LAUSR

INTRODUCTION Black patients are under-represented in randomized trials evaluating oral anticoagulants in non-valvular atrial fibrillation (NVAF). We sought to evaluate the effectiveness and safety of rivaroxaban versus warfarin in African Americans with NVAF. METHODS We performed an analysis using Optum® De-Identified Electronic Health Record (EHR) data from 1/1/2012-9/30/2018. We included adult African American patients with a diagnosis of NVAF who were anticoagulant-naïve during the 12-months prior to initiation of rivaroxaban or warfarin. Patients receiving rivaroxaban were 1:1 propensity score matched to warfarin patients. Our primary effectiveness and safety outcomes were the 2-year incidence rates (%/year) of stroke or systemic embolism (SSE) and major bleeding using an intention-to-treat approach. Cohorts were compared using doubly-robust Cox regression and reported as hazard ratios (HRs) with 95% confidence intervals (CIs). RESULTS We matched 4102 rivaroxaban and 4102 warfarin users with a median (interquartile range) available follow-up of 2.0 (0.9, 2.0) years. Median CHA2DS2-VASc and HASBLED scores were 3 (2, 4) and 2 (1, 3). Rivaroxaban use was associated with a lower risk of SSE (1.99 versus 2.48, HR = 0.77, 95%CI = 0.60-0.99), ischemic stroke (1.84 versus 2.37, HR = 0.76, 95%CI = 0.59-0.98) and major bleeding (4.22 versus 4.98, HR = 0.84, 95%CI = 0.70-0.99). No differences in intracranial hemorrhage or gastrointestinal bleeding were observed. Neither sensitivity analyses utilizing an on-treatment methodology nor inverse probability-of-treatment weighting showed significant differences in SSE or major bleeding between rivaroxaban and warfarin users. CONCLUSIONS Rivaroxaban appeared at least as effective and safe as warfarin when used to manage African American patients with NVAF in routine practice.