1. Informing local suicide prevention through co-design of a system dynamics model in Greater Western Sydney (Australia)
With a small research partnership grant from Western Sydney Uni and BMC, USyd, significant in kind contributions from Andrew Page, Mark Heffernan and Geoff McDonnell, and in partnership with Bill, Ian and the team at Went West, we embarked on an innovative project to develop a decision support tool that could inform your strategies for reducing suicide and suicidal behaviour in Western Sydney over the short and longer term.
Mark Heffernan Dr. Ante Prodan
Professor Andrew Page Dr. Geoff McDonnell
Dr. Jo-An Atkinson

2. The key question being addressed:
What suite of programs and services will deliver improved accessibility to mental health care and reductions in suicidal behaviour in Western Sydney now and into the future?
Over the last 6 months we have been co-developing a dynamic systems model to answer the key question of:
what combination of mental health service improvements and specific suicide prevention interventions will deliver the greatest impact on suicidal behaviour in Western Sydney.

3. Workshop
We kicked off the process by holding a participatory workshop with multidisciplinary stakeholders of the Regional Data Planning Group which include police, ambulance, service providers, program planners, and those with lived experience….guided them through a mapping exercise, and we drew on their valuable local knowledge to better understand the challenges Western Sydney faces in terms of suicidal behaviour, mental health services, access issues and suicide prevention efforts.

4. Model mapping to conceptualisation to computation
We then synthesized that map and the recorded discussions into a conceptual model of the problem.
We compiled relevant evidence and publicly available data and used it to inform the process of converting the conceptual model into a quantitative and testable computational model. Through follow-up meetings with the Regional Data planning group, with WS LHD, and with WW, we have received valuable feedback and additional data essential to parameterising and calibrating the model.

5. Primary mechanism Overall the model captures:
The pathway from vulnerability to distress to disorder;
Captures assessment of care needs and delivery of services
And recovery pathways
And importantly we have captured the dynamics of suicidal behaviour
This structure allows us to test the impact on suicidal behaviour (and other outcome indicators) of a range of interventions focussed on the quality, capacity, delivery and accessibility of mental health services.
It also allows us to integrate and explore the effectiveness of technology, infrastructure, peer and community support, trauma informed care and crisis management.

6. Interactive user interface – primary interface
Go to model - Explain interface:
Primary output graphs – multiple pages for alternative outcome indicators
Switches grouped – C for control panel (GP / GK training) takes you to the default parameter estimates / assumptions underlying the intervention – which can be changed to test the implications of alternative scenarios
Intervention mechanisms and info
Unfurl the model structure

7. Transparent intervention default parameter values that can be modified

8. Best overall combination: Post suicide attempt assertive aftercare PLUS community support PLUS new intervention to re-engage those lost to services PLUS increases to mental health service capacity, workforce and training (-2-)

9. Summary of key insights
Best combination of Suicide Prevention interventions:
Post-attempt assertive aftercare and community support interventions were forecast to have the greatest impact, averting 15% of attempted suicides and 23% of suicide cases over the 10-year period.
Best combination of Mental Health interventions:
Online/technology enabled care combined with interventions to re-engage those lost to services were forecast to have the greatest impact, averting approximately 17% of attempted suicides and 16% of suicide cases over the 10-year period.
Best combination of MH services interventions:
Increases in hospital staff (5% pa) and additional training with increases in secondary and non-secondary service capacity was forecast to have the greatest impact, averting approximately 15% of attempted suicides and 15% of suicide cases over the 10-year period.
Best overall combination of interventions to reduce suicidal behaviour in Western Sydney (using default settings):
The combined effect of post-suicide attempt assertive aftercare; community support interventions; interventions to re-engage those lost to services; additional hospital staffing and training; and increased mental health service capacity were forecast to have the greatest impact, averting 30% of attempted suicides and 35% of suicide cases over the 10-year period.
Intervention combinations have non-additive effects.

10. Has it made any difference
Impact in terms of deaths - too early to tell, “accurate” stats lag by 3 years. Suicides are unlikely to decrease in absolute numbers, but we may reduce the rate ( 10 per 100,000 in this region).
Impact in terms of changing the way we think about the problem – recognition of interdependent parts of the system, the “black hole” and swamping downstream services.
Impact in terms of change to organisational focus – PHN is looking to Co-Commission things that hadn’t been considered previously, and to set aside things that look less likely to be a good investment
Impact in terms of how prospective monitoring and evaluation will be done, in that the model will be update with routine data as it comes in and assumptions can be modified in line with evidence