Prediction models based on machine learning algorithms for COVID-19 severity risk

Abstract Background The World Health Organization has highlighted the risk of Disease X, urging pandemic preparedness. Coronavirus disease 2019 (COVID-19) could be the first Disease X; therefore, understanding the epidemiological experiences of COVID-19 is crucial while preparing for future similar...

Full description

Saved in:
Bibliographic Details
Main Authors: Hansong Zhang, Ying Wang, Yan Xie, Cuihan Wang, Yuqi Ma, Xin Jin
Format: Article
Language:English
Published: BMC 2025-05-01
Series:BMC Public Health
Subjects:
Online Access:https://doi.org/10.1186/s12889-025-22976-x
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849734092731449344
author Hansong Zhang
Ying Wang
Yan Xie
Cuihan Wang
Yuqi Ma
Xin Jin
author_facet Hansong Zhang
Ying Wang
Yan Xie
Cuihan Wang
Yuqi Ma
Xin Jin
author_sort Hansong Zhang
collection DOAJ
description Abstract Background The World Health Organization has highlighted the risk of Disease X, urging pandemic preparedness. Coronavirus disease 2019 (COVID-19) could be the first Disease X; therefore, understanding the epidemiological experiences of COVID-19 is crucial while preparing for future similar diseases. Methods Prediction models for COVID-19 severity risk in hospitalized patients were constructed based on four machine learning algorithms, namely, logistic regression, Cox regression, support vector machine (SVM), and random forest. These models were evaluated for prediction accuracy, area under the curve (AUC), sensitivity, and specificity as well as were interpreted using SHapley Additive exPlanation. Results Data were collected from 1,485 hospitalized patients across 6 centers, comprising 1,184 patients with severe or critical COVID-19 and 301 patients with nonsevere COVID-19. Among the four models, the SVM model achieved the highest prediction accuracy of 98.45%, with an AUC of 0.994, a sensitivity of 0.989, and a specificity of 0.969. Moreover, oxygenation index (OI), confusion, respiratory rate, and age were found to be predictors of COVID-19 severity risk. Conclusions SVM could accurately predict COVID-19 severity risk; thus, it can be prioritized as a prediction model. OI is the most critical predictor of COVID-19 severity risk and can serve as the primary and independent evaluation indicator.
format Article
id doaj-art-d0809406ec2c469fbdcdfc107b1c02f7
institution DOAJ
issn 1471-2458
language English
publishDate 2025-05-01
publisher BMC
record_format Article
series BMC Public Health
spelling doaj-art-d0809406ec2c469fbdcdfc107b1c02f72025-08-20T03:07:54ZengBMCBMC Public Health1471-24582025-05-0125111910.1186/s12889-025-22976-xPrediction models based on machine learning algorithms for COVID-19 severity riskHansong Zhang0Ying Wang1Yan Xie2Cuihan Wang3Yuqi Ma4Xin Jin5School of Mechanical Engineering, Tianjin UniversityDepartment of Nursing, Tianjin First Center HospitalDepartment of Liver Transplantation, Tianjin First Center HospitalTianjin Nankai Hospital, Tianjin Medical UniversitySchool of Mechanical Engineering, Tianjin UniversityMedical School of Tianjin UniversityAbstract Background The World Health Organization has highlighted the risk of Disease X, urging pandemic preparedness. Coronavirus disease 2019 (COVID-19) could be the first Disease X; therefore, understanding the epidemiological experiences of COVID-19 is crucial while preparing for future similar diseases. Methods Prediction models for COVID-19 severity risk in hospitalized patients were constructed based on four machine learning algorithms, namely, logistic regression, Cox regression, support vector machine (SVM), and random forest. These models were evaluated for prediction accuracy, area under the curve (AUC), sensitivity, and specificity as well as were interpreted using SHapley Additive exPlanation. Results Data were collected from 1,485 hospitalized patients across 6 centers, comprising 1,184 patients with severe or critical COVID-19 and 301 patients with nonsevere COVID-19. Among the four models, the SVM model achieved the highest prediction accuracy of 98.45%, with an AUC of 0.994, a sensitivity of 0.989, and a specificity of 0.969. Moreover, oxygenation index (OI), confusion, respiratory rate, and age were found to be predictors of COVID-19 severity risk. Conclusions SVM could accurately predict COVID-19 severity risk; thus, it can be prioritized as a prediction model. OI is the most critical predictor of COVID-19 severity risk and can serve as the primary and independent evaluation indicator.https://doi.org/10.1186/s12889-025-22976-xPrediction modelsMachine learning algorithmsSeverity riskCOVID-19
spellingShingle Hansong Zhang
Ying Wang
Yan Xie
Cuihan Wang
Yuqi Ma
Xin Jin
Prediction models based on machine learning algorithms for COVID-19 severity risk
BMC Public Health
Prediction models
Machine learning algorithms
Severity risk
COVID-19
title Prediction models based on machine learning algorithms for COVID-19 severity risk
title_full Prediction models based on machine learning algorithms for COVID-19 severity risk
title_fullStr Prediction models based on machine learning algorithms for COVID-19 severity risk
title_full_unstemmed Prediction models based on machine learning algorithms for COVID-19 severity risk
title_short Prediction models based on machine learning algorithms for COVID-19 severity risk
title_sort prediction models based on machine learning algorithms for covid 19 severity risk
topic Prediction models
Machine learning algorithms
Severity risk
COVID-19
url https://doi.org/10.1186/s12889-025-22976-x
work_keys_str_mv AT hansongzhang predictionmodelsbasedonmachinelearningalgorithmsforcovid19severityrisk
AT yingwang predictionmodelsbasedonmachinelearningalgorithmsforcovid19severityrisk
AT yanxie predictionmodelsbasedonmachinelearningalgorithmsforcovid19severityrisk
AT cuihanwang predictionmodelsbasedonmachinelearningalgorithmsforcovid19severityrisk
AT yuqima predictionmodelsbasedonmachinelearningalgorithmsforcovid19severityrisk
AT xinjin predictionmodelsbasedonmachinelearningalgorithmsforcovid19severityrisk