Current limitations in identifying and responding to suicide risk have necessitated a shift towards novel approaches utilising Artificial Intelligence (AI) platforms, which model patterns within existing data to generate algorithms that have the potential to assist clinicians in identifying and responding to suicide. This symposium presents on a two-year collaborative effort between On the Line Australia (OTLA) and researchers at Swinburne University of Technology (SUT) on the potential of the latest in AI research to enhance response to suicide risk by using. Integrating insights of people with lived experience, the research program aimed to co-design a Clinical Decision Support System (CDSS) tailored to the needs of a suicide helpline service.
The first presentation presents an overview of the co-production process, starting from planning the system infrastructure, exploring avenues for the integration of a voice analysis algorithm and the building of a chatbot. Integrating service requirements (e.g., data collection, suicide risks assessment, location-based referral recommendations) with end user feedback throughout the process, ensures the relevance, acceptability and the continuous improvement of a potential CDSS.
The second talk focuses on the hypothetical integration of a suicide severity prediction system into existing helpline call infrastructure. This system leverages information available in the caller’s voice, by analysing the unique voice signal signatures of each caller. It provides, to counselling staff, key metrics on degree of suicide severity and psychological distress, in close to real-time. With this significant step forward, we can begin to understand the nuanced nature of how suicide presents, in the full diversity of community members.
The third presentation describes how a chatbot can empathetically engage with callers to collect essential data (e.g. name, age, gender). This information can automatically be transferred to counsellors by the CDSS, reducing the amount of time needed to obtain this critical information.
The final presentation reflects on the lived experience perspectives on integrating such a hypothetical CDSS within a suicide helpline service. It discusses the potential impact on callers and their concerns as well as organisational, systems and professional complexities of introducing technology in what is a sensitive human interaction.
The collaboration between On the Line Australia, lived experience experts and SUT, highlights the importance of co-design approaches in providing a nuanced, responsive and respectful service for, and informed by, individuals at risk of suicide. However, balancing the potential benefits with potential risks to privacy and trust among individuals seeking support is crucial in ensuring successful and acceptable use of these novel approaches.