Reply to Letter: “Preoperative Aspirin-dosing Strategy and Mortality After Coronary Artery”
We thank Xue et al for their comments on our recent study published in Annals of Surgery1 about the association between preoperative aspirin dosing strategy and mortality after coronary artery bypass graft surgery. Xue et al suggest 2 potential confounders with our study. First, they suggest that the lack of reported perioperative hemoglobin (Hb) values could potentially confound our finding that low-dose aspirin (ASA) use within 24 hours of coronary artery bypass graft (CABG) surgery is independently associated with decreased early postoperative mortality. Supporting their assertion, they reference previous studies in which perioperative anemia was associated with increased CABG mortality. While we did not factor perioperative anemia into our analysis, we used both propensity matching and multivariate logistic regression on multiple variables, which have been previously validated as independent risk factors of postoperative mortality in both the EuroSCORE and Society of Thoracic Surgeons risk assessment. Indeed, these are the best known and most widely validated perioperative risk stratification tools for assessing mortality risk in cardiac surgery, and neither uses perioperative anemia as an independent predictor of risk.2–4 Furthermore, it is often hard to tease out whether perioperative anemia or transfusion used to treat the anemia is actually more contributory toward increased mortality. Indeed, one of the citations in the letter by Xue et al concluded that anemia without accompanying transfusion did not increase postoperative mortality.5 As previously noted in our study's limitations section, we agree that the lack of the transfusion data is a weakness of our study. Correlation between perioperative Hb level, transfusion rate, and preoperative ASA use would have been ideal and should merit inclusion in future studies.
Second, Xue et al suggest that additional cardiac-related variables should have been included in the multivariate and propensity matching analyses. In particular, they cite Gardner et al,6 suggesting that New York Heart Association III/IV status, previous myocardial infarction, preoperative intra-aortic balloon pump (IABP), preoperative diuretic use, and previous heart operation are major predictors of early mortality after CABG surgery. Not only had we already included New York Heart Association III/IV status and previous myocardial infarction in our analyses, but inclusion of diuretics or IABP as variables did not alter any of our conclusions (data not published). Finally, redo cardiac surgical patients were excluded in our study, thus redo surgery as a variable was not included in the analyses.