Development of an Early-Phase Local Model for Pandemics Using Public Health Data: Application to the COVID-19...

Area of Science:

• Epidemiology
• Mathematical Modeling
• Public Health

Background:

• Novel infectious pathogens present significant challenges for early-phase transmission modeling due to data scarcity and incomplete pathogen understanding.
• Regional variations in outbreaks require localized modeling approaches to capture specific dynamics.

Purpose of the Study:

• To develop and validate an early-phase local model for disease transmission dynamics using constrained public health data.
• To address data limitations and model misspecifications through advanced statistical techniques.
• To enable real-time data updates and parameter estimation for dynamic outbreak analysis.

Main Methods:

• Utilized a quasi-likelihood approach with a flexible error term to handle data limitations and potential model misspecifications.
• Implemented an online estimator with an iterative algorithm for real-time parameter estimation and data assimilation.
• Applied the model to early COVID-19 data from over 800 U.S. counties, incorporating infection counts and county-level risk factors.

Main Results:

• Successfully predicted disease spread in early-phase COVID-19 outbreaks.
• Assessed the impact of social behavior, demographics, and vaccination coverage on disease transmission dynamics.
• Demonstrated the framework's utility in analyzing suboptimal data conditions.

Conclusions:

• The developed framework enhances early outbreak analysis by effectively utilizing constrained public health data.
• The model provides valuable insights for informing local pandemic response strategies under data-limited scenarios.
• This approach improves the understanding of disease transmission dynamics at a local level.