A Prognostic Nomogram Model to Predict Deep Vein Thrombosis in Preoperative Pelvic Fracture Patients
Objective To investigate the risk and prognostic factors of deep vein thrombos(DVT)is in pelvic fracture patients before surgery and to develop a nomogram model to predict the risk of DVT.Methods A retrospective observational study was conducted,including 665 patients hospitalized in the emergency intensive care unit(EICU)with pelvic fractures from January 2017 to January 2022.The value of stress hyperglycemia ratio(SHR)was calculated based on the patient's immediate blood glucose and hemoglobin A1c(Hb A1c).According to the results of ultrasound examination,the subjects were divided into DVT group and non-DVT group.There were 236 patients in the DVT group,including 165 males and 71 females,aged from 18 to 89 years old,with an median age of 56 years old.There were 429 patients in the non-DVT group,including 312 males and 117 females,aged from 19 to 92 years old,with an median age of 49 years old.The differences of demographic characteristics,laboratory indicators,injury severity score(ISS)and other data were compared.Logistic regression analysis was used to screen for risk factors in the DVT group.Logistic regression analysis was applied to identify independent risk factors for DVT after pelvic fracture.Receiver operating characteristic(ROC)curve was used to analyze the predictive the clinical value of those parameters to predict preoperative DVT.Results Significant differences were observed between the DVT and non-DVT groups in age,ISS,hemoglobin,platelet,albumin,blood glucose,SHR,the 24-hour RBC transfusion volume and length of hospital stay(P＜0.05).Logistic regression analysis revealed that the following factors were independent risk factors for DVT:Age between 50 and 69 years(OR=2.02,95％CI:1.4～2.95,P＜0.001),age over 70 years(OR=4.74,95％CI:2.57～8.97,P＜0.001),ISS(OR=1.57,95％CI:1.11～2.23,P=0.012),hemoglobin levels of 100～124 g/L(OR=0.53,95％CI:0.36～0.80,P=0.002),hemoglobin levels over 124 g/L(OR=0.32,95％CI:0.21～0.50,P＜0.001),and SHR(OR=1.64,95％CI:1.12～2.41,P=0.011).ROC curve analysis demonstrated that the area under the ROC curve(AUC)for age,ISS,hemoglobin,and SHR in predicting DVT were 0.63,0.59,0.64,and 0.57,respectively,with an overall AUC of 0.72.A predictive nomogram model was constructed using these identified risk factors.The calibration chart of the model indicated a good agreement between the predicted probability and the actual occurrence probability,and decision curve analysis confirmed the nomogram's ability to predict the risk of DVT.Conclusion Age,ISS,hemoglobin levels,and SHR are independent risk factors for DVT in preoperative pelvic fracture patients.The nomogram prediction model,based on logistic regression,exhibits good predictive efficacy for DVT in this patient population.
万方数据
维普
