Abstract

The impacts of the environment on human health are significant and can include air, water, and noise pollution, exposure to toxic chemicals, food contamination, and climate change. Despite a growing understanding of environmental health impacts in international research literature, significant obstacles exist in comprehending key characteristics of these impacts. These obstacles include data volumes, data access rights, the time required to compile and compare data over regions and time, and the derivation of credible insights from the complex derived information. Business and industry, including local government representatives, must overcome these obstacles to understand the impact of the natural and built environment on human health and make important decisions regarding health policy, resource allocation, and intervention strategies.

In this presentation, we introduce the AusEnHealth platform which aims to address the barriers in environmental health data by providing an open data repository as well as methodology for producing environmental health indicators and health outcome weighted population vulnerability indices at the statistical area level 2 (SA2) geographical resolution across Australia. AusEnHealth’s extensive interdisciplinary data set repository and analysis techniques allow policy makers, health managers, and researchers to identify key population characteristics, environmental exposures, and track overall vulnerability changes in their area over time. This enables more effective strategies for the prevention or mitigation of economic and health burdens associated with the natural and built environment.

The presentation will also address the critical challenge of accessing required data for products such as AusEnHealth, This is a particular issue for commercially private or personally sensitive data sources in industry and health. We discuss our efforts to develop federated analysis approaches which allow data to remain with the data custodians and the analysis to be performed in the cloud. This requires novel modifications to the models and algorithms underpinning the analyses.

The work discussed in this presentation is led by Dr Aiden Price and Conor Hassan, members of the Centre for Data Science at QUT, Australia.