Research Impact: Mental Health Epidemiology Unit

The Mental Health Epidemiology Unit leads research on the design and analysis of studies of mental health outcomes, particularly studies of suicide and self-harm. We study this through the lens of epidemiology and biostatistics. Most of our research uses quantitative methods to study important problems in suicide prevention. We also undertake methodological work where we either develop new methods or consolidate information about existing methods for the suicide prevention community.

Key research projects

Real-time monitoring of suicide and self-harm

Coroners Courts and other organisations in most states and territories of Australia have set up suicide registers. These registers operate in real time, with coded data from police reports of suspected suicides entered into registers within 24 hours of a death. Several states now publish counts of suspected suicides each month. We are developing new methods to use this data for regular monitoring. Specifically, we have developed tools and techniques to identify whether suspected suicides have exceeded the expected range and whether there are suicide clusters in communities (a greater than expected number of suicides given the underlying rate in that area).

Impact: Our research is being used by the Coroners Court of Victoria to monitor suicide rates in real time. Using data from the Victorian Suicide Register, we provide the Court with monthly reports that identify any age or sex groups where the suicide rate has exceeded the expected limits and locations of any potential suicide clusters and the case numbers within the clusters. The Court uses this information to identify any emerging trends and establish any link between cases. It alerts coroners to inform their investigations, reports trends or clusters to government, and works with relevant stakeholders (e.g., the Suicide Prevention and Response Office at the Department of Health) so they can enact an early postvention response. The tools and techniques we have developed have also been adopted by NSW Health. The Suicide Monitoring and Reporting team now use real time monitoring data on suspected and confirmed suicides to assess when the number of suicides exceeds the expected limits. They have developed a real-time dashboard for internal use which disaggregates data by any combination of age, sex and location. This information is also being used by NSW Health to inform a suicide prevention response similar to the one in Victoria.

Preventing railway suicide

Our research has demonstrated that restricting access to the track (e.g., by removing level crossings and constructing trackside fencing) can result in reductions in railway suicides. We have also used coronial data to explore the circumstances surrounding railway suicides.

Impact: Findings about the effectiveness of railway suicide interventions have been used extensively to lobby industry and government for improvements to the railway, nationally and internationally. TrackSAFE presented our findings about the effectiveness of fencing on the railway to senior executives at the Victorian Department of Transport in 2022, which directly informed the construction of new fencing in 2023/24. In addition, we were involved in meetings with the CEO of the Level Crossing Removal Project in 2021 in which we briefed him about the reduction observed in railway suicide close to level crossing removal sites. Subsequently, in 2022 the Victorian government announced an expansion of the project – an extra 25 level crossing removals to occur by 2030. The expertise we have developed in the area has resulted in the Victorian Department of Transport requesting our advice about methods for assessing the impact of further infrastructure projects occurring on the railway.

New methods in suicide prevention research

There are many gaps in our understanding of how to conduct research in the suicide prevention field. For example, although suicide is a major public health problem, it is a rare event, and this has implications for the way suicide event data are analysed. New methods have been developed to address this and other issues in other areas of health, but these methods have not yet been used in suicide prevention. We have developed new methods and applied them in the suicide prevention field. In addition to prospectively monitoring for suicide clusters in real-time (see Example 1), we have applied methods for dealing with rare events to the suicide prevention field, have progressed ways of conducting interrupted time series studies for population-level suicide prevention interventions, and developed methods for extracting rates from studies of counts of suicide events in person time prior to conducting meta-analyses.

Impact: The impact of this work is predominately in the knowledge domain. The methodological papers we have published are well cited and the methods have been applied in other contexts. For example, our methods for dealing with population-level interventions have been used to evaluate changes in suicide rates in South Korea following broadcast of the television show “Squid Game”.