A discrete time-space geography for epidemiology: from mixing groups to pockets of local order in pandemic simulations.

The World Health Organization urges all nations to develop and maintain national influenza preparedness plans. Important components of such plans are forecasts of morbidity and mortality based on local social and geographic conditions. Most methodologies for simulations of epidemic outbreaks are implicitly based on the assumption that the frequency and duration of social contacts that lead to disease transmission is affected by geography, i.e. the spatial distribution of physical meeting places. In order to increase the effectiveness of the present methods for simulation of infectious disease outbreaks, the aim of this study is to examine two social geographic issues related to such models. We display how the social geographic characteristics of mixing networks, in particular when these significantly deviate from the random-mixing norm, can be represented in order to enhance the understanding and prediction of epidemic patterns in light of a possible future destructive influenza pandemic. We conclude that social geography, social networks and simulation models of directly transmitted infectious diseases are fundamentally linked.