N. Salet, A. Gökdemir, J. Preijde, C. H.van Heck, F. Eijkenaar
Abstract
Early recognition, which preferably happens in primary care, is the most important tool to combat cardiovascular disease (CVD). This study aims to predict acute myocardial infarction (AMI) and ischemic heart disease (IHD) using Machine Learning (ML) in primary care cardiovascular patients.
Introduction
Over the past 20 years, cardiovascular disease (CVD) has been the most common cause of death worldwide [1]. Most of this mortality can be attributed to ischemic heart disease, accounting for around 16% of fatalities.
Methods
We used anonymized electronic health record (EHR) data collected between 01-01-2011 and 31-03- 2021 on primary care cardiovascular patients who were enrolled in a cardiovascular risk management (CVRM) program in the Netherlands.
Results
The sample used for predicting AMI (model 1) included 13,079 patients of whom 59.3% had the male sex (Table 1).
Discussion
In this study, we aimed to predict AMI and IHD in primary care cardiovascular patients using machine learning (ML). We evaluated the predictive performance of two random forest models and made a head-to-head comparison with the commonly used SMART algorithm.
Conclusions
Our ML models showed high predictive performance and outperformed the existing SMART algorithm in predicting AMI and IHD in primary care cardiovascular patients.
Citation: Salet N, Gökdemir A, Preijde J, van Heck CH, Eijkenaar F (2024) Using machine learning to predict acute myocardial infarction and ischemic heart disease in primary care cardiovascular patients. PLoS ONE 19(7): e0307099. https://doi.org/10.1371/journal.pone.0307099
Editor: Hean Teik Ong, HT Ong Heart Clinic, MALAYSIA
Received: April 18, 2024; Accepted: June 28, 2024; Published: July 18, 2024
Copyright: © 2024 Salet et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: This study utilized existing data obtained upon request and subject to license restrictions from several sources. The database and source code are not publicly available due to commercial and privacy constraints. Public availability would compromise patient privacy and confidentiality, which are legally and ethically protected. However, (parts of) the code or data can be made available upon reasonable request. Specific questions about this can be directed to the first author or to EscuLine b.v. at thijs.debruijn@esculine.nl
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Source: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0307099#abstract0
