Derivation and Validation of a Risk Factor Model to Identify Medical Inpatients at Risk for Venous Thromboembolism

Background Venous thromboembolism (VTE) prophylaxis is recommended for hospitalized medical patients at high risk for VTE. Multiple risk assessment models exist, but few have been compared in large datasets. Methods We constructed a derivation cohort using 6 years of data from 12 hospitals to identify risk factors associated with developing VTE within 14 days of admission. VTE was identified using a complex algorithm combining administrative codes and clinical data. We developed a multivariable prediction model and applied it to three validation cohorts: a temporal cohort, including two additional years, a cross-validation, in which we refit the model excluding one hospital each time, applying the refitted model to the holdout hospital, and an external cohort. Performance was evaluated using the C-statistic. Results The derivation cohort included 155,026 patients with a 14-day VTE rate of 0.68%. The final multivariable model contained 13 patient risk factors. The model had an optimism corrected C-statistic of 0.79 and good calibration. The temporal validation cohort included 53,210 patients, with a VTE rate of 0.64%; the external cohort had 23,413 patients and a rate of 0.49%. Based on the C-statistic, the Cleveland Clinic Model (CCM) outperformed both the Padua (0.76 vs. 0.72, p = 0.002) and IMPROVE (0.68, p < 0.001) models in the temporal cohort. C-statistics for the CCM at individual hospitals ranged from 0.68 to 0.78. In the external cohort, the CCM C-statistic was similar to Padua (0.70 vs. 0.66, p = 0.17) and outperformed IMPROVE (0.59, p < 0.001). Conclusion A new VTE risk assessment model outperformed recommended models.