As Canada’s population continues to age the strain placed on our health care system will only increase. As the current health system is not prepared to handle this increased strain (Coyte 2000), innovative and practical solutions are called for. As this paper has shown, adoption of GIS theories and methods can mitigate this stress. The application of spatial network cluster analysis to patient addresses facilitates a move to more effective and efficient home support worker scheduling, leading to greater overall patient satisfaction and reduced cost to the healthcare provider.
      While the approach in the paper was designed with a place-specific and institutional context in mind, the underlying methods can be applied across a broader range of issues where the locations of patient address clusters are of principle concern. Indeed, a similar broad application of home support worker routing methods has been implemented across a range of place-based contexts (Cheng and Rich 1998; Bräysy, Dullaert, and Nakari 2007; Dohn et al. 2008). The application of the model outlined above, complete with its advantages over Euclidean based models, will surely provide similar benefits across a range of locations and contexts.
      Additionally, beyond applications to home support worker scheduling, the successful implementation of this spatial network cluster analysis in locating clusters of patient addresses provides opportunity for additional research into the underlying geospatial factors influencing human health. Such research can inform future policies on a range of health issues and further strengthen healthcare systems in the coming decades as the face increased demand with restricted funding.
      Though the application of spatial network cluster analysis in home support worker scheduling in its infancy, it represents a promising method to address the future strains facing healthcare. As this area of research moves forward a number of exciting opportunities exist to further enhance its ability to capture the networked space in which humans live and make decisions.

TOP

Opportunities for Future Research
      Moving forward we suggest that these home support service optimization techniques not be regarded in isolation from alternative approaches. Following the lead of Begur, Miller, and Weaver (1997) who integrated GIS and a heuristic decision support system, we see opportunity for spatial network cluster approaches to be integrated with other methods. For example, point clustering techniques do not exclude route finding methods and the two could be combined. A Dijkstra shortest route algorithm, popular in route scheduling, could be employed within an identified patient address cluster to calculate an optimal route for a home support service worker. Such creative application of a diverse set of techniques will most likely further enhance home support worker satisfaction, patient health and happiness, as well as reduce costs for health care providers.
      For such applications to be feasible however, better integration of network analysis tools into established software programs, such as ESRI’s ArcGIS platform, must occur. If patient clustering and route optimization techniques are to be integrated in practice then there isolation computationally must also be addressed.
      With a better integrated software platform from which to undertake analysis, greater amounts of data and data of a dynamic nature can be incorporated into analysis. Addressing one of the limitations of our study, the inclusion of traffic and road data, such as construction projects, at regular intervals can better inform patient address clustering.
      Additionally, following the work of Steenberghen, Aerts, and Thomas (2010), road intersection connectivity analysis should be integrated into spatial network cluster analysis. In combination with a better understanding of road travel costs this would go a long way towards capturing the heterogeneity of networked space rather than the largely homogenous road networks currently used.
      Following from this, we see opportunities to integrate qualitative data into network analysis to account for restrictions or resistances not captured through quantitative road data. As has been documented by Kwan (2002), people differ in the way they perceive and move around space. While our study was largely quantitative and aimed to improve home support care through scheduling efficiency we contend that integration of qualitative data will further enhance employee and patient satisfaction and lead to long term benefits for patient, employee, and administration.
      Lastly, culminating from all these possible research opportunities, we recognise that GIScientists must continue to learn and adapt their methods. As the field of spatial network cluster analysis moves beyond its theoretical and computational beginnings GIScientists must become more flexible in the application of these techniques in order to be able to address the needs of place-specific problems such as the clustering of VIHA’s patient addresses.
      There is tremendous opportunity for the expansion and development of spatial network cluster analysis theoretically, computationally, and practically. As we move forward into a future where our healthcare systems will face increasing pressure from a lack of funding and aging population, the integration of GIS and health offers an innovative solution to address future challenges. As we have shown in this paper, an integrated software approach utilizing spatial network cluster analysis is a promising avenue of research in the effort to improve our healthcare systems.

TOP