Rural, regional and remote health: A guide to remoteness classifications

Abstract:

Policy makers, researchers and the general community are interested in the ways that the lives of Australians vary according to where they live. For example: There has been an increasing concern over a number of years about perceived difficulties faced by Australians living outside major metropolitan centres in accessing services (DHAC & GISCA 2001). There has also been particular concern about possible differences in health, education, income and a range of other factors, between those living in and those living outside major metropolitan centres. For example, a newly released report, Rural, Regional and Remote Health: A Study on Mortality (AIHW 2003), showed that, during the period 1997–1999, the mortality rates for people in regional and remote areas were higher than for people in capital cities. Analyses of such differences depend on the ability to classify areas according to their remoteness. Three major remoteness classifications are currently used: the RRMA (Rural, Remote and Metropolitan Areas) classification the ARIA (Accessibility/Remoteness Index of Australia) classification (based on ARIA index values), and ASGC (Australian Standard Geographical Classification) Remoteness Areas (based on ARIA+ index values—an enhanced version of the ARIA index values). This publication reviews these three classifications, their methodologies, and their strengths and weaknesses, and describes how the classifications are applied to administrative and survey data. The major points made in this publication are as follows: The methodologies underlying the ARIA classification and ASGC Remoteness Areas provide a better measure of remoteness than does the methodology underlying the RRMA classification. Two approaches have been used to apply ARIA and ARIA+ index values to SLA boundaries—allocating a class to the SLA based on the unweighted mean of index values for gridpoints lying within the SLA, and population weighting the SLA based on the mean of the index values for CDs lying within the SLA. Each approach has its strengths and weaknesses. Concordances used to assign remoteness classes to SLAs can be somewhat imprecise due to boundary and population changes that occur between censuses. The validity of these remoteness classifications in a given application (say, describing statistics or allocating funding) is greatest when the issue of interest is affected only, or mainly, by remoteness. Caution is required when other influences (for example, socioeconomic status, health outcomes, Indigenous status and local town size) are thought to play a role in the issue of interest.