Evaluating the Effectiveness of Safety Plans for Mitigating Suicide Risk in Two Samples of Psychiatrically Hospitalized Military Veterans

Although safety plans (SPs), following the Stanley-Brown Safety Planning Intervention protocol, are required for suicidal veterans receiving treatment in the Veterans Health Administration (VHA), prior studies have shown that they are frequently incomplete or are not sufficiently personalized to the unique circumstances of each patient. In two studies, we examined SP completeness, SP quality (i.e., degree to which the SP was clear, actionable, and personalized), and SP fidelity (i.e., sum of completeness and quality). We also examined which SP steps were associated with a reduced likelihood of future psychiatric rehospitalizations (Study 1) and suicide attempts (Study 2) following hospital discharge. Participants were veterans admitted to two VHA acute inpatient psychiatric units for suicide risk (Study 1: N = 78; Study 2: N = 132). SPs were coded by independent raters on completeness, quality, and fidelity; step scores (e.g., Step 1 quality) were summed to create whole-plan scores (e.g., SP quality). In Study 1, 52.5% of participants had an SP and, in Study 2, 93.1% of participants had an SP. In Study 1, whole plan scores were not associated with subsequent psychiatric hospitalization status, but higher Step 2 (internal coping) fidelity scores were associated with decreased likelihood of rehospitalization (AHR = 0.05, 95% CI [0.30, 0.84], p = .008). In Study 2, higher whole-plan quality (AHR = 0.79, 95% CI [0.66, 0.95], p = .012) and fidelity (AHR = 0.84, 95% CI [0.71, 0.99], p = .040) scores were associated with a decreased likelihood of future suicide attempt. Step 1 (warning signs) quality (HR = 0.48, 95% CI [0.30, 0.76], p = .002) and fidelity scores (AHR = 0.57, 95% CI [0.37, 0.90], p = .016) were associated with a decreased likelihood of future suicide attempt. The association of SP characteristics differs by outcome of interest, and fidelity of internal coping strategies may contribute to preventing rehospitalizations, whereas quality and fidelity of warning signs may help prevent future suicide attempts. Overall, results suggest that mandating SPs without training and implementation strategies to ensure quality is not enough.