How complete is the evidence for thromboembolism prophylaxis in general medicine patients? A meta-analysis of...

Guidelines recommend pharmacologic prophylaxis for hospitalized medical patients at increased risk of thromboembolism. Despite recommendations, multiple studies demonstrate underutilization. Factors contributing to underutilization include uncertainty that prophylaxis reduces clinically relevant events, as well as questions about the best form of prophylaxis. We sought to determine whether prophylaxis decreases clinically significant events and to answer whether unfractionated heparin (UFH) or low molecular weight heparin (LMWH) is either more effective or safer.

The MEDLINE, EMBASE, CINAHL, and Cochrane databases were searched through June 2008. Relevant bibliographies and conference proceedings were reviewed and LMWH manufacturers were contacted.

Randomized trials comparing UFH or LMWH to control, as well as head-to-head comparisons of UFH to LMWH in general medicine patients.

End points of deep venous thrombosis (DVT), proximal or symptomatic DVT, pulmonary embolism, mortality, bleeding, and thrombocytopenia were extracted from individual trials. Pooled relative risks were calculated using random effects modeling.

We identified 8 trials comparing prophylaxis to control, and 6 trials comparing UFH to LMWH. Prophylaxis reduced DVT (relative risk [RR] = 0.55; 95% confidence interval [CI]: 0.36-0.92), proximal DVT (RR = 0.46; 95% CI: 0.31-0.69), and pulmonary embolism (RR = 0.70; 95% CI: 0.53-0.93). Prophylaxis increased the risk of any bleeding (RR = 1.54; 95% CI: 1.15-2.06) but not major bleeding. Across trials comparing LMWH to UFH, there were no differences for any outcome.

Among medical patients, pharmacologic prophylaxis reduced the risk of thromboembolism without increasing risk of major bleeding. The current literature does not demonstrate superior efficacy of UFH or LMWH.