Contact network structure and risk of infection

Abstract

This study investigates the effect of structural features of contact networks on transmission of infection in populations. We examined measures of network centrality that may be useful to identify individuals of high risk of infection during outbreaks using susceptible-infectious-recovered models. Centrality describes an individual’s position in a population; numerous parameters are available to assess this attribute. Here we use a number of centrality measures including degree, eigenvector centrality, betweenness, closeness and information centrality. Each of the measures of centrality was associated with risk of infection in the simulated outbreaks. Importantly, degree (which is the most readily measured) was at least as good as other network parameters in predicting risk of infection. Identification of more central individuals in populations may be used to inform surveillance and infection control strategies.