Nine RCTs were included in the review (16,701 patients, range 138 to 4,832). There were four phase II dose finding studies and five phase III non-inferiority studies. Risk of bias was considered as low in four studies and unclear in five, Three of the five unclear studies were graded as such due to a single unspecified domain and other domains were classed as low risk. All of the studies used blinded outcome assessors. Follow-up ranged from two weeks to 12 months.
Rivaroxaban significantly reduced major bleeding (RR 0.57, 95% CI 0.39 to 0.84; four RCTs, 8,709 patients) but not recurrent VTE or all-cause mortality. One RCT of ximelagatran (2,489 patients) reported non-significant reductions in all-cause mortality (RR 0.67, 95% CI 0.42 to 1.08) and major bleeding (RR 0.54, 95% CI 0.28 to 1.03) but not recurrent VTE. The other ximelagatran RCT (138 patients) reported few or no events. Neither apixaban (one RCT, 258 patients) nor dabigatran (two RCTs, 5,107 patients) significantly reduced major bleeding, recurrent VTE or all-cause mortality.
The adjusted indirect comparison showed no differences between rivaroxaban and dabigatran for acute VTE (RR 0.78, 95% CI 0.49 to 1.24), major bleeding (RR 0.75, 95% CI 0.41 to 1.34) or all-cause mortality (RR 0.96, 95% CI 0.59 to 1.58).
Results for secondary outcomes were reported. The funnel plots showed a low risk of publication bias for any analysis.