An analytical platform for the integration of vichealth indicators survey and spatial objectives contextual data
The overall vision of this project is to develop a set of aggregation and analytical tools capable of integrating VicHealth Indicators Survey data with objective spatial contextual data.
The AURIN Portal will be augmented with additional capabilities to provide researchers with an environment through which research questions on health and well-being of communities in Victoria can be interrogated, visualised and presented graphically at various spatial scales.
The key aim of this AURIN subproject is to develop an eResearch platform capable of integrating VicHealth Indicators Survey data with objective spatial contextual data. Unit level survey results are often statically aggregated by agencies and made available to researchers to provide a broader perspective. This project has developed a securitised solution that provides urban researchers with a dynamic, user-driven platform to access unit level survey data through which research questions on health and wellbeing of communities in Victoria can be interrogated, visualised and presented graphically at various spatial scales.
It enables urban researchers, government agencies and other stakeholders to generate a range of customised spatial indicators of health and wellbeing on–the-fly from across multiple datasets, thereby encouraging the exchange of knowledge about the effective use of health indicators, measure and monitor change in a community and develop appropriate action. Importantly the work demonstrates how advanced and practical solutions can be supported in utilisation of unit-level data from health organisations.
- An open source tool to enable interrogation and integration of VicHealth Indicators Survey with spatial contextual data
- Development of aggregation tools to create user-defined aggregate health indicators with mapping capability
- Embedded statistical and graphical interface to analyse and display VicHealth Survey data
- An export utility tool to export aggregated data outputs and graphics.
Case Study Scenarios
The functional requirements were evaluated through a workshop that discussed and refined the relevance and usefulness of the proposed tools to end users. Based on the recommendations, a scenario-based approach was adopted to develop a set of case studies. The scenarios focused on survey questions related to three key domains:
- feeling safe at night
- commuting patterns
- alcohol purchases
The key questions that underpin the case studies include: how does availability of street lighting related to feeling safe at night; how does alcohol purchase behavior of VicHealth survey respondents relate to the location of their nearest licensed premises; and how do commuting times relate to distance to public transport? It is noted that these other data sets (streetlights, alcohol outlets, public transport) were obtained from other non-VicHealth sources through the AURIN Portal. These scenarios resulted in generic solutions related to average distance to particular features of interest.
System Design and Tools
This project has developed a range of tools to aggregate unit record VicHealth Indicators Survey data at a small area level. In this project, the unit level data of VicHealth is not directly accessible. Instead, pull-based queries from secure, targeted AURIN service endpoints are asynchronously retrieved through services hosted at VicHealth. As shown in the diagram below, this system enables end-users to interrogate and integrate VicHealth Indicators Survey with spatial contextual data with an embedded statistical and graphical interface to analyse and display VicHealth Survey data.
An export utility tool is also developed to export aggregated data outputs and graphics. The AURIN system has been designed explicitly to avoid the possibility of compromises caused by exposing VicHealth data directly to the Internet. It has been designed in such a way to overcome the barriers and risk of exposing the data/systems to external software systems and the risks of incoming connections/queries from the Internet; and the dangers of disclosure risk control and information governance, and the potential to identify individuals from the survey.
Key to this work is the algorithms that calculate the average distance for all individuals that live in a particular localised region, e.g. a Statistical Area level (SA2/SA1) to a given feature of interest. A naïve approach to recording the distance is the direct (as the crow flies) distance between two points, however this ignores the local road network topology which can have a major impact on distance (e.g. crossing major transport routes or rivers). To tackle this, distance measure algorithms utilising the street network have been implemented and are supported.
Two key measures were coded, which include aggregation and Herfindahl index. The Herfindahl index measures the concentration of a self-reported evaluation on a particular survey question by survey respondents. The tables below demonstrate a simple calculation of HI for three areas. Higher Herfindahl index values show the predominance of a particular viewpoint on a survey question in an area.
This AURIN system aids and augments VicHealth Indicators Survey data and is invaluable for local governments in developing Municipal Public Health and Wellbeing Plans as required by Victorian legislation. Currently, the capacity and capability of such facility that enables the integration of survey data with different spatial datasets are limited in Australia. The development of an eResearch portal enables community planners and health practitioners to interrogate and integrate VicHealth Indicators Survey with spatial data, characterising the environmental context. It offers a significant step forward to more informed, engaged and integrated decision-making to improve health outcomes of the community.
Building geospatial intelligence to assist healthcare professionals to establish linkages between health outcomes of people, where they live; their residential environmental context, their socio-economic status, accessibility to amenities and healthcare facilities and other built environment characteristics, is the key focus of this project.
Examples of the outcomes generated from the AURIN eResearch portal
|Professor Prem Chetri
|Professor Richard Sinnott
University of Melbourne
Sinnott, R, Chhetri, P, Gong, Y, & Macualay 2015, ‘Privacy-oriented geospatial data linkage: a case study from the AURIN project’, eResearch NZ 2015, Queenstown, New Zealand
Chhetri, P, Stimson, R & Western, J 2011, ‘Using GIS to derive region-wide patterns of quality of life dimensions: Illustrated with data from the Brisbane-SEQ region’, Investigating Quality of Urban Life: Theory, Methods and Empirical Research, Springer.
McCafferty, S, Doherty, T, Sinnott, RO & Watt, J 2010, ‘Supporting research into depression, self-harm and suicide across Scotland’, Journal of the Philosophical Transactions of the Royal Society A.
Download the FACT SHEET: VicHealth Indicators