(PS8-18) A Meta-analysis of Prospective Associations of Suicide Risk Screening Tools with Suicide-related Outcomes

Abstract Text: Standardized suicide risk screening tools are increasingly required and used in healthcare settings. Many studies have sought to validate their use for detecting individuals at elevated risk of suicidal behavior, showing varied effects across different populations and care settings. Meta-analyses summarizing overall associations of leading suicide risk screeners with future suicidal behavior are scarce. This pre-registered meta-analysis aimed to summarize the prospective associations of widely recommended suicide risk tools with suicide-related outcomes, offering a comprehensive overview of their performance across diverse clinical contexts.
A systematic search of PsycINFO and MEDLINE from inception to September 16, 2024, identified studies on suicide risk screening tools and prospectively assessed suicide attempts, suicide deaths, or emergency visits/inpatient hospitalizations for suicide risk. Only English-language, peer-reviewed studies with quantitative suicide-outcome data in relation to at least one of the following tools were included: C-SSRS, ASQ, PHQ-9, SBQ-R, or PSS-3. These tools were selected based on their inclusion in a 2023 Joint Commission report, recognizing them as rigorously validated and widely adopted suicide risk screening tools. Two independent, doctoral-level reviewers extracted data, resolving discrepancies by consensus. Study quality was assessed using a modified version of the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2).
Nineteen studies were included, which reported distinct effects between ASQ, C-SSRS, and PHQ-9 item 9 on suicide-related outcomes. The PSS was not used in any studies, and only one study used the SBQ-R, leading to their exclusion due to insufficient data on this measure.
Among the three tools evaluated, the ASQ > 0 screener showed the strongest prospective association with all suicide-related outcomes (OR = 18.81, 95% CI [14.11, 25.09], p<.001), significantly outperforming the C-SSRS > 0 (OR = 6.74, 95% CI [2.59, 17.55], p < .001), C-SSRS > 3 (OR = 2.76, 95% CI [1.28, 5.94], p<.01), and the PHQ-9 item #9 > 0 (OR = 5.36, 95% CI [4.00, 7.18], p<.001). The ASQ > 0 also had a larger pooled effect than C-SSRS > 0 (p = .02), C-SSRS > 3 (p <.001), and PHQ-9 item 9 > 0 (p <.001) for a composite suicidal behavior outcome. Study quality (per QUADAS-2) was highly variable.
Overall, the ASQ demonstrated considerably larger effect sizes compared to the other tools. Medium-sized effects were observed for both the C-SSRS > 3, and large effects for both the C-SSRS > 0 and PHQ-9 item #9 > 0, indicating similar performance. While the ASQ emerged as the tool most strongly associated with suicide-related outcomes, this relatively small sample of studies with extractable effects highlights that additional research is needed to conduct more rigorous moderation analyses and robust between-measure comparisons in diverse clinical settings.