An entropy-based algorithm for detecting clusters of cases and controls and its comparison with a method using nearest neighbours.

A new method for detecting disease clustering based on entropy is presented. For this method cases and controls are plotted on a map. The map is divided into regions. The entropy of the space is calculated as the log of the number of possible ways of placing the cases and controls in the various regions given the total number of cases and controls and the number of cases and controls in each region. The power of the entropy technique is tested against the power of the nearest neighbour technique (NNT). The entropy method is shown to be substantially more powerful than the NNT when there is more than one cluster in the space or when the clusters are near the boundary of the space.