Comparison of major bleeding risk in patients with non-valvular atrial fibrillation receiving direct oral anticoagulants in the real-world setting: a network meta-analysis

Objective: To conduct a systematic literature review (SLR) and network meta-analysis (NMA) of real-world studies comparing major bleeding risk among patients with non-valvular atrial fibrillation (NVAF) on direct oral anticoagulants (DOACs) or warfarin. Methods: Systematic searches were conducted in MEDLINE and Embase for full-text articles published between January 1, 2003 and March 18, 2017. Eligible studies compared at least two of the following in a real-world setting: warfarin, apixaban, dabigatran, rivaroxaban, or edoxaban. A Bayesian NMA was conducted to estimate hazard ratios (HRs) for major bleeding using a random-effects model. Results: Eleven studies were included in the NMA. Nine studies included DOACs vs Warfarin comparisons, and four studies included DOACs vs DOACs comparisons (two studies included both comparisons). Median follow-up duration ranged from 2.6-31.2 months. No evidence was identified for edoxaban. Apixaban was associated with a significantly lower risk of major bleeding compared to other oral anticoagulants (warfarin HR = 0.58; 95% credible interval [CrI] = 0.48-0.69; dabigatran=0.73; 0.61-0.87; rivaroxaban = 0.55; 0.46-0.66). Dabigatran was associated with a significantly lower risk than warfarin (0.79; 0.71-0.88) and rivaroxaban (0.76; 0.67-0.85), and rivaroxaban was not statistically different from warfarin (1.05; 0.91-1.19). Sensitivity analyses with standard dose and sponsorship showed consistent results. Conclusion: DOACs were associated with lower or similar risk of major bleeding compared with warfarin in NVAF patients. Apixaban was associated with a significantly lower risk of major bleeding than other DOACs. Dabigatran was associated with a significantly lower risk of major bleeding compared to rivaroxaban and warfarin.