维持性血液透析患者认知衰弱风险预测模型的构建及应用  

Construction and application of a risk prediction model of cognitive frailty in maintenance hemodialysis patients

作　　者：韦微光 何联义[1] 唐喜宁 吕华 雷秋连 WEI Weiguang;HE Lianyi;TANG Xining;LYU Hua;LEI Qiulian(Hemodialysis Room,Nanning Second People's Hospital,Wuxiang Hospital,Guangxi Zhuang Autonomous Region,Nanning 530031,China)

机构地区：[1]南宁市第二人民医院五象医院血透室,广西南宁530031

出　　处：《中国当代医药》2024年第35期130-135,共6页China Modern Medicine

基　　金：广西壮族自治区中医药管理局自筹经费科研课题(GXZYZ20210094,GXZYA20230423);广西壮族自治区卫生健康委自筹经费科研课题(Z-A20241031)。

摘　　要：目的探究维持性血液透析(MHD)患者认知衰弱风险预测模型的构建及应用效果。方法选择南宁市第二人民医院2021年1月至2023年12月收治的507例MHD患者为研究对象,将2021年1月至2022年6月的284例患者纳入建模组,将2022年7月至2023年12月的223例患者纳入验证组。结合单因素和多因素logistic回归分析影响认知衰弱发生的独立危险因素,并以此为基础构建预测模型,绘制ROC曲线和校准曲线图,评估预测模型对MHD患者认知衰弱发生风险的区分度和准确度。结果单因素分析显示,年龄、生活自理能力、受教育程度、糖尿病、高血压、抑郁症、睡眂质量、社会支持、血清白蛋白、C反应蛋白、血红蛋白、血肌酐、甘油三酯、总胆固醇、低密度脂蛋白与MHD患者认知衰弱有关(P<0.05)。多因素logistic回归分析显示,年龄(β=1.012,OR=3.115,95%CI:1.002~5.986)、生活自理能力(β=1.251,OR=5.051,95%CI:1.025~9.851)、受教育程度(β=1.205,OR=5.125,95%CI:1.332~8.485)、糖尿病(β=1.346,OR=3.768,95%CI:1.524~6.892)、高血压(β=1.242,OR=3.158,95%CI:1.215~5.489)、抑郁症(β=1.258,OR=3.481,95%CI:1.333~4.859)、睡眂质量(β=1.269,OR=4.362,95%CI:1.105~7.154)、社会支持(β=1.344,OR=4.256,95%CI:1.514~8.058)、血清白蛋白(β=1.273,OR=1.328,95%CI:1.006~8.763)、C反应蛋白(β=1.132,OR=1.133,95%CI:1.001~1.987)、血红蛋白(β=1.458,OR=4.438,95%CI:1.014~8.859)、血肌酐(β=1.439,OR=4.415,95%CI:1.059~8.056)、甘油三酯(β=1.392,OR=3.116,95%CI:1.061~10.875)、总胆固醇(β=1.256,OR=4.125,95%CI:1.844~6.987)、低密度脂蛋白(β=1.248,OR=4.269,95%CI:1.829~6.556)是MHD患者认知衰弱的独立危险因素(P<0.05)。ROC曲线显示,灵敏度为97.14%,特异度为93.46%,诊断符合率为94.37%,AUC为0.945,模型预测维持性血液透析认知衰弱的校准曲线斜率与1接近。结论将认知衰弱风险预测模型应用于MHD患者中能够有效预测患者认知衰弱风险,可行性较好,建议推广。Objective To explore the construction and application of cognitive frailty in maintenance hemodialysis(MHD)patients.Methods A total of 507 MHD patients admitted to Nanning Second People's Hospital from January 2021 to December 2023 were selected as the study objects,284 patients from January 2021 to June 2022 were included in the modeling group and 223 patients from July 2022 to December 2023 in the validation group.Combined with univariate and multivariate logistic regression analysis,the independent risk factors affecting the occurrence of cognitive frailty were analyzed,and the prediction model was built on this basis,and the ROC curve and calibration curve were drawn to evaluate the differentiation and accuracy of the prediction model for the risk of cognitive frailty in MHD patients.Results Univariate analysis showed that Age,self-care ability,education,diabetes,hypertension,depression,quality of sleep,social support,serum albumin,C-reactive protein,hemoglobin,blood creatinine,triglycerides,total cholesterol,low-density lipoprotein is associated with cognitive frailty in patients with MHD(P<0.05).Multivariate logistic regression analysis showed that age(β=1.012,OR=3.115,95%CI:1.002-5.986),self-care ability(β=1.251,OR=5.051,95%CI:1.025-9.851),educational level(β=1.205,OR=5.125,95%CI:1.332-8.485),diabetes mellitus(β=1.346,OR=3.768,95%CI:1.524-6.892),hypertension(β=1.242,OR=3.158,95%CI:1.215-5.489),depression(β=1.258,OR=3.481,95%CI:1.333-4.859),sleep quality(β=1.269,OR=4.362,95%CI:1.105-7.154),social support(β=1.344,OR=4.256,95%CI:1.514-8.058),serum albumin(β=1.273,OR=1.328,95%CI:1.006-8.763),C-reactive protein(β=1.132,OR=1.133,95%CI:1.001-1.987),hemoglobin(β=1.458,OR=4.438,95%CI:1.014-8.859),blood creatinine(β=1.439,OR=4.415,95%CI:1.059-8.056),triglycerides(β=1.392,OR=3.116,95%CI:1.061-10.875),total cholesterol(β=1.256,OR=4.125,95%CI:1.844-6.987),low-density lipoprotein(β=1.248,OR=4.269,95%CI:1.829-6.556)were independent risk factors for cognitive frailty in MHD patients(P<0.05).The ROC curve sh

关 键 词：维持性血液透析 认知衰弱 预测模型 

分 类 号：R459.5[医药卫生—治疗学]