Abstract 17137: Safety and Effectiveness of Dabigatran Relative to Warfarin in Routine Clinical Practice - New Interim Results of Long-term Study Program

Abstract

Background: Dabigatran etexilate is a newer oral anticoagulant developed as an alternative to warfarin for stroke prevention in patients with non-valvular atrial fibrillation (NVAF). Although the efficacy and safety of dabigatran were established in a randomized trial, evaluation of its effects in routine clinical practice is ongoing. Objectives: Quantification of the comparative safety and effectiveness of dabigatran in routine care within two large US commercial health insurance databases (Truven MarketScan and Optum Clinformatics) Methods: Cohort design with propensity score (PS) matching to compare new initiators of dabigatran with warfarin between October 2010 and June 2013. Primary outcomes were stroke and major bleeding. Proportional hazards regression of time to outcome was conducted separately within each data source and results were pooled. Results: There were 22,736 PS matched dabigatran and warfarin initiators with NVAF pooled across data sources (MarketScan 18,726; Optum 4,010). The matching resulted in well balanced cohorts with no individual characteristic (including stroke and bleeding risk factors) having a standardized difference of greater than 0.1. The average follow-up for the as-treated analyses was 5 months for dabigatran and 4 months for warfarin. There were 69 strokes amongst dabigatran initiators and 84 strokes amongst warfarin initiators for a pooled HR of 0.71 (95% CI 0.51 – 0.98) (MarketScan HR = 0.65, 95% CI=0.45-0.93, Clinformatics HR = 1.05, 95% CI = 0.50-2.19). For the outcome of major hemorrhage, there were 429 events amongst dabigatran initiators and 480 events amongst warfarin initiators for a pooled HR of 0.77, 95% CI = 0.68 – 0.88) (MarketScan HR = 0.80, 95% CI = 0.70 - 0.92, Clinformatics HR = 0.55, 95% CI = 0.37 - 0.82). Pooled results in the full cohort were fairly consistent across numerous subgroup and sensitivity analyses, however, comparative conclusions within some subgroups are limited by small numbers and limited follow-up time. Conclusions: Analyses from this ongoing sequential long-term study program suggest a reduced risk of both stroke and major hemorrhage for dabigatran relative to warfarin. Continued follow-up will increase the sample size, permitting greater precision in effect estimates.