation to collect this data in a letter from the Minister of Health, Chairman of the Academy of Medical Royal Colleges and the President of our Society coupled with the inclusion in the Healthcare Commission’s performance ratings of coronary surgery outcomes and thrombolysis data collected for the Royal College of Physicians Myocardial Infarction National Audit Project.

    Understanding the value of risk models

    A variety of scoring systems are available for measuring variations in severity of illness. Although our specialty leads the way in risk stratification modelling there is increasing recognition that no scoring system will ever be completely predictive of outcome, particularly in high risk patients, for three reasons. Firstly, we do not yet fully understand the basis of the pathophysiological response to surgery or factors influencing an individual patient's physiological reserve. Secondly, some of the major risk factors are not easily quantifiable or definable and are therefore omitted from most scoring systems. A typical example would be the state of the coronary arteries. To quote Parsonnet 11, What may be identified as severe and diffuse disease by one surgeon may be considered relatively routine and non-intimidating by another. Thirdly, some high risk patients may be difficult to characterise and the statistical denominators are relatively small. Nevertheless, both the EuroSCORE and Bayes models provide a useful yardstick when examining mortality in groups of patients. The Parsonnet system, which is still often quoted, is becoming outdated. Current UK practice results in a mortality of around a third of that predicted by the Parsonnet score, but this will decrease with time - in part because practice is improving and in part because the weighting of the preoperative risk variables changes with time. Indeed, the essence of responsible surgical audit is to understand and attack the most influential risk factors in order to reduce their impact. As a result there is accumulating evidence that the influence of previously important risk factors is being reduced towards the mean (see page 211). Taken to a natural conclusion, the currently used variables will be of reduced value for risk stratification in a decade’s time. So we must get smarter. The variables that we measure now (gender, diabetes, hypertension etc.) are simply an expression of our genes, which are our underlying blueprint. Risk stratification will move towards the inclusions of clever molecular and genetic analyses of what influences a patient’s ability to withstand pathophysiological turmoil inflicted by surgical stress.

    Effective, conventional risk modelling must be an iterative process; it is this that has prompted some groups to explore alternative, more contemporary and locally appropriate scoring system. We should exercise caution in the construction and interpretation of local risk models and we must understand the balance between an accurate and a useful risk model. An accurate risk model accurately predicts an outcome. At one extreme an accurate local risk model could be constructed using data from a single individual or unit with poor results, reliably predict those results, and then be used to lull the participants into a false sense of security. Although accurate and helpful for predicting local outcome, such a model would have limited value for comparative purposes. A genuinely useful risk model is one that reliably predicts an outcome but is based on, and applicable to, a wider constituency, thereby facilitating reasonable and meaningful

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    21

    Intro

    du

    ctory sectio

    n

    comparisons between individuals or units. Local risk modelling has value in that it caters for immeasurable local influences, which may not pertain elsewhere. Thus a hierarchy of risk models at different levels, based on single units, several units in a region, along with a nationally based model, should allow accurate local analysis and comparisons while also providing the substrate for dissecting and understanding local variations in the process of surgical care. But again, to be of value, all models should rely on the same definitions for their variables.

    In-hospital mortality and quality of care

    The definition of quality in surgery is difficult and was first considered by Florence Nightingale in the UK and Ernst Codman in the USA, but both were ahead of their time and were to some extent ostracised for their efforts, which were perceived by some as threatening 12. The issue of how to measure quality remains perplexing, but it is an issue, which we must grasp if we are to retain the respect and confidence of those we seek to treat. The concept of quality in cardiac surgery should encompass the whole hospital journey, from the time the patient walks through the front door for his or her preliminary assessment to the time of discharge from the post surgical outpatient clinic and beyond. Individual surgical performance constitutes only a small, albeit important, part of this process. Surgeon specific mortality data collected by the Society of Cardiothoracic Surgeons clearly indicates that surgeons in the same unit tend to have similar mortalities, highlighting the importance of additional local influences on surgical outcomes. Such influences may include the socio-economic status of the catchment area, severity of cardiac illness, prevalence of co-morbidities, threshold of referral from both the general practitioner and the cardiologist, threshold of acceptance by the surgeons, standards of anaesthesia, surgery, and intensive care, adequacy of facilities and staffing levels, attitude to training, interpersonal relationships between staff, and architectural dispersion within the unit. Any one of these can influence surgical outcome and it is clear that hospital mortality is not necessarily a measure of overall quality but is simply a rough guide to the success of the surgical episode. An appreciation of quality begins when the patient has recovered from the trauma of surgery and asks, was it all worth it? If the answer is yes, then added value can be measured by how long the answer remains yes. A favourable answer to these questions depends as much, if not more, on appropriate referral and acceptance as it does on the technical quality of surgery and postoperative management.

    Innovative quality indicators

    Mortality is a very coarse measure of outcome and quality. It has the advantage of being easy to understand, measure and validate. But to the patient other outcomes, such as a stroke, may feel as bad or even worse. In order to spot changes or differences in practices analysis of other outcomes may be more sensitive. For example, re-opening for bleeding, insertion of an intra-aortic balloon pump, prolonged ventilation, or a new requirement for dialysis may, in a patient who survives, be deemed to represent a near miss. Measuring a constellation of such outcomes may provide a more sensitive measure of quality of practice. The downside is that early intervention, such as return to theatre for bleeding or insertion of an intra-aortic balloon pump, may be avoided to try and reduce the number of near misses. This issue is discussed by Geoff Berg on pages 260-261.

    Public reporting of outcome data

    Public reporting of outcome data remains a sensitive issue, but times and attitudes are changing. We must prepare ourselves for the abolition of anonymity and the public reporting of surgeon specific risk-stratified outcome data. Our specialty will lead the way, but only when we have a robust system that can withstand scrutiny. There are many talented individuals within our Society and I am sure that we will be able to develop a system of detailed reporting that is agreeable to surgeons, patients, politicians and the media alike.

    In this edition we take a major step forward and cross the Rubicon by publishing a list of surgeons’ names and indicate how many coronary operations they perform and whether their results meet standards agreed by the Society of Cardiothoracic Surgeons and the Department of Health. There will be those within and without the specialty who will be critical for different reasons. They don’t like the concept, they don’t like the methodology or they just think we haven’t gone far enough. We ask those potential detractors to be temperate in the expression of their comment and to bear in mind that the shift toward publishing data on individual surgeons is entirely voluntary and driven by a fragile professional consensus, without any external funding or support. If criticism is voiced in an unconstructive way the delicate balance between anxiety and the desire to be seen to be moving forward will be destabilised and the good will and trust of honest surgeons might be shattered.

    This is a complicated area, but the generic issues have been succinctly summarised in a Nuffield Trust publication Dying to know: Public release of information about quality of healthcare. This report concludes with some policy recommendations based on the review of US experience of public disclosure. They were presented to Mr Frank Dobson, the then Secretary of State for Health, in July 1999 and circulated to key policy makers and health advisers in the UK and are outlined below:

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    22

    Intr

    od

    uct

    ory

    sec

    tio

    n

    1. The intended purpose or purposes of public disclosure should be made clear to all stakeholders.

    2. Public disclosure should be seen as an evolutionary process, becoming progressively more sophisticated and comprehensive over time.

    3. Public disclosure should be seen as one component of clinical governance.

    4. Provider organisations should be a key audience for information about performance.

    5. The financial cost of implementing a national policy on public disclosure is likely to be significant and should be considered alongside the benefits.

    6. Specific educational initiatives for target audiences should be implemented alongside public disclosure.

    7. Health professionals and their representative bodies should be fully involved in the process of public disclosure.

    8. Both process and outcome measures of quality should be published.

    9. Outcome indicators must be risk adjusted.

    10. Public disclosure should be accompanied by a strategy for monitoring the benefits and unintended consequences.

    11. Public disclosure should be accompanied by possible explanations for the variations reported.

    12. A research and development programme focusing on the generation and evaluation of public performance data should be supported by the NHS R&D Directorate.

    Building a national resource

    The success of any collaborative venture depends on co-ownership by all parties. This database has grown as a result of the efforts of many contributing surgeons. So the database should be available and useful to those who have contributed. We have therefore agreed a mechanism for data release and publication. Firstly all graphs in this report are downloadable from www.scts.org in Acrobat™ format. Secondly, contributors may seek additional data for presentation or publication. Application forms and rules for data release, determined by the Society’s Executive Committee, are also available on the Society’s website.

    The Society of Cardiothoracic Surgeons continues to lead the medical and surgical specialties in the UK in the arena of data collection and outcome measures. This is not just fortuitous; it is the result of the vision, effort, generosity and openness of a large number of surgeons and allied professionals. The production of a report of this nature represents a huge commitment of effort at many levels. Although we have received considerable encouragement from a number of quarters this has stopped short of any form of direct financial assistance and we therefore thank Dendrite Clinical Systems for helping us initiate, develop and sustain this important endeavour.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    23

    Intro

    du

    ctory sectio

    n

    The Central Cardiac Audit Database

    In the early 1990s the Department of Health was inundated with requests for funding cardiac related audit projects. They therefore established a working group to look at ways cardiac data collection could be co-ordinated between interested parties. This group recommended that the datasets of interested specialist groups should be harmonised and mechanisms for on-line, real time data submission and viewing should be explored. The vision was that patient data would be entered at the time of a procedure and passed down the line in an encrypted format to a Central Cardiac Audit Database (CCAD). The patient’s progress would then be tracked through the database so that additional procedures would be flagged and mortality verified through the Office of National Statistics. This would provide accurate information on re-interventions (which might be performed in another hospital) and mortality allowing a greater understanding of natural disease processes and the impact of interventions. Thus the central server would hold merged data from contributing centres, which would be visible to contributors and provide benchmarking measures. These data would be available to the participating specialist associations for analysis and the production of specialty related reports. A pilot project with six centres was established in 1995 and reported in 2000. Reliable data collection was not achieved, mainly for technical reasons related to data transmission, but a number of lessons were learned.

    The project has since been placed under the umbrella of the National Clinical Audit Support Programme within the NHS Information Authority. The National Clinical Audit Support Programme is responsible for promoting clinical audit in cancer and coronary heart disease; it is currently directed by Mr Martin Old.

    CCAD has several modules. The congenital heart disease module collects and links data from cardiological and surgical interventions in children with heart disease. The Myocardial Infarction National Audit Project (MINAP) collects data on patients admitted to over 250 trusts in England and is administered on behalf of the Royal College of Physicians by the CCAD. The Society has now agreed to use the CCAD as the main route of data collection from local databases for the National Adult Cardiac Surgical Database in England and Wales.

    The CCAD will provide an internationally unique opportunity for understanding the impact of both surgical and non-surgical interventions for acquired heart disease. Its strength lies in the linkage with the Office of National Statistics for long-term mortality tracking which will help us understand who will benefit most from which intervention. Herein lies the added value over and above existing initiatives in this country and abroad.

    However, a process of this nature also brings dilemmas. If it is to retain the confidence of the public and the media there must be some element of independence from clinicians. Equally if it is to retain the confidence of clinicians, whose performance may be judged through such a system, it must be guided by clinicians, a principle agreed by the Secretary of State and Chief Medical Officer. Failure to observe this strategy will result in continued and unnecessary duplicate data collection with different groups expressing varying degrees of confidence in different initiatives.

    With responsibility for funding shifting to the Healthcare Commission, this principle has been enshrined in the tripartite governance arrangements between the Society, the Commission and the Department of Health (see page 321).

    In the meantime, the Society must continue with existing initiatives, which cover the whole of the United Kingdom and Ireland.

    Harmonising datasets for the CCAD

    There have been three forces at work to encourage us to make a few quite significant changes to the 1996 dataset. These changes were made at a meeting of clinical audit leads at a meeting in Birmingham in September 2002, with a view to implementation from April 2003.

    The first has been an international influence. The original dataset was a hybrid of existing datasets in the UK combined with the original dataset compiled by the Society of Thoracic Surgeons (STS) in the US. Since then the United States dataset has been modified based on their extensive experience. The new STS dataset has been adopted widely across the globe. In particular, it has been adopted by the European Association for Cardio-Thoracic Surgery as the European standard. It has also been adopted in Japan, Australasia, Canada and several South American countries. With this in mind we have modified the dataset to increase compatibility with our international colleagues, which will facilitate international comparisons. A key example of this is the development of the EuroSCORE risk stratification system, which is being widely adopted across Europe and which requires pre-operative variables not included in the previous SCTS, or current STS, datasets.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    24

    Intr

    od

    uct

    ory

    sec

    tio

    n

    The second influence has been data quality. Whilst the collection of pre-operative data has steadily improved, the collection of post-operative complication and morbidity data has been pretty poor. This is related to the difficulty of collecting information after the event and problems with defining some outcomes. So we are now down to four outcomes: in-hospital death, stroke, new dialysis and re-operations, all of which are not only relevant but relatively easy to define. This should make the data collection much simpler.

    The third substantial influence has come from a variety of national NHS information initiatives. The strategic aim within the NHS is to ensure that information can be transferred seamlessly, where properly authorised, across sectors. So, within this strategy the CHD Information Strategy in England aims to integrate a variety of information streams relating to the treatment of heart disease. These streams range from chest pain clinics and myocardial infarction information to waiting times for angiography and outcomes from surgery or angioplasty. In order to make this information useful to clinicians we have had to harmonise our old existing dataset with that of the more recent Myocardial Infarction National Audit Project (MINAP) and that of the British Cardiovascular Intervention Society and also the longstanding UK Heart Valve Registry. Although this will be painful for some centres, in the longer term it will facilitate patient tracking within the CCAD and outcomes analysis across the board.

    With greater reliance on electronic data transfer it is increasingly important that uniform clinical and technical standards are enforced across the NHS. In order to enable this, the following principles apply:

    • Where possible, existing standards will be adopted. New standards will be developed only where there are none already available, and where there is a defined business need.

    • Standards must be implementable, and therefore need to be tested in use before adoption, and must be agreed only where there are measures which can be used to test for conformance.

    • Higher authorities will normally take precedence. In other words, the standards set out in the Government Interoperability Framework, and the various e-Government guidelines will be adopted in preference to any health specific standards. Where there is a need for health specific standards, then established international standards will be adopted rather than developing new ones.

    To implement these principles four national Standards Boards have been established:

    • The Information Standards Board (ISB) will oversee the whole standards process, setting priority areas for standardisation and approving proposals. The ISB is accountable to the National Information Policy Board.

    • The Clinical Data Standards Board is responsible for agreeing clinical standards, in areas such as terms (e.g. SNOMED Clinical Terms) and clinical messages. They will be most concerned with the use of standards on the care of individual patients/clients and the human behavioural implications.

    • The Technical Standards Board is responsible for agreeing technical standards to be applied. • The Management Information Standards Board is responsible for agreeing dataset and

    other performance assessment data requirements. They are most concerned with the organisational and population aspects of the use of standards.

    Our new dataset as part of the harmonised CCAD dataset, will, where necessary, be passed through the appropriate approval process, but we will not allow this to hinder the progress of data submission to CCAD.

    From the Society’s professional standpoint the international considerations are the most important. We represent England, Scotland, Wales and the whole of Ireland and we therefore look to standardisation of the dataset across all these countries.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    25

    Intro

    du

    ctory sectio

    n

    Difficulties in data collection especially in adult congenital heart disease

    Mr Leslie Hamilton Consultant Paediatric and Adult Cardiac Surgeon, Freeman Hospital, Newcastle-upon-Tyne

    The challenges to collection of high-quality data on cardiac surgical patients are well known. However, patients with congenital heart disease pose particular problems. A major difficulty has been a lack of international agreement on what to call individual congenital heart defects meaning that coding systems could not be developed. From the surgical viewpoint this was finally overcome in 2000 when the proceedings of the International and Nomenclature Database Conferences for Paediatric Cardiac Surgery were published 13.

    A further difficulty is that congenital heart defects are in fact not isolated entities but part of a spectrum – for example Tetralogy merges into double outlet right ventricle which merges into transposition with ventricular septal defect. Thus it is sometimes difficult to classify accurately the diagnosis of an individual patient.

    Operative survival rates for children undergoing surgery for congenital heart defects in early childhood have improved over the years and we are now seeing an increasing number of adults with congenital heart disease. In addition, some congenital heart defects (e.g.s, atrial septal defect and coarctation) only become apparent in adulthood. Thus the population of adults with congenital heart disease coming for surgery are in two groups – those who have had previous surgery in childhood and require re-operation for a variety of reasons and those adults presenting de novo. The former group are usually dealt with by the paediatric cardiac surgeon who undertook the original surgery, whereas the latter group often have their investigations undertaken by adult cardiologists and are then referred to adult surgeons who in earlier years had training in congenital heart disease.

    This dichotomy poses a particular problem in data collection for adults with congenital heart disease. Those undergoing surgery in the adult practice of a paediatric cardiac unit will have their operations recorded in the National Paediatric and Congenital Database (CCAD – Central Cardiac Audit Database). Adult patients undergoing surgery in adult cardiac units (who are not connected to CCAD) have their data recorded in the Society of Cardiothoracic Surgeons National Database. Our challenge for the future is to merge these two databases into a single high-quality source of information on adults undergoing surgery for congenital heart disease.

  • The Society of Cardiothoracic Surgeons of Great Britain and Ireland National Adult Cardiac Surgical Database Report 2003

    26

    Intr

    od

    uct

    ory

    sec

    tio

    n

  • The So