In Reply Although the available data suggest that, in patients with atrial fibrillation, sex is not associated with serious oral anticoagulant-related bleeding, as suggested by Dr Samuelson Bannow and colleagues, we analyzed the interaction between diltiazem and apixaban/rivaroxaban according to sex. The cohort included 107 652 females with a mean age of 77.8 (SD, 7.2) years and 96 503 males with a mean age of 75.9 (SD, 6.6) years. The outcome was the composite of bleeding-related hospitalization or death with recent evidence of bleeding, the primary end point of our original analysis. We conducted the analysis as previously reported, with the exception that separate propensity scores were estimated for females and males, which ensured covariate balance between the diltiazem and metoprolol groups within each sex category. Both female and male patients treated with diltiazem had significantly increased risk for the composite outcome (Table). The 95% CIs for the rate difference and hazard ratio in males included the point estimates for females, indicating limited evidence for effect modification according to sex.
