轻度急性缺血性脑卒中患者早期发生卒中后认知障碍的风险预测列线图模型构建  

Construction of Nomogram Model for Predicting Risk of the Early Post-Stroke Cognitive Impairment in Patients with Mild Acute Ischaemic Stroke

作　　者：杨训娟 顾正义 丁晴 曹梅 金平 YANG Xunjuan;GU Zhengyi;DING Qing;CAO Mei;JIN Ping(Department of Neurology,Lu'an Hospital of Anhui Medical University,Lu'an 237000,China;Department of Rehabilitation,Affiliated Hospital of West Anhui Health Vocational College,Lu'an 237000,China;Department of Neurology,Affiliated Hospital of West Anhui Health Vocational College,Lu'an 237000,China)

机构地区：[1]安徽医科大学附属六安医院神经内科,安徽省六安市237000 [2]皖西卫生职业学院附属医院康复科,安徽省六安市237000 [3]皖西卫生职业学院附属医院康复科神经内科,安徽省六安市237000

出　　处：《实用心脑肺血管病杂志》2025年第4期38-44,共7页Practical Journal of Cardiac Cerebral Pneumal and Vascular Disease

基　　金：安徽省高等学校自然科学研究项目(2024AH051979)。

摘　　要：目的 构建轻度急性缺血性脑卒中(AIS)患者早期发生卒中后认知障碍(PSCI)的风险预测列线图模型,同时开发动态列线图。方法 回顾性选取2020年3月—2023年6月安徽医科大学附属六安医院收治的轻度AIS患者137例,收集患者临床资料,依据患者是否发生PSCI将其分为PSCI组和非PSCI组。采用多因素Logistic回归分析、最小绝对收缩和选择算子(LASSO)回归分析和支持向量机(SVM)分析探讨轻度AIS患者早期发生PSCI的影响因素并构建相应模型;采用ROC曲线分析三种模型预测轻度AIS患者早期发生PSCI的价值;然后基于最佳模型的影响因素,采用regplot包建立轻度AIS患者早期发生PSCI的风险预测列线图模型;采用Bootstrap法进行1 000次重复抽样,计算一致性指数;采用rms和survival包进行Hosmer-Lemeshow拟合优度检验以评价该列线图模型的拟合程度;绘制决策曲线分析该列线图模型的临床获益;采用Simple模型预测1 000例轻度AIS患者早期发生PSCI的风险,绘制临床影响曲线;采用DynNom包将列线图模型发布至网络以开发动态列线图。结果 随访过程中失访3例,共纳入134例轻度AIS患者,发生PSCI 71例(53.0%)。PSCI组与非PSCI组年龄、受教育程度、有饮酒史者占比、有糖尿病史者占比、有心房颤动史者占比、肌酐(Cr)、血尿素氮(BUN)、尿酸(UA)、HDL-C、血红蛋白(Hb)、空腹血糖(FPG)比较,差异有统计学意义(P<0.05)。多因素Logistic回归分析结果显示,年龄、受教育程度、糖尿病史、Hb、FPG是轻度AIS患者早期发生PSCI的独立影响因素(P<0.05);LASSO回归分析结果显示,年龄、受教育程度、糖尿病史、UA、Hb是轻度AIS患者早期发生PSCI的独立影响因素;SVM分析结果显示,年龄、受教育程度、饮酒史、糖尿病史、UA、HDL-C、Hb、FPG是轻度AIS患者早期发生PSCI的影响因素。ROC曲线分析结果显示,Logistic回归模型、LASSO回归模型、SVM模型预测轻度AIS患者�Objective To construct risk prediction nomogram model of the early post-stroke cognitive impairment(PSCI)in patients with mild acute ischaemic stroke(AIS),and to develop a dynamic nomogram.Methods A total of 137mild AIS patients admitted to Lu'an Hospital of Anhui Medical University from March 2020 to June 2023 were retrospective selected.Clinical data of subjects were collected,and patients were divided into PSCI group and non-PSCI group according to the PSCI.Multivariate Logistic regression analysis,least absolute shrinkage and selection operator(LASSO)regression analysis and support vector machine(SVM)analysis were used to explore the influencing factors of early PSCI in patients with mild AIS,and constructed corresponding models.The ROC curve was used to explore the predictive value of three models for early PSCI in patients with mild AIS.Based on the influencing factors of the optimal model,the risk prediction nomogram model of early PSCI in patients with mild AIS was constructed by regplot package.Bootstrap method was used to repeatedly sample 1000 times,and the consistency index was calculated.Hosmer-Lemeshow goodness of fit test was used to evaluate the fitting degree of the nomogram model by rms and survival packages.The decision curve was drawn to evaluate the clinical net income of the nomogram model.Simple model was used to predict the risk of PSCI in 1000 mild AIS patients and draw clinical impact curves.The nomogram model was published on the Web using the DynNom package to develop a dynamic nomogram.Results During the follow-up process,3 cases were lost,and a total of 134 patients with mild AIS were included,and 71 cases(53.0%)occurred PSCI.There were significant differences in age,educational level,proportion of patients with drinking history,proportion of patients with diabetes history,proportion of patients with atrial fibrillation history,creatinine(Cr),blood urea nitrogen(BUN),uric acid(UA),HDL-C,hemoglobin(Hb),and fasting plasma glucose(FPG)between PSCI group and non-PSCI group(P<0.05).Multiv

关 键 词：缺血性卒中 认知障碍 卒中后认知障碍 影响因素分析 列线图 

分 类 号：R743.3[医药卫生—神经病学与精神病学]