A framework for future national pediatric pandemic respiratory disease severity triage: The HHS pediatric COVID-19...

Area of Science:

• Computational epidemiology
• Pediatric infectious diseases
• Public health informatics

Background:

• COVID-19 remains a significant burden on the pediatric population due to persistent incidence and incomplete vaccination.
• Identifying at-risk pediatric patients for severe COVID-19 is crucial for effective healthcare resource allocation during surges.
• A lack of validated, nationwide computational tools hinders proactive risk identification in children.

Purpose of the Study:

• To develop computational models for identifying pediatric patients at risk of severe COVID-19 outcomes.
• To address two key clinical questions: risk for hospitalization in outpatients and risk for mechanical ventilation or cardiovascular interventions in hospitalized children.
• To leverage real-world data from the National COVID Cohort Collaborative (N3C).

Main Methods:

• A multi-agency, coordinated computational challenge was launched by HHS ASPR BARDA.
• Fifty-five computational models were evaluated across two distinct tasks.
• The National COVID Cohort Collaborative (N3C) provided the real-world data for model development and evaluation.

Main Results:

• Two winning models and three honorable mentions were selected from the evaluated submissions.
• The challenge successfully demonstrated the potential of computational modeling in addressing critical public health questions.
• This initiative represents a novel federal approach to leveraging data and competition for pandemic response.

Conclusions:

• The computational challenge provides a successful framework for government, research, and data repositories to collaborate on solutions during public health crises.
• This approach can accelerate the development and validation of tools for identifying at-risk pediatric populations.
• Future efforts can build upon this model to enhance preparedness and response to infectious disease threats.