The Canadian health system is increasingly under stress from an aging population and reduction of healthcare funding. Nowhere is this stress greater felt than in the home support field. Home support can be thought of as the development and implementation of both social and medical programs that assist outpatients so they remain in their homes. In order to optimize home support scheduling and mitigate the strain of increased demand, health authorities have increasingly turned to GIS. Predominant within the scheduling literature are route finding approaches which are based on a linear progression between locations. However, this approach is not ideal in situations where health authorities wish to provide their employees with high levels of autonomy and flexibility. Accordingly, in this paper we outline a new method of home support scheduling based on spatial network cluster analysis in partnership with an integrated software model to group patients by address along a road network. We then apply this model to the area administered by the Vancouver Island Health Authority (VIHA). We find that this approach successfully groups patients by address while avoiding the limitations of Euclidean based cluster methods. The advantages and limitations of this approach are then reviewed and directions for further research are discussed.
Matthew Bakker, Kelvin Co, Clayton Crawford, James Mack
