两种预测模型在脑卒中患者卒中后疲劳风险预测中的应用价值比较  

Comparison of the application value of two prediction models in the prediction of post-stroke fatigue risk in stroke patients

作　　者：杨金盘 马秋平[1] 刘裕君 张佳琳 YANG Jin-pan;MA Qiu-ping;LIU Yu-jun;ZHANG Jia-lin(Guangxi University of Chinese Medicine,Nanning,Guangxi 530200,China)

机构地区：[1]广西中医药大学,广西南宁530200

出　　处：《现代预防医学》2024年第7期1161-1165,共5页Modern Preventive Medicine

基　　金：广西壮族自治区中医药管理局自筹经费科研课题(GXZYA 20220065);广西中医药大学高层次人才培育创新团队建设项目(2022A010);广西中医药大学引进博士科研启动基金项目(2023BS055)。

摘　　要：目的探讨并比较分析脑卒中后疲劳(post-stroke fatigue,PSF)风险预测列线图和Web计算器程序在脑卒中恢复期患者PSF风险预测中的应用价值。方法前瞻性选取某家三甲医院脑病科2023年2—8月收治的急性期卒中患者,使用列线图和Web计算器程序预测每位卒中患者未来在恢复期发生PSF的风险概率。应用受试者工作特征曲线下面积(area under curve,AUC)、Hosmer-Lemeshow拟合优度检验、临床决策曲线分别评价两种模型的区分度、校准度和临床效用。结果共纳入282例卒中患者,其中128例发生PSF,PSF发生率为46.89%。列线图和Web计算器程序的AUC分别为0.687(95%CI:0.624~0.749)、0.743(95%CI:0.683~0.803)。Hosmer-Lemeshow拟合优度检验提示,列线图(χ2=8.357,P=0.213)和Web计算器程序(χ2=4.467,P=0.614)在卒中患者中的拟合度良好。临床决策曲线分析显示,列线图和Web计算器程序分别在阈概率0.35~0.81、0.33~0.88范围内有临床获益。结论列线图和Web计算器程序能有效预测卒中恢复期患者的PSF风险,Web计算器程序优于列线图。但未来Web计算器程序模型仍需进一步更新以平衡简便性和准确性,进而提高其临床适用性。Objective To explore and compare the application value of post-stroke fatigue(PSF)risk prediction nomogram and Web calculator program in post-stroke fatigue risk prediction in convalescent stroke patients.Methods The patients with acute stroke admitted to the Department of Encephalopathy of a tertiary hospital from February 2023 to August 2023 were prospectively selected,and the nomogram and Web calculator program were used to predict the risk probability of PSF in convalescent stage.The area under the receiver operating characteristic curve(AUC),Hosmer-Lemeshow goodness-of-fit test,and clinical decision curve were used to evaluate the discrimination,calibration,and clinical utility of the two models,respectively.Results A total of 282 patients were included,of which 128 had PSF,and the incidence of PSF was 46.89%.The AUC of the nomogram and Web calculator were 0.687(95%CI:0.624-0.749)and 0.743(95%CI:0.683-0.803),respectively.The Hosmer-Lemeshow goodness-of-fit test showed that the nomogram(χ2=8.357,P=0.213)and Web calculator program(χ2=4.467,P=0.614)were well fitted in stroke patients.The clinical decision curve analysis showed that the nomogram and Web calculator program had clinical benefits in the range of threshold probability of 0.35-0.81 and 0.33-0.88,respectively.Conclusion The nomogram and Web calculator program can effectively predict the PSF risk of stroke convalescent patients,and the Web calculator program is superior to the nomogram.However,the future Web calculator program model still needs to be further updated to balance simplicity and accuracy,thereby improving its clinical applicability.

关 键 词：脑卒中患者 卒中后疲劳 预测模型 外部验证 

分 类 号：R743.3[医药卫生—神经病学与精神病学]