The spatial resolution of epidemic peaks

Area of Science:

• Epidemiology
• Mathematical Modeling
• Public Health

Background:

• Novel respiratory pathogens strain healthcare resources like intensive care units.
• Predicting epidemic peak incidence is crucial for evaluating public health interventions.
• Current models lack clarity on the interplay between epidemic trajectories, population mobility, and fine-grained spatial transmission dynamics.

Purpose of the Study:

• To investigate the impact of spatial resolution on epidemic trajectory modeling.
• To determine the relationship between spatial scale, population mobility, and predicted peak incidence.
• To provide a framework for utilizing high-resolution spatial incidence data in meta-population models.

Main Methods:

• Developed a spatially-explicit stochastic meta-population model with variable spatial resolution.
• Simulated the spread of an influenza-like pathogen across diverse population densities.
• Varied mobility assumptions using Latin-Hypercube sampling to assess its influence on epidemic dynamics.

Main Results:

• Peak incidence varied significantly with spatial resolution, despite constant cumulative attack rates.
• Identified critical resolution thresholds; models below these thresholds overestimated population-wide peak incidence.
• The influence of spatial resolution was most pronounced in lower-density and lower-mobility populations.

Conclusions:

• A fundamental spatial resolution exists for each pathogen-population interaction.
• Modeling at or above this fundamental resolution enables accurate prediction of city-scale epidemic peak incidence.
• Accurate spatial incidence data, when used within appropriate models, can significantly improve epidemic forecasting.