老年急性脑梗死患者静脉溶栓后不良结局的预测模型的研究  

A model for predicting the probability of poor outcome at 3 months after intravenous thrombolysis for elderly patients with acute cerebral infarction

作　　者：徐伟 李辉萍[2] 王振[1] 贺国华 胡珏 宋康平 童洋萍 李芳毅 郭洪权 刘新峰 Xu Wei;Li Huiping;Wang Zhen;He Guohua;Hu Jue;Song Kangping;Tong Yangping;Li Fangyi;Guo Hongquan;Liu Xinfeng(Department of Neurology,Affiliated Changsha Central Hospital,Hengyang Medical School,University of South China,Changsha 410004,China;Department of Rehabilitation,The People's Hospital of Hunan Province(The First Affiliated Hospital of Hunan Normal University),Changsha 410005,China;Department of Neurology,Jinling Hospital,the First School of Clinical Medicine,Southern Medical University,Nanjing 210002,China)

机构地区：[1]南华大学衡阳医学院附属长沙中心医院(长沙市中心医院)神经内科,长沙410004 [2]湖南省人民医院(湖南师范大学附属第一医院)康复科,长沙410005 [3]南方医科大学第一临床学院附属金陵医院(东部战区总医院)神经内科,南京210002

出　　处：《中华老年医学杂志》2022年第11期1303-1309,共7页Chinese Journal of Geriatrics

基　　金：湖南省卫生健康委科研计划项目(202103070273,20201928);湖南省自然科学基金资助项目(2021JJ70091);长沙市科技计划项目经费资助(kzd2001068);南华大学附属长沙中心医院资助项目(YNKY202211)。

摘　　要：目的探讨老年急性脑梗死(ACI)患者静脉溶栓(IVT)治疗3个月后不良结局的预测因子,构建预测不良结局的列线图预测模型。方法回顾性队列研究。收集2016年1月至2021年4月在我院接受IVT治疗的346例老年ACI患者的临床、实验室和影像学资料。将3个月后改良Rankin量表(mRS)>2分定义为预后不良,采用Logistic回归分析筛选老年ACI患者IVT治疗后不良结局的独立预测因子,然后利用R软件建立相应的列线图模型。使用ROC曲线、校准曲线以及决策曲线分析分别评价列线图模型的区分度、校准度以及临床应用价值。结果346例入选者中,109例预后不良,预后不良率31.5%。Logistic回归分析结果显示,症状性出血转化(OR=15.647,95%CI:8.913~27.454)、卒中严重程度(中度卒中OR=3.322,95%CI:1.414~7.811;中-重度卒中OR=8.169,95%CI:4.102~16.258;重度卒中OR=9.653,95%CI:5.440~17.121)、卒中相关性肺炎(OR=2.239,95%CI:1.134~4.420)、心力衰竭(OR=2.758,95%CI:1.424~5.336)是预测老年ACI患者IVT治疗3个月后不良结局的独立预测因子(均P<0.05)。基于以上预测因子构建的列线图模型,其受试者工作特征曲线下面积(AUC)为0.85(95%CI:0.80~0.89),表明该模型具有良好的区分度;校准曲线显示平均绝对误差为0.020,表明该模型具有良好的校准度;决策曲线分析表明该模型具有良好的临床应用价值。结论基于症状性出血转化、卒中严重程度、卒中相关性肺炎、心力衰竭这四个预测因子构建的列线图模型可以预测老年ACI患者IVT治疗3个月后不良结局,预测准确度高,具有较高的临床应用价值。Objective To explore independent predictors for poor outcome at 3 months in elderly patients with acute cerebral infarction(ACI)treated with intravenous thrombolysis(IVT),and to develop a nomogram-based predictive model.Methods This was a retrospective cohort study.Clinical,laboratory and imaging data of 346 elderly patients with ACI treated with IVT from January 2016 to April 2021 in our hospital were collected.Poor outcome was defined as a modified Rankin Scale(mRS)score>2 at 3 months after the stroke.Logistic regression analysis was used to screen for independent factors predicting poor outcome in elderly ACI patients treated with IVT,and a corresponding nomogram model was developed using the R software.The ROC curve,calibration plots and decision curve analysis were used to evaluate discrimination,calibration and clinical application value of the nomogram model.Results Among 346 candidates,109 developed a poor outcome,representing a rate of 31.5%.Logistic regression analysis showed that symptomatic hemorrhagic transformation(OR=15.647,95%CI:8.913-27.454),stroke severity(moderate stroke,OR=3.322,95%CI:1.414-7.811;moderate-severe stroke,OR=8.169,95%CI:4.102-16.258;severe stroke,OR=9.653,95%CI:5.440-17.121),stroke-associated pneumonia(OR=2.239,95%CI:1.134-4.420),and heart failure(OR=2.758,95%CI:1.424-5.336)were independent predictors for poor outcome at 3 months in elderly ACI patients treated with intravenous thrombolysis(all P<0.05).With the area under curve(AUC-ROC)value at 0.85(95%CI:0.80-0.89),the nomogram model,which was composed of the above four predictors,demonstrated good discrimination.On the calibration plot,the mean absolute error was 0.020,indicating that the model had good calibration.Decision curve analysis revealed that the model had good clinical application value.Conclusions The nomogram model composed of symptomatic hemorrhagic transformation,stroke severity,stroke-associated pneumonia and heart failure may predict poor outcome at 3 months in elderly ACI patients treated with IVT,with high pr

关 键 词：脑梗死 静脉溶栓 预后 列线图 

分 类 号：R743.3[医药卫生—神经病学与精神病学]