HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Add to Personal Folders Download to citation manager Permission Requests Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Seccareccia, F. Articles by D’Errigo, P. Search for Related Content PubMed Articles by Seccareccia, F. Articles by D’Errigo, P. Related Collections Coronary disease

Eur J Cardiothorac Surg 2006;29:857-858
© 2006 Elsevier Science NL

Letter to the Editor

Reply to Hekmat et al.

^a National Centre of Epidemiology, Surveillance and Health Promotion—Istituto Superiore di Sanità, Via Giano della Bella, 34, I-00161 Rome, Italy
^b Department of Epidemiology—ASL RME, Rome, Italy

Received 6 February 2006; accepted 8 February 2006.

^_* Corresponding author. Tel.: +39 06 49904234; fax: +39 06 49904230. (Email: fulvias{at}iss.it).

Key Words: Coronary artery bypass graft • Outcome • Risk-adjustment • Mortality

Starting from the title chosen for this letter, we understand that the objective of our work was completely misunderstood [1].

Therefore, although already specified in the paper, we need to clarify that the risk adjustment model used in this analysis did not have the objective to create a new risk score, but that to apply risk adjustment procedures to profile providers, allowing the comparison of their performance with an internal reference. In this case, dealing with explicative (and not predictive) models, the relevance of parameters like HL dramatically decreases. Moreover, considering the dataset dimension (34310 records), it is possible that even a very small difference between observed and expected number of deaths determines a HL statistically significant.

Although the risk function produced in this study does not represent a ‘new scoring system’ and has not the objective to launch a ‘new risk stratification model’ to definitively replace out-of-date systems, we tried to apply the logistic EuroSCORE [2] on our database and we found a ROC value of 0.78 but a very poor calibration (HL p < 0.001).

As reported in the ‘Discussion’ section of the paper and well supported by literature [3], we underlined that "... any risk adjustment function to be utilized for the purpose of comparison between Centres or population must be time and population specific. This statement implies a substantial difference between ‘predictive models’ aimed to predict the occurrence of outcome, and ‘explicative models’ aimed to control confounding in comparison of occurrence of outcome between Centres or population. Even a very valid risk score having a strong association with the concerned outcome could be irrelevant for the purpose of risk adjustment if the distribution of its values is homogeneous between Centres [1]."

Finally, the meaning of the sentence already reported in the Editorial comment by Menicanti [4] and now drawn on by Hekmat, regarding the supposed statistical ability in ‘resuscitating dead patients’, remains mysterious. Our humble scientific approach modestly tries to reduce biases of observation but it is, unfortunately, unable to resuscitate patients. What our study described in detail, in each Centre, were the observed number of deaths as compared to those expected through the application of the best risk-adjustment model. ‘Expected number of deaths’ indicates how many deaths would have occurred in a Centre if it had the same ‘case mix’ of the average population. The ratio ‘observed/expected deaths’ applied to the observed overall mortality rate of the standard population yields the values of the risk adjusted mortality rate (RAMR).

The odd, besides inappropriate, method used by Menicanti [4] to compute the number of "patients in good shape that are dead for statistics and ... patients dead but for statistics in good shape!" (consisting in multiplying the RAMR by the number of patients for each Centre!!!) is wrong and not allowed by any mathematical rule.

Even tough criticisms on a scientific work are extremely useful in promoting quality research, provided they are based on scientific ground!

References

1. Seccareccia F, Perucci CA, D’Errigo P, Arca M, Fusco D, Rosato S, Greco D, On behalf of the Research Group of the Italian CABG Outcome Study The Italian CABG Outcome Study: short-term outcomes in patients with coronary artery bypass graft surgery. Eur J Cardiothorac Surg 2006;29(1):56-62.[Abstract/Free Full Text]
2. Roques F, Michel P, Goldstone AR, Nashef SA. The logistic EuroSCORE. Eur Heart J 2003;24(9):882-883.[Free Full Text]
3. Ivanov J, Tu JV, Naylor CD. Ready-made, recalibrated, or remodeled? Issues in the use of risk indexes for assessing mortality after coronary artery bypass graft surgery. Circulation 1999;99(16):2098-2104.[Abstract/Free Full Text]
4. Menicanti LA. Editorial comment. Eur J Cardiothorac Surg 2006;29(1):63-64.[Free Full Text]

This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Add to Personal Folders Download to citation manager Permission Requests Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Seccareccia, F. Articles by D’Errigo, P. Search for Related Content PubMed Articles by Seccareccia, F. Articles by D’Errigo, P. Related Collections Coronary disease