A tailored machine learning approach for mortality prediction in severe COVID-19 treated with glucocorticoids
The International Journal of Tuberculosis and Lung Disease · 2024
DOI: 10.5588/ijtld.24.0169 ↗Kurz gefasst
Fachartikel in der Zeitschrift The International Journal of Tuberculosis and Lung Disease, 2024. Verfasst von F Salton und 10 weiteren Autor:innen. Beteiligte Einrichtung: Pulmonology Unit, University Hospital of Cattinara, Trieste, Italy;, Department of Medical Surgical and Health Sciences, University of Trieste, Tries…. Hauptthemen laut MEDLINE: Machine Learning, Glucocorticoids, COVID-19, COVID-19 Drug Treatment. Die Arbeit wurde bisher 4-mal zitiert (Crossref).
Abstract (Originalsprache)
BACKGROUND
The impact of severe COVID-19 pneumonia on healthcare systems highlighted the need for accurate predictions to improve patient outcomes. Despite the established efficacy of glucocorticoids (GCs), variable patient responses are observed, and the existing clinical scores are limited in predicting non-responders. We propose a machine learning (ML) based approach to predict mortality in COVID-19 pneumonia treated with GCs.
METHODS
This is an ML-driven retrospective analysis involving 825 patients. We leveraged XGBoost to select the most appropriate features from the initial 52, including clinical and laboratory data. Six different ML techniques were compared. Shapley additive explanation (SHAP) values were used to describe the influence of each feature on classification. Internal validation was performed.
RESULTS
Nine key predictors of death were identified: increasing C-reactive protein (CRP), decreasing arterial partial pressure of oxygen to fraction of inspired oxygen ratio (PaO 2 /FiO 2 ), age, coronary artery disease, invasive mechanical ventilation, acute renal failure, chronic heart failure, PaO 2 /FiO 2 earliest value, and body mass index. Random forest achieved the highest test area under the receiver operating characteristic curve at 0.938 (95% CI 0.903–0.969). SHAP values highlighted age and PaO 2 /FiO 2 improvement as the most influential features; the latter showed a higher impact than CRP reduction over time.
CONCLUSION
The proposed ML algorithm effectively predicts the risk of hospital death in COVID-19 pneumonia patients undergoing GCs. This approach can be adapted to datasets measuring similar clinical variables.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Substanzen
Autor:innen und Institutionen
- Pulmonology Unit, University Hospital of Cattinara, Trieste, Italy;, Department of Medical Surgical and Health Sciences, University of Trieste, Trieste, Italy.
- Department of Mathematics, Informatics and Geosciences, University of Trieste, Trieste, Italy.
- Department of Medical Surgical and Health Sciences, University of Trieste, Trieste, Italy.
- Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy;, Division of Internal Medicine, Multimedica Istituto di Ricovero e Cura a Carattere Scientifico, Milan, Italy.
Zitieren
APA
Salton, F., Rispoli, M., Confalonieri, P., De Nes, A., Spagnol, E., Salotti, A., Ruaro, B., Harari, S., Rocca, A., Manzoni, L., & Confalonieri, M. (2024). A tailored machine learning approach for mortality prediction in severe COVID-19 treated with glucocorticoids. The International Journal of Tuberculosis and Lung Disease. https://doi.org/10.5588/ijtld.24.0169
Vancouver
Salton F, Rispoli M, Confalonieri P, De Nes A, Spagnol E, Salotti A, et al. A tailored machine learning approach for mortality prediction in severe COVID-19 treated with glucocorticoids. The International Journal of Tuberculosis and Lung Disease. 2024. doi:10.5588/ijtld.24.0169
BibTeX
@article{Salton2024,
author = {Salton, F and Rispoli, M and Confalonieri, P and De Nes, A and Spagnol, E and Salotti, A and Ruaro, B and Harari, S and Rocca, A and Manzoni, L and Confalonieri, M},
title = {A tailored machine learning approach for mortality prediction in severe COVID-19 treated with glucocorticoids},
journal = {The International Journal of Tuberculosis and Lung Disease},
year = {2024},
doi = {10.5588/ijtld.24.0169}
}
EUVIMED