BACKGROUND:Nonsteroidal anti-inflammatory drug (NSAID) use may reduce the incidence of post-cardiothoracic surgery (CTS) atrial fibrillation (AF). The cerebrovascular and cardiovascular safety of using NSAIDs for post-CTS AF has not been determined. OBJECTIVE:To evaluate whether NSAIDs could reduce the incidence of post-CTS atrial fibrillation without increasing patients' risk of stroke or myocardial infarction (MI). METHODS:Patients (n = 555) undergoing CTS from the Atrial Fibrillation Suppression Trials I, II and III were evaluated in this nested cohort study. Demographic, surgical and medication use characteristics were prospectively collected as part of the AFIST trials. Endpoints included post-CTS atrial fibrillation, stroke, MI and the need for red blood cell transfusion. Multivariable logistic regression was used to control for potential confounders and calculate adjusted odds ratios with 95% confidence intervals. RESULTS:The population was 67.8 +/- 8.6 years old and 77.1% male with 127 (22.9%) patients receiving an NSAID postoperatively. Overall, 14.6% underwent valve surgery, 6.1% had prior AF, 12.6% had heart failure and 84.0% and 44.1% received postoperative beta-blockade and prophylactic amiodarone. NSAID use was associated with reductions in the adjusted odds of post-CTS atrial fibrillation (0.54 (0.32-0.90)) and the need for RBC transfusions (0.63 (0.41-0.97)). No elevation in the odds of developing stroke (1.10 (0.21-5.66)) or MI (1.70 (0.40-7.10)) was observed. LIMITATIONS:Patients were not randomized to receive NSAIDs versus a control. We may not have had adequate power to evaluate stoke or MI in this analysis. CONCLUSIONS:NSAIDs decreased the odds of developing post-CTS atrial fibrillation, further supporting the hypothesis of inflammation as a trigger for post-CTS atrial fibrillation. The need for RBC transfusions was also reduced with NSAID use. We may have been underpowered to evaluate stroke or MI incidence, but the qualitative elevations in these variables suggest more safety data is needed before NSAIDs can be routinely recommended.