Twelve RCTs (n=1,951) were included.
All of the studies used proper methods to generate the randomised treatment allocation, and appeared to adequately conceal treatment allocation. One of the 12 studies blinded both the patients and investigators to treatment allocation. Five studies had 100% clinical follow-up, while 7 studies had 90% or greater follow-up. Further information on the quality of the included studies is available on the Annals of Internal Medicine website (accessed 07/07/2005). See Web Address at end of abstract.
Symptomatic VTE at the end of heparin treatment.
There was no significant difference between symptomatic events with LMWH compared with UFH (1.4% versus 2.4%; OR 0.63, 95% CI: 0.33, 1.18); there was no evidence of statistical heterogeneity. There was no significant difference between LMWH and UFH in patients presenting with symptomatic PE (1.7% versus 2.3%; OR 0.72, 95% CI: 0.35, 1.48) and patients presenting with asymptomatic PE in the context of symptomatic deep venous thrombosis (1.2% versus 3.2%; OR 0.53, 95% CI: 0.15, 1.88).
Symptomatic VTE at 3 months.
There was no significant difference in symptomatic events with LMWH compared with UFH (3.0% versus 4.4%; OR 0.68, 95% CI: 0.42, 1.09); there was no evidence of statistical heterogeneity. There was no significant difference between LMWH and UFH in patients presenting with symptomatic PE (3.3% versus 4.3%; OR 0.72, 95% CI: 0.4, 1.28) and patients presenting with asymptomatic PE in the context of symptomatic deep venous thrombosis (3.5% versus 3.8%; OR 1.07, 95% CI: 0.4, 2.91).
All-cause mortality.
There was no significant difference in the rate of all-cause mortality between patients receiving LMWH and those receiving UFH, either at the end of treatment (1.4% versus 1.2%; OR 1.20, 95% CI: 0.59, 2.45) or at 3 months (4.7% versus 6.1%; OR 0.77, 95% CI: 0.52, 1.15).
Bleeding.
There was no significant difference between patients receiving LMWH and those receiving UFH in the incidence of major bleeding (1.4% versus 2.3%; OR 0.67, 95% CI: 0.36, 1.27) or minor bleeding (6.8% versus 5.5%; OR 1.08, 95% CI: 0.73, 1.59). There was no evidence of statistical heterogeneity in either analysis.
Different LMWH preparations.
There was no evidence of differences in the efficacy or safety outcomes between different LMWH preparations (data not shown).
The primary outcome did not differ with the removal of individual studies, nor with the removal of low-quality studies (studies with incomplete follow-up). There were no important differences in the results of the fixed-effect analysis compared with the random-effects analysis.
The funnel plot was reported to show no evidence of publication bias (plot not shown).