This study examines a novel way to prevent suicide among Veterans with serious mental illness (SMI). It will assess the efficacy of PREVAIL-VA, a 3-month intervention of 12, one-on-one sessions between a Peer Specialist (PS) and a Veteran that involve semi-structured conversations focused on hope, belongingness, support, and safety. PSs are Veterans with SMI trained to use their own experience and recovery to help other Veterans with SMI. PSs have improved mental health outcomes in other research, but this study would be the first to test their efficacy for suicide prevention in VHA. The project will compare outcomes of Veterans at risk for suicide receiving usual care, to PREVAIL-VA. This work is responsive to national calls for Veterans with SMI to receive support that is evidence-based, improves recovery (not just symptoms), and is tailored to individual needs. If successful, PREVAIL-VA could be adopted by the 1400 Peer Specialists employed in VA, greatly increasing the delivery of evidence-based services to Veterans at risk for suicide.

BACKGROUND: Suicide is a top priority for the VA, claiming about 6,000 Veterans each year. VA suicide services primarily focus on immediate response to increased suicide risk but lack a focus on recovery, conceptualized by VHA as an internal process of hope, healing, empowerment, and social reconnection. The lack of a recovery focus in suicide prevention is a gap in VHA care quality. VHA Peer Specialists (PSs)-Veterans with serious mental illness (SMI) who are trained to use their experience to help other Veterans with SMI as full-fledged employees (~1,400 VA-wide)-could improve the quality of suicide prevention by making these services more recovery oriented. Based on multiple studies showing that PSs improve a range of outcomes for Veterans with SMI, the VA National Strategy for Preventing Veteran Suicide calls for the development of peer-to-peer services to help those at risk for suicide for their ability to "impart hope and motivation for achieving recovery; provide support for addressing specific stressors, and help foster a sense of meaning and purpose."

SIGNIFICANCE: Administrative data shows that VHA PSs are already working with Veterans at risk for suicide. If PREVAIL-VA is effective, it could greatly increase the delivery of evidence-based services Veterans at risk for suicide receive and enhance the services by VHA Peer Specialists.

INNOVATION: No study has tested a PS-delivered suicide prevention service in the VHA in a rigorous trial.

SPECIFIC AIMS: The proposed study is a Hybrid Type 1 randomized trial with two aims: Aim 1: Deliver PREVAIL-VA and compare recipients of PREVAIL-VA and Usual Care on suicide-related and recovery outcomes. Aim 2: Collect qualitative data on PREVAIL-VA helpfulness, and implementation barriers and facilitators using the Consolidated Framework for Implementation Research (CFIR) and rapid qualitative analysis.

METHODS: This application proposes to test in two VA health systems-Pittsburgh and Maryland-a promising PS-based approach to helping VA patients with a high risk of suicide, called PREVAIL-VA. Adapted from a civilian version with VA PS and suicide stakeholder input, PREVAIL-VA (Peers for Valued Living) provides 12, one-on-one sessions over about three months between a PS and a Veteran that involve semi-structured conversations focused on hope, belongingness, and safety. Session context is flexible, driven by the needs of the individual and by the ILSM (Invite, Learn, Share, Motivate) structure that guides how PSs interact with those whom they work. A civilian pilot study was promising; preliminary findings of a larger civilian trial showed small improvements. Veterans with documented suicide risk at each site (n=153) will be randomized to PREVAIL-VA or Usual Care (total N=306). Each site will have two trained PSs, who will each deliver PREVAIL-VA to ~40 Veterans over ~3 years (~4 Veterans at any one time). PSs will be trained and receive an hour of group supervision weekly by the PREVAIL-VA developer and licensed psychologists, who will also be available for emergencies. Sessions will be taped and 15% rated for fidelity on standardized measures. Each Veteran will be assessed at baseline, post-intervention, and six months post-intervention on the primary outcome of suicide risk with additional primary outcomes focusing on recovery domains such as self-rated community integration; sense of hope, meaning, and purpose. Secondary outcomes will include suicide attempts, depression, belonging, and acute care visits for suicide-related reasons. The primary analytic strategy will be a generalized mixed-effect model approach including intervention group, time, and time by group interaction terms. Relevant covariates will include site, fidelity ratings, treatment attendance, VA service use, and demographic variables. At each site, all PSs and a subset of PREVAIL-VA Veterans, VHA clinicians whose patients received PREVAIL-VA, and clinical leaders will participate in qualitative data collection (periodic reflection, interview, focus group) about helpfulness of PREVAIL-VA and relevant CFIR-based implementation factors that could inform its future adoption. Qualitative data will be analyzed with a 'rapid analysis' approach.

NEXT STEPS/Implementation: If effective, we will develop an implementation toolkit and work with partners to teach PREVAIL-VA to all PSs.

Inclusion Criteria:

There are two Inclusion criteria:

• Moderate to high suicide risk according to one of the following occurring in the prior 3 months:

  • a completed Comprehensive Suicide Risk Evaluation (CSRE) indicating intermediate or high acute or chronic risk
  • a high-risk suicide PRF
  • a suicide attempt indicated on a Suicide Behavior or Overdose Report (SBOR)
  • a suicide safety plan completed in the past 6 months

AND

• Current suicidal ideation according to the 9th item of the Patient health questionnaire (PHQ-9) > 0 at enrollment

  • While a positive 9th item usually precedes the CSRE, because the CSRE could have been administered in the 3 months prior to enrollment, the purpose of requiring a positive 9th item is to ensure more recent suicidal ideation.
  • It may be more difficult to show an intervention effect if the investigators enroll participants with no suicidal ideation at baseline.

Exclusion Criteria:

• Inability to provide informed consent for any reason, including due to acute psychosis or mania, dementia, or active guardianship or durable power of attorney

  • These criteria will be determined according to medical record review and, if clarification needed, consultation with a treating provider.
  • The investigators will also screen for decision-making capacity using the Blessed Orientation, Memory, Concentration (BOMC) Test and a brief quiz about the study.
  • This quiz will involve an iterative process of querying the participants' understanding of consent information with a 10-item true/false test and providing feedback until an acceptable level of understanding is achieved (must get 100% correct after 3 tries to enroll).

• Current or planned receipt of residential or intensive outpatient treatment

• Plan to transfer or terminate care in the next 3 months

• Homicidal ideation in the past 6 months or violent behavior in the past 12 months

The PI will create de-identified, study-specific datasets. All variables presented in a publication will be included in the dataset. Some loss of information might occur in this process given the need to remove PHI. The PI will replace social security and medical station numbers with study-specific numbers. The PI will drop date of birth and replace age with age categories, per PHI requirements for people 85+years of age. Service dates will be replaced with study specific time indicators.

The analytical datasets and statistical code used in the publication will be retained for 6 years, in accordance with VA record retention policy. The crosswalk between the analytical dataset(s) and the public release datasets will be maintained so that a VA-approved auditor or the PI could conduct or facilitate validation if needed. By sharing our data set, it will allow other investigators to independently confirm and validate our findings.

Investigators must demonstrate a clear rationale for how the data will be used.

Once Veterans are enrolled and complete the baseline assessment, research staff will open a sealed envelope containing the Veteran's random assignment to PREVAIL-VA or treatment as usual in a 1:1 ratio at each site. All randomized Veterans will be included in intent-to- treat analyses. Although Veterans and the PSs will know to which condition Veterans have been assigned, re- search assistants responsible for all assessments will be blinded to study group assignment.