Abstract:

Objective To explore the independent risk factors for deep vein thrombosis（DVT） in patients with multiple injuries，develop a risk prediction model for DVT and evaluate the effect. Methods By retrospective analysis method，425 multiple trauma patients from a tertiary hospital trauma center in Beijing from 2022 to 2023 were collected，and the patients were divided into DVT and non-DVT groups according to ultrasound results，and the variables were screened by univariate and binary logistic regression analysis. An visualized nomogram was constructed by R software，and the Bootstrap method was applied for the internal validation and model adjustment. Results The incidence of DVT in patients with multiple injuries in this study was as high as 26.82%. Logistic regression analysis showed that age，combined pelvic and acetabular fracture，combined lower extremity fracture，blood transfusion，and fibrinogen degradation products（FDP）>5 ug/ml within 24~48 h of admission were the independent risk factors for the occurrence of DVT. A risk prediction model for DVT in patients with multiple trauma was constructed and evaluated based on the above five factors，which predicted DVT with an AUC of 0.777，a Brier score of 0.049，a sensitivity of 0.728，and a specificity of 0.704. In the validation set，the AUC for predicting DVT was 0.749，the Brier score was0.070，the sensitivity was 0.553，and specificity was 0.826. Conclusion The model constructed in this study provides a simple method to calculate the DVT risk of patients with multiple trauma，and the indicators can be automatically extracted through the information system，providing a reference for early clinical screening of high-risk DVT patients with multiple trauma.

Key words: Multiple Trauma, Deep Vein Thrombosis, Risk Prediction Modeling, Risk Factors