[Meta-analysis on randomized trials comparing the results of low-molecular weight heparins to those of fractioned heparins in the prevention of deep venous thrombosis].

The main reason for using low molecular weight heparin (LMWH) is its capacity to keep the antithrombotic effect of standard non-fractioned heparin (NFH) while reducing its hemorrhagic power. Various studies comparing LMWH to NFH gave contradictory results, so we carried out a meta-analysis of randomized trials comparing these two treatments in prevention of deep-vein thrombosis (DVT) (and in prevention of deep-vein thrombosis or pulmonary embolism). From a selection of ten trials we could not show any significant difference in the DVT rates, as well as for the whole of surgical indications as for the subgroups of abdominal and orthopedic surgery. The weighted global difference between the DVT rates is so low that, in spite of the use of meta-analysis, the power is clearly insufficient. The hemorrhagic risk is not statistically reduced as well for the whole of surgical indications as for the subgroups.