Abstract W P309: Hemorrhagic Risk Associated with Warfarin Initiation among Patients with Atrial Fibrillation: Assessment via a Large Insurance Database

Abstract

Objective: Studies suggest that warfarin may increase the risk of stroke in the first month after initiation among atrial fibrillation (AF) patients. This study assess whether there is a similar occurrence with major bleeds. We compared rates of major bleeds (hospitalizations for bleeding) in the first 30 days and >30 days after warfarin initiation between AF patients treated with warfarin and AF patients not on oral anticoagulants (OACs) in a large database. Methods: We identified adult AF patients via one inpatient or two outpatient AF diagnosis claims from the MarketScan insurance database with 95 million enrollees from 1/1/2008 to 3/31/2012. AF patients who started warfarin based on prescription claims, 1/1/2009-12/31/2010 and who didn’t use OACs 6-month prior were defined as warfarin initiators. Comparators were AF patients not on OACs from 7/1/2008-12/31/2010 and were 1:1 matched to warfarin initiators on age, sex and date of AF diagnosis. Patients were followed until major bleeding, discontinuation/switch of therapy, disenrollment, or end of study. We report crude rates of major bleeding and adjusted hazard ratios (HRs). Result: We studied 45,052 AF patients with an average follow-up of 1.9 years. Warfarin initiators and comparators were the same age (73.0) and sex (50% female) but had slightly lower CHADS2 scores (1.7 vs. 1.8) and ATRIA bleed scores (2.2 vs. 2.4). Compared to patients not on OACs, the rates of bleeding were higher among warfarin initiators during the first 30 days (adjusted HR=1.70) and after 30 days (adjusted HR=1.77) (Table) (p for interaction=0.86). The bleeding rate was higher in the first 30 days than after 30 days for the warfarin initiators, but this was also observed for the comparators (p=.02). Conclusion: Compared with AF patients not on OACs, warfarin appears to increase the risk of major bleeding similarly during the first 30 days and thereafter even though the rate was higher in the first 30 days for both warfarin initiators and comparators.