Thirteen randomised controlled trials (RCTs) were included (total number of patients not stated).
Over the 90-day treatment period, the RRs of recurrent VTE for LMWH, compared with UFH patients, were 0.39 (95% confidence interval, CI: 0.3, 0.4, p < 0.05) and 1.7 (95% CI: 0.3, 3.7, p > 0.05, non significant) for level 1 and 2 studies, respectively, in favour of LMWH.
The RRs of major bleeding were 0.42 (95% CI: 0.2, 0.9, p < 0.05) and 0.85 (95% CI: 0.3, 1.9, p > 0.05, non significant) for level 1 and 2 studies, respectively, in favour of LMWH; similarly, for minor bleeding the RRs were 1.16 (95% CI: 0.7, 1.9, p > 0.05, non significant) and 1.03 (95% CI: 0.5, 1.8, p > 0.05, non significant) for level 1 and 2 studies, respectively, in favour of LMWH.
The overall mortality rate was 3.2% in the LMWH group and 5.9% in the UFH group (risk reduction of 49%, p = 0.01) in favour of LMWH. The mortality rates were 13.5% in cancer patients treated with LMWH, and 28.4% in those treated with UFH (risk reduction of 67%, p = 0.01). In contrast, the mortality rate in non-cancer patients was not significantly different between the LMWH- and UFH-treated groups.