Preoperative NTproBNP and Perioperative Major Cardiovascular Events in Noncardiac Surgery

M. Adnan Nadir, Research Fellow,

Other Contributors:

December 07, 2009
We read the paper by Choi et al with great interest. [1] Clearly NTproBNP (and CRP) appeared to predict perioperative major cardiovascular events (PCME) better than the Revised Cardiac Risk Index (RCRI). The most traditional way to depict that one predictor is better than other is to use the c-statistic or the area under the ROC curve (AUC). Although the abstract states that NTproBNP and CRP performed better than RCRI by ROC analysis, the paper does not tell us the actual c-statistic or AUC for either NTproBNP alone or CRP alone. This information would be useful in order to be able to compare directly the c-statistic of NTproBNP (or CRP) alone and RCRI alone. The c-statistic for the NTproBNP alone will help us to judge whether NTproBNP alone is as good as adding NTproBNP to the RCRI which is what the authors advocate. There would need to be a major advantage for combining NTproBNP and the RCRI over using NTproBNP alone for this to become practical reality since combining two separate risk scores is more cumbersome for the practicing clinicians than using either one alone. Instead of discussing the possibility that NTproBNP could replace the RCRI, the authors present the c-statistic data only for adding NTproBNP (and CRP) to the RCRI. We would urge the authors to now report c-statistic for NTproBNP alone (and CRP alone) so that it can be judged whether we really need to add NTproBNP (or CRP) to the RCRI (as they suggest) or whether NTproBNP (or CRP) can somehow replace the RCRI itself. 1 Choi JH, Cho DK, Song YB, et al. Preoperative NT-proBNP and CRP Predict Perioperative Major Cardiovascular Events in Noncardiac Surgery. Heart 2009.

Conflict of Interest

None declared