Disease X epidemic control using a stochastic model and a deterministic approximation: Performance comparison with...

Area of Science:

• Epidemiology
• Mathematical Modeling
• Decision Science

Background:

• Infectious disease spread can be modeled using deterministic or stochastic approaches.
• Deterministic models offer an approximation but cannot capture population discreteness.
• Model selection is crucial from a decision-making perspective.

Purpose of the Study:

• To investigate the impact of model choice (stochastic vs. deterministic) on vaccination policy decisions for an emerging disease.
• To evaluate decision-making performance under varying parameter uncertainty and sample sizes.
• To understand the trade-offs between model complexity and decision outcomes.

Main Methods:

• A stochastic SIR model and its deterministic approximation were used to model Disease X in a closed population.
• Decision-making scenarios included using either model, with or without parameter uncertainty.
• The impact of different sample sizes for parameter draws and model runs was assessed.

Main Results:

• The choice of model significantly influences the selected vaccination policies.
• Optimal performance was achieved using the stochastic model with known parameters and large sample sizes.
• For small sample sizes, the deterministic model may outperform the stochastic model due to stochastic effects.

Conclusions:

• Model selection, parameter uncertainty, and sample size are interconnected in infectious disease modeling.
• Optimizing stochastic models requires careful consideration of these interacting factors.
• Resolving parameter uncertainty can be more beneficial than switching to a stochastic model in certain cases.