ABSTRACT
I. INTRODUCTION
The detection of clusters of disease in space, time, or in both space and time is important to epidemiologist in trying to explain various health phenomena in a population. An aggregation of cases may yield clues as to the causative mechanism of the disease in question. Space clustering is a non-uniform distribution of the cases over the area relative to the underlying population. The existence of endemic pockets of disease can be recognized through a study of incidence rates for the suspect area. Time clustering is a non-uniform distribution of the cases over the duration of the study. Pure temporal clustering can be identified by a study of incidence rates. Space-time clustering is an interaction between the places of onset and the times of onset of the disease, cases which are close in space tending to be close in time. Such interaction is regarded as evidence for contagion or infection of the disease. The presence of time-space clustering usually indicates some causal environmental factor, which is local in both time and space, such as an infection. Many of the classic studies consider only clustering in one dimension, space or time. For example, the investigation of Snow (1855) of a cholera outbreak in London might be regarded as a demonstration of spatial clustering of cases around the Broad Street pump. However, it is only in recent years that space-time interaction has been used as a measure of clustering, and, in fact, has been the subject of considerable research.
