Predicting Suicide Acts Among Crisis Line Callers: Screening Recent Suicidal History or Using a Comprehensive Scale?

Abstract

Objective: Crisis lines face challenges in identifying individuals at high suicide risk. We aimed to compare two routine methods for predicting suicide acts at different time points.

Methods: In this prospective cohort study, we recruited and monitored 8859 callers from the Beijing Psychological Support Hotline. We evaluated their suicide risk through two strategies during their index calls: (a) one screening about suicidal ideation, plan, or behavior in the last two weeks and (b) the Comprehensive Suicidal Risk Assessment Scale, which included more risk factors, such as depression, hopelessness, and psychological distress. We monitored their suicidal behaviors for one year through telephone interviews.

Results: The number of callers who attempted suicide or died by suicide within 24 hours (1.2%), 30 days (3.8%), 180 days (6.9%), or 365 days (9.1%) was: 102, 341, 615, and 802, respectively. The sensitivities, specificities, and positive predictive values for screening for recent suicidal history were higher than those of the comprehensive scale for predicting suicidal acts within each time point. However, with the increasing duration of the follow-up, the screening had a poorer predictive ability than the comprehensive scale (predicting suicide acts within 270 days: Youden's index, 34.5% vs. 36.2%.; the Area under the Receiver Operator Characteristic Curve, 67.2% vs. 68.1%), especially among those without a history of suicide attempts.

Conclusions: Screening for recent suicidal history is valid for predicting suicidal acts within six months; however, screening is insufficient for predicting long-term suicidal acts compared to assessing more suicide risk factors.