Assessing the heterogeneity of disease spread through a community.

Care needs to be exercised in attempts at obtaining a description of the spread of disease merely by fitting a mathematical model to infectious disease data and adjusting the model until it adequately fits the observed epidemic curve. It is always necessary to perform separate statistical tests of the underlying assumptions of an epidemic model before attempting to use such a model to obtain epidemiologically meaningful insights into the mechanism of disease spread. Methods for such tests are presented and illustrated with reference to epidemics of respiratory diseases that occurred on the island of Tristan da Cunha in the South Atlantic between 1964 and 1968.