Comparison of the Effectiveness and Safety of Apixaban, Dabigatran, Rivaroxaban, and Warfarin in Newly Diagnosed Atrial Fibrillation

No studies have performed direct pairwise comparisons of the effectiveness and safety of warfarin and the new oral anticoagulants (NOACs) apixaban, dabigatran, and rivaroxaban. Using 2013 to 2014 claims from a 5% random sample of Medicare beneficiaries, we identified patients newly diagnosed with atrial fibrillation who initiated apixaban, dabigatran, rivaroxaban, warfarin, or no oral anticoagulation therapy in 2013 to 2014. Outcomes included the composite of ischemic stroke, systemic embolism (SE) and death, any bleeding event, gastrointestinal bleeding, intracranial bleeding, and treatment persistence. We constructed Cox proportional hazard models to compare outcomes between each pair of treatment groups. The composite risk of ischemic stroke, SE, and death was lower for NOACs than for warfarin: hazard ratio (HR) 0.86, 95% confidence interval (CI) 0.76 to 0.98 for apixaban; 0.73, 95% CI 0.63 to 0.86 for dabigatran; and 0.82, 95% CI 0.75 to 0.89 for rivaroxaban, all compared with warfarin. There were no differences in effectiveness across NOACs. The risk of any bleeding was lower with apixaban than with warfarin, but higher with rivaroxaban than with warfarin. Apixaban (HR 0.69, 95% CI 0.60 to 0.79) and dabigatran (HR 0.79, 95% CI 0.69 to 0.92) were associated with lower bleeding risk than rivaroxaban. Treatment persistence was highest for apixaban (82%), and lowest for dabigatran and warfarin (64%) (p value <0.001). Compared with warfarin, NOACs are more effective in preventing stroke but their risk of bleeding varies, with rivaroxaban having higher risk than warfarin. Altogether, apixaban had the most favorable effectiveness, safety, and persistence profile.