急性心肌梗死患者并发急性脑梗死的危险因素及列线图预测模型构建  

Risk Factors and Line Chart Prediction Model Construction for Patients with Acute Myocardial Infarction Complicated with Acute Cerebral Infarction

作　　者：宋岳涵 张小琼[2] 孙雪 SONG Yuehan;ZHANG Xiaoqiong;SUN Xue(Department of Family Medicine,The Second Hospital of Baoding,Baoding,Hebei 071000,China;Department of Cardiovascular Medicine,The Second Hospital of Baoding,Baoding,Hebei 071000,China;Department of Function,The Second Hospital of Baoding,Baoding,Hebei 071000,China)

机构地区：[1]保定市第二医院全科医学科,河北保定071000 [2]保定市第二医院心血管内二科,河北保定071000 [3]保定市第二医院功能科,河北保定071000

出　　处：《岭南心血管病杂志》2025年第1期7-12,20,共7页South China Journal of Cardiovascular Diseases

摘　　要：目的对急性心肌梗死(acute myocardial infarction,AMI)患者并发急性脑梗死(acute cerebral infarction,ACI)的危险因素进行分析,并构建列线图预测模型。方法回顾性选择2020年2月至2024年2月收集的986例AMI患者的临床资料进行研究,随机分为训练集(690例)及验证集(296例),使用训练集构建列线图预测模型,使用验证集对模型进行验证,依据AMI患者是否并发ACI发生分为AMI+ACI组(68例)及AMI组(622例)。采用多因素Logistic回归分析AMI患者并发ACI发生的影响因素,利用R软件构建AMI并发ACI发生的列线图预测模型,采用受试者工作特征曲线(receiver operating characteristic curve,ROC)、霍斯默-莱梅肖(Hosmer-Lemesho,H-L)拟合优度检验及决策性曲线检验预测模型的效能。结果AMI+ACI组患者的年龄≥70岁、糖尿病、原发性高血压(高血压)、心房颤动(房颤)及血肌酐浓度>133μmol/L患者比例高于AMI组,差异有统计学意义(P<0.05)。多因素回归分析结果显示,年龄(OR=5.007,95%CI:2.625~9.550)、糖尿病(OR=5.793,95%CI:3.013~11.136)、高血压(OR=5.881,95%CI:3.030~11.416)、房颤(OR=4.492,95%CI:2.147~9.399)及血肌酐(OR=7.368,95%CI:3.823~14.200)是AMI并发ACI发生的影响因素(P<0.05),训练集及验证集的ROC曲线下面积分别为0.900(95%CI:0.875~0.921),0.874(95%CI:0.831~0.910),显示模型区分度较好,H-L拟合优度检验结果分别为χ2=3.139,P=0.791,χ2=2.767,P=0.837,显示预测模型的准确度好,决策曲线显示预测模型的临床净获益较高。结论年龄、糖尿病、高血压、房颤、血肌酐是AMI患者并发ACI的影响因素,据此构建的列线图预测模型具有较好的区分度及准确度,临床净获益较高。Objectives To analyze the risk factors of acute myocardial infarction(AMI)patients with concurrent acute cerebral infarction(ACI)and construct a column chart prediction model.Methods Clinical Data of 986 AMI patients collected from February 2020 to February 2024 were retrospectively selected for the study and randomized into training set(690 patients)and validation set(296 patients).A column-line graph prediction model was constructed using the training set,and the model was validated using the validation set.Based on whether AMI patients had concurrent ACI,they were separated into an AMI+ACI group(68 cases)and an AMI group(622 cases).Clincal data of patients were collected.Multivariate Logistic regression was applied to analyze the influencing factors of ACI in AMI patients.R soft⁃ware was applied to construct a column chart prediction model for AMI patients combined with ACI occurrence.Receiver operating characteristic curve(ROC),Hosmer-Lemesho(H-L)goodness of fit test,and decision curve were applied to test the effectiveness of the predictive model.Results The numbers of patients with age≥70 years old,diabetes,hypertension,atrial fibrillation and serum concentration of creatinine>133μmol/L in AMI+ACI group were higher than those in AMI group(P<0.05).Multivariate regression found that age(OR=5.007,95%CI:2.625-9.550),diabetes(OR=5.793,95%CI:3.013-11.136),hypertension(OR=5.881,95%CI:3.030-11.416),atrial fibrillation(OR=4.492,95%CI:2.147-9.399)and serum concentration of creatinine(OR=7.368,95%CI:3.823-14.200)were the influencing factors of AMI with ACI(P<0.05).The areas under the ROC for the training and validation sets were 0.900(95%CI:0.875-0.921)and 0.874(95%CI:0.831-0.910),respectively,indicating good model discrimination.The H-L goodness of fit test result wasχ2=3.139,P=0.791,χ2=2.767,P=0.837,indicating good accuracy of the predictive model.The decision curve showed that the clinical net benefit of the predictive model was high.Conclusions Age,diabetes,hypertension,atrial fibrillation,and serum concent

关 键 词：心肌梗死 脑梗死 危险因素 列线图预测模型 

分 类 号：R542.22[医药卫生—心血管疾病]