Suicide safety plan engagement: Temporal dynamics and momentary predictors

Suicide safety planning is an evidence-based intervention, yet little is known about when plans are used in daily life or preceding states. Given the documented dynamic nature of suicide ideation, this study examined proximal predictors of safety plan use using intensive longitudinal methods. Fifty-two undergraduate students screened for past-two-week suicide ideation completed a Stanley-Brown safety plan and a 10-day ecological momentary assessment protocol (five prompts/day). Suicide ideation facets (wish to live, wish to die, suicidal desire, suicidal intent), negative affect, perceived stress, and safety plan use were assessed. Lagged generalized linear multilevel models and autoregressive models tested whether within-person elevations predicted subsequent use. Forty-six percent of participants used their safety plan at least once, reporting high perceived helpfulness and likelihood of future use. Lagged suicidal intent (OR = 1.38, p = .034), perceived stress (OR = 3.43, p = .021), and lagged safety plan use (OR = 3.20, p < .001) predicted a greater likelihood of subsequent safety plan use. Wish to live, wish to die, suicidal ambivalence, suicidal desire, and negative affect were not significant. Safety plan use showed modest temporal clustering (ρ = 0.21, SE = 0.02). Safety plan use appeared state-dependent, increasing in response to acute elevations in suicidal intent and perceived stress rather than reflecting a general function of suicidal thinking, highlighting temporally proximal intervention targets.