Validation of graph-theoretical methods for pattern identification in public health datasets

Pattern identification issues are commonly used in public health practice to identify disease clusters and tendencies towards clustering. The basic building blocks or units for such patterns may be individuals or geographical units, but the key factor is the association between units in terms of time, space or other complex links. A range of methods has been developed for cluster detection but these methods are not designed to handle complex pattern searching. This paper describes early work in developing a novel method of tackling this problem, using graph theoretical techniques developed for computational chemistry. A modified version of the maximum common subgraph isomorphism method was used to search and retrieve enumeration districts (EDs) using 27 user-defined patterns from a set of 106 EDs. The results were then checked manually to ensure that all the appropriate and no additional patterns and EDs were retrieved. The program successfully retrieved all the relevant patterns and EDs and did not retrieve any patterns not specified by the query patterns. This study demonstrates the applicability of using graph theory for identifying and retrieving patterns in public health datasets.