Abstract 158: Exposure to Warfarin and the Risks of Stroke and Bleeding Events among Patients with Non-Valvular Atrial Fibrillation: Real-World vs. Clinical Trial

INTRODUCTION: While recent randomized controlled trials (RCTs) have examined safety and efficacy of apixaban versus warfarin in stroke prevention among patients with nonvalvular atrial fibrillation (NVAF), real-world (RW) event rates may differ from the RCT setting. HYPOTHESIS: In a managed care, NVAF population exposure to warfarin is associated with a decreased risk of stroke and increased risk of bleed relative to non-exposure, but both stroke and bleed rates remain higher than those observed in RCT settings. METHODS: Using a retrospective cohort design, we selected patients aged ≥18 years with diagnosis of NVAF during 2007- 2009 from a Medco population of both U.S. commercial and Medicare health plans. The first claim with an AF diagnosis was defined as the study index date. Pharmacy claims were used to define periods of warfarin exposure following the study index date. Stroke and bleed events occurring during follow-up were identified using diagnosis codes on the medical claims and categorized according to occurrence during warfarin exposure or no exposure and according to patient baseline CHADS2 risk to address differences in population risk. RESULTS: Of the 37,878 patients meeting the study criteria, the mean age was 75 years (SD=8.6), and 53% were male. Among the 66% of patients with at least one prescription for warfarin, the total person-time of warfarin exposure was 41,515 years. Compared to rates occurring in two recent RCTs of warfarin exposed (ARISTOTLE) and warfarin unexposed (AVERROES; aspirin treated) patients, stroke and bleed event rates for NVAF patients in the real-world were higher overall and within each CHADS2 group. CONCLUSIONS: While in the real world warfarin use was associated with a reduction in stroke rates and an increase in bleed rates, the rate of events in real world patients relative to their RCT counterparts were higher. The absolute impact of novel agents, such as apixaban, in the real world may be greater than in RCT setting if the relative risk reductions versus alternatives from RCTs persist in the real world. ![Graphic][1] [1]: /embed/inline-graphic-1.gif