Effectiveness and safety of non-vitamin K antagonist oral anticoagulants among nonvalvular atrial fibrillation patients with prior bleeding events

Abstract Background Among non-valvular atrial fibrillation (NVAF) patients with a history of bleeding, there is a reluctance to use oral anticoagulants (OACs) due to concerns about the risk of bleeding associated with OACs. However, lack of OAC treatments for NVAF patients is associated with a higher risk of stroke and mortality. Non-vitamin K antagonist OAC (NOACs) have been approved for the prevention of stroke in NVAF patients. There are limited data comparing the risk of stroke/systemic embolism (S/SE) and major bleeding (MB) between patients prescribed NOACs and with a history of bleeding. Purpose This study used multiple United States data sources to evaluate the risk of S/SE and MB among NVAF patients with prior bleeding events who were prescribed NOACs. Methods This retrospective observational study used data from CMS Medicare and four commercial databases–covering >180 million beneficiaries. The study selected adult NVAF patients who were prescribed apixaban, dabigatran, or rivaroxaban (01JAN2013–30JUN2019) and had a prior bleeding event which was defined as a hospitalization with a bleeding diagnosis (intracranial hemorrhage [ICH], gastrointestinal [GI] bleeding, or other bleeding sites) prior to or during the index treatment episode. After 1:1 propensity-score-matched (PSM) in each database between NOACs (apixaban-dabigatran, apixaban-rivaroxaban, and dabigatran-rivaroxaban), the resulting patient records were pooled. S/SE and MB (identified by inpatient claims) were captured during the follow-up period, which was defined as the time between the day after the index treatment date and treatment discontinuation or switch, death, end of study period, or end of medical and pharmacy enrollment. Hazard ratios of S/SE and MB were calculated using Cox proportional hazards models. Results Of the overall NVAF population treated with NOACs, 6.2% had a prior bleeding event (ICH: 13.5%; GI: 61.8%; Other: 24.6%). After PSM, a total of 11,106 apixaban-dabigatran, 30,665 apixaban-rivaroxaban, and 11,148 dabigatran-rivaroxaban pairs were matched. Apixaban was associated with a lower risk of S/SE compared to dabigatran and rivaroxaban, and dabigatran was associated with a similar risk of S/SE compared to rivaroxaban. Apixaban was associated with a lower risk of MB compared to dabigatran and rivaroxaban, and dabigatran was associated with a lower risk of MB compared to rivaroxaban (Figure). Conclusions In this subgroup of NVAF patients with a history of bleeding,