Abstract: BACKGROUND: For elderly patients with hip fractures, improving the ability to screen for deep vein thrombosis of the lower extremities and implementing individualized prevention programs are important measures to improve the prognosis of patients and reduce the burden on patients' families and socioeconomics.
OBJECTIVE: To investigate the risk factors of deep venous thrombosis in elderly patients with hip fracture after operation, and to establish a risk model to predict the occurrence of deep venous thrombosis. 
METHODS: Clinical data of 325 hospitalized patients with hip fracture over 70 years old from January 2017 to December 2019 were retrospectively analyzed, and they were divided into deep venous thrombosis group (n=50) and non-deep venous thrombosis group (n=275) according to the results of deep vein color Doppler ultrasound examination of lower extremities. The univariate and multivariate Logistic regression analyses were used to identify the independent risk factors for postoperative deep venous thrombosis in elderly patients with hip fracture. Receiver operating characteristic curve was used to analyze the diagnostic efficacy of independent risk factors and combined models for deep venous thrombosis. Furthermore, R Studio software was used to construct a column chart to predict the risk of postoperative deep venous thrombosis of lower limbs in elderly patients with hip fracture, and calibration curves were made to verify the accuracy of the model. 
RESULTS AND CONCLUSION: (1) The univariate analysis results showed that age, fracture type, the time between injury to operation, anesthesia, operation time, preoperative basic diseases, potassium, cholesterol, creatinine, and total protein were significantly different between the two groups (P < 0.05). (2) Multivariate Logistic regression analysis showed that fracture type of intertrochanteric fracture of femur, time from injury to operation, presence of diabetes before operation, and cerebrovascular disease were independent risk factors for postoperative deep venous thrombosis in elderly patients with hip fracture 
(P < 0.05). (3)  Receiver operating characteristic curve of each independent risk factor and combined model exhibited that the time from injury to operation, the presence of diabetes, cerebrovascular disease, and the combined model had predictive ability, but the type of hip fracture did not have predictive ability, and the combined model had better predictive ability than the single independent risk factor. (4) The column chart has good diagnostic efficiency and accuracy.

Key words: advanced age, hip fracture, deep vein thrombosis, regression analysis, risk factor, column chart