超高龄髋膝关节术后谵妄及心血管并发症风险预测  

Risk prediction of delirium and cardiovascular complications after hip and knee surgeries in elderly patients

作　　者：陈晓玲[1] 钟永洌 刘巧梨[1] 李娜[1] 张志奇[1] 廖威明[1] 黄桂武 Chen Xiaoling;Zhong Yonglie;Liu Qiaoli;Li Na;Zhang Zhiqi;Liao Weiming;Huang Guiwu(Department of Joint Surgery,the First Affiliated Hospital of Sun Yat sen University,Guangzhou 510080,China)

机构地区：[1]中山大学附属第一医院关节外科,广州510080

出　　处：《中华关节外科杂志(电子版)》2024年第5期575-584,共10页Chinese Journal of Joint Surgery(Electronic Edition)

基　　金：中山大学5010基金2013002(K0402002)

摘　　要：目的探讨超高龄患者髋膝关节术后发生谵妄以及心血管并发症的危险因素及构建可视化风险预测模型。方法回顾性收集2018年10月至2021年10月于中山大学附属第一医院行髋膝关节手术的超高龄患者118例,纳入患者年龄≥80岁,初次行髋/膝关节置换手术,无严重的髋关节炎、股骨头坏死或者是髋部骨折的患者,排除行关节翻修术,术前存在肺部感染或泌尿系感染或严重肝肾功能不全,合并严重内科疾病。采用多因素logistic回归分析术后谵妄和心血管并发症的危险因素。应用R语言软件建立预测风险的列线图并构建线上动态预测模型并部署至Shinyapps.io网站;利用受试者工作特征曲线下面积(AUC)、校准曲线及临床决策曲线(DCA)评估模型的预测性能。结果研究显示,对于术后发生谵妄并发症,两组患者的年龄(t=-2.164,P<0.05)、高血压(F=4.635,P<0.05)、慢性阻塞性肺气肿(F=22.861,P<0.05)、术后血液白蛋白水平(Alb)(t=3.539,P<0.05)、术前血红蛋白水平(t=2.366,P<0.05)、静息疼痛(t=-3.180,P<0.05)以及活动疼痛(t=-2.240,P<0.05)比较,差异具有统计学意义;而对于术后发生心血管并发症而言,两组患者的术后血液白蛋白水平(t=2.069,P<0.05)、糖尿病(F=-5.696,P<0.05)、静息疼痛(t=-3.758,P<0.05)以及活动疼痛(t=-2.406,P<0.05)比较,差异具有统计学意义(P<0.05)。多因素logistic回归分析结果显示,术后静息疼痛[比值比(OR)=3.140,95%置信区间(CI)(0.170,2.118)]、慢性阻塞性肺气肿[OR=20.673,95%C(I1.333,4.724)]以及高血压[OR=14.101,95%C(I0.268,5.023)]是术后谵妄并发症发生的独立危险因素(均为P<0.05);而术后静息疼痛[OR=5.522,95%C(I0.557,2.861)]和糖尿病[OR=5.220,95%C(I0.026,3.280)]是术后心血管并发症发生的独立危险因素(均为P<0.05)。研究同时建立了预测模型[AUC=0.903,95%置信区间(CI)(0.810,0.995)],模型内验证C-index=0.903,提示该模型具有一定的预测能力。结论本研Objective To explore the risk factors of delirium and cardiovascular complications in ultra-senile patients after hip and knee arthroplasties,and to build a visual risk prediction model.Methods The clinical data of 118 super-elderly patients who underwent hip and knee arthroplasty in the First Affiliated Hospital of Sun Yat-sen University from October 2018 to October 2021 were retrospectively collected.The patients who were age over 80 years and underwent primary hip or knee arthroplasty were included,while the patients with severe hip arthritis,femoral head necrosis or hip fracture,revision surgery,infection in lungs or urinary system,severe hepatic or renal dysfunction,and combining other severe diseases were excluded.The grouping was based on whether postoperative complications occurred,and they were divided into control group and complication group.The clinical data of the two groups were compared,and the risk factors for postoperative delirium and cardiovascular complications were analyzed by univariate and multivariate logistic regression.R language software was used to establish an alignment chart for predicting risk and construct an online dynamic prediction model and deploy it to the Shinyapps.io website.The predictive performance of the model was evaluated using the area under the curve(AUC),calibration curve,and clinical decision curve analysis(DCA).Results This analysis showed that for postoperative delirium complications,age(t=-2.164,P<0.05),hypertension(F=4.635,P<0.05),chronic obstructive emphysema(F=22.861,P<0.05),postoperative blood albumin(Alb)level(t=3.539,P<0.05),preoperative hemoglobin level(t=2.366,P<0.05),rest pain(t=-3.180,P<0.05)and activity pain(t=-2.240,P<0.05),the differences of these factors were statistically significant(P<0.05).As for postoperative cardiovascular complications,there were statistically significant differences in postoperative blood albumin levels(t=2.069,P<0.05),diabetes mellitus(F=-5.696,P<0.05),rest pain(t=-3.758,P<0.05)and activity pain(t=-2.406,P<0.05)between the

关 键 词：老年人 80以上 关节成形术 置换 谵妄 心血管疾病 预测 

分 类 号：R687.4[医药卫生—骨科学]