老年髋部骨折患者术后低蛋白血症风险预测模型的构建与应用  

Construction and application of hypoproteinaemia risk prediction model in elderly hip fracture postoperation patients

作　　者：周文娟 李可[2] 周建兰 叶寒丽 王慈丽 ZHOU Wenjuan;LI Ke;ZHOU Jianlan;YE Hanli;WANG Cili(Yushan Yuangjiaqi Hospital(Yushan People's Hospital),Shangrao 334700,China;Jiangxi People's Hospital,Nanchang 330000,China)

机构地区：[1]江西省玉山县黄家驷医院(玉山县人民医院),江西省上饶市334700 [2]江西省人民医院,江西省南昌市330000

出　　处：《护理实践与研究》2025年第3期384-390,共7页Nursing Practice and Research

摘　　要：目的基于Logistic回归分析构建老年髋部骨折患者术后低蛋白血症风险预测模型,并验证其预测效果。方法选择医院2020年2月-2023年2月收治的132例老年髋部骨折患者为调查对象,根据术后是否发生低蛋白血症分为病例组与对照组,采用一般资料调查表进行资料收集,并采用Logistic回归分析老年髋部骨折患者术后低蛋白血症的相关因素,建立列线图风险预测模型,检验模型预测效果。结果132例患者中,术后发生低蛋白血症的有48例,发生率为36.36%;Logistic回归分析结果显示,年龄、合并症、术中失血量、术后引流量、术前白蛋白均为老年髋部骨折患者术后低蛋白血症的重要影响因素(P<0.05)。受试者工作特征曲线下面积(Area Under Curve,AUC)为0.924,95%CI:0.888~0.961,最佳临界值为0.950,灵敏度为93.75%,特异度为92.86%,Kappa值为0.745。结论本研究在Logistic回归分析明确的危险因素基础上,建立的老年髋部骨折患者术后低蛋白血症列线图预测模型拟合程度和区分能力良好,且准确度较高,可作为预测老年髋部骨折患者术后发生低蛋白血症风险的有效评估工具。Objective To construct a postoperative hypoproteinaemia risk prediction model for elderly hip fracture patients based on Logistic regression analysis,to verify the predictive effect.Methods 132 cases of elderly hip fracture patients admitted to the hospital from February 2020 to February 2023 were selected as the survey object,to divided into case group and control group according to whether postoperative hypoproteinemia occurred or not,and the general information questionnaire was used for data collection,the Logistic regression analysis was used to analyse the relevant factors of postoperative hypoproteinemia in elderly hip fracture patients,and a column chart risk prediction model was established.A column-line graph risk prediction model was established to test the predictive effect of the model.Results Among the 132 patients,48 had postoperative hypoproteinemia,with an incidence rate of 36.36%.Logistic regression analysis showed that age,comorbidities,intraoperative blood loss,postoperative drainage volume,and preoperative albumin were important factors influencing postoperative hypoproteinemia in elderly patients with hip fracture(P<0.05).The Area Under Curve(AUC)of subjects'work characteristics was 0.924,95%CI:0.888~0.961,with an optimal critical value is 0.950,sensitivity is 93.75%,the specificity is 92.86%,and the Kappa value is 0.745.Conclusion The postoperative hypoproteinemia column chart prediction model established in this study on the basis of logistic regression analysis of clear risk factors has a Sound degree of fit and differentiation ability,and a high degree of accuracy,which could provide an effective tool for predicting the risk of postoperative hypoproteinemia in elderly hip fracture patients.

关 键 词：老年髋部骨折 术后 低蛋白血症 风险预测模型 构建与验证 

分 类 号：R471[医药卫生—护理学]