Predicting Early Unplanned Readmission to Critical Care Units Using Machine Learning and Metaheuristic Hyperparameter Optimization: Evidence from a Chilean Public Hospital
4 affiliations
- 1Institute for Health Care Research, Faculty of Nursing, Universidad Andrés Bello, Santiago 8370146, Chile
- 2Transportation and Logistics Center, Department of Engineering Sciences, Universidad Andrés Bello, Santiago 7500735, Chile
- 3Departamento de Ingeniería Industrial, Universidad del Bío-Bío, Concepción 4051381, Chile
- 4Instituto de Estadística, Pontificia Universidad Católica de Valparaíso, Valparaíso 2340031, Chile
Abstract
Background: Early unplanned readmission to critical care units is associated with adverse outcomes and consumes scarce capacity twice, yet the decision to transfer a patient out of the unit is still made without objective criteria. Machine learning may support this decision, but metaheuristic hyperparameter search remains largely unexplored here. Methods: We analyzed 371 patients discharged from the critical care unit of a Chilean public hospital, of whom 41 were readmitted within seven days of transfer to a general ward. Five models (logistic regression, support vector classification, decision tree, random forest, and XGBoost) were trained on clinical and nursing variables available at transfer, using nested stratified five-fold cross-validation with preprocessing fitted within the training folds. Bayesian optimization and a genetic algorithm were compared under identical folds. Results: Under nested cross-validation, discrimination was moderate (AUC 53.3–75.2%), and no single model performed best across all metrics. Logistic regression achieved the highest AUC (75.2%) with the lowest variability across folds, followed by XGBoost (75.0%), whose results varied markedly across folds. The genetic algorithm yielded higher AUC than Bayesian optimization for most tree-based models. Days in the critical care unit, age, and unit of origin ranked highest in impurity-based importance. Conclusions: This proof-of-concept study suggests that routinely collected clinical and nursing information available at transfer may inform early readmission prediction. As the sample was case-enriched and derived from a single center, recalibration to the natural readmission prevalence and prospective external validation are required before clinical implementation.
Related
- Development and Validation of a Machine Learning Model for Predicting Three-Month Unplanned Readmission in Patients with Moderate-to-Severe Chronic Obstructive Pulmonary Disease: A Retrospective Cohort Study
- Deep learning and metaheuristic optimization for predicting drug release from nanocarrier systems: a data-driven framework for formulation development
- Analysis of risk factors for unplanned readmission following acute appendicitis surgery in children.
- Predicting the Risk of Unplanned Hospital Encounters in Hemodialysis Patients: A Comparison Between Clinicians and a Machine Learning Prediction Algorithm
- Predicting unplanned readmissions in neurological patients: A large-scale AI-driven analysis using CHAID decision trees
BibTeX
@article{barrigagallegos2026predicting,
title = {{Predicting Early Unplanned Readmission to Critical Care Units Using Machine Learning and Metaheuristic Hyperparameter Optimization: Evidence from a Chilean Public Hospital}},
author = {Fredy Barriga-Gallegos and Gonzalo Ríos-Vásquez and Antonia Ilabaca and Hanns de la Fuente-Mella and Verónica Badilla-Morales and Naldy Febré Vergara},
journal = {Journal of Clinical Medicine},
year = {2026},
month = oct,
doi = {10.3390/jcm15218328},
url = {https://www.mdpi.com/2077-0383/15/21/8328},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Predicting Early Unplanned Readmission to Critical Care Units Using Machine Learning and Metaheuristic Hyperparameter Optimization: Evidence from a Chilean Public Hospital
AU - Fredy Barriga-Gallegos
AU - Gonzalo Ríos-Vásquez
AU - Antonia Ilabaca
AU - Hanns de la Fuente-Mella
AU - Verónica Badilla-Morales
AU - Naldy Febré Vergara
PY - 2026
DA - 2026/10/26
JO - Journal of Clinical Medicine
DO - 10.3390/jcm15218328
UR - https://www.mdpi.com/2077-0383/15/21/8328
AB - Background: Early unplanned readmission to critical care units is associated with adverse outcomes and consumes scarce capacity twice, yet the decision to transfer a patient out of the unit is still made without objective criteria. Machine learning may support this decision, but metaheuristic hyperparameter search remains largely unexplored here. Methods: We analyzed 371 patients discharged from the critical care unit of a Chilean public hospital, of whom 41 were readmitted within seven days of transfer to a general ward. Five models (logistic regression, support vector classification, decision tree, random forest, and XGBoost) were trained on clinical and nursing variables available at transfer, using nested stratified five-fold cross-validation with preprocessing fitted within the training folds. Bayesian optimization and a genetic algorithm were compared under identical folds. Results: Under nested cross-validation, discrimination was moderate (AUC 53.3–75.2%), and no single model performed best across all metrics. Logistic regression achieved the highest AUC (75.2%) with the lowest variability across folds, followed by XGBoost (75.0%), whose results varied markedly across folds. The genetic algorithm yielded higher AUC than Bayesian optimization for most tree-based models. Days in the critical care unit, age, and unit of origin ranked highest in impurity-based importance. Conclusions: This proof-of-concept study suggests that routinely collected clinical and nursing information available at transfer may inform early readmission prediction. As the sample was case-enriched and derived from a single center, recalibration to the natural readmission prevalence and prospective external validation are required before clinical implementation.
SN - 2077-0383
ER -