The magnitude of an iatrogenic disorder: a systematic review of the incidence of venous thromboembolism for general medical inpatients.

Venous thromboembolism (VTE) is recognized as a potential serious complication of hospitalization, and anticoagulant prophylaxis is widely recommended for medical inpatients. The purpose of this review was to systematically inspect the literature to determine the incidence of VTE without anticoagulant prophylaxis for hospitalized general medical patients. A systematic review was performed using MEDLINE and Cochrane Collaboration searches of the English-language literature. References of the retrieved articles were reviewed for additional studies, and pharmaceutical companies were contacted directly to identify unpublished studies. Studies were eligible if they reported the incidence of symptomatic or asymptomatic VTE in hospitalized general medical patients and included a group which did not receive prophylaxis. Eight trials were identified. Study design, quality, and results were heterogeneous. The incidence of symptomatic VTE without prophylaxis was examined in three trials; the incidence was 2.3% in the only randomized trial, and was 0.8% and 6.2% in two observational studies. The incidence of asymptomatic VTE ranged from 3.7% to 26.0% (median 12.8%). Risk factors identified were older age, prior VTE, and obesity. Though there is little high quality evidence, the best estimate of the incidence of hospital-acquired VTE for general medical patients is approximately 2% for patients at moderate to high risk and <1% for patients at low risk. Further studies are needed to more precisely define a low risk group for whom anticoagulant prophylaxis is unnecessary.