Risk of nonmajor bleeding upon use of direct oral anticoagulants for preventing and treating venous thromboembolism: A network meta-analysis

INTRODUCTION: Direct oral anticoagulants (DOACs) are associated with bleeding. Patients often stop taking DOACs due to nonmajor bleeding, which may lead to venous thromboembolism (VTE) recurrence. We aimed to determine the risk of nonmajor bleeding using different DOACs to prevent and treat VTE. METHODS: PubMed, Embase, Web of Science, and Cochrane Library databases were searched from inception until January 6, 2021. The incidence of clinically relevant nonmajor bleeding and minor bleeding was investigated. In frequentist-based network meta-analysis, we analyzed the odds ratio (OR) with 95% CI and the surface under the cumulative ranking curves (SUCRA). RESULTS: Twenty-seven randomized controlled trials (RCTs) (involving 64,493 patients) were included. For preventing VTE, the risk for clinically relevant nonmajor bleeding was lowest for apixaban, followed by that for low-molecular weight heparin (LMWH), dabigatran, edoxaban, and rivaroxaban. The risk for minor bleeding was lowest for apixaban, followed by that for rivaroxaban, LMWH, dabigatran, and edoxaban. For treating VTE, the risk for clinically relevant nonmajor bleeding was also lowest for apixaban, followed by that for edoxaban, vitamin K antagonists (VKAs), and rivaroxaban. The risk for minor bleeding was lowest for apixaban, followed by that for rivaroxaban and VKAs. CONCLUSIONS: Regardless of whether it was used for preventing or treating VTE, apixaban had the lowest risk of nonmajor bleeding. This suggests that apixaban may have a lower risk of nonmajor bleeding than other anticoagulants and may help provide some clinical reference for choosing a more appropriate drug for the patient.