SaTScan was developed by Martin Kulldorff to scan for temporal, spatial, and spatial temporal clusters. It places circles or ellipses of continuously varying size over a spatial study area and can add time as a continuously varying third-dimension scan. The program offers a wide variety of scanning models. Paraphrasing Kulldorff, SaTScan calculates a Poisson-based model according to a known population at risk, a Bernoulli model which allows for cases and controls, a space-time permutation model that needs only case data, an ordinal model, an exponential model for survival analysis, and a normal model for continuous data. Either the data may be aggregated to a geographic region or each case may have unique coordinates. The end result is quite intuitive and includes the location of a cluster in space and time and the significance of the cluster based on a Monte Carlo simulation. Although analysis is easy to do and interpret, input and output are unnecessarily cumbersome. SaTScan has no direct interface with any statistical, database, or GIS program, but it requires their use.