Fractional poisson--a simple dose-response model for human norovirus

Area of Science:

• Virology
• Epidemiology
• Mathematical Modeling

Background:

• Norovirus (NoV) infection risk is modeled using dose-response relationships.
• Previous models like beta-Poisson incorporated virus aggregation and host susceptibility variations.
• Updating these models with new human challenge data is crucial for accurate risk assessment.

Purpose of the Study:

• To update Norovirus (NoV) dose-response models using combined human challenge data.
• To evaluate the existing beta-Poisson model and propose a simpler, more efficient alternative.
• To improve the understanding of NoV infection probability at varying exposure levels.

Main Methods:

• Utilized combined old and new Norovirus (NoV) human challenge data.
• Examined the beta-Poisson dose-response model, including parameters for virus aggregation and host susceptibility.
• Developed and validated a new fractional Poisson model with fewer parameters.

Main Results:

• The fractional Poisson model simplifies host susceptibility by using a single parameter for the fraction of susceptible hosts.
• This new model accounts for virus aggregation similarly to the beta-Poisson model.
• The fractional Poisson model showed comparable deviance to the beta-Poisson model but with fewer parameters, favored by the Akaike information criterion.

Conclusions:

• The fractional Poisson model is computationally simple and well-suited for Norovirus (NoV) human challenge data.
• It provides a statistically favored alternative to the beta-Poisson model for dose-response analysis.
• Further low-dose Norovirus (NoV) data is recommended to refine risk assessment for environmental exposures.