Background: Non-valvular atrial fibrillation (NVAF) is associated with ischemic stroke risk in the aging population. Observational studies have indicated beneficial effects of direct-acting oral anticoagulant (DOAC) against ischemic stroke compared to warfarin. This study aimed to investigate ischemic stroke incidence and bleeding risk in patients on DOAC therapy. Methods: Using the database of Korean Health Insurance Review and Assessment-Aged Patient Sample 2015, we conducted a retrospective cohort study. Study subjects with NVAF diagnosis and prescribed anticoagulants were enrolled. Propensity score (PS) matching by age, sex, comorbidities, and medications were used. The clinical outcomes were major adverse cerebro-cardiovascular events (MACCEs, ischemic stroke/systemic embolism, myocardial infarction, cardiac death) and bleeding events. A cox proportional hazard model analysis was performed to compare the outcomes with hazard ratio (HR) and 95% confidence interval (CI). Results: Total 4,773 elderly patients with NVAF were initially included. Four PS-matched groups including rivaroxaban vs. warfarin-only (n=1,079), dabigatran vs. warfarin-only (n=721), rivaroxaban vs. dabigatran (n=721), and switchers of warfarin to rivaroxaban vs. warfarinonly (n=287) were analyzed. Every group showed statistically similar results of MACCEs and bleeding events, except for the group of rivaroxaban vs. dabigatran. Rivaroxaban users showed higher risks of bleeding events than dabigatran users (HR 2.25, 95% CI 1.01-4.99). Conclusion: In the elderly patients with NVAF, efficacy and safety outcomes among oral anticoagulants including DOACs and warfarin were similar, while rivaroxaban are more likely to have higher bleeding risks than dabigatran. Further research using large size sample is needed.
