急性心肌梗死后心脏破裂的风险预测模型构建与外部验证  

Construction and external validation of a risk-prediction model for cardiac rupture after acute myocardial infarction

作　　者：罗雅菲 戴凯 李志兵 熊芊卉 熊瑶 宋志芳[1] 熊文俊 田清山 LUO Yafei;DAI Kai;LI Zhibing;XIONG Qianhui;XIONG Yao;SONG Zhifang;XIONG Wenjun;TIAN Qingshan(department of Cardiology,The First Affiliated Hospital of Nanchang University,Nanchang,330063,China;Department of Cardiology,Affiliated Peopled Hospital of Nanchang University)

机构地区：[1]南昌大学第一附属医院心内科,南昌330063 [2]南昌大学附属人民医院心内科

出　　处：《临床心血管病杂志》2022年第2期107-112,共6页Journal of Clinical Cardiology

基　　金：国家自然科学基金(No:82000234);江西省自然科学基金(No:20202BABL206004)。

摘　　要：目的:回顾性分析急性心肌梗死(AMI)后心脏破裂(CR)患者的临床特征及危险因素,并构建实用风险预测模型并进行独立的外部验证。方法:从南昌大学第一附属医院及南昌大学附属人民医院数据库中选取2013年1月—2020年9月的急性ST段抬高型心肌梗死(STEMI)CR患者126例和非CR患者339例,将南昌大学附属人民医院纳入的患者作为建模组(CR组:86例,非CR组:228例),南昌大学第一附属医院纳入的患者作为验证组(CR组:40例,非CR组:111例)。采用单因素分析和多因素logistic回归分析得出CR的危险因素,根据各危险因素的加权赋分建立CR风险预测模型,通过受试者工作特征(ROC)曲线下面积(AUC)及Hosmer-Lemeshow检验检测模型的区分度和校准度,并在验证组中检验该模型的价值。结果:建模组中多因素logistic回归分析结果显示:年龄≥63岁(OR=2.446,95%CI:1.017~5.881,P=0.046)、女性(OR=2.614,95%CI:1.155~5.917,P=0.021)、收缩压≤120 mmHg(OR=2.569,95%CI:1.121~5.887,P=0.026)、心率≥100次/min(OR=2.532,95%CI:1.055~6.073,P=0.037)、中性粒细胞百分比/白蛋白(NPAR)≥2.2(OR=3.328,95%CI:1.463~7.570,P=0.004)、肌酐≥106μmol/L(OR=2.744,95%CI:1.178~6.392,P=0.019)、Ca^(2+)≤2.2 mmol/L(OR=3.059,95%CI:1.369~6.835,P=0.006)、行急诊PCI(OR=0.442,95%CI:0.198~0.989,P=0.047)、口服β受体阻滞剂(OR=0.268,95%CI:0.118~0.607,P=0.002)是CR的独立影响因素。基于这9个影响因素进行加权后建立了预测模型,总分为9分,患者评分≤4分归为低风险和评分≥5分归为高风险。本研究预测模型的AUC(0.906,95%CI:0.865~0.948,P<0.001),Hosmer-Lemeshow检验(χ^(2)=14.297,P>0.05),外部验证中AUC(0.785,95%CI:0.696~0.875,P<0.001),Hosmer-Lemeshow检验(χ^(2)=1.7,P=0.989),结果表明本研究模型对CR有较好的预测价值。结论:高龄、女性、收缩压低、心率快、NPAR升高、肌酐升高、Ca^(2+)浓度降低是AMI后CR的危险因素,行急诊PCI、口服β受体阻滞剂是其保护因素,基Objective:To explore the clinical characteristics and risk factors of patients with cardiac rupture(CR)after acute myocardial infarction(AMI)through retrospective research,and construct a practical risk-prediction model with independent external validation.Methods:A total of 126 CR patients with AMI were selected from the databases of the First Affiliated Hospital of Nanchang University and the Jiangxi Provincial People’s Hospital Affiliated to Nanchang University from January 2013 to September 2020,and 339 patients with non-CR were enrolled.The patients included in the Jiangxi Provincial People’s Hospital Affiliated to Nanchang University were used as the modeling group(CR group:n=86,non-CR group:n=228),and the patients in the First Affiliated Hospital of Nanchang University were treated as the verification group(CR group:n=40,non-CR group:n=111).The risk factors of CR were obtained by univariable and multivariable logistic regression analysis.According to the weighted scores of each independent risk factor,the CR risk-prediction model was established.The discrimination and calibration of the model were detected by the area under the curve(AUC)of the receiver operating characteristic(ROC)and Hosmer-Lemeshow test,respectively.The value of the model was examined by the validation group.Results:The results of the multivariate logistic regression analysis showed that the age≥63 years old(OR=2.446,95%CI:1.017-5.881,P=0.046),female(OR=2.614,95%CI:1.155-5.917,P=0.021),systolic blood pressure≤120 mmHg(OR=2.569,95%CI:1.121-5.887,P=0.026),heart rate≥100 times/min(OR=2.532,95%CI:1.055-6.073,P=0.037),neutrophil percentage to albumin ratio(NPAR)≥2.2(OR=3.328,95%CI:1.463-7.57,P=0.004),creatinine≥106μmol/L(OR=2.744,95%CI:1.178-6.392,P=0.019),calcium concentration≤2.2 mmol/L(OR=3.059,95%CI:1.369-6.835,P=0.006),emergency PCI(OR=0.442,95%CI:0.198~0.989,P=0.047)and oral beta blockers(OR=0.268,95%CI:0.118-0.607,P=0.002)were independent risk factors for CR after AMI.Based on the weighting of these nine risk factors,

关 键 词：急性心肌梗死 心脏破裂 风险预测模型 危险因素 

分 类 号：R542.2[医药卫生—心血管疾病]