Low-dose rivaroxaban versus warfarin in Asian patients after surgical bioprosthetic valve replacement: a propensity score–matched and inverse probability-weighted real-world study

Evidence for rivaroxaban during the early postoperative period after surgical bioprosthetic valve replacement remains limited, particularly in Asian patients receiving low-dose rivaroxaban. We compared the effectiveness and safety of low-dose rivaroxaban versus dose-adjusted warfarin within the initial 3 months after surgery. In this single-center retrospective cohort study, we included consecutive adults who underwent surgical bioprosthetic valve replacement between August 2021 and December 2023 and received postoperative oral anticoagulation with low-dose rivaroxaban (15 or 10 mg once daily) or warfarin. Propensity score matching was used as the primary analysis, and inverse probability of treatment weighting was performed as a sensitivity analysis. The primary effectiveness outcome was thromboembolic events; the primary safety outcome was clinically relevant bleeding. Time-to-event outcomes were analyzed using Cox proportional hazards models, and absolute risk differences were reported descriptively. A total of 367 patients were included (178 rivaroxaban, 189 warfarin). After 1:1 PSM, 125 pairs were analyzed. During the 3-month follow-up, thromboembolic events occurred in 1/125 patients (0.8%) in each group, with an absolute risk difference of 0.0% points. Clinically relevant bleeding occurred in 7/125 patients (5.6%) in the rivaroxaban group and 11/125 patients (8.8%) in the warfarin group, with an absolute risk difference of -3.2% points. The composite outcome occurred in 8/125 (6.4%) and 12/125 (9.6%) patients, respectively, with an absolute risk difference of -3.2% points. The confidence interval for thromboembolic events was very wide, indicating substantial imprecision. Sensitivity analysis using IPTW yielded directionally consistent results. During the first 90 days after surgical bioprosthetic valve replacement, no statistically significant differences in thromboembolic or clinically relevant bleeding outcomes were detected between low-dose rivaroxaban and warfarin. However, thromboembolic events were very rare and the effectiveness analysis was substantially underpowered. These findings should be considered exploratory and hypothesis-generating, and larger prospective studies are needed to provide more precise estimates.