Epidemics on networks with large initial conditions or changing structure

TL;DR

This paper extends edge-based compartmental models to accurately analyze epidemic spread on networks with large initial infection proportions and dynamic structures, addressing limitations of previous models.

Contribution

It introduces a generalized model that accounts for finite initial conditions and changing network structures, enhancing applicability to real-world epidemic scenarios.

Findings

Models accurately capture effects of large initial infections

Generalization to dynamic networks is feasible

Models can inform intervention strategies

Abstract

Background: Recently developed techniques to study the spread of infectious diseases through networks make assumptions that the initial proportion infected is infinitesimal and the population behavior is static throughout the epidemic. The models do not apply if the initial proportion is large (and fail whenever R_0<1), and cannot measure the impact of an intervention. Methods: In this paper we adapt "edge-based compartmental models" to situations having finite-sized initial conditions. Results: The resulting models remain simple and accurately capture the effect of the initial conditions. It is possible to generalize the model to networks whose partnerships change in time. Conclusions: The resulting models can be applied to a range of important contexts. The models can be used to choose between different interventions that affect the disease or the population structure.