Oral direct factor Xa inhibitor versus enoxaparin for thromboprophylaxis after hip or knee arthroplasty: Systemic review, traditional meta-analysis, dose–response meta-analysis and network meta-analysis.

Objectives To analyze the efficacy and safety of direct factor Xa inhibitors for thromboprophylaxis after total hip or knee replacement. To delineate the dose response effect of direct factor Xa inhibitors. To compare the efficacy between any two direct factor Xa inhibitors. Design Systemic review, traditional meta-analysis, dose-response meta-analysis and network meta-analysis. Data source PubMed, EMBASE and Cochrane Library. Study selection Randomized controlled trials of rivaroxaban, apixaban, betrixaban, darexaban and edoxaban were compared with enoxaparin for thromboprophylaxis after total hip or knee replacement. Two reviewers independently checked the quality of RCTs. Another two investigators independently extracted data. The primary efficacy outcomes (composite of deep venous thrombosis, non-fatal pulmonary embolism and death of all causes) and the primary bleeding outcomes (major bleeding and non-major but clinically relevant bleeding) were summarized for meta-analysis. Stata software was used for traditional meta-analysis and dose–response meta-analysis, and Winbugs software was used for network meta-analysis. Results Twenty trials with 38,507 subjects in the intention-to-treat population were included. Compared with enoxaparin, the risk of total venous thromboembolism was lower with rivaroxaban (relative risk 0.70, 95% confidence interval 0.60 to 0.81), apixaban (0.62, 0.47 to 0.81), and edoxaban (0.62, 0.39 to 0.97) and similar to darexaban (0.96, 0.84 to 1.11) and betrixaban (1.28, 0.97 to 1.68). Compared with enoxaparin, the risk of major or clinically relevant non-major bleeding was higher with rivaroxaban (1.52, 1.14 to 2.02), lower with betrixaban (0.34, 0.14 to 0.84) and similar to apixaban (0.88, 0.73 to 1.05), darexaban (0.85, 0.66 to 1.09) or edoxaban (1.30, 0.72 to 2.33). The risk of major and clinically relevant non-major bleeding of rivaroxaban had a linear relationship with its treatment doses; the risk of total venous thromboembolism of betrixaban and darexaban had linear relationships with their respective treatment doses. There was no linear nor non-liner relationships between the effect of apixaban and its treatment dose. The ranking of total venous thromboembolism risk from low to high was: rivaroxaban, apixaban, edoxaban, enoxaparin, darexaban, and betrixaban. The ranking of major and clinically relevant non-major bleeding from low to high was: betrixaban, enoxaparin, darexaban, edoxaban, apixaban, and rivaroxaban. Conclusions Direct oral factor Xa inhibitors are more effective to prevent venous thromboembolism after total hip or knee replacement. Their anticoagulant effect was not necessarily compromised with a higher bleeding risk. Rivaroxaban, apixaban and edoxaban showed a better anticoagulant effect, as compared with enoxaparin. Rivaroxaban had a higher bleeding rate, while apixaban and edoxaban did not show significantly higher bleeding risks. [ABSTRACT FROM AUTHOR]