On-treatment follow-up in real-world studies of direct oral anticoagulants in atrial fibrillation: Association with treatment effects

BACKGROUND: Numerous observational studies support the safety and effectiveness of the direct oral anticoagulants (DOAC) for stroke prevention in atrial fibrillation (AF), but these data are often limited to short duration of follow-up. We aimed to assess the length of on-treatment follow-up in the accumulated real-world evidence and the relationship between follow-up duration and estimates of DOAC effectiveness and safety. METHODS: We searched the literature for observational studies reporting comparative effectiveness and safety outcomes of DOACs versus warfarin. In random-effects meta-analyses, we assessed associations of specific DOACs vs warfarin for stroke/systematic embolism (SE) and major bleeding. In meta-regression analyses, we assessed the correlation between the reported on-treatment follow-up with the effect sizes for stroke/SE and major bleeding outcomes. RESULTS: In 45 eligible observational studies, the average on-treatment follow-up was <1 year for all DOACs. In meta-analyses, all DOACs showed significantly lower risks of stroke/SE, but only dabigatran and apixaban showed lower risks for major bleeding compared to warfarin. There was no correlation between follow-up duration and magnitude of stroke/SE reduction for any of the DOACs. Longer follow-up correlated with greater major bleeding reduction for dabigatran (p = 0.006) and rivaroxaban (p = 0.033) as compared to warfarin, but it correlated with smaller major bleeding reduction for apixaban (p = 0.004). CONCLUSIONS: The numerous studies of DOAC effectiveness and safety in the routine AF practice pertain to short treatment follow-up. Study follow-up correlates significantly with DOAC-specific vs warfarin associations for major bleeding.