Rapid Responses to:

RESEARCH: Paul Aylin, Alex Bottle, and Azeem Majeed Use of administrative data or clinical databases as predictors of risk of death in hospital: comparison of models BMJ 2007; 334: 1044 [Abstract] [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

In-patient death is only one of many possible performance measures

Chris P. Gale   (23 May 2007)

SHFA- value for money

Damien G Reid, Alberto Gregori, and Diana Beard   (30 May 2007)

Benchmarking with administrative or clinical databases: Serious pitfalls.

Alain Braillon   (1 June 2007)

In-patient death is only one of many possible performance measures 23 May 2007
Chris P. Gale, Academic Clinical Lecturer in Cardiology Academic Unit of Cardiovascular Medicine,Leeds Institute of Genetics Health and Therapeutics,LS2 9JT Send response to journal: Re: In-patient death is only one of many possible performance measures	Although I do not disagree with Aylin’s findings that routinely collected administrative data may be used to predict mortality risk with similar discrimination to clinical databases, one must consider the potential implications of abandoning clinical databases for administrative ones. Firstly, the models by Aylin et al only evaluated in-patient death. Mortality after discharge from hospital is also important. Surely this was possible with the advent of HES / ONS linkage? Clinical databases with linkage to ONS are able to track mortality. Secondly, death is only one of many possible measures of performance. That is, it is still possible to perform badly and have hospital survivors. Alternative measures of performance are pertinent to the audit of the management of patients with acute coronary syndromes (ACS). ACS guidelines (developed from randomised trials that provide evidence for the use of key interventions in patients with ACS) offer recommendations of ‘best care’ for patients with ACS 1-3 and are supported by the National Service Framework (NSF) standards of care for patient with coronary heart disease (CHD). Clinical databases such as the Myocardial Infarction National Audit Project (MINAP) allow improvements in hospital performance to be clearly documented.4;5 MINAP collects information on patients admitted with ACS to all acute hospitals in England and Wales (n=228) and monitors audit standards established by the NSF for CHD.6-9 By means of 108 data fields, each patient entry offers details of the patient journey, and (if applicable) date of death (from linkage to ONS). Indeed, through MINAP it is possible to demonstrate variations in performance over and above that of mortality (such as the attainment of pre-specified thrombolysis time targets or proportion of patients discharged on appropriate secondary prophylaxis).10 As such, the MINAP database is central to the evaluation of the management of ACS in England and Wales and in tern is pivotal in the appraisal of good medical practice. Finally, the rejection of clinical databases will remove the ability to perform contemporary ‘real world’ research. Only through research can we test hypotheses, develop superior performance indicators, and produce practices of care that improve the management and outcomes of our patients. Reference List 1. Antman EM, Anbe DT, Armstrong PW, Bates ER, Green LA, Hand M et al. ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction--executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1999 Guidelines for the Management of Patients With Acute Myocardial Infarction). Circulation 2004;110:588-636. 2. Braunwald E, Antman EM, Beasley JW, Califf RM, Cheitlin MD, Hochman JS et al. ACC/AHA guideline update for the management of patients with unstable angina and non-ST-segment elevation myocardial infarction-- 2002: summary article: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on the Management of Patients With Unstable Angina). Circulation 2002;106:1893-900. 3. Bertrand ME, Simoons ML, Fox KA, Wallentin LC, Hamm CW, McFadden E et al. Management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J 2002;23:1809-40. 4. Department of Health. The national service framework for coronary heart disease. Delivering better heart services. Progress report: 2003. London, DOH. 2003. 5. Birkhead JS, Walker L, Pearson M, Weston C, Cunningham AD, Rickards AF. Improving care for patients with acute coronary syndromes: initial results from the National Audit of Myocardial Infarction Project (MINAP). Heart 2004;90:1004-9. 6. Birkhead JS. Responding to the requirements of the national service framework for coronary disease: a core data set for myocardial infarction. Heart 2000;84:116-7. 7. Department of Health. National service framework for coronary heart disease. Modern standards and service models. London, DOH . 2000. 8.http://www.rcplondon.ac.uk/college/ceeu/ceeu_ami_home.htm. 2006. 9. Birkhead, J., Pearson, M., and Norris, RM. The national audit of myocardial infarction: a new development in the audit process. Journal of Clinical Excellence 4, 379-385. 2002. 10. Gale CP, Roberts AP, Batin PD, Hall AS. Funnel plots, performance variation and the Myocardial Infarction National Audit Project 2003-2004. BMC.Cardiovasc.Disord. 2006;6:34. Competing interests: Member of the MINAP Academic Group
SHFA- value for money 30 May 2007
Damien G Reid, Chair SHFA, Consultant Geriatrician Hairmyres Hospital, East Kilbride, G75 8RG, Alberto Gregori, and Diana Beard Send response to journal: Re: SHFA- value for money	We write in response to the article by Aylin et al, in which a comparison is made between administrative data and clinical databases as predictors of death in hospital. The authors make reference to the per patient cost of the Scottish Hip Fracture Audit (SHFA). Also, in ‘This Week’ the editors headline the comparison of £60 per record for the SHFA database to £1 per record for routine hospital episode statistics. We consider this comparison to be inappropriate. Aylin et al compare three clinical databases to administrative databases yet do not compare the costs of these systems. They appear to arbitrarily select the SHFA as a cost comparator with no explanation of the role of the SHFA. The SHFA is a prospective clinical audit, the aim of which is to improve the care of hip fracture patients in Scotland. It is not a database established to predict in-hospital mortality. The audit has nurses in each hospital employed to examine and report on the management of hip fracture patients. The core dataset records fifty -five variables and the current time-limited audit of medical reasons for delay to theatre has a further thirty-nine parameters. We monitor compliance with the Scottish Intercollegiate Guidelines Network Guideline No. 56 (1) and the national standard Older Patients in Acute Care (2) on a monthly basis and report on the previous month’s activity to clinical, management and executive teams at a hospital and board level. We report directly each month to the Scottish Executive Health Department and undertake site visits to hospitals where problems have been identified or help has been requested. We encourage the clinical teams to use the information we collect and have responded to over 40 ad hoc requests for analysis of the data in the last year. Our audit nurses continue to have a clinical role, acting as a liaison service by conducting interviews with each patient (or their carer) at four months after discharge. They may refer patients to orthopaedic clinics or local falls services. While we acknowledge the cost benefit of using routinely collected administrative data for the prediction of in-hospital mortality, we do not consider it comparable to prospectively collected clinical audit data. By providing robust and comparative data about multiple aspects of a complex patient journey, a properly conducted clinical audit can exert upward pressures on the standard of care and ultimately improve the outcomes for our patients. Yours faithfully 1. Scottish Intercollegiate Guidelines Network Guideline No. 56. Prevention and Management of Hip Fracture in Older People. www.sign.ac.uk 2. NHS Quality Improvement Scotland. Older Patients in Acute Care. www.nhshealthquality.org Competing interests: None declared
Benchmarking with administrative or clinical databases: Serious pitfalls. 1 June 2007
Alain Braillon, Public Health University hospitals of Amiens, 80000 Amiens, France Send response to journal: Re: Benchmarking with administrative or clinical databases: Serious pitfalls.	Alyn et al stated with too much faith that routinely collected administrative data can be used to predict risk with similar discrimination to clinical databases. However, there is strong evidence suggesting that comparisons of rates of risk-adjusted clinical outcomes as measures of performance are the Holy Grail, whatever the database is. The validity can be biased by the exclusion or inconsistent reporting of significant risk factors in predictive equations. Moreover, even elementary risk-adjustment variables, such as ASA score, are limited by inter observer inconsistency.1 Lastly, the logistic regression to fit the models is not explicative, it is just descriptive, and needs a prospective validation. Use of risk-adjusted clinical outcomes to test comparability among hospitals and to guide payment decisions needs serious. 1. Mak PH, Campbell RC, Irwin MG. The ASA (American Society of Anesthesiologists) Physical Status Classification: inter-observer consistency. Anaesth Intensive Care 2002;30:633-40. Competing interests: None declared