Spatial Heterogeneity in Infectious Disease Epidemics

Infectious disease epidemics in populations are inherently spatial—infectious agents are spread by contact from an infectious host to a susceptible host nearby.Among-host differences can determine which hosts suffer disease and the population dynamics of infectious disease epidemics.From the perspective of the infectious agent, a host is a habitat patch; among-host differences that are epidemiologically important are related to the concepts of compositional and configurational heterogeneity in landscape ecology. Heterogeneous mixing in epidemiology encompasses factors that determine who comes into contact with whom; it is analogous to configurational heterogeneity in landscape ecology. Other sorts of heterogeneity are analogous to compositional heterogeneity, including among-host differences in the duration of an infection, susceptibility to infection, or the amount of an infectious agent that is dispersed from an infected host. In real epidemics, compositional heterogeneity and configurational heterogeneity can introduce an overwhelming amount of complexity. Mathematical modeling provides a method for understanding epidemic processes and for taming the complexity. The idea of epidemic distance is introduced as a way of comparing and contrasting two different epidemic processes, and it is used to compare and contrast some of the mathematical models used to understand the role of space and spatial heterogeneity in epidemiology. In understanding real epidemics, the notion of parsimony is a guiding principle—heterogeneity should be weighed and ignored whenever possible. Several case studies are presented in which compositional and configurational heterogeneity are shown to be important.