Prediction of Suicide Attempts Among Adolescents in Acute Mental Health Crisis

OBJECTIVE: Suicide attempts among adolescents in acute mental health crisis are frequent but difficult to predict. Our study aimed to examine whether suicide attempts and patient states at baseline predicted subsequent suicide attempts, and whether suicide lethality increased in repeated attempts. METHODS: We assessed 119 adolescent inpatients at intake using self-report measures (Interpersonal Needs Questionnaire and Fearlessness About Death) and a behavioral measure (Implicit Association Test Death/Suicide). For adolescents with a recent suicide attempt, we used clinician-rated measures of suicide lethality and intent (Risk-Rescue Rating Scale and Suicide Intent Scale). Using logistic regression models, we examined whether these variables prospectively predicted suicide attempts during a six-month follow-up, and we compared the lethality of first-time and repeated attempts. RESULTS: Eighteen participants attempted suicide during follow-up. The combination of elevated unmet interpersonal needs and fearlessness about death predicted subsequent suicide attempts (OR = 2.81, p= .004) substantially better than either variable alone. Implicit association with death, suicidal intent, and lethality did not predict suicide attempts, and repeated attempts did not increase in lethality. CONCLUSIONS: The self-reported measures performed better than the behavioral and clinician-rated measures in prospective prediction of suicide attempts. The lack of predictive ability of suicide lethality, particularly the absence of escalation in repeated attempts, underscores the importance of adolescents' reports of distress over medical severity in understanding suicidal behavior.