Development and validation of a prognostic model based on comorbidities to predict COVID-19 severity: a...

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Prognosis for COVID-19 patients remains uncertain.
• Development of a predictive risk model is crucial for managing patient outcomes.
• Existing models may not fully capture disease severity progression.

Purpose of the Study:

• To derive and validate the Gal-COVID-19 scores, a novel risk model.
• To predict disease severity, hospitalization, ICU admission, and mortality in COVID-19 patients.
• To provide clinicians with a tool for risk stratification and decision-making.

Main Methods:

• Retrospective cohort study of RT-PCR confirmed COVID-19 patients in Galicia, Spain.
• Extraction of demographic and comorbidity data from electronic health records.
• Logistic regression models used to predict severity, with calibration and discrimination assessed.

Main Results:

• The Gal-COVID-19 model demonstrated good predictive performance for hospitalization (AUC 0.77), ICU admission (AUC 0.83), and mortality (AUC 0.89).
• Key predictors included age, gender, and chronic comorbidities (e.g., cardiovascular disease, diabetes, hypertension).
• Incidence of infection was 0.39% with significant rates of hospitalization (23.8%), ICU admission (2.7%), and mortality (5.2%).

Conclusions:

• The Gal-COVID-19 scores offer reliable risk estimates for COVID-19 severity.
• The model can assist clinicians in identifying and prioritizing high-risk individuals.
• Facilitates informed decision-making for healthcare authorities in managing the pandemic.