Efficacy, safety, and bleeding risk factor analysis of oral anticoagulants in AF patients ≥ 65 years of age: a multicenter retrospective cohort study

Background Stroke prevention in elderly patients with atrial fibrillation (AF) is challenging and requires a balance between thromboembolic prevention and bleeding. The comparison of novel oral anticoagulants (NOACs) and warfarin in clinical practice in elderly Asian patients has not been well studied. The purpose of this study was to evaluate the efficacy and safety of NOACs versus warfarin in elderly patients with AF in conjunction with data from real-world observational studies. Methods This was a retrospective multicenter cohort study conducted in 4 centers in China, where patient information and clinical events were collected through an average of 15 months of follow-up and case queries. Clinical outcomes included major bleeding, minor bleeding, total bleeding, thrombosis, and all-cause mortality. Results A total of 3450 elderly patients with AF were enrolled. 2656 patients were treated with at least 1 NOAC (dabigatran, rivaroxaban, apixaban, or edoxaban), and 794 patients were treated with warfarin. After correcting for confounders, NOACs significantly reduced the risk of minor bleeding [OR 0.70 (95% CL, 0.49-1.01),P = 0.049] and all-cause mortality [OR 0.57( 95% CI, 0.44-0.75),P < 0.001] compared with warfarin, however, major bleeding events [OR 1.51 (95% CL, 0.98-2.42),P = 0.075] and thrombotic events [OR 0.79 (95% CL, 0.57-1.13),P = 0.187] were not significantly different. There was no heterogeneity between clinical outcomes of NOACs and warfarin in subgroup analyses of age (65-74, 75-84, >= 85 years), sex (male, female), BMI (>= 25, < 25), comorbidities (including hypertension, diabetes and no hypertension, no diabetes), except in female subgroup, where NOACs significantly reduced the risk of minor bleeding [OR 0.56 (95% CL, 0.34-0.91),P = 0.018] and increased the risk of major bleeding [OR 2.28 (95% CL, 1.12-5.14),P = 0.032] compared with warfarin. Conclusion Compared with warfarin, NOACs significantly reduced the risk of minor bleeding, all-cause mortality, and there were no statistically significant differences in major bleeding or thrombotic events. NOACs were not more effective than warfarin in thrombotic and bleeding events, regardless of the subgroup analyses on age, male, BMI and comorbid hypertension and diabetes.