Comparison of three risk assessment methods for venous thromboembolism prophylaxis.

Venous thromboembolism (VTE) is a major problem for hospitalized patients. Nevertheless, adherence to VTE prophylaxis guidelines is suboptimal, partly because of physicians' neglect due to excessive workload. Simplified risk assessment methods (RAMs) facilitate timely risk stratification and thromboprophylaxis. However, the accuracy of such RAMs has not been extensively studied. Using the prospectively collected data from hospitalized patients of Masih-Daneshvari Hospital, we tested the accuracy of the Goldhaber RAM and the Harinath and St. John RAM for VTE prophylaxis, compared with the eighth edition of the American College of Chest Physicians (ACCP) recommendations. We evaluated 1091 patients. With reference to the ACCP recommendations, both RAMs had high specificities for detection of patients at risk of VTE (97.33 and 99.11%, respectively); however, we found significant interdepartment differences. The Goldhaber RAM had superior accuracy in medical oncology patients (P = 0.03), whereas the Harinath and St. John method was superior among surgical patients (P < 0.001). Overall accuracies of Goldhaber RAM for appropriate VTE risk assessment and for proper detection of at-risk patients were close to 60%. Corresponding figures were close to 70% for the Harinath and St. John method. Simplified VTE prophylaxis RAMs are valuable, especially for transmitting electronic alerts and for timely risk assessment and thromboprophylaxis. Both of the studied RAMs had high specificities and positive-predictive values, minimizing the risk of overprophylaxis. Improving the sensitivity of such RAMs can help for timely risk assessment for a greater array of real-world patients.