{"title":"Predicting In-Hospital Mortality in Patients With Acute Myocardial Infarction: A Comparison of Machine Learning Approaches","authors":"Hamidreza Soleimani, Soroush Najdaghi, Delaram Narimani Davani, Parham Dastjerdi, Parham Samimisedeh, Hedieh Shayesteh, Babak Sattartabar, Farzad Masoudkabir, Haleh Ashraf, Mehdi Mehrani, Yaser Jenab, Kaveh Hosseini","doi":"10.1002/clc.70124","DOIUrl":null,"url":null,"abstract":"<div>\n \n \n <section>\n \n <h3> Background</h3>\n \n <p>Acute myocardial infarction (AMI) remains a leading global cause of mortality. This study explores predictors of in-hospital mortality among AMI patients using advanced machine learning (ML) techniques.</p>\n </section>\n \n <section>\n \n <h3> Methods</h3>\n \n <p>Data from 7422 AMI patients treated with percutaneous coronary intervention (PCI) at Tehran Heart Center (2015–2021) were analyzed. Fifty-eight clinical, demographic, and laboratory variables were evaluated. Seven ML algorithms, including Random Forest (RF), logistic regression with LASSO, and XGBoost, were implemented. The data set was divided into training (70%) and testing (30%) subsets, with fivefold cross-validation. The class imbalance was addressed using the synthetic minority oversampling technique (SMOTE). Model predictions were interpreted using SHapley Additive exPlanations (SHAP).</p>\n </section>\n \n <section>\n \n <h3> Results</h3>\n \n <p>In-hospital mortality occurred in 129 patients (1.74%). RF achieved the highest predictive performance, with an area under the curve (AUC) of 0.924 (95% CI 0.893–0.954), followed by XGBoost (AUC 0.905) and logistic regression with LASSO (AUC 0.893). Sensitivity analysis in STEMI patients confirmed RF's robust performance (AUC 0.900). SHAP analysis identified key predictors, including lower left ventricular ejection fraction (LVEF; 33.24% vs. 43.46% in survivors, <i>p</i> < 0.001), higher fasting blood glucose (190.38 vs. 132.29 mg/dL, <i>p</i> < 0.001), elevated serum creatinine, advanced age (70.92 vs. 61.88 years, <i>p</i> < 0.001), and lower LDL-C levels. Conversely, BMI showed no significant association (<i>p</i> = 0.456).</p>\n </section>\n \n <section>\n \n <h3> Conclusion</h3>\n \n <p>ML algorithms, particularly RF, effectively predict in-hospital mortality in AMI patients, highlighting critical predictors such as LVEF and biochemical markers. These insights offer valuable tools for enhancing clinical decision-making and improving patient outcomes.</p>\n </section>\n </div>","PeriodicalId":10201,"journal":{"name":"Clinical Cardiology","volume":"48 4","pages":""},"PeriodicalIF":2.4000,"publicationDate":"2025-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/clc.70124","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Clinical Cardiology","FirstCategoryId":"3","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/clc.70124","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"CARDIAC & CARDIOVASCULAR SYSTEMS","Score":null,"Total":0}
引用次数: 0
Abstract
Background
Acute myocardial infarction (AMI) remains a leading global cause of mortality. This study explores predictors of in-hospital mortality among AMI patients using advanced machine learning (ML) techniques.
Methods
Data from 7422 AMI patients treated with percutaneous coronary intervention (PCI) at Tehran Heart Center (2015–2021) were analyzed. Fifty-eight clinical, demographic, and laboratory variables were evaluated. Seven ML algorithms, including Random Forest (RF), logistic regression with LASSO, and XGBoost, were implemented. The data set was divided into training (70%) and testing (30%) subsets, with fivefold cross-validation. The class imbalance was addressed using the synthetic minority oversampling technique (SMOTE). Model predictions were interpreted using SHapley Additive exPlanations (SHAP).
Results
In-hospital mortality occurred in 129 patients (1.74%). RF achieved the highest predictive performance, with an area under the curve (AUC) of 0.924 (95% CI 0.893–0.954), followed by XGBoost (AUC 0.905) and logistic regression with LASSO (AUC 0.893). Sensitivity analysis in STEMI patients confirmed RF's robust performance (AUC 0.900). SHAP analysis identified key predictors, including lower left ventricular ejection fraction (LVEF; 33.24% vs. 43.46% in survivors, p < 0.001), higher fasting blood glucose (190.38 vs. 132.29 mg/dL, p < 0.001), elevated serum creatinine, advanced age (70.92 vs. 61.88 years, p < 0.001), and lower LDL-C levels. Conversely, BMI showed no significant association (p = 0.456).
Conclusion
ML algorithms, particularly RF, effectively predict in-hospital mortality in AMI patients, highlighting critical predictors such as LVEF and biochemical markers. These insights offer valuable tools for enhancing clinical decision-making and improving patient outcomes.
背景:急性心肌梗死(AMI)仍然是全球主要的死亡原因。本研究利用先进的机器学习(ML)技术探讨AMI患者住院死亡率的预测因素。方法分析2015-2021年德黑兰心脏中心7422例AMI患者经皮冠状动脉介入治疗(PCI)的资料。评估了58个临床、人口统计学和实验室变量。实现了随机森林(Random Forest, RF)、LASSO逻辑回归和XGBoost等7种ML算法。数据集被分为训练子集(70%)和测试子集(30%),进行五次交叉验证。利用合成少数派过采样技术(SMOTE)解决了类不平衡问题。模型预测采用SHapley加性解释(SHAP)进行解释。结果129例患者住院死亡,占1.74%。RF的预测效果最好,曲线下面积(AUC)为0.924 (95% CI 0.893 - 0.954),其次是XGBoost (AUC 0.905)和LASSO logistic回归(AUC 0.893)。STEMI患者的敏感性分析证实了RF的稳健性能(AUC 0.900)。SHAP分析确定了关键预测因素,包括低左室射血分数(LVEF);33.24% vs. 43.46%, p < 0.001),较高的空腹血糖(190.38 vs. 132.29 mg/dL, p < 0.001),血清肌酐升高,高龄(70.92 vs. 61.88岁,p < 0.001),以及较低的LDL-C水平。相反,BMI无显著相关性(p = 0.456)。结论ML算法,尤其是RF算法,能够有效预测AMI患者的住院死亡率,突出了LVEF和生化指标等关键预测指标。这些见解为加强临床决策和改善患者预后提供了有价值的工具。
期刊介绍:
Clinical Cardiology provides a fully Gold Open Access forum for the publication of original clinical research, as well as brief reviews of diagnostic and therapeutic issues in cardiovascular medicine and cardiovascular surgery.
The journal includes Clinical Investigations, Reviews, free standing editorials and commentaries, and bonus online-only content.
The journal also publishes supplements, Expert Panel Discussions, sponsored clinical Reviews, Trial Designs, and Quality and Outcomes.