GIS projects and activities in rural health

Health issues contain many complex underlying relationships between population, environment, economic and social factors. Geographic Information System (GIS) software is an ideal tool to analyse such relationships, and has increasingly been capturing the attention of public health researchers and policy makers.

CUCRH has advocated for geographic information systems and spatial analysis to be applied to pressing rural and remote health issues since 2000.  In 2001 and 2002 exploratory work was conducted on patterns of diabetes hospitalizations in the Northern Goldfields region.  We have been assisted through a collaboration with A/Prof Bert Veenendaal of the Department of Spatial Science at Curtin University of Technology.

Several other projects were established since that initial foray into GIS, including several student projects.  Some of these are described below.  In late 2004 CUCRH sponsored a forum to bring researchers, policy makers and other users of health information together to discuss applications of this technology.

Through Ann Larson, CUCRH is a member of ARCRNSISS, an ARC Research Network on Spatially Integrated Social Science.  This network brings together 80 of the countries experts on spatial analysis.  A strong hub interested in health research and another on regional development contributes to CUCRH access to methodological tools and expertise.

Estimating access to health care for small areas

Most measures of access use large geographical areas – they count up the number of doctors or hospital beds in a statistical division and divide by the number of people.  The problem is that within those large areas are the communities where people live.  Some of them have good access and others live far away from services. 

CUCRH, in partnership with Western Australian Centre for Remote and Rural Medicine and Curtin’s Department of Spatial Science, developed a method to measure access that is more realistic.  It uses the smallest geographic areas available and assumes that people can go to any service that is within 100km, not just the service in their community.  We trailed the method geographic information system software and a rural GP workforce survey compiled by WACRRM.  The results give a much more nuanced view of where there may be shortages.  The results are published in an article by Jessica Scott, Ann Larson, Felicity Jefferies and Bert Veenendaal in volume 14 of Australian Journal of Rural Health.

Solutions to the modifiable area unit problem in rural and remote regions

Louise Fogarty of the Department of Spatial Sciences, Curtin University of Technology tackled this problem in her prize winning honours thesis.  Rural and remote areas present unique challenges because of the great size and diversity of standard areal units such as postcodes, statistical local areas and census collection districts.  When the topic of interest can be studied using point data, such as with the data on hospitalisations for diabetes, would an alternative method of areal aggregation create more homogenous units?  Such a solution would reduce the modifiable area unit problem (MAUP) in which conclusions drawn from the data are dependent on the areal units used.  Ms Fogarty did a comprehensive literature review on the MAUP and its possible solutions and explored the utility of one solution for the Pilbara. The results suggested that further exploration in defining more meaningful areal units would be useful to explore the diversity of health status and services in the region. 

First Western Australian forum on GIS in Health Research convened by Combined Universities Centre for Rural Health

The forum, convened by Combined Universities Centre for Rural Health and held at the Telethon Institute for Child Health Research in Perth on 8 November 2004, attracted some 80 participants from many different sectors of the health and spatial sciences.  Conference partners included ICHR, Spatial Sciences Department of Curtin University of Technology, Western Australian Department of Health, School of Population Health of The University of Western Australia and the Northern Rivers University Department of Rural Health.

Professor Graeme Hugo (Director of the National Centre for Social Applications of GIS at University of Adelaide) gave the keynote address summarising principal applications.  He observed that aspects of GIS which are particularly useful for understanding health issues include the mapping of data, integration of data, capacity to overlay population, environmental, health and service variables, development of complex new variables, modelling capacity, ability to include locational concepts, and the ability to develop meaningful regions of analysis.  Professor Hugo discussed how, in the area of health research, these methods lead to three main areas of application: 1) the understanding of health problems, or spatial epidemiology, 2) modelling and planning of health service provision and delivery, and 3) organising centralised data from different sources, using spatial referencing as the key to integration.

The breadth of projects discussed at the forum reflects a strong interest in this growing area of research.  Main topic areas covered by the presentations were health services and service utilisation, spatial patterns of disease, and spatial data usage and techniques.
Following is a brief description of the presentations given under each topic heading:

Health services and service utilisation

Steve Spiker and Mark Peel, Epidemiology Branch, HIC Department of Health. “Building a healthy GIS: Promotion and Centralisation, Web technologies and advanced spatial analysis”.

Jessica Scott, Combined Universities Centre for Rural Health “Small area floating catchment method to identify areas of medical workforce shortage”.

Prof. Geoff Solarsh, Monash University School of Rural Health “The effect of distance on Primary Health Care usage patterns in rural South Africa”.

Spatial Distribution of Disease

Dr Geoff Morgan, Northern Rivers University Department of Rural Health. “Spatial environmental epidemiology and socio demographics data in NSW –  case studies of Ischaemic Heart Disease and Childhood Leukaemia.”

Dr Jane Freemantle, Telethon Institute of Child Health Research. “Where do Western Australian infants and children die?”

Spatial Data usage and techniques

A. Prof. Bert Veenendaal, Department of Spatial Sciences, Curtin University of Technology. “Using GIS to explore remoteness and accessibility to the health services in Western Australia”.

Robert Tanton, Analysis Branch, Australian Bureau of Statistics. “Introduction to using Socio-Economic Indexes for Areas (SEIFA) “.

Diana Rosman, Data Linkage Unit, HIC Department of Health. “DLU’s geocoding processes and results”.

In addition to formal presentations, a number of lunchtime discussion sessions were held, which gave participants the opportunity to network with other researchers in similar fields. Topics included “GIS and GPs”, “Health and the built environment”, “Constructing and using geographic indexes”, “Spatial analysis issues” and “Knowing enough to be dangerous – spatial analysis for epidemiologists”.