| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2006;29:989-990
© 2006 Elsevier Science NL
Denver Children's Hospital, University of Colorado, Denver, CO 80262, USA
* Corresponding author. Tel.: +1 303 8616624; fax: +1 303 7648022. (Email: lacour-gayet.francois{at}tchden.org).
The article [1] from the Great Ormond Street (GOS) group raises a precise question: ‘Does the Aristotle Score Predict Outcome in Congenital Heart Surgery?’ The answer from the authors is ‘The Basic Aristotle score is only weakly associated with post-operative mortality in this GOS series.’ In fact, the Aristotle score showed a statistically significant correlation with mortality in their series with a p-value of 0.02 or 0.03. This seems acceptable for a score that was not created to predict mortality. The Aristotle Basic Score (ABS) was designed [2], based on a subjective approach, to define a constant called complexity, which is calculated as the sum of mortality, morbidity, and technical difficulty for 145 procedures in our specialty. Complexity, that we believe is a global constant for a given patient, serves as the basis to compare performance between centers or surgeons.
In contrast, the RACHS 1 was designed specifically to predict mortality as quoted in their initial article [3]. ‘The aim of RACHS 1 was to develop a consensus-based method of risk adjustment for in-hospital mortality among children younger than 18 years after surgery for congenital heart disease.’ All recent studies show that the two scores have excellent correlation with hospital mortality, particularly when applied to large multicenter databases, as shown here for the ABS applied to the congenital UK database (Fig. 1 ).
|
When commenting on risk adjustment in CHS, Blackstone [4] concluded: ‘Unfortunately, risk adjustment tends to be particularly incomplete when there are rare or multiple measured, unmeasured, or unevaluated risk factors present, so the search for adequate unconfounded quality measures should go on ... until the data speaks for themselves.’ The question is whether we should wait passively or try to devise a reasonable instrument for measuring quality even if it is not perfect? Our answer, both for Aristotle [5] and RACHS, is that we should build today a risk adjustment method, even if it will be partially based on subjectivity.
Recently, the RACHS and the Aristotle groups have decided to combine their efforts to produce a mortality score, based on outcome data from more than 50,000 patients from the EACTS and STS databases, using a minimum of subjective probability. Nevertheless, such a score will not predict mortality in a given institution but will give an average value of mortality for a given procedure which every center or surgeon could use for comparison to their own results.
The article on Aristotle published in this issue alluded to performance. GOS is a leading center in Europe. The article reports that their complexity is calculated at 7.88 and their performance at 7.52. It would be interesting to recalculate these values with reintroduction of off-pump procedures (included in the STS/EACTS databases). These usually account for 25% of procedures, characterized by a low complexity but quite a significant mortality (BT shunt mortality is 7.2% in EACTS/STS official data).
The authors’ remarks regarding different scenarios of performance are not realistic. Surgeon A, who has a 5% mortality while restricting himself to simple procedures (ASD, VSD, etc.) would be the worst center within the STS/EACTS databases. Surgeon B, with an average complexity of 10 and 10% mortality, would be by very far the world's best center. Surgeon C, with a 20 complexity, does not exist (the ABS being limited to 15). Moreover, the rationale behind the proposed equation ‘Performance = Complexity/Mortality’ is unclear in that it minimizes the impact of complexity to a degree that nearly eliminates any risk adjustment. Surgeons A and B, when compared to our databases, correspond to the weakest and the strongest centers; they cannot have the same performance. In comparison, using multiplication and survival, the performance will be 4.75 (5 x 0.95) for Surgeon A and 9 for (10 x 0.90) for Surgeon B. This seems fair enough. Additionally, mortality zero cannot be evaluated by their equation.
Evaluating quality of care in CHS is an ongoing endeavor and the GOS center has provided an important contribution. Predicting mortality in CHS is a hazardous exercise. Retrospective analysis of performance and counter-performance of a group or a surgeon is a positive approach. This should allow active changes which include retraining, restructuring, and particularly directing the most complex patients to the best prepared centers. The goal of the Aristotle score is to precisely evaluate performance. The ABS provides a reasonable basis that will be improved with access to observed data. The introduction of the comprehensive Aristotle score allows identification [6–8] of the most challenging patients, still at high risk for mortality, morbidity, and elevated cost for the health care payers.
 |
References |
|---|
 Top 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 |
