A path-specific SEIR model for use with general latent and infectious time distributions

Area of Science:

• Epidemiology
• Mathematical Biology
• Biostatistics

Background:

• Current Bayesian SEIR models often assume exponential distributions for latent and infectious periods.
• These assumptions may not accurately reflect the complex nature of infectious disease progression.
• Limited data availability for latent periods poses challenges for existing models.

Purpose of the Study:

• To develop an alternative SEIR model that accommodates general distributions for exposed and infectious periods.
• To remove the necessity of exponential distribution assumptions in SEIR modeling.
• To provide a more realistic framework for analyzing infectious disease dynamics, particularly when latent period data is incomplete.

Main Methods:

• Introduction of a path-specific SEIR (PS SEIR) model.
• The PS SEIR model tracks individual progression through exposed and infectious compartments.
• Demonstration of the PS SEIR model as a stochastic analog to deterministic SEIR models.

Main Results:

• The PS SEIR model allows for flexible, non-exponential distributions of latent and infectious periods.
• Simulation results show improved performance compared to population-averaged models.
• A novel analysis of the 2006 Iowa mumps epidemic using the PS SEIR model was conducted.

Conclusions:

• The PS SEIR model offers a more realistic and flexible approach to infectious disease modeling.
• This method enhances accuracy when dealing with variable incubation and infectious periods.
• The model provides a valuable tool for epidemiological studies and public health interventions.