Program evaluation and decision analytic modelling of universal suicide prevention taining (safeTALK) in secondary schools

BACKGROUND: Universal suicide education and awareness training in schools are promising suicide prevention initiatives. This study aims to evaluate a suicide awareness training (safeTALK) and to model potential return on investment (ROI) on a population basis. SafeTALK, comprises a 3-h education session, and has been delivered to secondary school students (aged 15-16 years) in Mackay, located in the Australian state of Queensland. METHODS: Evaluation consisted of two phases, ex-post and ex-ante. Phase I was a pre-post, follow-up analysis using a mixed-method questionnaire administered immediately prior (Time 1), immediately after (Time 2), and 4 weeks after training (Time 3). Phase II involved decision analytic modelling comparing safeTALK to the status quo. ROI was modelled using Markov chains for a hypothetical population of students aged 15-19 years in Mackay (n = 2561; suicide rate 78.1 per 100,000), Queensland (n = 296,287; 10.2) and Australia (n = 1,421,595; 8.3). Model parameters, including rates of hospitalised self-harm and suicide, cost implications and effectiveness of safeTALK were drawn from published literature. The baseline model adapted a health and justice system's perspective, with an alternative model incorporating a societal perspective. All costs were adjusted to reflect AU$2017-2018. RESULTS: Students reported seeking help mostly from friends (79%) or parents (68%); in the last 6 months 61% considered another student's behaviour as suicidal, but only 21% reported asking about this. The main barriers to help-seeking were (i) being too embarrassed, (ii) shy or (iii) being judged. Students who attended safeTALK gained suicide-related knowledge (p < 0.001), confidence (p < 0.001), willingness (p = 0.006), and likelihood of seeking help (p = 0.044) and retained these up until follow-up assessment 4 weeks later with the exception of seeking help. From a health and justice system's perspective, the model estimated a cumulative return of AU$1.45 per AU$1 invested in safeTALK in Mackay; AU$0.19 in Queensland; AU$0.15 across Australia. From a societal perspective, ROI increased to AU$31.21, AU$4.05 and AU$3.28, respectively. CONCLUSION: Results strengthen the premise that safeTALK is feasible to implement within a school setting. The economic case for implementation of safeTALK is promising on a population basis, especially in high-risk communities, but further research is required to confirm the study results