Venous Thromboembolism Prophylaxis in Acute Medically Ill Patients: A Retrospective Cohort Study

Objective: Current guidelines recommend pharmacologic prophylaxis for medical patients at high risk for venous thromboembolism. We aimed to assess the benefit and safety of venous thromboembolism prophylaxis in acutely ill medical patients hospitalized. Methods: We prepared a retrospective cohort study in a tertiary hospital in Israel with patients hospitalized in medical departments with an admission lasting more than 48 hours during 2014-2017. Primary outcome: 30-day mortality. Secondary outcomes: 90-day incidence of pulmonary embolism, symptomatic deep vein thrombosis, and major bleeding. Propensity-weighted logistic multivariate analysis was performed. Results: A total of 18,890 patient-unique episodes were included in the analysis. Of them, 3206 (17.0%) received prophylaxis. A total of 1309 (6.9%) died, 540/3206 (16.8%) of those who received venous thromboembolism prophylaxis and 769/15,864 (4.9%) of those who did not. Prophylaxis was not associated with a reduction in mortality, multivariable-adjusted odds ratio propensity-weighted (OR) 0.99 (95% confidence interval [CI], 0.84-1.14). One hundred forty-two patients (0.7%) developed venous thromboembolism, 44/3206 (1.4%) of those who received prophylaxis and 98/15,864 (0.6%) of those who did not. Prophylaxis was not associated with reduction in venous thromboembolism in the whole cohort, multivariable-adjusted propensity-weighted OR 1.09 (95% CI, 0.52-2.29). Prophylaxis was associated with an increase in major bleeding (multivariable-adjusted propensity-weighted OR 1.24; 95% CI, 1.04-1.48). Conclusion: The current practice of routinely administering venous thromboembolism prophylaxis to medically ill patients considered at high risk for thrombosis resulted in a high risk for bleeding without a clear clinical benefit, and should be reassessed.