Abstract
OBJECTIVE: The development and validation of a nomogram to predict the risk of venous thromboembolism VTE in depressed patients requiring hospitalization. METHODS: A retrospective cohort of 726 depressed inpatients who were treated between December 2023 to October 2025 was analyzed in this study. Computer-generated random numbers were used to randomly assign the patients (to training 70% and validation 30% cohorts). Analysis was performed as logistic regression with backward stepwise selection, which is based on Akaike information criterion. Variance inflation factor was used to test multicollinearity. ROC, calibration and decision curve analysis were used to assess model performance. RESULTS: Predictors of age, male gender, hypertension, antidepressants, antipsychotics, D-dimer ≥ 0.5 μg/mL, and length of hospital stay were independent variables. Good discrimination (AUC 0.827 training; 0.822 validation), good calibration, and good clinical utility of the nomogram were observed. CONCLUSION: The tested nomogram allows stratifying risk of VTE in patients with depression and could be used to inform specific thromboprophylaxis in the psychiatric ward.