Study designs for dependent happenings.

In 1916, Sir Ronald Ross defined "dependent happenings" as events where the number affected in a unit of time depends on the number already affected. That is, the incidence depends on the prevalence, a characteristic of many infectious diseases. Because of this dependence, interventions against infectious diseases can have not only direct protective effects for the person receiving an intervention, but also indirect effects resulting from changes in the intensity of transmission in the population. This paper develops the conceptual framework for four types of study designs that differentiate and account for direct and indirect effects of intervention programs in dependent happenings.