摘要
目的 本研究的目的是识别影响急性心肌梗死(AMI)患者的临床主要风险因素,并建立死亡风险预测模型预测AMI患者的短期死亡率。方法 从MIMIC-Ⅲ数据库中收集符合纳入标准的AMI患者数据。采用随机森林算法进行变量重要性排序。结合多元logistic回归识别AMI相关死亡危险因素,并以列线图表示结果。结果 最终纳入患者4610例,90天(d)内死亡的患者共894例。研究结果显示AMI患者的死亡风险因素为年龄,肌钙蛋白T,室性心动过速,心室颤动,心肌梗死史,APSⅢ,冠状动脉搭桥手术和经皮冠状动脉介入治疗。与全球急性冠状动脉事件登记评分(GRACE),APSⅢ,序贯器官衰竭评分(SOFA)相比,本文构建的风险模型AUC最高(训练集:0.826,测试集:0.818)。Hosmer-Lemeshow拟合优度检验和标准曲线结果一致。NRI和IDI值均表明该风险模型具有显著的预测能力,DCA结果表明该风险模型具有良好的净效益,可供临床应用。结论 随机森林算法结合多元logistic回归可识别AMI相关危险因素,据此构建的模型具有良好的死亡风险预测能力,对改善AMI患者的预后有一定的指导意义。
Objective The purpose of this study was to identify the factors influencing the mortality in patients with acute myocardial infarction(AMI) and to establish a mortality risk prediction model to predict the short-term mortality of AMI patients.Methods Collect AMI patient data that met the inclusion criteria from the MIMIC-Ⅲ database.Variable importance selection was determined using the random forest algorithm.Multiple logistic regression was used to determine AMI-related mortality risk factors,with the results represented as a nomogram.Results 4610 patients were eventually enrolled and 894 died within 90 days.The results of this study indicated that age,troponin T,Venricular tachycardia,ventricular fibrillation,history of myocardial infarction,APSⅢ,bypass,and percutaneous coronary intervention surgery were risk factors in AMI patients.Compared with GRACE,APSⅢ,and SOFA scores,our constructed risk model had the highest AUC(0.826 and 0.818 in the training and validation groups,respectively).The Hosmer-Lemeshow goodness-of-fit test and standard curve both produced very consistent results.Both the NRI and IDI values indicated that the risk model had significant predictive power,and DCA results indicated that the risk model had good net benefits for clinical application.Conclusions Random forest algorithm combined with multivariate logistic regression can determine AMI-related risk factors.The model constructed on this basis has good mortality risk prediction ability and has a certain guiding significance for improving the prognosis of AMI patients.
作者
黄韬
杨瑞
郑帅
许丰硕
韩迪迪
乔萌萌
吕军
Huang Tao;Yang Rui;Zheng Shuai;Xu Fengshuo;Han Didi;Qiao Mengmeng;Lyu Jun(Department of Clinical Research,The First Affiliated Hospital of Jinan University,Guangzhou,510630,China;不详)
出处
《中国循证心血管医学杂志》
2022年第4期406-410,共5页
Chinese Journal of Evidence-Based Cardiovascular Medicine
