A predictive model for disease severity among COVID-19 elderly patients based on IgG subtypes and machine learning
Zhenchao Zhuang1, Yuxiang Qi2, Yimin Yao1
1Department of Laboratory Medicine, The First Affiliated Hospital of Zhejiang Chinese Medical University (Zhejiang Provincial Hospital of Chinese Medicine), Hangzhou, China.
Frontiers in Immunology
|December 18, 2023
Summary
A new model predicts severe coronavirus disease 2019 (COVID-19) in elderly patients using logistic regression and key variables. This tool aids clinicians in assessing COVID-19 severity and improving outcomes for older adults.
Area of Science:
- Geriatric Medicine
- Infectious Diseases
- Computational Biology
Background:
- Elderly individuals face a high mortality rate from coronavirus disease 2019 (COVID-19) due to increased pneumonia risk.
- Existing models lack the ability to predict COVID-19 severity in the elderly based on immunoglobulin G (IgG) subtypes.
Purpose of the Study:
- To develop and validate a novel algorithm for distinguishing severe COVID-19 cases among elderly patients.
- To identify key predictors for forecasting COVID-19 severity in this demographic.
Main Methods:
- Retrospective analysis of laboratory data from 103 elderly patients with confirmed SARS-CoV-2 infection.
- Development of predictive models using machine learning algorithms, including logistic regression, with data split into training, testing, and external validation cohorts.
- Model performance evaluated using Area Under the Curve (AUC), calibration curves, Decision Curve Analysis (DCA), and Shapley Additive Explanations (SHAP).
Main Results:
- A logistic regression model with four key variables was identified as the best predictor of COVID-19 severity.
- The model achieved high performance: AUC of 0.889 (training), 0.824 (testing), and 0.74 (external validation).
- Excellent calibration and significant clinical utility were demonstrated by the model.
Conclusions:
- A validated model effectively distinguishes between severe and non-severe COVID-19 cases in the elderly.
- This predictive tool can assist clinicians in evaluating disease severity and mitigating adverse outcomes in elderly COVID-19 patients.
