Risk factors for severe COVID-19 differ by age for hospitalized adults

Area of Science:

• Infectious Diseases
• Medical Informatics
• Epidemiology

Background:

• Effective risk stratification for hospitalized COVID-19 patients is crucial for resource allocation and patient management.
• Existing risk models often lack optimization for distinct age groups (younger vs. older adults) and may not reflect current treatment advancements.

Purpose of the Study:

• To develop and validate updated risk models for predicting severe COVID-19 outcomes (mechanical ventilation or death) in hospitalized adults.
• To assess the performance of these models across different age groups and identify key predictors.

Main Methods:

• Retrospective analysis of 6906 hospitalized adults with COVID-19 from a multi-state community health system.
• Development of risk models using clinical data available within the first hour of admission or positive SARS-CoV-2 test.
• Evaluation of model performance using Area Under the Receiver Operating Characteristic curve (AUROC) for predicting outcomes within 7 days.

Main Results:

• Age-stratified models achieved high predictive accuracy (AUROC 0.81-0.82) for 7-day outcomes.
• Significant differences in risk factor importance were observed between younger (<50 years) and older (≥50 years) adult populations.
• Vital signs and laboratory results were more predictive than sex or chronic comorbidities for hospitalized patients.

Conclusions:

• Updated, age-stratified risk models enhance the prediction of severe outcomes for hospitalized COVID-19 patients.
• These models provide a valuable tool for clinical decision-making and resource management, tailored to different age demographics.
• Early clinical data are sufficient for developing accurate risk stratification models for COVID-19 patients.