基于心脏超声指标的心房颤动血栓栓塞风险列线图预测模型研究  

Study on Risk Nomogram Prediction Model of Thromboembolism in Atrial Fibrillation Based on Cardiac Ultrasound Indicators

作　　者：王一川 陈素雨 宋玉娟[1] WANG Yi-chuan;CHEN Su-yu;SONG Yu-juan(Ultrasonic Medicine Teaching and Research Department,Heze Medical College,Shandong 274000,China)

机构地区：[1]菏泽医学专科学校超声医学教研室,山东274000

出　　处：《影像技术》2023年第6期57-63,共7页Image Technology

摘　　要：目的:基于心脏超声指标探讨心房颤动并发血栓栓塞的影响因素,并构建其风险列线图预测模型,以期为临床医生提供指导和决策支持。方法:回顾性选择2019年1月至2020年12月菏泽医学专科学校超声医学教研室收治的1128例经心电图和超声心动图检查确诊为心房颤动患者作为研究对象,根据心房颤动患者是否并发血栓栓塞分为两组,未并发血栓栓塞患者作为对照组(n=944),并发血栓栓塞患者作为研究组(n=184)。采集患者心脏超声指标,进行随访观察,记录2年内发生血栓栓塞事件状况,多变量逻辑回归分析构建预测模型,绘制风险列线图,评估模型灵敏度、特异度和ROC曲线下面积(AUC)。结果:研究组患者年龄、吸烟占比、LAD、RA、LAA、LVEDD、血清hs-CRP水平高于对照组,研究组LVEF低于对照组,组间比较,差异存在统计学意义(P<0.05);年龄、吸烟、LAD、LVEDD、LASd、LAA、LVEF、hs-CRP为心房颤动并发血栓栓塞的独立影响因素(P<0.05);列线图模型内部验证结果显示:CI为0.953,列线图模型对心房颤动并发血栓栓塞预测值与实际值对比,差异不存在统计学意义(P>0.05),对疾病预测校正曲线与理想曲线趋近;列线图模型预测心房颤动并发血栓栓塞最佳截断值为0.151,AUC为0.954,特异度、灵敏度分别为0.863、0.925。结论:年龄、吸烟、LAD、LVEDD、LASd、LAA、LVEF、hs-CRP是心房颤动并发血栓栓塞的独立影响因素,构建基于心脏超声指标的心房颤动血栓栓塞风险列线图预测模型具有重要意义,预测准确度与特异度高,有望成为一种评估心房颤动患者风险的有效方式。Objective:To explore the influencing factors of atrial fibrillation complicated with thromboembolism based on cardiac ultrasound indicators,and to construct the risk prediction nomogram prediction model,so as to provide guidance and decision support for clinicians.Methods:Retrospective selection of 1128 patients diagnosed with atrial fibrillation through electrocardiogram and echocardiography admitted to the Ultrasound Medicine Teaching and Research Department of Heze Medical College from January 2019 to December 2020 as the research subjects.According to whether atrial fibrillation patients complicated with thromboembolism,they were divided into two groups.Patients with atrial fibrillation without thromboembolism were taken as the control group(n=944),and patients with atrial fibrillation complicated with thromboembolism were taken as the research group(n=184).Collect patients'cardiac ultrasound indicators,conduct follow-up observation,record the situation of thromboembolism events within 2 years,use multivariate Logistic regression analysis to build a prediction model,draw a risk nomogram,and evaluate the sensitivity,specificity and area under the Receiver operating characteristic(AUC)of the model.Results:The age,smoking proportion,LAD,RA,LAA,LVEDD,and serum hs-CRP levels of the study group patients were higher than those of the control group,while the LVEF of the study group was lower than that of the control group.There was a statistically significant difference between the groups(P<0.05);age,smoking,LAD,LVEDD,LASd,LAA,LVEF,hs CRP were the independent influencing factors of atrial fibrillation complicated with thromboembolism(P<0.05);the internal validation results of the nomogram model showed that the CI was 0.953.Compared with the actual value,the predictive value of the nomogram model for atrial fibrillation complicated with thromboembolism had no statistical significance(P>0.05),and the disease prediction correction curve was close to the ideal curve;The optimal cutoff value of nomogram model for predicti

关 键 词：心脏超声指标 心房颤动 血栓栓塞 风险 列线图 预测模型 

分 类 号：R540.45[医药卫生—心血管疾病]