Suicide is the 2nd leading cause of death for teenage youth in Canada, and every year tens of thousands of young people engage in suicidal behaviors. Many young people manifest red-flag behaviors in the digital realm before these incidents occur. The faceless nature of online communication often emboldens youth to reveal details about their mental state: leaving tell-tale signs or "bread crumbs" of their suicidal intentions or behaviors.

This research will examine the content of adolescents social media entries to find red flags and detect patterns in social media interactions of adolescents that could be predictive of subsequent suicide risk. The study participants will be patients admitted to Child and Adolescent Psychiatry Emergency unit at BCCH, subdivided into two groups: those admitted due to suicidal behaviors, and those admitted for non-suicidal behaviors. The text of social media activity for the month prior to emergency admission of the two groups will be collected, anonymized, and analyzed using text-analytic algorithms. The objective of the study is to find patterns and indicators of social media entries, prior to admission, that would have been predictive of suicidality.

The implications of successful outcome of this project for mental health care of children and adolescents reaches well beyond the scope of this study. An objective method to predict risk of suicidal behaviors in youth has application in almost all pediatric clinical settings. The outcomes of this project will also serve as the foundation for further utilization of social sensing technology to identify, predict, and prevent many other mental health crises in children and adolescents.

Inclusion Criteria:

• use social media
• patient of BCCH

Exclusion Criteria:

• no use of social media

We aim to recruit a total of 200 subjects in the four different arms of the study. First arm: 50 adolescents admitted to CAPE because of potentially lethal suicidal behavior. This group consists of patients whose suicidal behavior could have ended their lives or significantly compromised their health in the absence of timely interception or intervention. We use this cohort as a proxy for "would be" completers. Second arm: 50 adolescents admitted to CAPE because of non-lethal suicidal behavior. This group includes patients who are admitted because of suicidal ideation without any attempt, or suicidal behaviors that would not have ended their lives or caused significant compromise in their health. Third arm: 50 adolescents admitted to CAPE due to mental health crises other than suicidal behavior. Fourth arm: 50 adolescents admitted to BCCH for reasons not related to mental health.