股骨头坏死行全髋关节置换术后输血的危险因素分析及列线图预测模型建立

Analysis of risk factors of blood transfusion after total hip arthroplasty for necrosis of femoral head and establishment of nomogram prediction model

目的:探讨股骨头坏死行全髋关节置换术后输血的危险因素,构建此类患者术后输血风险的列线图预测模型。方法:回顾性研究2019年1月至2020年12月因股骨头坏死行初次全髋关节置换的237例患者的病例资料,根据术后是否输血分为输血组60例,未输血组177例,术后输血率为25.3%。收集两组性别、体重指数(BMI)、麻醉方式、术前血红蛋白水平、手术时长、出血量等资料,运用多因素logistic回归分析法分析导致术后输血的危险因素,而后利用R软件构建股骨头坏死全髋关节置换术后输血的列线图预测模型,并绘制受试者操作特征(ROC)曲线来评估列线图模型区分度。结果:多因素分析显示,骨质疏松、术前血红蛋白降低、手术时长、出血量、术后1 d血钙降低是股骨头坏死行全髋关节置换术后输血的危险因素。通过R软件绘制列线图预测模型及ROC曲线,得到曲线下面积(AUC)为0.891,表明该预测模型有较好的区分度,预测术后输血风险与实际发生风险一致性较好。结论:基于骨质疏松、术前血红蛋白降低、手术时长、出血量、术后1 d血钙降低5项独立危险因素构建的预测股骨头坏死患者行全髋关节置换术后输血风险的列线图模型具有良好的区分度和准确度。

Objective: To investigate the risk factors of blood transfusion after total hip arthroplasty for femoral head necrosis, and to construct a nomogram prediction model for the risk of blood transfusion in such patients. Methods: Data of 237 patients undergoing total hip replacement due to femoral head necrosis from January 2019 to December 2020 were retrospectively analyzed. According to whether blood transfusion after operation, 60 patients were included in the transfusion group and 177 patients were included in the nontransfusion group. The postoperative blood transfusion rate was 25.3%. Data such as gender, body mass index(BMI), anesthesia, preoperative hemoglobin, operation time and blood loss were collected, and multivariate logistic regression analysis was used to analyze the risk factors of blood transfusion, and then R software was used to construct a nomogram prediction model of blood transfusion after total hip replacement of femoral head necrosis. ROC curve was used to evaluate the degree of differentiation of the nomogram model.Results: Multivariate analysis showed that operation time, osteoporosis, reduced preoperative hemoglobin, blood loss and blood calcium one day after surgery were risk factors for postoperative blood transfusion. The nomogram prediction model and ROC curve(AUC=0.891) were drawn by R software, indicating that the prediction model had a good degree of discrimination, and the prediction of postoperative blood transfusion risk was in good agreement with the actual risk. Conclusions: A nomogram model based on five independent risk factors of osteoporosis, reduced preoperative hemoglobin, operation time, blood loss and blood calcium one day after surgery constructed to predict the risk of blood transfusion in patients with femoral head necrosis after total hip replacement has good discrimination and accuracy.