Background: The real-life benefits and risks of the direct oral anticoagulants (DOAC) for non-valvular atrial fibrillation (NVAF) have been discussed.
Purpose: To compare the 1-year risk of clinical outcomes for new users of dabigatran (D) or rivaroxaban (R) vs. vitamin K antagonists (VKA) in NVAF.
Methods: Cohorts of new users of D, R or VKA for NVAF in 2013 identified and followed for 1 year in the 66 million persons nationwide French claims and hospitalisation database (SNIIRAM). Two NVAF populations were defined: i) a specific population (long-term disease or hospitalisation with atrial fibrillation diagnosis without valvular disease history); ii) a sensitive population (specific population plus patients with a probable NVAF using a disease score). For each comparison (D vs. VKA, R vs. VKA), patients were matched 1:1 on gender, age, date of the first drug dispensing, and high-dimensional propensity score, including major arterial thrombosis and bleeding risk factors. The 1-year incidence of outcomes was compared using Cox proportional hazard risk model during drug exposure period.
Results: Of 371 539 new anticoagulant users in 2013, 103 101 (27 060 D, 31 388 R, 44 653 VKA) were included in the specific population, 144 220 (37 222 D, 46 882 R, 60 116 VKA) in the sensitive population. For the specific population, mean age was 75.2 years, 54.1% male, 82.5% CHA2DS2-VASc score ≥2, 9.8% HAS-BLED score >3 with significant differences between groups which were normalized after matching. For D vs VKA, 20 489 patients were matched per arm, and 23 053 per arm for R vs VKA. The 1-year cumulative incidence of outcomes and comparison between groups for matched populations are presented in the table. All results were similar for the sensitive population.
Conclusions: This study shows a significantly overall better benefit-risk of DOAC vs. VKA including death and intracranial bleeding without increased risk of gastrointestinal bleeding.