Introduction: Non-vitamin K oral anticoagulants (NOACs) were shown to have better outcomes than warfarin for non-valvular atrial fibrillation (NVAF). Given limited local real-world data, this study aims to evaluate the safety and efficacy of NOACs versus warfarin for NVAF in Singapore.
Methods: This single-centre retrospective cohort study included 439 patients ≥ 21 years old that were newly prescribed with oral anticoagulants (OACs) for NVAF in 2015. Follow-ups for patients upon OAC initiation lasted either for 2 years or until the occurrence of bleeding or thromboembolism event or death (whichever was earlier). Primary endpoints included major bleeding and stroke, while secondary endpoints included overall bleeding and thromboembolic events. Time-to-events was evaluated via Kaplan-Meier survival analysis. Data on time in therapeutic range (TTR) and compliance were analysed.
Results: Patients were assigned to 4 groups: warfarin (157, 35.8%), rivaroxaban (154, 35.1%), apixaban (98, 22.3%) and dabigatran (30, 6.8%). With a mean age of 70.8 (±10.8) years old, the population were predominantly males (56.5%) and comprised Chinese (73.8%), Malays (18.7%) and others (7.5%). The rates of stroke per year were 0.7%, 1.7%, 2.2% and 0% for warfarin, rivaroxaban, apixaban and dabigatran, respectively (P=0.411), whereas those of major bleeding were 2.7%, 1.4%, 2.2% and 0% (P=0.560). As compared to warfarin, no significant differences were observed for risks of stroke and of major bleeding for rivaroxaban (adjusted hazard ratio (HR) 4.19, 95% confidence interval (CI) 0.68-26.05, P=0.124 and adjusted HR 0.43, 95% CI 0.12-1.59, P=0.207) and apixaban (adjusted HR 5.33, 95% CI 0.85-33.34, P=0.074 and adjusted HR 1.54, 95% CI 0.39-6.15, P=0.538). Mean TTR was 68.8% (±24.3%) for warfarin. Compliance rates for rivaroxaban, apixaban, and dabigatran were 56.6%, 59.2%, and 44.8%, respectively (P=0.177).
Conclusion: NOACs were associated with similar stroke and major bleeding rates as warfarin for NVAF.