Establishment and Verification of a Nomogram Model to Predict the Risk of Postoperative Venous Thromboembolism in Patients with Thoracolumbar Fracture
Objective:To construct and verify a Nomogram model for predicting the risk of postoperative venous thromboembolism(VTE)in patients with thoracolumbar fracture,and thus providing a reference for prevention and treatment of VTE.Methods:A total of 252 patients with thoracolumbar fracture admitted to our hospital from June 2020 to February 2023 were enrolled.Lower extremity vascular ultrasound examination was performed at 14 days postoperatively.According to the presence and absence of VTE,the patients were divided into VTE group(n=43)and non-VTE group(n=209).The clinical data of the two groups were as-sessed,and the influencing factors of postoperative VTE in patients with thoracolumbar fracture were analyzed by least absolute shrinkage and selection operator(LASSO)-Logistic regression.A Nomogram prediction mod-el was constructed,and the model was evaluated by calibration curve and receiver operating characteristic(ROC)curve.Results:The ratio of patients≥60 years old,hyperlipidemia,general anesthesia,lumbar and back fascia injury,bone mineral density≥-2.5g/cm2 and body mass index in VTE group were significantly higher than those in non-VTE group,the time from fracture to admission,the time from fracture to operation and the operation time were significantly longer than those in non-VTE group.Hemoglobin(Hb)and albumin(ALB)in VTE group were significantly lower than those in non-VTE group(P<0.05).The regression screening variables by LASSO regression were age,lumbar and back fascia injury,pooled osteoporosis,body mass index,time from fracture to admission,ALB,Hb,and operation time.Age,lumbar and back fascia in-jury,pooled osteoporosis,body mass index,time from fracture to admission,ALB,operation time,and Hb were factors that influence the incidence of postoperative VTE in patients with thoracolumbar fractures(P<0.05).Based on the results obtained from LASSO-logistic regression,a Nomogram prediction model was con-structed.The ROC curve showed that the model had good predictive value for predicting postoperative VTE in patients in both the training and validation sets.The calibration results showed a high consistency degree of a-greement between the predicted incidence of postoperative VTE and actual events in both sets.Conclusion:The factors influencing the occurrence of postoperative VTE in patients with thoracolumbar fractures were age,lumbar and back fascia injury,pooled osteoporosis,body mass index,time from fracture to admission,ALB,Hb,and operation time.Based on these factors,a Nomogram prediction model was constructed to predict the occurrence of postoperative VTE with good consistency and efficacy.
万方数据
维普
