3 - (SYM 38) Implementation of a Suicide Risk Assessment Protocol During an Outpatient Real-time Monitoring Study: Findings and Practical Considerations

Background: Although real-time monitoring methods have the potential to improve understanding and prevention of suicide, one inevitable outcome of frequently assessing suicidal thoughts and behaviors is that researchers will be faced with decisions about how to intervene with at-risk individuals. In this presentation, we will report findings and practical considerations for implementing a suicide risk assessment protocol during an intensive longitudinal study of suicidal adults and adolescents.

Methods: Adolescents (n = 320) and adults (n = 320) who recently experienced a suicide attempt or high levels of suicidal intent were recruited from an inpatient psychiatric unit or emergency room. For six months post-discharge, participants wore a biosensor and completed frequent smartphone EMA surveys (6x/day for the first three months post-discharge, then 1x/day for the next three months). Following a risk monitoring protocol developed by our team (Fortgang et al., in prep), when participants indicated elevated suicidal intent on EMA surveys, we received an automated risk alert message.

Results: Data cleaning and coding is ongoing. From April 2021 to November 2022, 271 risk alerts (n=184 alerts from adult participants, n=87 from adolescents) were received from 71 unique participants (n=38 adults, n=33 adolescents), averaging between 3-4 alerts a week. Contact was made either via call, text, or both for about 61% of all risk alerts. Steps taken on these risk contacts included (note that multiple strategies were often used and percentages are not cumulative): asked questions to further assess suicide risk (50% of contacts), reminded participants to use and/or reviewed their safety plan (67%), and asked participants’ parents to check-in with their child (20%). In two cases, the risk team called emergency services to help participants who were at imminent suicide risk. In eight cases, the participant had initiated emergency services or admitted to the hospital before we spoke with them about their risk alert.

Conclusions: Results demonstrate the feasibility of implementing a risk monitoring protocol for high-risk outpatient participants. Risk alerts were received from a minority of participants and occurred relatively infrequently given the size of the study. Contact with participants allowed us to connect them with resources, including potentially life-saving treatment in two cases. Practical considerations and recommendations for implementing similar protocols will be discussed.