Meta-analysis of low molecular weight heparin in the prevention of venous thromboembolism in general surgery

Background: Low molecular weight heparins (LMWHs) have become routine thromboprophylaxis in general surgery. However, their actual clinical effect, its magnitude relative to that of unfractionated heparin (UFH), and the optimal dose are still debated. Methods: A meta‐analysis was performed of all available randomized trials in general surgery comparing LMWH with placebo or no treatment, or with UFH. Results: Comparison versus placebo or no treatment confirmed that the significant reduction in asymptomatic deep vein thrombosis (DVT) obtained with LMWH (n = 513; relative risk (RR) 0·28 (95 per cent confidence interval 0·14–0·54)) was associated with a significant reduction in clinical pulmonary embolism (n = 5456; RR 0·25 (0·08–0·79)) and clinical venous thromboembolism (VTE) (n = 4890; RR 0·29 (0·11–0·73)), and a trend towards a reduction in overall mortality rate. Comparison versus UFH showed a trend in favour of LMWH, with a significant reduction in clinical VTE (P = 0·049), a trend also found for cancer surgery. LMWH at doses below 3400 anti‐Xa units seemed to be as effective as, and safer than, UFH, while higher doses yielded slightly superior efficacy but increased haemorrhagic risk, including that of major haemorrhage. Conclusion: Asymptomatic DVT may be regarded as a reliable surrogate endpoint for clinical outcome in studies investigating thromboprophylaxis in general surgery. LMWH seems to be as effective and safe as UFH. Determination of the optimal dose regimen of LMWH for this indication requires further investigation. © 2001 British Journal of Surgery Society Ltd