3

Considerations for Program Development and Oversight and Grantee-Level Implementation and Performance Metrics

Building on the program examples presented in the preceding session, the next workshop session focused on considerations for program development and oversight, as well as grantee-level implementation and performance metrics. These closely connected topics were addressed together in a single session to foster cohesive dialogue on design, delivery, and oversight. The session opened with a series of brief presentations that provided a foundation for two moderated panel discussions. The first panel discussion focused on lessons learned from the programs described in the previous session. In the second panel discussion, other invited experts offered perspectives on questions related to program development, oversight, and implementation. The session concluded with a round of audience Q&A.

With emphasis on the public health approach, this chapter presents practical strategies for designing and overseeing non-clinical suicide prevention programs implemented at the community level. Drawing on foundational presentations and two panel discussions, participants explored how programs can be tailored to local capacity, needs, and cultural context; supported through proactive technical assistance; and evaluated using realistic metrics that reflect incremental progress. Additional topics included the development of theories of change and logic models, the use of dashboards and storytelling to communicate impact, equitable funding structures, and approaches to cultivating strong grantee networks and clarifying inclusion criteria.

FOUNDATIONAL PRESENTATIONS

Foundational presentations covered a range of topics. Following an introductory talk, presenters addressed the public health approach to comprehensive suicide prevention, models for effective community-based programs, the development of logic models, strategies for providing comprehensive technical assistance, and the use of data dashboards to support program implementation and oversight.

Theory of Change, Theory of Action, and Performance Metrics

Daniel Friend (Mathematica; member, workshop planning committee) kicked off the session with a presentation assembled in collaboration with Elly Stout (Education Development Center [EDC]; member, workshop planning committee). Friend opened by highlighting the importance of grounding programs in a cohesive theory of change and identifying realistic performance metrics. He noted that context is key, and it is important to consider who is being served by a program. A cohesive theory of change links program activities to their intended outcomes. Friend explained that the structure begins with activities—the mechanisms through which change is introduced—and then flows through a logical sequence of outcome stages:

• Change outcomes refer to proximal effects, such as shifts in attitudes, knowledge, or skills.
• Short- and intermediate-term outcomes reflect more distal changes, such as sustained use of knowledge and skills.
• Long-term outcomes represent broader population health improvements.

Friend used a psychoeducational workshop focused on resilience-building as an illustrative example of a cohesive theory of change (see Figure 3-1). A successful activity of this kind might reduce stigmatizing attitudes and improve social support (change outcomes), which in turn could increase help-seeking behaviors (intermediate outcome), ultimately contributing to a reduction in suicidal behaviors (long-term outcome).

Friend emphasized the distinction between a theory of change (which explains how change is achieved) and a theory of action (which ensures implementation quality). The latter is supported by performance metrics—quantitative indicators that track progress toward a particular target and guide decision-making (see Figure 3-2). Friend stressed that metrics are a cornerstone of continuous quality improvement. For instance, if a program has a goal of recruiting 100 veterans per month but is only enrolling 12, this indicates a need for strategic adjustment, such as changing

recruitment strategies. Realistic and achievable metrics allow for responsive management and improvement and help to ensure accountability to various stakeholders.

Drawing from experience across programs, Friend identified several common performance metrics that can be adapted based on program context:

• Recruitment and enrollment
• Dose (extent of services received)
• Initial engagement
• Service uptake
• Satisfaction and acceptability to participants
• Fidelity (adherence to the program model)
• Feasibility
• Sustainability
• Cost

Together, all of these metrics not only help ensure accountability but also serve as key inputs into program implementation as well as guiding program evaluation, Friend concluded.

The Public Health Approach and Comprehensive Suicide Prevention

Alex Crosby (Morehouse School of Medicine) provided an in-depth presentation on the public health approach to comprehensive suicide prevention, emphasizing its systematic methodology and application to reducing suicidal behaviors across diverse populations.

Crosby began by introducing the public health framework, noting its four core components: assessing the problem, identifying causes, developing and testing interventions, and disseminating and implementing successful strategies (see Figure 3-3). This structured method allows public health practitioners to systematically understand suicide-related issues and effectively intervene across different community contexts.

Assessing the problem involves developing understanding of the scope and nature of suicide by examining who is affected, where incidents occur, and the broader circumstances surrounding suicidal behaviors, Crosby explained. Identifying causes involves examining “why” questions, such as why suicide might be more prevalent in one population than another—e.g., urban versus rural, males versus females. After gaining understanding around aspects of the “why,” the next step is developing and testing programs and policies tailored to address these factors. Crosby offered as an example developing an intervention to reduce alcohol misuse because substance misuse, particularly alcohol misuse, is a major risk factor for

suicidal behavior. The final step in the public health approach entails expanding the reach of successful programs and disseminating lessons from such programs.

Crosby highlighted historical trends, noting suicide rates in the United States dating back to the 1930s. He pointed out that during the 1920s and 1930s, the Great Depression significantly influenced higher suicide rates, demonstrating the critical impact of economic factors on suicidal behaviors. Crosby stressed the importance of recognizing the multifaceted nature of suicide, cautioning against attributing rising suicide rates to any single factor, but rather understanding it as resulting from complex interactions of multiple determinants.

Examining recent trends, Crosby presented data showing increased suicide rates across various demographic groups, including age, sex, and racial/ethnic categories. Between 1999 and 2023, suicide rates notably increased across almost all age groups for both males and females, except for males aged 75 and older, where rates remained relatively stable (see Figure 3-4). Similarly, data disaggregated by race and ethnicity indicated increasing suicide rates among White, Hispanic, Black, and Asian American populations, and particularly high increases among American Indian and Alaska Native populations. Crosby emphasized this widespread increase to underline the need for broad-based suicide prevention efforts (see Figure 3-5).

Crosby also highlighted suicide rates among veterans (see Figure 3-6) compared to non-veterans, revealing ongoing challenges despite slight

improvements in recent years. This underscored the importance of targeted suicide prevention strategies tailored specifically to the veteran community.

Crosby explained that it is important to consider the impacts of suicidal behaviors beyond mortality. Suicide-related hospitalizations, emergency department visits, reported attempts, and serious ideation occur with progressively greater frequency. While approximately 46,000 suicides occurred in 2021, millions more experienced serious suicidal ideation or attempts, representing critical intervention opportunities to prevent suicides, Crosby noted (see Figure 3-7). He added that those at risk of suicide are often dealing with multiple issues that can be at the individual level, at the family or peer level, at the community level, or at the societal level. For example, an individual might be facing housing insecurity, chronic illness, and hunger or food insecurity simultaneously. Intervening on just one of these issues might be enough to decrease their risk for suicidal behavior, Crosby suggested.

Crosby outlined the Centers for Disease Control and Prevention’s (CDC’s) Comprehensive Suicide Prevention program, describing key components such as strong leadership, multisectoral partnerships, and the use

of data-driven strategies. He discussed CDC’s approach to identifying disproportionately affected populations—populations that experience rates of suicide greater than the general population, which include veterans, tribal populations, rural communities, LGBTQ+ individuals, and youth.

To illustrate comprehensive suicide prevention strategies, Crosby described CDC’s Suicide Prevention Resource for Action, which identifies specific strategies and approaches supported by evidence (see Table 3-1). Crosby pointed out evidence indicating that communities implementing multiple strategies simultaneously experienced greater success compared to those employing single-strategy interventions.

In describing upstream suicide prevention, Crosby referred to the visual metaphor illustrated in Figure 3-8. He likened suicide risk to individuals falling through a gap in a bridge into a river leading to a waterfall downstream, describing different points along the river where interventions might occur. As shown in the figure, the furthest upstream strategy is to repair the gap in the bridge, an approach Crosby equated with strategies such as strengthening economic supports and building coping and problem-solving skills. Midstream interventions might focus on reducing substance misuse, while crisis response is often necessary downstream. Crosby emphasized that public health approaches, including those discussed throughout the workshop, address all of these levels simultaneously. However, he stressed that upstream strategies tend to be the most effective and cost-efficient. Even so, midstream and downstream interventions remain vital, as not everyone can be reached early.

TABLE 3-1 Comprehensive Suicide Prevention Strategies and Approaches

SOURCE: Presentation by Alex Crosby on April 29, 2025; Centers for Disease Control and Prevention (2022).

Concluding his presentation, Crosby reiterated that suicidality is a critical, multifaceted public health issue. He stressed the necessity of a comprehensive, evidence-informed approach, emphasizing that effectively addressing suicide requires interventions at multiple levels—individual, family, community, and societal. Crosby affirmed that tailored strategies responsive to specific community needs and protective factors are essential for meaningful reductions in suicidal behaviors.

Models for Effective Community Suicide Prevention

Stout presented an overview of models and tools that can support the design and implementation of effective community-based suicide prevention efforts. She emphasized the importance of helping communities adapt concepts and strategies to fit their specific needs, assets, and levels of readiness.

Stout opened by noting that community suicide prevention has long been a priority in national strategies, including the 2012 National Strategy for Suicide Prevention and more recent updates. While national strategies outline broad goals and objectives, Stout explained, relatively few models are available that translate those goals into actionable guidance for communities. She highlighted three frameworks that offer communities clear direction on both what to implement and how to implement it.

The first framework Stout mentioned was the CDC Suicide Prevention Resource for Action, which outlines evidence-informed strategies for reducing suicide risk. As described earlier in Crosby’s presentation on the public health approach and comprehensive suicide prevention, this resource identifies multiple areas for intervention that, when combined, are associated with greater impact (see Table 3-1). Stout clarified that CDC’s resource focuses on broad strategies—such as strengthening economic supports or creating protective environments—rather than prescribing specific programs. The goal, she said, is to help communities prioritize and combine strategies based on their own context.

Stout next discussed the Suicide Prevention Resource Center’s (SPRC’s) Comprehensive Approach to Suicide Prevention,¹ which predates the CDC model but includes many of the same elements (see Figure 3-9). She described the visual representation of the model as a puzzle, with interconnected components that collectively support a comprehensive approach. Importantly, she noted that this is not a checklist; communities are not expected to implement every element. Rather, the model can help community stakeholders identify where they are currently investing and where additional focus may be needed.

To complement the SPRC framework, Stout introduced SPRC’s Strategic Planning Approach, which offers a step-by-step process for tailoring community suicide prevention strategies (see Figure 3-10). The process includes assessing suicide-related risk and protective factors in the local community, identifying populations at highest risk, and taking stock of the local context, including assets and opportunities. From there, community leaders can select appropriate strategies from the CDC or SPRC models and identify culturally appropriate programs or practices for implementation. Evaluation is the final step in the cycle, allowing communities to assess whether they are making measurable progress toward reducing suicidal behaviors.

Stout then turned to a third model—EDC’s Community-Led Suicide Prevention approach—which builds on the CDC and SPRC frameworks and incorporates insights from a 2017 National Action Alliance for Suicide Prevention report on transforming communities. That report reviewed global examples of community-based suicide prevention efforts and identified seven key elements that contribute to success (National Action Alliance for Suicide Prevention, 2017):

1. Unity
2. Strategic planning

3. Integration
4. Sustainability
5. Data
6. Fit
7. Communication

Stout highlighted two of these components due to their relevance to discussions at this workshop. She explained that the unity component involves bringing community partners together and building consensus. With regard to the data component, she noted that it involved considering the local context and pointed to the White Mountain Apache surveillance system discussed earlier in the workshop (see Chapter 2) as an example of a culturally appropriate data system that also supports sustainability and effective communication.

Stout concluded by underscoring the value of grounding grant programs and community initiatives in structured, evidence-informed models. Doing so, she said, helps ensure that strategies are aligned with community needs and priorities, and gives communities a way to assess whether they are making progress toward meaningful outcomes. She encouraged participants to explore publicly available resources from CDC (Centers for Disease Control and Prevention, 2022), EDC (Education Development Center, 2025), the National Action Alliance for Suicide Prevention (2017), SPRC (2020a,b), U.S. Department of Health and Human Services (2024), and U.S. Department of Veterans Affairs (2018).

Developing Logic Models for Community-Based Suicide Prevention

Corbin Standley (American Foundation for Suicide Prevention [AFSP]²) presented an overview of logic models and their role in guiding evaluation of community-based interventions. He described logic models as visual roadmaps that explain the relationship between a program’s activities and its intended outcomes, provide a framework for both implementation and evaluation and making explicit the program’s underlying logic. He added that logic models are typically one-page, iterative documents that evolve as programs mature and data become available.

Standley explained that logic models often include five core components:

1. Goals, which define the overarching purpose of the program;
2. Learning objectives and program objectives;
3. Processes and activities carried out as part of the program;
4. Outcomes, including short-term and long-term changes; and
5. Context, which accounts for cultural and community considerations.

He underscored the importance of cultural humility and cultural appropriateness, working with the community to define problems needing to be addressed along with desired solutions.

To help orient the audience to terminology, Standley introduced a general template showing the relationship between inputs, activities, outputs, outcomes, and long-term impacts (see Figure 3-11). The framework shows how inputs such as funding, personnel, and materials support activities like training or outreach, which lead to outputs (e.g., number of presentations or attendees) and short-term outcomes—measurable changes you anticipate immediately following a program or intervention, such as knowledge gain or skill development. These, in turn, contribute to longer-term outcomes like reduced stigma, increased help-seeking behavior, and ultimately reduced suicide rates.

Standley then walked through a more detailed breakdown of each logic model component, offering specific guiding questions and prompts for program planners (see Figure 3-12A). For example, when defining inputs, program planners might ask what materials, training, and technology are needed. To illustrate how these principles are applied in practice, Standley shared a logic model used by the AFSP for its flagship education program, Talk Saves Lives (see Figure 3-12B). Inputs for the program include a scripted PowerPoint, an implementation guide, and evaluation tools. Activities include training volunteer presenters and delivering presentations in communities. Outputs track the number of trainings, attendees, and participant satisfaction. Short-term outcomes include increased knowledge of suicide prevention, improved attitudes and beliefs, greater self-efficacy, and reduced barriers to talking about suicide. While long-term impacts such as reduced suicide rates are more difficult to measure directly, they remain central goals.

Comprehensive Technical Assistance to Support Program Outcomes

Carrie Farmer (RAND Corporation) offered reflections on the role of technical assistance as a strategic tool for helping community-based programs monitor their performance and achieve success. Echoing earlier remarks, Farmer noted that grantees and community-based organizations can have mixed capacities for data collection, evaluation, and reporting depending on how big or small they are; availability of financial,

technological, and other resources; staff expertise; and connections to other expertise.

Farmer explained that traditional technical assistance typically involves providing on-demand assistance for specific problems as they arise over the course of program design, implementation, and evaluation. In contrast, comprehensive technical assistance is more proactive and takes a long-term approach, with a goal of building knowledge and capacity for grantees to collect and evaluate outcomes on their own, thus reducing their reliance on technical assistance in the future.

A comprehensive approach can help align grantees around the measurement of priority outcomes the funder is interested in, thus ensuring relevant data are collected, Farmer noted. It can also build a cohort of programs or grantees so they can learn from each other and share best practices. Comprehensive technical assistance may include one-on-one coaching, in-depth assessment of grantee needs, development of individualized support plans, and active tracking of grantee progress over time.

Farmer pointed to the technical assistance provided by SAMHSA’s Service Members, Veterans, and their Families (SMVF) Technical Assistance Center³ as a promising comprehensive support model (see Figure 3-13). She described what is entailed in each of the model’s four phases:

1. Phase 1 focuses on orienting each grantee to the intervention through a cohort-based model that ensures fidelity to the program’s core components. This phase emphasizes level-setting and understanding the context of each specific grantee.
2. Phase 2 consists of a series of web-based preparatory sessions that introduce foundational concepts. Through presentations by subject matter experts, grantees receive structured education on key topics such as evidence-based practices, logic model development, and data management—including how to collect, interpret, and store data effectively.
3. Phase 3 brings grantees together for an Implementation Academy. During the Academy, grantees deepen their understanding of the specific issue(s) they are working to address, strengthen their ability to implement it successfully, and learn from lessons shared by peers and experts. This phase helps organizations identify suitable evidence-based practices, understand implementation challenges, and develop strategic implementation plans.

4. Phase 4 consists of individualized technical assistance. This ongoing support helps grantees troubleshoot challenges, adapt practices as needed, and sustain implementation quality over time.

Farmer underscored the value of the model’s heavy focus on implementation, noting that desired outcomes will not be achieved if the intervention is not implemented well from the start.

Service Members, Veterans, and Their Families Technical Assistance Center

The SMVF Technical Assistance Center’s approach to comprehensive technical assistance might serve as a useful model for the Sergeant Parker Gordon Fox Suicide Prevention Grant Program, Farmer noted. She suggested several specific elements that could strengthen the Fox program’s technical assistance approach. These include offering virtual workshops on topics such as evaluation design, data collection, data analysis, and communication. Farmer emphasized the importance of one-on-one technical assistance, as grantee needs can vary widely—some organizations may require support in identifying appropriate metrics, while others may need guidance in navigating institutional review board processes. She also recommended the development of evaluation support tools, such as a curated catalogue of process and outcome measures, as well as the creation of a community of practice that would allow grantees to learn from one another.

Design of Actionable Dashboards for Supporting Program Implementation and Oversight

Itzhak Yanovitzky (Rutgers University) provided an overview of how dashboards can serve as actionable tools in support of public health initiatives. Drawing from his expertise in behavioral and social change communication, he discussed how dashboards—when properly designed and implemented—can be leveraged not only for monitoring and dissemination but also for strategic planning, goal setting, coordination, and system learning.

He emphasized that while dashboards are often used to track program performance or report outcomes, they hold broader potential to advance decision-making and catalyze action. Actionable dashboards can permit real-time access to curated data, enable users to visualize trends and detect bottlenecks, support implementation decisions, and connect stakeholders across systems. As illustrated in Figure 3-14, dashboard design involves a range of decisions regarding data, users, and purpose—from monitoring and communication to alignment and learning.

However, Yanovitzky cautioned that building a dashboard alone does not guarantee its usefulness or use. Simply providing access to data does not mean people will use it—or use it well. He outlined several factors that determine whether a dashboard actually influences behavior or supports action. These include the characteristics of the data itself (e.g., timeliness, granularity), user capacity and needs, the purpose of the dashboard (strategic vs. tactical), the design elements (interactivity, customization), the context in which the dashboard is being introduced (e.g., policy environment, user norms), and whether it is integrated into existing workflows.

Dashboards, Yanovitzky noted, often fail when developed in isolation from their intended users. To overcome this challenge, he proposed an “actionability by design” approach—a process that brings together program staff, data scientists, and end users to co-create dashboards that meet real-world needs and achieve three overlapping goals: usability, usefulness, and integration (see Figure 3-15). Yanovitzky explained that usability refers to whether a dashboard is intuitive, efficient, and accessible. Usefulness reflects whether it delivers timely, context-relevant information that supports decision-making. Integration requires that dashboards become embedded into daily workflows and organizational culture. If a dashboard is not used consistently, it will not achieve its intended purpose, he stated.

Yanovitzky described “actionability by design” as a collaborative, iterative process that begins with identifying shared problems and mapping pathways for action. From there, the team defines data needs, develops the dashboard collaboratively, and periodically refines it while addressing potential unintended consequences.

Yanovitzky also stressed that dashboards must be tailored to the specific context and application; they are not templates. He shared a concrete example from New Jersey, where a statewide opioid dashboard was developed through inter-agency collaboration. The project began with public town halls and evolved into a tool that allowed for cross-county learning, resource mapping, and alignment of services across government systems. He concluded his presentation by reiterating that such an impact is only possible when the dashboard design process is grounded in early and sustained collaboration with stakeholders.

PANEL DISCUSSIONS AND AUDIENCE Q&A

After the foundational presentations, the session moved to a series of two moderated panel discussions. The first panel discussion focused on lessons learned from the examples of programs supporting non-clinical community-based suicide prevention efforts discussed earlier in the day. The second panel discussion provided the opportunity for other invited experts to offer additional insights. Audience Q&A followed the panel discussions.

Lessons Learned from Examples of Non-Clinical Community-Based Suicide Prevention Efforts

Colin Walsh (Vanderbilt University; member, workshop planning committee) served as moderator for both panel discussions summarized in this chapter. The first panel discussion was with the individuals who had given the presentations summarized in Chapter 2: Mary Cwik (Johns Hopkins University), Novalene Alsenay Goklish (Johns Hopkins University), Brandi Jancaitis (Virginia Department of Veterans Services), Richard McKeon (SAMHSA), and David Rozek (University of Texas Health Science Center at San Antonio). Topics included structuring programs to minimize grantee burden, fostering collaboration and peer learning, tailoring outreach and respecting community readiness, planning for evaluation and communication of findings, and reflections on grant design and adaptation.

Structuring Programs to Minimize Grantee Burden

Walsh began the moderated panel discussion by asking panelists to share lessons learned around how to structure programs to minimize burden on grantees. Jancaitis noted that this is an important question that she and her team grapple with daily. As their grant sizes are relatively small ($75,000 to $175,000 per grantee, which may barely fund a full-time staff member depending on level of staffing), grantees have to be creative in how they use that investment. Jancaitis added that the small awards drew grantees that might have smaller infrastructures, so expectations for impact needed to be realistic. She noted that they begin providing technical assistance before the grant is awarded to help grantees with awareness of resources and where to go for data to help define the need, as well as capacity to put together a basic logic model. In addition to pointing to best practices, technical assistance providers answered questions as prospective grantees put together their applications and continued to layer resourcing in between their reporting cycles. Jancaitis emphasized that a core lesson learned was that technical assistance has to start from the beginning and continue throughout the program.

Rozek also commented on the work that is needed to support prospective grantees prior to receiving an award. He and his team approach support for prospective grantees as building a partnership and examining how they might help with infrastructure or barriers to starting the work. As an example of the latter, he noted that some organizations might have hiring restrictions that require funding to have been received before a position can be posted. Given that there might be a lag time of three months or more between the start of the grant period and receipt of funding, this can present challenges if funding is not for a multi-year period. Rozek noted that they

Tailoring Outreach and Respecting Community Readiness

Walsh commented on the emerging themes of the cyclical or iterative nature of assistance and support, identifying barriers directly with grantees and working creatively to overcome them, and shaping programs in partnership with grantees. These themes, he noted, offered a useful lead-in to the next question—how to cultivate applicants, particularly in communities that programs may want to reach but that have not historically received strong support for efforts of this kind.

McKeon reflected on the early days of the GLS program and the landscape of suicide prevention at the time it was launched. “You have to understand what the environment was like back in 2004,” he said. “Garrett Lee Smith was the first nationally available funding source for suicide prevention in the United States, so there was a hunger for resources in states and communities across the country.” At that time, eligible applicants included states, tribes, and their designees. McKeon noted that states quickly became aware of the opportunity, in part due to infrastructure already in place through SPRC’s contact with state suicide prevention coordinators. Additional effort was needed to raise awareness among tribal communities, however, he added, requiring explicit attention as the program moved forward.

For the Face the Fight initiative, understanding the current suicide prevention landscape was seen as important to informing their work, Rozek stated. They provided a grant to the RAND Corporation to conduct a landscape analysis for current and emerging work in this area, which was released just prior to the workshop (Ramchand et al., 2025). This analysis will equip Face the Fight to identify potentially scalable programs that may be a fit for their mission.

Jancaitis responded that the organizations her team seeks to bring into the funding portfolio generally fall into two broad categories. The first group is deeply embedded in the military community, with strong military culture and military- or veteran-connected staff. While these organizations tend to be well positioned to serve veteran populations, they may need additional support in areas such as mental health, suicide prevention, and integration with broader service systems. The second group consists of more mainstream organizations with well-established mental health programming but limited experience working with military-connected populations.

Military bases and other military facilities are one possible avenue for reaching the former community, Jancaitis noted, but it is important to consider community needs and approach with cultural sensitivity, particularly in populations where help-seeking can be particularly difficult or even taboo.

Another point Jancaitis emphasized was that outreach and marketing are often undervalued by funders. “You can’t build it and assume they will

Strengthening Data Capacity: Best Practices for Collection, Use, and Infrastructure

For the final question of the panel discussion, Walsh invited panelists to share best practices for supporting grantees in handling data, a central cross-cutting theme of the workshop. He asked them to reflect on topics such as identifying common data elements, using data repositories, and building capacity for meaningful use and interpretation.

Walrath opened the discussion by highlighting the value of common data elements across grant programs. She stressed that common data elements do not imply inflexibility, but rather a baseline that supports both local and cross-site evaluation while allowing room for additional, tailored measures. Walrath emphasized the importance of collecting and analyzing implementation data alongside outcomes, explaining that without a clear understanding of what was delivered and at what scale, impact is difficult to interpret. She also shared an example from the GLS program, which uses a centralized data repository that allows all grantees to enter and access data within a shared system. While such infrastructure requires resources, Walrath noted, it offers substantial benefits in terms of consistency, accessibility, and insight.

Crosby expanded on the idea of core data elements, drawing on examples from infectious disease reporting. He pointed to the CDC’s system for weekly notifiable disease surveillance, where each state is required to report on a core data set but may also choose to add context-specific elements, and underscored the value of applying a similar model in suicide prevention—balancing national consistency with local flexibility. Crosby also emphasized the need for community sensitivity in data collection and cited the importance of ensuring that underserved populations are adequately represented. As an example, he noted early gaps in COVID-19 race and ethnicity data that limited the ability to tailor effective interventions. He urged suicide prevention programs to collect complete and representative data to avoid similar blind spots.

Yanovitzky offered three additional considerations. First, he emphasized that data should be actionable; they must be collected and used in support of decision-making and learning. Second, he encouraged programs to expand their definition of impact to include broader elements such as capacity-building, and to engage stakeholders in shaping how that impact is defined and measured. Third, he underscored the opportunity to build data infrastructure across communities and programs. He shared that, in his own work, data repositories are a requirement of foundation-funded grants, and that universities can often support long-term public access. He advocated for open data platforms and cross-state collaboration, noting the potential to create valuable shared resources that do not currently exist at the federal level.

Akinsanya added that both qualitative and quantitative data should be recognized as important. In her program’s federally funded work, they often rely on qualitative data from grantees due to restrictions on other forms of data collection. Grantees have reported that their involvement in evaluation has strengthened community partnerships, increased program credibility, improved services for veterans, and enhanced organizational viability. These outcomes, she emphasized, reflect meaningful benefits that may not be captured through quantitative indicators alone.

Walrath returned to the conversation to emphasize two final points: first, that programs should avoid duplicative data collection by leveraging existing sources when available, and second, that data should only be collected if they will be used. She reiterated that limited resources should be focused on generating useful, actionable insights that inform practice.

Audience Q&A

Walsh offered audience members the opportunity to ask questions of participants from both panels. Topics raised by audience members included clarifying program goals to inform data collection, inclusion criteria for community-based programs, finding funding and collaborative partners, and supporting community engagement in the pre-award phase.

Clarifying Program Goals to Inform Data Collection

Tanha Patel (CDC Foundation) raised a point about the importance of clarifying a program’s intent before establishing data collection protocols. While much of the discussion had centered on common data elements, she emphasized that without a clear understanding of a funder’s goals, it is difficult to determine which data are meaningful. Stout agreed, suggesting that just as grantees are expected to develop logic models, funders should do the same. These models can help define expected short- and intermediate-term outcomes and guide evaluation planning. Farmer added that upfront alignment between data needs at the program and grantee levels can help reduce burden while ensuring data are useful for both local and aggregate evaluation. Yanovitzky echoed the importance of feedback loops in this process. Too often, he said, grantees are treated like contractors, with little opportunity to shape future research agendas. Instead, funders should systematically gather grantee insights—such as through “exit polling”—to inform future funding priorities.

Inclusion Criteria for Community-Based Programs

The conversation then turned to screening tools and eligibility criteria. Michelle Kuntz (U.S. Department of Veterans Affairs Office of Suicide