Modeling an epidemic in an imaginary small town

Bell, G.

The course of an epidemic in an imaginary small town has been simulated with an agent-based model. The reproduction number R of the virus could be counted directly, and was roughly, but not precisely, exponentially distributed. The number of secondary infections was greater for an infection which was itself one of many secondary infections because of environmental heterogeneity, which created variance of R among sites and could drive the spread of infection, even when global R < 1. Different kinds of intervention were deployed to curtail the spread of infection. Measures applied to the general population, such as closing down sites and services or regulating individual behaviour, did not reduce the total number of individuals infected during the epidemic unless they were maintained until the virus became extinct. This was primarily because measures taken to reduce indirect transmission tended to increase direct transmission, and vice versa. Consequently, the overall effect of any combination of interventions was much less than the sum of their separate effects. On the other hand, the quarantine of infected or exposed individuals was effective in driving the virus to extinction and caused a permanent and substantial reduction in the number of cases.