Development And Verification Of Prognostic Factors And Predictive Models For Postoperative Hip Fracture In The Elderly

Objective:To explore the independent risk factors affecting the postoperative prognosis of elderly patient swith hip fracture,and to develop and verify the postoperative prognosis model of elderly hip fracture.Methods:The data of elderly patients with hip fracture diagnosed and treated in the Department of Geriatric Orthopaedics and Ersha Orthopaedics of Guangdong Hospital of traditional Chinese Medicine from May 2014 to May 2019 were collected and screened according to the criteria of this study.according to the different medical centers of the patients,they were divided into modeling development group and internal verification group.The medical records of all patients were recorded and analyzed,and the survival condition of the patients one year after operation was obtained by telephone follow-up.The patients in the modeling development group were divided into death group and survival group according to whether they survived one year after operation,and the baseline data were compared with SPSS software.The data of patients in the modeling and development group were analyzed by COX multivariate regression analysis with R software,the independent risk factors affecting postoperative prognosis were identified and the independent risk factors were analyzed and compared by Kaplan-Meie survival analysis,and a prediction model was developed based on risk factors.Calculate the Cray index of the prediction model,draw the ROC curve and model calibration diagram,and preliminarily evaluate the prediction ability of the model.The R software was used to analyze and evaluate the patient data of the internal verification group,and the internal verification of the prediction model was completed.Results:A total of 1120 effective cases were obtained in this study(modeling group(n=864)and internal verification group(n=256).The male-to-female ratio was 323～797,the average age was(82.49±7.14)years,the age span was 65～104 years,and the 1-year survival rate was 83.39%.The results of COX multivariate regression analysis showed gender(HR:1.836,95%CI:1.2661-2.662),age(HR:0.550,95%CI:3.5850～30.956),ASA grade(HR:5.877,95%CI:2.7609-12.512),number of basic medical diseases(HR:4.511,95%CI:2.4146～8.426)and preoperative waiting time(HR:1.803).95%CI:0.9937～3.270),preoperative Hb value(HR:5.270,95%CI:3.4157 to 8.131)and preoperative Alb value(HR:4.512,95%CI:3.0728～6.625)are independent risk factors affecting the postoperative prognosis of elderly patients with hip fracture.The C-index of prediction model for postoperative prognosis of elderly patients with hip fracture was 0.864,the sensitivity of the model was 95.83%,and the specificity was 92.36%.After internal verification,the C-index of this prediction model is recalculated to 0.811.Conclusion:Gender,age,ASA grade,number of basic medical diseases,waiting time before operation,HB value and Alb value before operation are independent risk factors affecting the postoperative prognosis of elderly patients with hip fracture.The prediction model of postoperative prognosis of hip fracture in the elderly based on the above risk factors has high predictive efficiency and can provide reference for judging the postoperative prognosis of patients.