Protocol for a stepped-wedge, cluster randomized controlled trial of the LifeSpan suicide prevention trial in four communities in New South Wales, Australia

Background: Despite increasing investment in suicide prevention, Australian suicide rates have increased steadily in the past decade. In response to growing evidence for multicomponent intervention models for reducing suicide, the LifeSpan model has been developed as the first multicomponent, evidence-based, system-wide approach to suicide prevention in Australia. The LifeSpan model consists of nine evidence-based strategies. These include indicated, selective and universal interventions which are delivered simultaneously to community and healthcare systems over a 2-year implementation period. This study will evaluate the effectiveness of the LifeSpan model in reducing suicide attempts and suicide deaths in four geographically defined regions in New South Wales, Australia.

Methods: We outline the protocol for a stepped-wedge, cluster randomized controlled trial. Following a 6-month transition phase, the trial sites will move to the 2-year active implementation phase in 4-monthly intervals with evaluation extending a minimum of 24 months after establishment of the full active period. Analysis will be undertaken of the change attributable to the invention across the four sites. The primary outcome for the study is the rate of attempted suicide in the regions involved. Rate of suicide deaths within each site is a secondary outcome.

Discussion: If proven effective, the LifeSpan model for suicide prevention could be more widely delivered in Australian communities, providing a valuable new approach to tackle rising suicide rates. LifeSpan has the potential to significantly contribute to the mental health of Australians by improving help-seeking for suicide, facilitating early detection, and improving aftercare to reduce re-attempts. The findings from this research should also contribute to the evidence base for multilevel suicide prevention programs internationally.