急性心肌梗死发生风险列线图预测模型的构建与验证  

Construction and validation of the nomogram prediction model for the risk of acute myocardial infarction

作　　者：徐以康 马晶茹[1] 刘蕾 王嘉钰 赵欣然 XU Yi-kang;MA Jing-ru;LIU Lei;WANG Jia-yu;ZHAO Xin-ran(Department of Cardiology,Second Affiliated Hospital of Shenyang Medical College,Liaoning 110001,China;School of Nursing,Liaoning University of Traditional Chinese Medicine)

机构地区：[1]沈阳医学院附属第二医院心内科,辽宁110001 [2]辽宁中医药大学护理学院

出　　处：《预防医学论坛》2025年第1期25-30,45,共7页Preventive Medicine Tribune

基　　金：辽宁省沈阳市科学技术计划项目(22-321-33-100);辽宁省自然科学基金科技援疆、援藏医疗专项(2023-MS-19)。

摘　　要：目的分析急性心肌梗死(AMI)患者的危险因素,构建AMI发生风险的列线图预测模型并对其效能进行验证。方法回顾性收集2020年1月-2022年12月某三甲医院1457例缺血性心脏病患者的临床资料,其中AMI 683例、心绞痛774例,采用随机数字表法按7∶3的比例将患者分为训练组(1019例)和验证组(438例)。比较训练组和验证组患者临床资料,基于赤池信息准则(AIC),单因素和多因素logistic双向逐步回归法筛选模型变量,构建AMI相比心绞痛发生风险的列线图预测模型,利用受试者工作特征曲线(ROC)及曲线下面积(AUC)、校准曲线和临床决策曲线(DCA)综合评估模型的预测价值、校准度和临床实用性。结果在训练组中,高密度脂蛋白胆固醇(HDL-C)水平高(OR=0.52,95%CI:0.34~0.78,P=0.002)是AMI发生的保护因素,低密度脂蛋白胆固醇(LDL-C)水平高(OR=1.20,95%CI:1.03~1.40,P=0.019)、总胆固醇(TC)水平高(OR=1.16,95%CI:1.03~1.31,P=0.013)、脉压差(PP)大(OR=1.02,95%CI:1.02~1.03,P<0.001)、尿素氮(BUN)水平高(OR=1.08,95%CI:1.02~1.14,P=0.010)、吸烟(OR=2.20,95%CI:1.68~2.87,P<0.001)是AMI的危险因素(均P<0.05)。训练组和验证组的AUC分别为0.705(95%CI:0.673~0.736)和0.703(95%CI:0.655~0.752),Hosmer-Lemeshow检验显示模型拟合度良好(训练组P=0.687,验证组P=0.548),预测值接近于真实值,临床决策曲线均提示本模型具有临床净获益。结论构建的AMI发生风险预测模型具有良好的预测性能,有助于医务人员快速识别AMI高风险人群。Objective To analyze the risk factors in patients with acute myocardial infarction(AMI)and to develop and validate the performance of a nomogram prediction model for assessing the risk of AMI occurrence.Methods The clinical data of 1457patients admitted to a tertiary hospital in Shenyang city from January 2020to December 2022were retrospectively collected.Patients were divided into the AMI group(683cases)and the angina group(774cases)based on whether they had AMI.Using the random number table method,they were divided into a training group(1019cases)and a validation group(438cases)in a 7∶3ratio,and the clinical data of the 2groups of patients were compared.Based on the Akaike information criterion(AIC),single factor and multivariate logistic two-way stepwise regression methods were used to screen model variables and construct the nomogram prediction model for the risk of AMI compared with angina pectoris.The predictive value,calibration degree,and clinical practicality of the model were comprehensively evaluated using the receiver operating characteristic curve(ROC)and area under the curve(AUC),calibration curve,and clinical decision curve analysis(DCA).Results In the training group,higher levels of the high-density lipoprotein cholesterol(HDL-C)(OR=0.52,95%CI:0.34-0.78,P=0.002)were protective factors for AMI(P=0.002),while higher levels of the low density lipoprotein cholesterol(LDL-C)(OR=1.20,95%CI:1.03-1.40,P=0.019),higher levels of the total cholesterol(TC)(OR=1.16,95%CI:1.03-1.31,P=0.013),larger pulse pressure difference(PP)(OR=1.02,95%CI 1.02-1.03,P<0.001),higher levels of the blood urea nitrogen(BUN)(OR=1.08,95%CI:1.02-1.14,P=0.010),and smoking(OR=2.20,95%CI:1.68-2.87,P<0.001)were risk factors for AMI(all P<0.05).The area under the working characteristic curve(AUC)of the training group and validation group were 0.705(95%CI:0.673~0.736)and 0.703(95%CI:0.655-0.752),respectively.The Hosmer-Lemeshow test showed good fit of the model(P=0.687for the training group and P=0.548for the validation group),and the pr

关 键 词：急性心肌梗死 心绞痛 列线图 发病风险 预测模型 

分 类 号：R542.22[医药卫生—心血管疾病] R195.1[医药卫生—内科学]