Mapping and visualizing the location HIV service providers: An exploratory spatial analysis of Toronto neighborhoods

Abstract

Efforts have been made to identify, reduce and ultimately eliminate health disparities, yet variation in access to health services continues to be an important concern. As with large American cities, Toronto has been particularly hard hit by the AIDS epidemic, representing 68% of Ontario's HIV diagnoses (Health Canada, 2000). The accessibility of healthcare in terms of the geographic location and spatial distribution of health services are important factors in healthcare utilization. In this descriptive paper we map the location of HIV-related services and use exploratory spatial data analysis to visualize and examine the distribution of HIV service providers. In examining the location of HIV service providers we map the minimum distance to the nearest service provider. Our analyses also map and analyze five separate categories of HIV-related services. These include: (1) Diagnostic and preventive services; (2) Health and social services for initial HIV diagnosis; (3) Emotional and social support; (4) Emergency services; and (5) Medical and end-of-life services. While our findings point to significant clustering of some types of HIV-related services (such as emergency and preventive services), other services are more evenly distributed across Toronto (this includes medical and end-of-life services). Our findings point to the need for policy makers and researchers to integrate mapping, GIS and spatial analytic techniques into their analyses of the neighborhoods and subsequently the populations in those neighborhoods that are underserved in terms of accessibility of some categories of HIV-related services.

Acknowledgments

This research was supported in part by the National Library of Medicine's Biochemical Health Informatics Research Training Grant 2-TIS-LM07089-12.

Notes

Prior to conducting the spatial analysis, the structure of the spatial weights matrix must be assigned. This defines a neighboring structure and determines which census tracts will be considered neighbors of other census tracts. We use the queen criterion which defines neighbors as those that share a common boundary and common corner. All statistical computations were done using the GeoDa software package (Anselin, 2004) and all mapping was done using ArcView 3.2 interface with GeoDa and ArcMap.

Throughout the paper we will use the terms HIV service provider(s), HIV-related services (s), AIDS service organization(s), and HIV service(s) interchangeably.