Risk factors and prediction model for periprosthetic femoral fracture after total knee arthroplasty in the elderly / 局解手术学杂志

ABSTRACT

Objective To investigate the risk factors of periprosthetic femoral fracture(PFF)after total knee arthroplasty(TKA)in the elderly and to construct a predictive model for the prevention of PFF after clinical operation.Methods The clinical data of 537 elderly patients who underwent TKA in the orthopedic department of the First Affiliated Hospital of Air Medical University from October 2016 to October 2021 were retrospectively analyzed.The occurrence of PFF during the follow-up period was statistically analyzed and the clinical data were collected.Binary Logistic regression was used to analyze the risk factors of PFF after TKA in the elderly,and a predictive model of PFF after TKA in the elderly was constructed based on the risk factors.Receiver operating characteristic(ROC)curve and Hosmer-Lemeshow(H-L)were used to test the discrimination and calibration of prediction model.Results The patients were followed up for 12 to 72 months after discharge,with a median time of 47 months.During the follow-up period,31 patients(5.77%)developed PFF.Age,osteoporosis,Parkinson's disease and anterior femoral notch(AFN)were the risk factors for PFF after TKA in the elderly(P<0.05),and cross fixation of prosthesis and bone cement fixation were the protective factors(P<0.05).The results of H-L test showed that the risk prediction model of PFF after TKA in the elderly had good calibration(P>0.05).ROC curve analysis showed that the risk prediction model of PFF after TKA in the elderly has high discrimination(area under the curve was 0.858,95%CI:0.826 to 0.887),the sensitivity was 83.87%,the specificity was 88.34%.Conclusion The risk of PFF after TKA in the elderly is high,and prevention should be carried out according to the high risk factors.The prediction model constructed based on the high risk factors has good prediction efficiency.