A Machine Learning Framework for COVID-19 Mortality Risk Prediction Using Clinical and Demographic Data
Keywords:
Deep Learning, Medical Prognosis, COVID-19 Risk, Machine Learning, Predictive HealthcareAbstract
The COVID-19 pandemic revealed an immediate need for trustworthy mortality prediction tools for effective allocation of health care resources. This paper presents the creation and validation of a machine learning model for predicting mortality risk using clinical and demographic information from the Mexican national registry (1,048,575 records, February 2020-October 2021). In this study we utilized six different algorithms: Logistic Regression, Support Vector Machine, Random Forest, AdaBoost, Decision Tree, and Deep Neural Networks. The models were taught 21 clinically relevant variables (including comorbidities, treatment, and demographic factors). For model assessment, we implemented stratified 80-20 training/testing splits with 5-fold cross-validation, and measured accuracy, F1-score, precision, and recall. The best value of accuracy ( 0.9181 ) was obtained by the Deep Neural Networks, and by Logistic Regression with value 0.9107 and a better value of interpretability. The model performance difference was only 0.7 percentage points which implies that the interpretable models strike a good deal for clinical usage. The more interpretable a model is; the more a clinician is likely to trust it. Importance of certain features measures ( pneumonia, hospitalization type, age ) were found to be predictors of interest, clinically plausible. The framework shows the value of clinically useful interpretable ML models and offers support for their use in sensitive medical decision making
Downloads
References
[1] World Health Organization, "WHO Director-General’s opening remarks at the media briefing on COVID-19 - 11 March 2020," World Health Organization, 2020.
[2] E. Dong, H. Du, and L. Gardner, "An interactive web-based dashboard to track COVID-19 in real time," The Lancet Infectious Diseases, vol. 20, no. 5, p. 533–534, 2020.
[3] B. Armocida, B. Formenti, S. Ussai, F. Palestra, and E. Missoni, "The Italian health system and the COVID-19 challenge," The Lancet Public Health, vol. 5, no. 5, p. e253, 2020.
[4] C. Huang et al., "Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China," The Lancet, vol. 395, no. 10223, p. 497–506, 2020.
[5] F. Zhou et al., "Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: A retrospective cohort study," The Lancet, vol. 395, no. 10229, pp. 1054–1062, 2020.
[6] M. A. O. Ahmed, R. Alotaibi, Y. A. Satar, N. Gaber, N. F. Omran, and O. Reyad, "Fast detection of acute lymphoblastic leukemia through stacked pre-trained ensemble learning and efficient segmentation," Arabian Journal for Science and Engineering, p. 1–14, 2025.
[7] M. A. O. Ahmed, Y. A. Satar, E. M. Darwish, and E. A. Zanaty, "Synergistic integration of multi-view brain networks and advanced machine learning techniques for auditory disorders diagnostics," Brain Informatics, vol. 11, no. 1, p. 3, 2024.
[8] E. W. Steyerberg and Y. Vergouwe, "Towards better clinical prediction models: Seven steps for development and an ABCD for validation," European Heart Journal, vol. 35, no. 29, p. 1925–1931, 2019.
9] J. H. Chen and S. M. Asch, "Machine learning and prediction in medicine—beyond the peak of inflated expectations," New England Journal of Medicine, vol. 376, no. 26, p. 2507–2509, 2016.
[10] S. Rajaraman et al., "Pre-trained convolutional neural networks as feature extractors toward improved malaria parasite detection in thin blood smear images," PeerJ, vol. 6, e4568, 2018.
[11] S. Wang et al., "A deep learning algorithm using CT images to screen for COVID-19," MedRxiv, 2020.
[12] W. Liang et al., "Early triage of critically ill COVID-19 patients using deep learning," Nature Communications, vol. 11, no. 1, p. 3543, 2020.
[13] L. Yan et al., "An interpretable mortality prediction model for COVID-19 patients," Nature Machine Intelligence, vol. 2, no. 5, p. 283–288, 2020.
[14] L. Wynants et al., "Prediction models for diagnosis and prognosis of COVID-19: Systematic review and critical appraisal," BMJ, vol. 369, m1328, 2020.
[15] R. K. Gupta et al., "Systematic evaluation and external validation of 22 prognostic models among hospitalised adults with COVID-19," European Respiratory Journal, vol. 57, no. 6, p. 2003498, 2021.
[16] C. Rudin, "Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead," Nature Machine Intelligence, vol. 1, no. 5, p. 206–215, 2019.
[17] R. Caruana et al., "Intelligible models for healthcare: Predicting pneumonia risk and hospital 30-day readmission," Proc. ACM SIGKDD Int. Conf. Knowl. Discov. Data Min., p. 1721–1730, 2015.
[18] S. M. Lundberg and S. I. Lee, "A unified approach to interpreting model predictions," Adv. Neural Inf. Process. Syst., vol. 30, p. 4765–4774, 2017.
[19] E. G. Price-Haywood, J. Burton, D. Fort, and L. Seoane, "Hospitalization and mortality among black patients and white patients with COVID-19," New England Journal of Medicine, vol. 382, no. 26, p. 2534–2543, 2020.
[20] Z. Obermeyer, B. Powers, C. Vogeli, and S. Mullainathan, "Dissecting racial bias in an algorithm used to manage the health of populations," Science, vol. 366, no. 6464, pp. 447–453, 2021.
[21] M. A. O. Ahmed, Y. AbdelSattar, and I. Abbas, "Expected risk minimization and robust preventive inference of transfer learning for COVID-19 diagnosis within chest X-rays," Sohag Journal of Sciences, vol. 8, no. 1, p. 75–82, 2023.
[22] M. A. O. Ahmed, I. Abbas, and Y. AbdelSatar, "HDSNE: A new unsupervised multiple image database fusion learning algorithm for lung infection diagnosis in chest X-ray images," BMC Medical Imaging, vol. 23, no. 1, p. 134, 2023.
[23] A. Karlinsky and D. Kobak, "Tracking excess mortality across countries during the COVID-19 pandemic with the World Mortality Dataset," Elife, vol. 10, e69336, 2021.
[24] S. Barquera, L. Hernández-Barrera, B. Trejo-Valdivia, T. Shamah, I. Campos-Nonato, and J. Rivera-Dommarco, "Obesity in Mexico: Prevalence and trends in adults," Salud Pública de México, vol. 62, no. 6, p. 682–692, 2020.
[25] L. Moncada, M. Flores, C. Alpuche-Aranda, et al., "Health-care policies during the COVID-19 pandemic in Mexico," Health Policy OPEN, vol. 5, p. 100100, 2023.
[26] O. Mendez, C. Alpuche, S. Doubova, et al., "Twenty-three years of public policy towards universal health coverage in Mexico: A time-series analysis," The Lancet Regional Health - Americas, vol. 52, p. 101271, 2025.
[27] C. Cisterna, M. Rinaldi, N. Amoroso, and R. Bellotti, "A novel method for extreme imbalance based on an ensemble of logistic regression for COVID-19 mortality prediction," Scientific Reports, vol. 12, no. 1, p. 22437, 2022.
[28] T. W. Tulu et al., "Machine learning-based prediction of COVID-19 mortality using immunological and metabolic biomarkers," BMC Digital Health, vol. 1, no. 1, p. 6, 2023.
[29] H. Gong, M. Wang, H. Zhang, M. F. Elahe, and M. Jin, “An explainable AI approach for the rapid diagnosis of COVID-19 using ensemble learning algorithms,” Front. Public Health, vol. 10, p. 874455, 2022, doi:10.3389/fpubh.2022.874455.
[30] F. López Seguí, F. García Cuyàs, X. Sanz, et al., "Machine learning-based mortality prediction in patients with complex chronic conditions: A retrospective study," PLOS ONE, vol. 18, no. 5, e0285311, 2023.
[31] H. M. R. U. Rehman et al., “A systematic literature study of machine learning techniques based on intrusion detection,” J. Big Data, vol. 12, Art. no. 264, 2025.
[32] M. Nizri, "COVID-19 dataset," Kaggle, 2022.
[33] Secretaría de Salud México, "Datos abiertos: Dirección General de Epidemiología," Gobierno de México, 2023.
Downloads
Published
Issue
Section
Categories
License
Copyright (c) 2026 Engineering Systems and Intelligent Technologies (ESIT)
This work is licensed under a Creative Commons Attribution 4.0 International License.
Engineering Systems and Intelligent Technologies (ESIT) content is published under a Creative Commons Attribution License (CCBY). This means that content is freely available to all readers upon publication, and content is published as soon as production is complete.
Engineering Systems and Intelligent Technologies (ESIT) seeks to publish the most influential papers that will significantly advance scientific understanding. Selected articles must present new and widely significant data, syntheses, or concepts. They should merit recognition by the wider scientific community and the general public through publication in a reputable scientific journal.
