Heparin-based treatment to prevent symptomatic deep venous thrombosis, pulmonary embolism or death in general medical inpatients is not supported by best evidence.

Prevention of venous thromboembolism (VTE) in medical patients is controversial. In contrast to surgical patients, the evidence supporting the use of heparin-based treatment for prevention of VTE (HVTEp) may not justify current guidelines. This study aims to determine whether current clinical guidelines for HVTEp are appropriate for medical patients. We searched medical databases for original randomised placebo-controlled studies of HVTEp in medical patients, excluding those with stroke and in intensive care. From 401 potentially relevant studies, we selected eight, which included over 16 000 patients. HVTEp decreased the incidence of all deep venous thromboses (DVT): 4.3% in the placebo group versus 2.3% in the treatment group, P = 0.002, number needed to treat, 50. However, this treatment effect was not seen for symptomatic DVT: 1.2% versus 0.9%, P = 0.18, odds ratio (OR) 0.72 (0.45-1.16). Similarly, HVTEp did not decrease the incidence of pulmonary embolism (PE): 0.54% versus 0.27%, P = 0.3, OR 0.57 (0.21-1.53), or fatal PE: 0.1% versus 0.0%, P = 0.3, OR 0.2 (0.01-4.11). Furthermore, HVTEp did not decrease total mortality: 5.63% versus 5.39%, P = 0.92, OR 0.96 (0.78-1.18). The use of HVTEp in hospitalised general medical patients does not result in a significant reduction in symptomatic DVT, PE, fatal PE or total mortality. The best evidence does not support the recommendations of the current clinical guidelines.