Computerized screen successfully identifies youth suicide risk
Click Here to Manage Email Alerts
A computerized suicidal risk screen appeared able to successfully identify suicide attempt risk among youth, according to study results published in JAMA Psychiatry.
“Too many of our teens are dying by suicide, many of whom have never received any mental health services,” Cheryl A. King, PhD, of the department of psychiatry at the University of Michigan, told Healio Psychiatry. “Their suicide risk was never recognized. Our goal was to develop a screen that accurately predicts a teen's risk for suicide attempt for use with all teens who visit medical emergency rooms.”
The investigators aimed to develop and independently validate CASSY, a novel computerized adaptive screen for suicidal youth, as a universal suicide risk screen in medical EDs. In the first study of this prognostic study, they prospectively enrolled 2,075 adolescent participants with a mean age of 15.1 years at 13 geographically diverse U.S. EDs in the Pediatric Emergency Care Applied Research Network. Participants provided suicide risk survey data and participated in 3-month follow-ups via telephone. Further, they derived algorithms for the CASSY via three fixed Ask Suicide-Screening Questions (ASQ) items as anchors, as well as additional items that varied in number and content across individuals. In the second study, they collected data of 2,754 participants with a mean age of 15 years at 14 Pediatric Emergency Care Applied Research Network EDs and one Indian Health Service hospital. They independently validated algorithms in a prospective cohort of adolescent participants who also participated in 3-month telephone follow-ups. Presentation at an ED served as the exposure, and a suicide attempt between ED visit and 3-month follow-up, measured according to participant and/or parent report, served as the main outcome and measure.
Results showed excellent predictive accuracy for suicide attempt of the CASSY algorithms in study one, with a mean number of adaptively administered items of 11. The CASSY demonstrated a sensitivity of 83% at a specificity of 80%. In the study two validation sample, it also demonstrated excellent accuracy. In the current study, it had a sensitivity of 82.4% for suicide attempt prediction at the 80% specificity cutoff, which was established in the first study.
“The CASSY demonstrated a high [area under the curve] for the prediction of a [suicide attempt], with an excellent balance of sensitivity and specificity, and is suitable for administration in busy EDs,” King and colleagues wrote. “Important next steps will be to measure the CASSY’s test-retest reliability and develop triage recommendations and conduct implementation studies.”