J Thromb Thrombolysis. 2025 Nov 27. doi: 10.1007/s11239-025-03215-x. Online ahead of print.
ABSTRACT
Several risk assessment models (RAMs) guide standardized prophylaxis to prevent venous thromboembolism (VTE) in trauma patients. The Caprini, Trauma Embolic Scoring System (TESS), and Greenfield Risk Assessment Profile (RAP) have been validated individually, but their predictive powers have not been directly compared in the general trauma population. This study evaluated the discriminatory ability of these three RAMs in trauma patients at an ACS-verified Level I Trauma Center. A retrospective review was performed of adult trauma patients in a single institution over one year. Demographic and clinical data were used to calculate Caprini, TESS, and RAP scores. The primary outcome was inpatient VTE. Logistic regression models – both combined and separate – generated receiver operating characteristic (ROC) curves for each score. Caprini served as the reference for comparing discriminatory ability. Among 1,276 patients, 33 (2.6%) developed inpatient VTE. Caprini, TESS, and RAP scores predicted VTE with odds ratios of 1.07 (95% CI 1.04-1.10), 1.39 (95% CI 1.23-1.56), and 1.20 (95% CI 1.12-1.29), respectively. ROC c-statistics were similar: Caprini 0.75 (95% CI 0.68-0.82), TESS 0.73 (95% CI 0.64-0.83), and RAP 0.70 (95% CI 0.60-0.79). Caprini, TESS, and RAP RAMs showed comparable moderate discriminatory ability (c-statistic > 0.70) in predicting inpatient VTE among trauma patients. No model was superior, suggesting any of these RAMs may guide standardized VTE prophylaxis in this population.
PMID:41307791 | DOI:10.1007/s11239-025-03215-x
