Five open-label RCTs (n=1,158) were included in the review.
Type of randomisation (sequence generation and allocation concealment) was clear in two studies. Allocation concealment was clear in three studies. Outcome assessment was open in all trials. Three studies reported use of a blinded adjudication committee for evaluation of outcomes. All five trials reported withdrawals and drop outs; one study had a withdrawal/termination rate of 52%. Intention-to-treat analysis was reported adequately in two studies. In all, three of the five studies were considered to have a high methodological quality. Sample sizes ranged from 35 to 676 participants.
LMWH significantly reduced risk of venous thromboembolism recurrence (RR 0.53, 95% CI 0.36 to 0.76; n=1,158 participants) compared to VKA. Heterogeneity was not observed. Studies that evaluated anticoagulation for six months showed significant statistical evidence in favor of LMWH; the three-month therapy was not statistically significant. There was no statistically significant difference between LMWH and VKA in risk of a major bleeding event (five studies, n=1,158 participants) or minor bleeding events (four studies, n=1,012 participants) or all-cause mortality (three studies, n=1,022 participants). All results were similar for fixed-effect and random-effects models.