Differences in safety and efficacy of oral anticoagulants in patients with non-valvular atrial fibrillation: A Bayesian analysis

Background: Novel oral anticoagulants are the cornerstone of therapy for non-valvular atrial fibrillation patients to lower the risk of ischaemic stroke. Given the lack of head-to-head comparisons among oral anticoagulants, a Bayesian analysis was used to evaluate their safety and efficacy based on studies from real-world practice. Methods: The PubMed, Embase, Cochrane and Web of Science databases were searched for relevant studies. Bayesian analyses were conducted to estimate hazard ratios (HR) and 95% credible intervals (CrI) for the safety and efficacy of oral anticoagulants. Results: In the 22 studies included in our analysis, novel oral anticoagulants exhibited a clear advantage over warfarin in preventing ischaemic stroke, haemorrhagic stroke and, especially, intracranial haemorrhage. Incidence of major bleeding was lowest for apixaban, followed by dabigatran, warfarin and rivaroxaban. Gastrointestinal bleeding risk was lowest for apixaban, followed by warfarin, and was slightly lower for dabigatran than for rivaroxaban with no statistical difference (HR 1.05, 95% CrI 0.99-1.11). Ischaemic stroke risk was lowest for rivaroxaban, followed by apixaban, dabigatran and warfarin (HR 1.13, 95% CrI 1.07-1.20). Conclusion: In real-world practice, apixaban may represent the optimal treatment in terms of safety and efficacy for patients with non-valvular atrial fibrillation. For patients with high risk of ischaemic events but low bleeding risk, rivaroxaban may be preferable.