绝经期女性冠心病发病风险预测模型的构建与验证  

Construction and validation of a risk prediction model for coronary heart disease in postmenopausal women

作　　者：徐巧茜 刘乘光 陈楚寒 郑景辉[2] Xu Qiaoxi;Liu Chengguang;Chen Chuhan;Zheng Jinghui(Guangxi University of Chinese Medicine,Guangxi Province,Nanning 530001,China;不详)

机构地区：[1]广西中医药大学,南宁530001 [2]广西中医药大学附属瑞康医院心内科

出　　处：《疑难病杂志》2023年第8期845-851,861,共8页Chinese Journal of Difficult and Complicated Cases

基　　金：国家自然科学基金(82160875);广西中医药大学高层次人才团队培育项目--中医药系统生物学创新研究团队(2021005);2019~2021年广西一流学科建设开放课题(2019XK152);中西医结合广西一流学科(桂教科研〔2022〕1号)。

摘　　要：目的明确绝经期女性冠心病(CAD)发病的影响因素,并基于上述影响因素构建绝经期女性冠心病风险预测模型,同时验证其预测效能。方法通过计算机检索国外4个电子数据库(Cochrane Library、Embase、PubMed、Web of Science)和4个中文电子数据库(中国知网、万方数据库、中国生物医学文献数据库、维普数据库),收集建库到2022年12月公开发表的文献。使用RevMan5.4软件运用Meta分析方法对纳入文献5次或以上的影响因素进行频次统计与分析,计算影响绝经期女性CAD发病的相关因素的综合危险度OR,将有统计学意义(P<0.05)的因素纳入构建Logistic回归预测模型并进行外部验证;同时使用2017年9月-2022年7月在广西中医药大学附属瑞康医院心内科住院的303例绝经期女性患者的临床资料绘制列线图模型并验证。结果Meta分析结果显示,合并糖尿病、合并高血压、吸烟、体质量指数(BMI)、冠心病家族史、合并血脂异常、总胆固醇(TC)、低密度脂蛋白(LDL)、血小板平均容积(MPV)、血小板分布宽度(PDW)等10项因素(X_(1)~X_(10))是绝经期女性CAD发病的独立危险因素(P<0.05),构建绝经期女性CAD发病风险预测模型:Logit(P)=-0.26+0.61X_(1)+0.65X_(2)+0.88X_(3)+0.7X_(4)+0.87X_(5)+1.26X_(6)+0.56X_(7)+0.27X_(8)+0.94X_(9)+1.13X_(10),经验证该模型有一定的准确性(AUC=0.630),列线图模型准确度较高(AUC=0.732),经校准度检验及决策曲线分析后均提示模型具有较高的临床应用价值。结论该预测模型具有较好的准确度,可以为临床早期筛查绝经期女性CAD发生风险提供循证依据。Objective Through 34 clinical studies,the influencing factors of coronary heart disease(AD)incidence in postmenopausal women were identified,and a risk prediction model for coronary heart disease in postmenopausal women was constructed based on the above influencing factors,while verifying its predictive efficacy.Methods Retrieve 4 foreign electronic databases(Cochrane Library,Embase,PubMed,Web of Science)and 4 Chinese electronic databases(CNKI,Wanfang Database,China Biomedical Literature Database,and VIP Database)through computer retrieval,and collect literature from the public development table established until December 2022.Using RevMan5.4 software and meta-analysis method,frequency statistics and analysis were conducted on the influencing factors included in the literature 5 or more times,to calculate the comprehensive risk OR of the relevant factors affecting the onset of CAD in menopausal women.Statistically significant(P<0.05)factors were included in the construction of a logistic regression prediction model and external validation was conducted;Using the clinical data of 303 menopausal female patients admitted to the Department of Cardiology at Ruikang Hospital Affiliated to Guangxi University of Traditional Chinese Medicine from September 2017 to July 2022,a column chart model was drawn and validated.Results Meta analysis showed that 10 risk factors(X1~X10),including diabetes,hypertension,smoking,body mass index(BMI),family history of coronary heart disease,Dyslipidemia,total cholesterol(TC),low-density lipoprotein(LDL),mean platelet volume(MPV),and platelet distribution width(PDW),were independent risk factors for CAD in menopausal women(P<0.05),Constructing a risk prediction model for CAD in menopausal women:Logit(P)=-0.26+0.61X_(1)+0.65X_(2)+0.88X_(3)+0.7X_(4)+0.87X_(5)+1.26X_(6)+0.56X_(7)+0.27X_(8)+0.94X_(9)+1.13X_(10).The model has been verified to have certain accuracy(AUC=0.630),and the column chart model has a high accuracy(AUC=0.732).After calibration test and decision curve analysis,it indicat

关 键 词：冠心病 绝经期女性 META分析 危险因素 预测模型 

分 类 号：R541.4[医药卫生—心血管疾病]