非瓣膜性心房颤动患者发生缺血性脑卒中风险列线图预测模型的构建和评价  

Construction of nomogram predictive model for ischemic stroke in patients with nonvalvular atrial fibrillation

作　　者：罗立[1] 黄素兰[1] 代彤宇 葛良清 LUO Li;HUANG Sulan;DAI Tongyu;GE Liangqing(The First People's Hospital of Changde City,Changde 415000,China)

机构地区：[1]常德市第一人民医院,415000

出　　处：《心肺血管病杂志》2021年第4期309-314,共6页Journal of Cardiovascular and Pulmonary Diseases

基　　金：常德市科学技术局技术研究与开发资金项目(2019S194);湖南省卫生健康委科研计划项目(20200224);常德市科学技术局技术研究与开发资金项目(2018S031)。

摘　　要：目的:研究非瓣膜性心房颤动(NVAF)患者发生缺血性卒中的危险因素,构建列线图预测模型并验证其价值和临床意义。方法:纳入2019年1月至2019年12月,我院住院并诊断为NVAF患者共325例,其中观察组共123例(合并急性缺血性卒中),对照组共202例(单纯心房颤动且未合并急性缺血性卒中)。收集两组患者的一般临床资料,用Logistic回归分析发生缺血性脑卒中患者的危险因素。然后将所有患者随机分为建模组(n=228)与验证组(n=97)。在回归模型的基础上建立列线图模型,并应用ROC曲线和校准曲线评价列线图模型的区分度和校准度,建立临床决策曲线评价其临床应用价值。结果:本研究中123例患者发生缺血性卒中,发生率为37.8%。NLR(OR=1.48,95%CI:1.18〜1.84)、同型半胱氨酸(OR=1.09,95%CI:1.02〜1.17)、D-Dimer(OR=1.88,95%CI:1.28〜2.77)、cTnI是否升高(OR=2.05,95%CI:1.02〜4.12)、NT-proBNP(OR=1.04,95%CI:1.00〜1.09)和LAD(OR=1.10,95%CI:1.04〜1.16)是NVAF患者发生缺血性卒中的危险因素(P<0.05)。应用ROC曲线评价列线图模型的预测能力,应用ROC曲线评价列线图模型的预测能力,在建模组中ROC曲线的AUC值为0.88,95%CI::0.834-0.919,灵敏度0.839、特异度0.731、阳性预测值0.646、阴性预测值0.886;在验证组中ROC曲线的AUC值为0.805,95%CI:0.693〜0.884,灵敏度0.611、特异度0.742、阳性预测值0.710、阴性预测值0.759,P<0.05;说明建立的模型有较好的预测价值和区分度。列线图风险模型校正曲线提示在预测值和实际值之间具备较好的一致性。临床决策曲线分析显示该列线图模型具有一定的临床实用价值。结论:NLR、同型半胱氨酸、D-Dimer、cTnI是否升高、NT-proBNP和LAD大小是NVAF患者发生缺血性卒中的危险因素。构建列线图模型能更直观、简洁为NVAF患者提供个体化的缺血性卒中风险预测。Objective:To study the risk factors for ischemic stroke in patients with non-valvular atrial fibrillation(NVAF),construct a nomogram prediction model and verify its value and clinical application.Methods:A total of 325 patients admitted to our hospital and diagnosed with NVAF from January 2019 to December 2019 were included,including 123 patients in the study group(complicated with acute ischemic stroke)and 202 patients in the control group(simple atrial fibrillation without complicated with acute ischemic stroke).General clinical data were collected for Logistic regression analysis of risk factors for patients with NVAF combined with ischemic stroke.They were randomly divided into two groups,the development group(n=228)and the validation group(n=97).Based on the regression model,a nomogram model was established,and ROC curve and calibration curve were used to evaluate the differentiation and calibration degree of the nomogram model,and the clinical decision curve was established to evaluate its clinical application value.Results:In this study,123 patients had ischemic stroke,with an incidence of 37.8%.NLR(OR=1.48,95%CI:1.181.84),homocysteine(OR=1.09,95%CI:1.02-1.17),D-dimer(OR=1.88,95%CI:1.28-2.77),cTnI(OR=2.05,95%CI:1.02-4.12),NT-proBNP precursor(OR=1.04,95%CI:1.00-1.09)and LAD(OR=1.10).The independent risk factors for ischemic stroke in patients with NVAF(P<0.05)were 95%CI:1.04-1.16.ROC curve was used to evaluate the predictive ability of the nomograph model.The ROC curve was used to evaluate the predictive ability of the nomograph model.In the development group,the AUC value of the ROC curve was 0.88,95%CI:0.834-0.919,sensitivity 0.839,specificity 0.731,positive predictive value 0.646,and negative predictive value 0.886.In the validation group,the AUC value of ROC curve was 0.805,95%CI:0.693-0.884,sensitivity 0.611,specificity 0.742,positive predictive value 0.710,negative predictive value 0.759,P<0.05.These indicate that the established model had good predictive value and differentiation degree.The risk mode

关 键 词：非瓣膜性心心房颤动动 缺血性脑卒中 列线图 

分 类 号：R54[医药卫生—心血管疾病]