Development and validation of a cytokine-driven predictive model for COVID-19 severity

开发和验证基于细胞因子的COVID-19严重程度预测模型

阅读:1

Abstract

BACKGROUND: This study aims to identify key factors influencing the severity of COVID-19 and to develop a predictive model for assessing disease severity. METHODS: A retrospective analysis was conducted on 164 patients with COVID-19 admitted to the First Affiliated Hospital of Dali University from July 2022 to February 2024, and the integrity of electronic medical records and laboratory databases of all included patients was systematically reviewed. Based on established diagnostic criteria for COVID-19 severity, patients were categorized into two groups: mild/ordinary (94 cases) and severe/critical (70 cases). Lasso regression and multivariate logistic regression were employed to identify independent risk factors, and a predictive model was constructed using a column chart. Internal validation of the model was conducted with receiver operating characteristic (ROC) curves, 10-fold cross-validation and calibration curves. The clinical utility of the model was evaluated using decision curve analysis. RESULTS: Univariate analysis revealed significant differences between the mild/ordinary and severe/critical groups in levels of IL-4, IL-6, IL-8, IL-10, IL-17, IFN-γ, WBC count, neutrophils (N), lymphocytes (L), red blood cells (RBC), hematocrit (HCT), R-CV, R-SD, CRP, PCT, DBI, AST, TP, ALB, urea, creatinine (CREA), uric acid (UA), sodium (Na), magnesium (Mg), calcium (Ca), and total cholesterol (TC) (all P < 0.05). Lasso regression identified IL-6, IL-8, IL-17, IFN-γ, N, R-CV, TP, ALB, and urea as predictive variables. Multivariate logistic regression confirmed that IL-6, IL-8, and IL-17 were independent risk factors for COVID-19 severity. The model demonstrated an area under the curve (AUC) of 0.88, with a sensitivity of 74.3% and specificity of 94.7%. The 10-fold AUCs are 0.691, 0.939, 0.657, 1.00, 0.81, 0.797, 0.75, 0.847, 0.969, and 1.00, with an average AUC of 0.856 (95% confidence interval: 0.791–0.922). The calibration curve showed strong agreement between predicted and actual values. Clinical decision curve analysis shows that the model offers optimal clinical utility, with a net benefit higher than alternative strategies when the probability threshold is between 15% and 95%. CONCLUSIONS: IL-6, IL-8, and IL-17 are identified as independent risk factors for the severity of COVID-19. The predictive model based on these factors provides a reliable tool for assessing disease severity in COVID-19 patients.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。