Medically-ill patients are at increased risk for developing venous thromboembolism (VTE) due to multiple risk factors. Evidence-based practice guidelines recommend pharmacologic modalities such as unfractionated heparin (UFH), low molecular weight heparins (LMWH; i.e.enoxaparin, dalteparin, nadroparin) and pentasaccharide fondaparinux for the prevention of VTE. We conducted systematic review and meta-analysis to evaluate the role of different interventions in prevention of VTE in medically-ill patients. We searched MEDLINE, EMBASE and the Cochrane Controlled Trials Register from 1981–2006 for randomized controlled trials reflecting contemporary antithrombin strategies against in-hospital VTE. Search terms included heparin, enoxaparin, dalteparin, nadroparin, and fondaparinux. Eligible studies included medically-ill patients with risk factors for VTE followed for 7–21 days. We constructed a meta-analysis using Comprehensive Meta Analysis, version 2.0 using the Der-Simonian-Laird random effects model as a sensitivity analysis. Primary end points included the incidence of DVT or PE. Secondary end points included minor and major bleeding, and death from VTE. Eight trials representing 9914 patients were selected for analysis. Four trials compared a LMWH to placebo and four trials compared a LMWH to UFH. Fondaparinux was considered an ultra LMWH. The mean age of represented patients was 72.7±4.89 years, with a mean weight of 68±2.57 kg. The results of primary end points are; 1. DVT-LMWH compared to placebo reduced the odds of DVT by 42% (OR=0.58, 95% CI [0.45–0.75], p≤0.001) with an absolute risk reduction (ARR) of 2.21% and number needed to treat (NNT) of 46. Compared to either placebo or UFH, LMWH reduced the odds of DVT by 37% (OR=0.63, 95% CI [0.50–0.79], p≤0.001) with an ARR of 1.28% and NNT of 78. 2. PE-There was no difference in the incidence of PE with either LMWH, placebo (p=0.16) or UFH (p=0.73). A 36% reduction in the odds PE or DVT (OR=0.64, 95% CI [0.49–0.82], p≤0.01, ARR=1.44%, NNT=70) was observed using a LWMH instead of UFH or placebo. The latter results were driven by the PRIME and PREVENT studies using enoxaparin and dalteparin respectively. The results of secondary end points; 1. Minor bleeding-LMWH use was associated with an increased risk of minor bleeding (OR=1.64, 95% CI [1.18–2.29], p=0.003) compared to placebo with an absolute risk increase (ARI) of 2.24% and a number needed to harm (NNH) of 45; where MEDNOX (enoxaparin study) was responsible for this difference. 2. Major bleeding-There was no difference in major bleeding with LMWH compared to placebo (p=0.18) or UFH (p=0.42). 3. VTE mortality-LMWH was similar to placebo (p=0.63) and UFH (p=0.47) for the reduction of VTE related death. LMWH also increased the risk of any bleeding or VTE-related death (OR=1.57, 95% CI [1.17–2.10], p=0.003, ARI=2.45%, NNH=41) compared to placebo which also was driven by an increase in minor bleeding from enoxaparin use. In conclusion, although there is a significant reduction in DVT when utilizing a LMWH rather than UFH or placebo, there is however, no difference in PE events. Furthermore, it is important to consider the increased incidence of minor bleeding with LMWH (NNH 43) which may offset the reduction in DVT (NNT 45). The objective of future studies should focus on the identification of potential subgroups of patients with higher risk of VTE and lower risk of bleeding to improve outcomes and optimize safety.

Disclosures: The study was supported by a grant from GSK.; Member of speaker’s bureau of GSK.

Author notes


Corresponding author