Supporting a community-based domestic, family and sexual violence consortium to build evidence of service recommendations using de-identified linked administrative data.

Abstract

Objective
To support a South Australian community-based service consortium to test assumptions about individuals and families experiencing domestic, family and sexual violence (DFSV) accessing their services, to build evidence to support recommendations of a submission to a Royal Commission, using de-identified linked administrative data.

Methods
The Better Evidence Better Outcomes Linked Data platform (BEBOLD) was used to construct a base population of individuals who had accessed DFSV consortium services within a selected timeframe, and a comparison group of individuals who had accessed similar-funded services outside of the consortium in the same timeframe. De-identified system contact information for the two populations was then explored to understand differences or similarities in demographic and other characteristics, as well as system service usage pattens. De-identified linked datasets related to specialist homelessness services, child protection, hospital admissions and emergency department presentations were used.

Results
Using de-identified system contact information in the BEBOLD platform, we were able to demonstrate that the two groups had both similarities and differences in their demographics, presenting needs, service-system usage patterns and relevant associated system outcomes (such as child removals by child protection). These findings were able to evidence some of the key recommendations that the consortium had written in their Royal Commission submission. This provided quantitative data to support their suggestions of key issues, and the changes required within the DFSV system landscape. We were also able to show similarities and differences between populations accessing services in metropolitan and regional/rural areas to explore whether geographical location had an impact on service system usage, outcomes and population characteristics.

Conclusion
Using de-identified linked data, we assisted a community-based consortium supporting people experiencing DFSV to test their assumptions about key issues affecting their client population. This evidence was then used to support consortium recommendations to the Royal Commission on DFSV regarding how systems can better support people experiencing DFSV in South Australia.