Plausible models for propagation of the SARS virus

TL;DR

This paper evaluates different models for SARS virus spread using Hong Kong data, finding that simple models are inadequate and that a small world network model better captures the complex, fluctuating outbreak patterns including super-spreader events.

Contribution

It introduces a small world network model for SARS propagation that more accurately reflects observed data and outbreak variability than traditional models.

Findings

Simple random models are insufficient.

Standard epidemic models do not account for data variability.

Small world network models can reliably simulate SARS outbreaks.

Abstract

Using daily infection data for Hong Kong we explore the validity of a variety of models of disease propagation when applied to the SARS epidemic. Surrogate data methods show that simple random models are insufficient and that the standard epidemic susceptible-infected-removed model does not fully account for the underlying variability in the observed data. As an alternative, we consider a more complex small world network model and show that such a structure can be applied to reliably produce simulations quantitative similar to the true data. The small world network model not only captures the apparently random fluctuation in the reported data, but can also reproduces mini-outbreaks such as those caused by so-called ``super-spreaders'' and in the Hong Kong housing estate Amoy Gardens.