| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2002;74:1748-1749
© 2002 The Society of Thoracic Surgeons
a Papworth Hospital, Cambridge CB3 8RE, United Kingdom
b Service de Chirurgie Cardiovasculaire, CHU de Fort de France, 97200 Fort de France, Martinique, French West Indies
e-mail: f.r.fwi{at}wanadoo.fr
To the Editor:
In an editorial by Grunkemeier and colleagues [1], readers were encouraged to read carefully the review article by Shahian and others [2] on the controversial issue of cardiac surgery report cards. We would like to make the following comments.
Both articles identify accurately and comprehensively the weaknesses of risk assessment systems in predicting death and warn about the danger inherent in total reliance on such systems. Of course, no risk assessment system is perfect, but Grunkemeier’s comment that a C-index of 0.8 is "only 60% of the way between worthless and perfect" is somewhat unfair. Perfection is not possible in risk prediction. If it were, it would no longer be prediction of risk but foretelling the future. Indeed, if we were ever able to predict death with 100% discrimination, no patient would ever die after cardiac operations, because we would have that advance knowledge and therefore simply not operate on the doomed.
There are two approaches to quality assurance in cardiac surgery and in medical care in general. The first is the obvious and easy path of report cards, or, as they are called in Europe, league tables or hit parades. Whatever the name, this is essentially publication of outcome data. Both articles eloquently express the concerns that many surgeons have about this approach, with its inherent risk of gaming, statistical play, and potential damage to providers (and, indeed, patients) when surgeons and hospitals are seen to fail, whether such perceived failure is due to chance, statistical quirk, risk model instability, or true underperformance. The alternative is that of robust, peer-reviewed, local quality monitoring with intelligent interpretation of data and the timely institution of action plans for any underperformance. The limitation of this approach is the loss of transparency and perceived reduction in the freedom of information and public accountability that many political forces are demanding of health care. Whichever path is chosen, robust risk assessment, properly constructed by the profession, is an essential part of the process and presents a substantial advance on crude outcome statistics alone.
Finally, we are disappointed that in two major and comprehensive articles on the subject of risk assessment and performance in cardiac surgery, with a combined reference list of more than 155 articles, EuroSCORE [3] was not mentioned once.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
