The Development of a Health and Geodemographic GIS
BACKGROUND
In the past few years there has been a significant change in the way that many state governments in Australia have addressed the issue of provision of public sector services. There has been an increasing trend towards outsourcing, privatisation and the creation of business units. The major aims of these changes in method of service provision is to increase efficiency and so reduce the cost of providing the services.
An important component of the provision of public sector services is to ensure they are available and accessible to those sections of the community most in need. Associated with privatisation and the drive for greater public sector efficiency is the risk that those in most need of services are disadvantaged. Hence governments must strive to ensure that all sections of the community have relatively good access to public services.
The Victorian government has recognised the potential for new technology and improved information systems to be used as tools for the assessment of levels of service and to identify areas that may be disadvantaged. In 1994, the Victorian Government commissioned a report from Professor Hugo of the University of Adelaide on the issue of Locational Disadvantage, (Hugo, 1995). This report lead to the development of a prototype GIS system for assessing locational disadvantage for the Ovens-Murray Statistical Division in north-eastern Victoria (Nicolson et al, 1995).
Following on from those initial studies, the Centre for Social Applications of GIS was approached to develop a GIS for the Hume Region of the Human Services Department. This system was intended to include the components of the original Locational Disadvantage GIS and in addition, is to include specific health information for the Hume Region.
DEVELOPMENT OF THE GIS
This report describes the development of a GIS for the Hume Region based on the Locational Disadvantage GIS, and the hospital data in the Victorian Inpatient Minimum Database (VIMD).
The GIS was developed as three separate projects using ARC/INFO for the manipulation and creation of spatial data sets and ARCVIEW2/3 for the data display and end-user functionality.
GEODEMOGRAPHIC GIS
This component of the project was to create a series of data layers about the general characteristics of the population that reside in the study region. The data layers were created from data collected at the 1991 ABS census and aggregated to Census Collector Districts (CCD's).
The information provided ranged from the distribution of persons aged 65 or more to the proportion of the population for each CD earning less than the national median weekly income. The data collection and display processes for this section of the project are relatively straight forward and a number of commercially available packages exist to produce the same results.
Accessibility Analysis
Access to services, in this case hospitals, is taken for granted by most of the population in urban areas. Whereby in many rural areas, access to hospitals and other medical services involves travelling quite large distances. In an attempt to measure relative access to hospitals an index was developed using grid modelling techniques in ARC/INFO.
The index was created by treating the entire study region as a continuous surface of grid cells and by calculating the distance from each grid cell in the surface to the selected hospital. The process was repeated for each hospital in the region and the resultant grids were summed to produce an overall value for each cell. In total 21 hospitals were processed and the final index was created by dividing the values into 10 equal areas. The map below indicates the relative access zones across the Hume region.
Health GIS
The third component of the project involved the use of the Victorian Inpatient Minimum Database for the Hume Region. For the project 66,000 data records from the 1994/95 financial year were made available. To ensure confidentiality names and addresses were removed, each record had a postcode and Statistical Local Area code to enable data aggregation.
Although not the ideal units, the Australia Post postcode areas were used to display the data. Data at the census collection district would have provided a greater level of spatial detail but is not available. From the database it was possible to query the records and produce chloropleth maps of selected variables. These data queries and displays were performed using ARCVIEW2.
Another method for displaying the data is to use weighted lines to represent flows of patients from or to a particular hospital. These plots are created by drawing a weighted line (weighting on the basis of inpatients).
The final project delivered to the Dept of Human Services (Hume Planning Region) included a customised ARCVIEW project to allow users to create their own plots. The customised project was developed using the AVENUE scripting language.
