Xiaojun Wu , Shiyu Wang , Haoning Cui , Xianghui Zheng , Xinyu Hou , Zhuozhong Wang , Qifeng Li , Qi Liu , Tianhui Cao , Yang Zheng , Jian Wu , Bo Yu
{"title":"Machine learning-based return-to-work assessment system for acute myocardial infarction patients within 12 months","authors":"Xiaojun Wu , Shiyu Wang , Haoning Cui , Xianghui Zheng , Xinyu Hou , Zhuozhong Wang , Qifeng Li , Qi Liu , Tianhui Cao , Yang Zheng , Jian Wu , Bo Yu","doi":"10.1016/j.hrtlng.2025.04.020","DOIUrl":null,"url":null,"abstract":"<div><h3>Background</h3><div>Returning to work is a critical indicator of recovery after acute myocardial infarction (AMI), and accurate identification of patients with low return-to-work rates is critical for timely intervention.</div></div><div><h3>Objectives</h3><div>To develop a machine learning (ML) model for predicting the return-to-work in AMI patients.</div></div><div><h3>Methods</h3><div>A retrospective study of data from 539 AMI patients was conducted using the Incidence Rate of Heart Failure After Acute Myocardial Infarction With Optimal Treatment database. Patients were randomly divided into training cohort and validation cohort (7:3). Seven ML algorithms were used to establish a prediction model for the training cohort. Model performance is evaluated by the area under the receiver operating characteristic curve (AUC), accuracy, F1 score, and Brier score.</div></div><div><h3>Results</h3><div>This study included 539 AMI patients (median [IQR] age, 50.0 [45.0, 54.0] years; 505 (93.7 %) were male, and 431 (80.0 %) returned to work within one year after discharge. The best-performing model was eXtreme gradient boosting, which achieved an AUC of 0.821 (95 % CI, 0.736–0.907), an accuracy of 0.802 (95 % CI, 0.733–0.861), and an F1 score of 0.873. The return-to-work score and stratification established based on this model can effectively distinguish patients into low, medium, and high probability groups (33.3 % vs. 60.0 % vs. 91.7 %, <em>P</em> < 0.001). The model was deployed on an open website <span><span>https://amirtw.streamlit.app/</span><svg><path></path></svg></span>, providing a convenient evaluation and analysis tool for medical staff.</div></div><div><h3>Conclusion</h3><div>A new return-to-work ML model was developed, which may help identify patients with low return-to-work rates and may become an effective management tool for AMI patients.</div></div>","PeriodicalId":55064,"journal":{"name":"Heart & Lung","volume":"73 ","pages":"Pages 19-25"},"PeriodicalIF":2.4000,"publicationDate":"2025-04-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Heart & Lung","FirstCategoryId":"3","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0147956325000883","RegionNum":4,"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
Returning to work is a critical indicator of recovery after acute myocardial infarction (AMI), and accurate identification of patients with low return-to-work rates is critical for timely intervention.
Objectives
To develop a machine learning (ML) model for predicting the return-to-work in AMI patients.
Methods
A retrospective study of data from 539 AMI patients was conducted using the Incidence Rate of Heart Failure After Acute Myocardial Infarction With Optimal Treatment database. Patients were randomly divided into training cohort and validation cohort (7:3). Seven ML algorithms were used to establish a prediction model for the training cohort. Model performance is evaluated by the area under the receiver operating characteristic curve (AUC), accuracy, F1 score, and Brier score.
Results
This study included 539 AMI patients (median [IQR] age, 50.0 [45.0, 54.0] years; 505 (93.7 %) were male, and 431 (80.0 %) returned to work within one year after discharge. The best-performing model was eXtreme gradient boosting, which achieved an AUC of 0.821 (95 % CI, 0.736–0.907), an accuracy of 0.802 (95 % CI, 0.733–0.861), and an F1 score of 0.873. The return-to-work score and stratification established based on this model can effectively distinguish patients into low, medium, and high probability groups (33.3 % vs. 60.0 % vs. 91.7 %, P < 0.001). The model was deployed on an open website https://amirtw.streamlit.app/, providing a convenient evaluation and analysis tool for medical staff.
Conclusion
A new return-to-work ML model was developed, which may help identify patients with low return-to-work rates and may become an effective management tool for AMI patients.
期刊介绍:
Heart & Lung: The Journal of Cardiopulmonary and Acute Care, the official publication of The American Association of Heart Failure Nurses, presents original, peer-reviewed articles on techniques, advances, investigations, and observations related to the care of patients with acute and critical illness and patients with chronic cardiac or pulmonary disorders.
The Journal''s acute care articles focus on the care of hospitalized patients, including those in the critical and acute care settings. Because most patients who are hospitalized in acute and critical care settings have chronic conditions, we are also interested in the chronically critically ill, the care of patients with chronic cardiopulmonary disorders, their rehabilitation, and disease prevention. The Journal''s heart failure articles focus on all aspects of the care of patients with this condition. Manuscripts that are relevant to populations across the human lifespan are welcome.