Machine learning approaches for predicting heart failure readmissions

Vista sencilla de ítem
dc.contributor.authorPikatza Huerga, Amaia
dc.contributor.authorAlmeida, Aitor
dc.contributor.authorQuirós López, Raúl
dc.contributor.authorLarrea, Nere
dc.contributor.authorLegarreta Olabarrieta, María José
dc.contributor.authorZulaika Zurimendi, Unai
dc.contributor.authorGarcía, Rodrigo Damián
dc.contributor.authorGarcía Gutiérrez, Susana
dc.date.accessioned2025-12-02T15:22:43Z
dc.date.available2025-12-02T15:22:43Z
dc.date.issued2025-07-06
dc.date.updated2025-12-02T15:22:43Z
dc.description.abstractPurpose: This study aims to develop and evaluate machine learning (ML) models to predict the likelihood of hospital readmission within 30 days after discharge for patients with heart failure (HF). The goal is to compare the predictive accuracy of ML models with traditional methods such as those based on Cox proportional hazards and logistic regression, to improve clinical outcomes and reduce hospital costs. Methods: We conducted a prospective cohort study of patients discharged from five hospitals following admission for HF. Data were collected on variables including sociodemographic characteristics, medical history, admission details, patient-reported outcomes, and clinical parameters. ML techniques were employed to analyse the data and predict readmission risk, incorporating strategies to handle class imbalance and missing data. Model performance was assessed based on accuracy, sensitivity, specificity, area under the receiver operating characteristic curve (AUC), and F1 score. Results: Ensemble methods with Synthetic Minority Over-sampling Technique balancing and bagging improved the predictive performance of ML models compared with traditional models. The best-performing ensemble model, using decision trees, Gaussian Naïve Bayes, and neural networks, achieved an AUC of 0.81. In contrast, Cox and logistic regression models showed significantly poorer performance (AUC of 0.58 and 0.50, respectively). SHapley Additive exPlanations analysis revealed that frailty, anxiety, and depression were critical in predicting readmission. Conclusion: ML models, particularly those using ensemble methods, significantly outperform traditional models in predicting short-term readmission for patients with HF. These findings highlight the potential of ML to improve clinical decision-making and resource allocation in HF management.en
dc.description.sponsorshipThis research has been funded by the Carlos III Health Institute through project PI15/01343 (co-funded by the European Regional Development Fund/European Social Fund Investing in your future), the Basque Government Health Department (2015111003), and the Research Committee of Galdakao-Usansolo Hospital (OSIBG18/203).en
dc.identifier.citationPikatza-Huerga, A., Almeida, A., Quiros, R., Larrea, N., Jose Legarreta, M., Zulaika, U., Garcia, R., & Garcia, S. (2025). Machine learning approaches for predicting heart failure readmissions. Postgraduate Medical Journal, 101(1202), 1351-1360. https://doi.org/10.1093/POSTMJ/QGAF102
dc.identifier.doi10.1093/POSTMJ/QGAF102
dc.identifier.eissn1469-0756
dc.identifier.issn0032-5473
dc.identifier.urihttps://hdl.handle.net/20.500.14454/4516
dc.language.isoeng
dc.publisherOxford University Press
dc.rights© The Author(s) 2025
dc.subject.otherAcute heart failure
dc.subject.otherBagging
dc.subject.otherExplainability
dc.subject.otherMachine learning
dc.subject.otherReadmission prediction
dc.titleMachine learning approaches for predicting heart failure readmissionsen
dc.typejournal article
dcterms.accessRightsopen access
oaire.citation.endPage1360
oaire.citation.issue1202
oaire.citation.startPage1351
oaire.citation.titlePostgraduate Medical Journal
oaire.citation.volume101
oaire.licenseConditionhttps://creativecommons.org/licenses/by/4.0/
oaire.versionVoR
Archivos
Bloque original
Mostrando 1 - 1 de 1
Miniatura
Nombre:
pikatza_machine_2025.pdf
Tamaño:
795.21 KB
Formato:
Adobe Portable Document Format
Descargar
Colecciones
Artículos