Real-time decision-making during emergency disease outbreaks.

Area of Science:

• Epidemiology
• Mathematical Modeling
• Infectious Disease Control

Background:

• Mathematical and simulation models are crucial for informing policy during infectious disease outbreaks.
• Early outbreak stages present significant parameter uncertainty, limiting model prediction accuracy.

Purpose of the Study:

• To evaluate the effectiveness of real-time decision-making in selecting optimal control interventions during infectious disease outbreaks.
• To compare early-stage policy recommendations with hindsight-based strategies using historical outbreak data.

Main Methods:

• Simulated real-time decision-making by fitting epidemic models to weekly, spatially-explicit infection data from historical outbreaks (foot-and-mouth disease, UK 2001; Miyazaki, Japan 2010).
• Compared forward simulations of alternative control interventions with policy recommendations derived from complete outbreak data.

Main Results:

• The optimal control policy identified using all outbreak data was also discernible early in the outbreak, despite projection variability.
• Improved understanding of infected farm locations, rather than transmission parameter accuracy, enhanced prediction of control intervention performance.
• Estimating undetected infectious premises depends on transmission parameter uncertainty.

Conclusions:

• Real-time model fitting and projection are essential for evaluating control interventions throughout an outbreak.
• Models are valuable at outbreak onset for informing policy.
• State-dependent interventions that adapt to new information are crucial for effective disease management.