摘要
Abstract
Objective:To analyze the influencing factors of postoperative frailty in elderly patients with lower extremity arteriosclerosis obliterans,and to establish the relevant decision tree model.Methods:The clinical data of 260 elderly LEASO patients admitted to the hospital from January 2022 to December 2024 were retrospectively selected.The patients were divided into the frailty group(n=67)and the control group(n=193)according to whether frailty occurred after the operation.Logistic regression was used to analyze the influencing factors of postoperative frailty in elderly LEASO patients,and a related decision tree prediction model was constructed.Results:Among the 260 elderly patients with LEASO,67 cases developed frailty after surgery,with an incidence rate of 25.77%.Logistic regression analysis results indicated that age≥70 years[OR=4.184,95%CI(2.003,8.740)],solitude[OR=3.155,95%CI(1.609,6.190)],operation duration>90 min[OR=2.212,95%CI(1.148,4.261)],preoperative hemoglobin<110 g/L[OR=2.257,95%CI(1.126,4.525)],and preoperative albumin<35 g/L[OR=4.382,95%CI(2.164,8.875)]were risk factors for postoperative frailty in elderly patients with LEASO(P<0.05).The decision tree model constructed based on these risk factors had three layers,13 nodes,and 7 terminal nodes.The model selected preoperative albumin,age,living situation,preoperative hemoglobin,and operation duration as nodes,among which preoperative albumin level was the most important factor for postoperative frailty in elderly patients with LEASO.The model validation results showed that the area under the curve(AUC)was 0.803[95%CI(0.740,0.861)].Conclusions:Age≥70 years old,preoperative hemoglobin<110 g/L,solitude,operation duration>90 min,and preoperative albumin<35 g/L are risk factors for postoperative frailty in elderly LEASO patients.The decision tree model constructed based on these factors has good predictive efficacy.关键词
老年/下肢动脉硬化闭塞症/衰弱/危险因素/决策树/预测模型Key words
elderly/lower extremity arteriosclerosis obliterans/frailty/risk factors/decision tree/prediction model
