Efficacy and safety of extended thromboprophylaxis for medically ill patients. A meta-analysis of randomised controlled trials.

Compelling evidence suggests that the risk of pulmonary embolism (PE) and deep-vein thrombosis (DVT) persists after hospital discharge in acutely-ill medical patients. However, no studies consistently supported the routine use of extended-duration thromboprophylaxis (ET) in this setting. We performed a meta-analysis to assess efficacy and safety of ET in acutely-ill medical patients. Efficacy outcome was defined by the prevention of symptomatic DVT, PE, venous thromboembolism (VTE) and VTE-related mortality. Safety outcome was the occurrence of major bleeding (MB) and fatal bleeding (FB). Pooled odds ratios (ORs) and 95 % confidence intervals (95 %CI) were calculated for each outcome using a random effects model. Four RCTs for a total of 28,105 acutely-ill medical patients were included. ET was associated with a significantly lower risk of DVT (0.3 % vs 0.6 %, OR 0.504, 95 %CI: 0.287-0.885) and VTE (0.5 % vs 1.0 %, OR: 0.544, 95 %CI: 0.297-0.997); a non-significantly lower risk of PE (0.3 % vs 0.4 %, OR 0.633, 95 %CI: 0.388-1.034) and of VTE-related mortality (0.2 % vs 0.3 %, OR 0.687, 95 %CI: 0.445-1.059) and with a significantly higher risk of MB (0.8 % vs 0.4 %, OR 2.095, 95 %CI: 1.333-3.295). No difference in FB was found (0.06 % vs 0.03 %, OR 1.79, 95 %CI: 0.384-8.325). The risk benefit analysis showed that the NNT for DVT was 339, for VTE was 239, and the NNH for MB was 247. Results of our meta-analyses focused on clinical important outcomes did not support a general use of antithrombotic prophylaxis beyond the period of hospitalization in acutely-ill medical patients.