Fifteen RCTs (n=15,111) were included in the review. Sample sizes ranged from 75 to 6,706 participants. Trials were randomised by computer, centrally, by telephone, or by a pre-determined random order (e.g. stratified by study centre). Mean follow-up durations, where reported, ranged from 1.0 to 3.1 years.
Warfarin versus placebo (four RCTs): There was no significant difference in the number of thromboembolic events between patients receiving warfarin and those receiving placebo. Trials assessing major bleeding showed a statistically significantly greater frequency in the warfarin group compared to placebo, OR 3.01 (95% CI: 1.31 to 6.92).
Warfarin versus antiplatelet agents (nine RCTs): There was a significant difference in the prevention of systemic embolism, with fewer systemic embolisms occurring in patients receiving warfarin compared to those receiving antiplatelet, OR 0.50 (95% CI: 0.33 to 0.75). There was no significant difference in risk of major bleeding between the two treatment groups.
Warfarin versus low-dose warfarin with or without aspirin (five RCTs): There were no statistically significant differences in the number of thromboembolic events in trials comparing warfarin with low-dose warfarin (four RCTs), or low-dose warfarin with aspirin (two RCTs). There was a significantly greater risk of major bleeding in patients receiving warfarin compared to low-dose warfarin (OR 2.88, 95% CI: 1.09 to 7.60; four RCTs), but no difference between warfarin and low-dose warfarin with aspirin (two RCTs).
There was no significant statistical heterogeneity among trials for any analyses. There was no evidence of publication bias among trials comparing warfarin and antiplatelet agents, but analysis for other comparisons were unclear due to the small number of trials.