A medical GIS approach to defining ‘rural’ for medical education research and policy

Abstract

Background While medical education researchers have acknowledged the difficulty of developing a definition of rural, authors have also suggested that definitions of rural should be suited to their research purpose (du Plessis et al. 2001, Couper 2003). Objectives One component of the Learners and Locations project undertaken at the Health Research Unit at MUN focused on developing a definition of rural for medical education research. In this paper, we describe the L & L community classification system and compare it to the Statistical Area Classification (SAC) categories developed by Statistics Canada. Methods Statistics Canada and Canadian National Road Network data, along with the GIS software ArcGIS, was used to divide Newfoundland and Labrador communities into the 11 categories developed in the L & L project: Metropolis (Pop. > 1,000,000); Very Large City (Pop. 500,000–1,000,000); Large City (Pop. 100,000–500,000); Medium City (Pop. 50,000–100,000); Small City (Pop. 10,000–50,000, <200km from larger city); Small Rural City (Pop. 10,000–50,000, 200-500km from larger city); Small Remote City (Pop. 10,000–50,000, >500km); Rural Close Community (Pop. <10,000, <60km from Small/Small Rural/Small Remote City); Rural Urban Community (Pop. <10,000, 60-200km from Metropolis/Very Large/Large/Medium City); Rural Community (Pop. <10,000, 60–200km from city); Remote Community (Pop. < 10,000, >200km from city). NL communities’ SAC classifications were then examined alongside corresponding L & L project classifications. Results Within all NL communities classified under SAC type 1, the L & L classification system identified 3 different types of communities (Large city, Rural Urban, Small City); within SAC type 3, 4 different types of communities were identified (Rural Close, Rural Community, Small Remote City, Small Rural City). Within SAC type 5, 3 types of communities were identified (Remote Community, Rural Close Community, Rural Community); within SAC type 6, 2 types of communities were identified (Remote Community, Rural Community); within SAC type 7, 2 types of communities were identified (Remote Community, Rural Community). Conclusions The L & L classification system disaggregates NL communities in a meaningful way. The potential usefulness of this system in other areas across Canada warrants further investigation.