An average of 17.5 Veterans died by suicide daily in 2021 (1). Of these Veterans, 61.9% had not used Veterans Health Administration (VA) services in 2020 or 2021 (1). Thus, Veterans not using VA services are a critical subgroup to target in suicide prevention efforts.

VA has developed and implemented a multicomponent suicide prevention program that is unmatched by other healthcare system in the public or private sectors (2–5). In particular, VA has transitioned to implementing a public health approach to suicide prevention that calls on VA and non-VA clinical and community organizations to implement suicide prevention, intervention, and postvention services, as detailed in the VA’s National Strategy for Preventing Veteran Suicide (6).

The National Strategy (6) describes four pillars of suicide prevention: 1) Healthy and Empowered Veterans, Families, and Communities, 2) Clinical and Community Preventive Services, 3) Treatment, Recovery, and Support Services, and 4) Surveillance, Research, and Evaluation. VA implemented a new national initiative with these pillars in mind, namely, Community-Based Interventions for Suicide Prevention (CBI-SP). Through the CBI-SP initiative and embracing collaborations with community partnerships, VA has expanded its capacity to prevent suicide among Veterans not connected to VA care. The CBI-SP model encourages establishing community-based suicide prevention coalitions that target priority areas aimed at increasing: 1) identification of Veterans and family members within the community and increasing screening for suicide risk; 2) connectedness within the community and during improved care transitions; and 3) community-wide lethal means safety and safety planning.

Aligning with the CBI-SP initiative, the VA Patient Safety Center of Inquiry for Suicide Prevention (PSCI-SPC), a national suicide prevention research center sponsored by the VA National Center for Patient Safety, developed and tested the VA COmmunity LeArning CollaboraTive (CO-ACT). CO-ACT employs a learning collaborative model as one method to engage and partner with clinical and nonclinical community organizations to expand suicide prevention programming (7, 8). Prior literature details a collaborative learning model developed by the Institute for Healthcare Improvement, known as the Breakthrough Series (9). This model provides a structured and time-limited process (6- to 16-months) for healthcare organizations to learn from each other and subject matter experts. Participating teams then utilize quality improvement methods to translate that knowledge into action at their respective organizations. The Zero Suicide Institute also uses a learning collaborative method to assist organizations in implementing the Zero Suicide Model (10, 11). However, suicide prevention learning collaboratives beyond Zero Suicide are limited, and none target preventing suicide among military Veterans, a population with substantially higher suicide rates than non-Veteran adults. In 2021, after adjusting for sex and age, the rate of Veteran suicide deaths was 71.8% greater than non-Veteran adults (1).

The CO-ACT suicide prevention learning collaborative draws heavily from existing prevention models. Namely, the Mental Health Intervention Spectrum is rooted in the National Academy of Medicine’s (formerly named Institute of Medicine) prevention model and outlines strategies from promotion to recovery (12–14). While the Substance Abuse and Mental Health Services Administration (SAMHSA) prevention classification system was originally created to target substance misuse, it’s use is applicable to suicide prevention efforts (15). Using a Veteran suicide prevention lens, the learning collaborative focused on Promotion and Prevention strategies (12, 13, 16). Promotion efforts broadly target the general public. Prevention is divided into three stages that target increasingly specific populations through strategies that include: universal prevention (e.g., targets the general population, such as through suicide prevention awareness and education), selective prevention (e.g., targets individuals at increased risk for suicide), and indicated prevention (e.g., targets individuals at high risk for suicide due to a past suicide attempt or current suicidal ideation; 13).

CO-ACT utilizes the integrated—Promoting Action on Research Implementation in Health Services (i-PARIHS) framework (17). i-PARIHS framework identifies three components as key to successfully implement evidence-based and best practices: context, recipients, and the innovation. Considering these key components prior to initiating implementation can increase the likelihood of success. Context refers to the work setting in which implementation will occur and relates to multiple levels (i.e., local, organizational, external health system). An innovation that does not fit into a clinic’s ongoing workflow is unlikely to be successfully implemented. The recipient refers to the individuals impacted by implementation at the individual and collective team levels. Recipients must understand the implementation and be able to implement the innovation into practice. For example, if a selected implementation requires independently licensed providers, the recipients implementing it must be clinically licensed. The innovation is the specific practice or change the team aims to implement and should be clearly defined and provide an advantage over current practices. Receptivity to implement an innovation is impacted both by evidenced-based and practice-based knowledge. Harvey and Kitson (2015) described how the compatibility of a proposed change can be enhanced by “aligning external explicit evidence [for an innovation] with local priorities and practice…” (p. 4) (17). Attending to these three key components can facilitate rapid implementation by preventing stuck points and improving the likelihood of implementation success.

CO-ACT has four aims: 1) create sustained organizational change so that organizations can adequately respond to Veteran suicide risk; 2) provide technical assistance, support, and education in building a suicide prevention program within community organizations; 3) build relationships and collaborations between the VA and community organizations to support suicide prevention; and 4) build relationships between community organizations to strengthen a Veterans suicide prevention safety net in the community. From 2020–2021, CO-ACT was pilot tested with clinical and non-clinical organizations across the broader Denver and Colorado Springs, Colorado region. The aim of the current paper is to detail CO-ACT’s structure and processes. In doing so, the authors seek to increase knowledge of mechanisms of learning collaboratives for Veteran suicide prevention, specifically methods that facilitate rapid implementation of needed Veteran suicide prevention strategies into community organizations.

Colorado has notably high rates of suicide. In 2021, Veterans in Colorado had a significantly higher rate of suicide compared to the national Veteran suicide rate and the national general population suicide rate (1). To expand and improve upon existing suicide prevention efforts across the state in its pilot test, CO-ACT employed community mapping to understand the landscape of community organizations serving Veterans in the greater Denver area and support organization recruitment. Further, snowball recruitment was used to identify and invite other clinical and non-clinical organizations serving Veterans to participate. Local VA and community leaders provided recommendations and/or connected the CO-ACT team to these potential organizations. In 2020, 13 clinical and non-clinical community organizations were invited and participated in a pilot test of CO-ACT. Five of these organizations provided mental health services to Veterans and/or Military Service Members, while the remaining eight provided non-healthcare services [see Figure 1 for additional details, as reported in DeBeer et al., 2023b (8)]. At the discretion of each organization, each organization’s team size ranged from one to four members.

CO-ACT was adapted for delivery using an entirely virtual learning collaborative (i.e., videoconference calls via Zoom™ and Microsoft Teams™) following early COVID-19 pandemic in-person meeting restrictions. CO-ACT occurred over 16 months, and consisted of 6 quarterly collaborative group meetings and 15 monthly individual organizational team facilitation calls. Implementation activities were guided by the Institute for Healthcare Improvement (IHI) model (9), existing prevention models, and suicide prevention programs/frameworks (6, 11, 18–24). The iPARIHS framework (17), in conjunction with Plan-Do-Study-Act (PDSA) cycles (25), guided implementation and quality improvement planning and action.

Program Components (i.e., the specific suicide prevention programming practices teams select to implement at their respective organizations) included suicide prevention, intervention, and postvention service options identified in one or more of four existing and prominent Veteran, active military, or civilian suicide prevention suicide prevention programs or models. Namely, these programs/models comprised of the Department of Veterans Affairs Suicide Prevention Program (6, 18), the Defense Suicide Prevention Program (DoD) (20, 21), the Zero Suicide program (11, 22), and the Division of Violence Protection National Center for Injury Prevention and Control-Centers for Disease Control (CDC) and Prevention program (23, 24). These four suicide prevention programs/models were compared to identify all possible suicide prevention strategies beneficial to VA and community suicide prevention programming (2). Once all distinct program components were identified, similar strategies were combined and re-defined to ensure integrity of the original definition.

This resulted in a comprehensive list of the suicide prevention program components detailed in the CO-ACT toolkit. Each program component was categorized according to two existing prevention frameworks. Specifically, the Mental Health Intervention Spectrum (adapted from the National Academy of Medicine Continuum of Care Model (12–14, 16) and the SAMHSA Prevention Model (15). The intent with labeling each program component according to these frameworks was to assist organizations in understanding the spread of their program components over different types of prevention. Consistent with Reason’s Swiss Cheese Model, prevention practices are strongest when they are spread over multiple types (26, 27).

Program components were also categorized as either a best practice or an evidence-based practice (28–33). Best practices are generally accepted treatments, techniques, or methods by health care experts that are used by professionals as appropriate treatments for certain disorders that have proven helpful over time and that are conferred as being measurable, replicable, and notably successful (31, 33), In contrast, an evidence-based practice is one that has undergone scientific evaluation to validate its effectiveness and which is recognized in peer-reviewed scientific journals (28–33).

Each program component in the CO-ACT toolkit was labeled according to the Zero Suicide classification to assist teams in determining how practices implemented through CO-ACT fit with the Zero Suicide framework (11). Zero Suicide categorizes suicide prevention components as: Lead (i.e., interventions promoting suicide prevention at a systems level), Train (i.e., training staff to promote high quality care services), Identify (i.e., identifying those at risk using a screening tool), Engage (i.e., using suicide care management plans for those at risk), Treat (i.e., using evidenced-based treatments and other best practices to reduce risk), Transition (i.e., using best practices for care coordination to ensure the Veteran remains connected to services), and Improve (i.e., using quality improvement to continually improve suicide prevention practices) (11).

The current paper described the structure and processes of CO-ACT, a Veteran suicide prevention focused learning collaborative (7) that employs a quality improvement framework to enact a public health approach toward reducing suicide. CO-ACT uses a learning collaborative model (9) to implement best and evidence-based suicide prevention program practices applicable to Veterans (6, 11, 18–24) in community organizations providing clinical or non-clinical services to Veterans. By utilizing a continuous quality improvement process (i.e., PDSA cycles) (25) and the i-PARIHS model (17) to guide implementation efforts, this learning collaborative was designed to enact sustainable Veteran suicide prevention programming into community organizations (8). With this intention, priority was given to implementing programming that was likely to be sustained after the learning collaborative ended. For example, before implementing a training for staff in suicide risk assessment, a new training policy was first implemented on who should be trained, when, and how often staff should be retrained.

The format and structure of the learning collaborative allowed teams to have a framework from which to: 1) conceptualize the current state of their program; 2) determine how to move forward in building an internal Veteran suicide prevention program; and to 3) utilize existing internal and external resources to support a continuous quality improvement process. As reported in DeBeer and colleagues (2023), organizational teams found these processes to be acceptable and feasible, and facilitated significant implementation of suicide prevention programming across all partnering organizations (8). Further, as reported in DeBeer and Colleagues (2024), participating in CO-ACT resulted in an increase in the quantity and quality of relationships/partnerships between participating community organizations and the VA (34).

In contrast to how learning collaboratives typically focus on healthcare systems (9), the current collaborative included non-healthcare organizations. This was done to be consistent with the VA’s public health approach to suicide prevention (6), reaching Veterans at opportunities outside of healthcare interactions. Decisions were also made during the course of this pilot to consolidate program materials as a means of improving their usability. Specifically, the initial CO-ACT toolkit was conceptualized as a toolkit with a companion workbook. However, informal feedback from organizational teams and co-investigators suggested these materials should be package into one document. Consequently, the workbook was integrated into the toolkit to provide a single resource for partner organizations to utilize during CO-ACT.

VA CO-ACT facilitation leads collaboratively worked with each organization’s team in an effort to overcome challenges to implementing recommended suicide prevention programming. There were two commonly encountered barriers. First, organizations often lacked necessary resources (e.g., staffing shortages) and/or had staffing with significant constraints on their time. Second, not all organizations had data platforms available to easily track the implementation of suicide prevention programming, and tracking is an essential component to engaging in a continuous quality improvement process.

In light of encouraging findings from this pilot (8), subsequent tests are applying the same structure and processes of CO-ACT to a learning collaborative focused only on VA Community Care organizations (i.e., health care organizations the VA contracts with to provide Veteran health care services). Learning collaboratives can serve as a powerful vehicle for enhancing suicide prevention practices across organizations and offer promise as a solution to address Veteran suicide deaths in community settings.

The original contributions presented in the study are included in the article. Further inquiries can be directed to the corresponding author.

JM: Conceptualization, Writing – original draft, Writing – review & editing. PR: Conceptualization, Writing – original draft, Writing – review & editing. EB: Writing – review & editing. CB: Writing – review & editing. LM: Writing – review & editing. KB: Writing – review & editing. EV: Writing – review & editing. CH: Writing – review & editing. AP: Writing – review & editing. JH: Writing – review & editing. NM: Writing – review & editing. SB: Writing – review & editing. KW: Writing – review & editing. MP: Writing – review & editing. MM: Writing – review & editing. BK: Writing – review & editing. JG: Writing – review & editing. AB: Writing – review & editing. JS: Writing – review & editing. TA: Writing – review & editing. JB: Conceptualization, Writing – review & editing. BD: Conceptualization, Writing – original draft, Writing – review & editing.