SIR-Model for Households

TL;DR

This paper introduces a novel ODE-based household model for disease spread that explicitly accounts for household size and infection dynamics, enabling analytical computation of key epidemiological metrics.

Contribution

It develops a new compartmental ODE model incorporating household size effects and validates it against agent-based simulations for realistic data.

Findings

Good agreement between ODE and agent-based models

Large households significantly boost disease transmission

Analytical formulas for reproduction number and peak prevalence

Abstract

Households play an important role in disease dynamics. Many infections happening there due to the close contact, while mitigation measures mainly target the transmission between households. Therefore, one can see households as boosting the transmission depending on household size. To study the effect of household size and size distribution, we differentiated the within and between household reproduction rate. There are basically no preventive measures, and thus the close contacts can boost the spread. We explicitly incorporated that typically only a fraction of all household members are infected. Thus, viewing the infection of a household of a given size as a splitting process generating a new, small fully infected sub-household and a remaining still susceptible sub-household we derive a compartmental ODE-model for the dynamics of the sub-households. In this setting, the basic reproduction number as well as prevalence and the peak of an infection wave in a population with given households size distribution can be computed analytically. We compare numerical simulation results of this novel household-ODE model with results from an agent--based model using data for realistic household size distributions of different countries. We find good agreement of both models showing the catalytic effect of large households on the overall disease dynamics.