• Summary
• Downloads
• Explanatory Notes
• Related Information
• Past Releases

SPECIAL ARTICLE

UNLOCKING THE POWER OF STATISTICS: MENTAL HEALTH SERVICES USES AND OUTCOMES

As Australia's national statistical agency, the ABS has more than a century of experience in data analysis, and it has built an enviable international reputation for the quality and thoroughness of its statistical methods.

Over the last few years, the ABS partnered with other government departments and agencies to develop innovative solutions to answer important policy questions, with an emphasis on unlocking the potential of existing administrative data to provide both new insights as well as improving the quality and usefulness of existing data.

The ABS is planning to expand further its use of existing government data, and it is now well-placed to use the experiences detailed in this article as a stepping stone to larger and more complex projects in the future.

The following project outlines how a partnership between the ABS, the National Mental Health Commission and the Department of Health has produced a new dataset to address a key policy need. This has been achieved through meshing the ABS's skills and capabilities - in data integration and methodology - with non-ABS datasets, providing an expanded and improved evidence base for policy and planning.

MESHING CENSUS DEMOGRAPHICS WITH ADMINISTRATIVE HEALTH DATA

During 2014-15, the ABS harnessed three existing Australian Government data sources to create a new source of information about the people who use government subsidised mental health-related services and medications, including which people were likely to make more use of these services or medications.

This project provided greater insight into the lives of these people without compromising the privacy of individuals.

PROJECT DRIVERS

In 2014 the National Mental Health Commission was asked by government to conduct a review of mental health programs and services.

The review was to look at the efficiency and effectiveness of programs and services supporting people with mental health problems to see if these programs, services and medications were helping people to lead a contributing life and engage productively in the community. Additionally, advice was sought to see if services were reaching those most in need, in particular those living in regional, rural and remote Australia.

The National Mental Health Commission asked for assistance from the ABS in developing a sound evidence base about mental health services in Australia, so that they could use that information to underpin the review and its recommendations.

While there was a great deal of information around the number of subsidised mental health-related prescriptions and services provided each year in Australia prior to this project, little was known about the demographic and socioeconomic characteristics of the people using them.

To support the review the ABS already had a range of mental health data from existing surveys, such as the 2007 National Mental Health and Wellbeing Survey and the 2011-13 Australian Health Survey.

However, none of the existing data sources had the detail around the use of mental health services and medication usage which was available from administrative data sources such as the Medicare Benefits Schedule or the Pharmaceutical Benefits Scheme.

By using the very basic demographics from both the Medicare and Pharmaceutical Benefits data as a key, the ABS was able to unlock a range of richer demographic data from the Census, adding information on education, employment, housing, income, country of birth, proficiency in spoken English, and need for assistance and family relationships to each of the Medicare and Pharmaceutical Benefits Scheme datasets.

HOW THE DATA WAS INTEGRATED

The Department of Health provided the ABS with the anonymised Medicare and Pharmaceutical Benefits datasets, while the Department of Human Services extracted and provided the associated anonymised demographic data.

These data were then combined with Census data using probabilistic linkage techniques - essentially finding best matches based on a person's sex, age and location. At the end of this process, around 70% of the Medicare and Pharmaceutical Benefits records were assigned a richer range of demographic characteristics thanks to Census data.

This new Mental Health Services – Census Integrated Dataset not only provided the National Mental Health Commission with additional information to support the review, it also expanded the suite of existing mental health data sources.

This new dataset is another outcome of the Census Data Enhancement Project, an ABS initiative that integrates data from the Census of Population and Housing with other datasets to efficiently create new enhanced data sources for statistical, policy and research purposes.

CHALLENGES

This project involved a number of challenges for the ABS including:

• becoming familiar with the requirements of the National Health Act 1953, and the Health Insurance Act 1973 in relation to the use of Medicare and Pharmaceutical Benefits data to ensure that legislative requirements were met
• working with large datasets that weren't primarily designed to produce statistics
• bringing together data from three different sources in a way that ensured the outputs from the combined dataset were both informative and high quality
• working within the very tight timeframe imposed by the review.