Abstract
Background
Research has revealed persistent disparities in meeting the needs of racially minoritized youth identified with suicide risk in schools. An evidence-based trauma-informed suicide prevention practice, such as SAFETY-A (Safe Alternatives for Teens and Youth-Acute), may reduce unmet need and promote equitable care outcomes. Yet, it is highly challenging to implement practice innovations in under-resourced non-specialty settings, such as schools.
Method
We conducted a multi-level assessment of barriers to implementing SAFETY-A in school districts serving predominantly immigrant families of color. School-based providers (N = 17) and caregivers and students (N = 10) were interviewed about their perceptions of the feasibility and acceptability of SAFETY-A within their school and community context. Immersion and crystallization analytical methods were used to identify implementation barriers that align with the Health Equity Implementation Framework (HEIF).
Results
Eleven themes were identified across multiple determinant levels. Findings show that implementation barriers are interrelated across determinant levels. Major barriers were related to the lack of resources in schools; the historical context of system over-involvement with communities of color that contribute to (l)earned mistrust; and cultural and linguistic differences working with families.
Conclusions
Results aligned with HEIF domains implicated in maintaining health disparities. Implementation strategies for SAFETY-A in schools should be responsive to these determinants of disparities.
Plain Language Summary Title
Community Partner Perceptions of Barriers to Implementing a Strengths-Based, Trauma-Informed Suicide Prevention Intervention in Schools
Implementing a trauma-informed family intervention may help reduce disparities in meeting the needs of racially minoritized youth identified with suicide risk in schools. We interviewed community partners (school-based providers, caregivers, and students) to identify barriers that may affect the implementation of such an intervention, SAFETY-A (Safe Alternatives for Teens and Youth-Acute). Participants reported multiple barriers, most of which were related to the lack of resources in schools; the historical context of system over-involvement with communities of color that contribute to (l)earned mistrust; and cultural and linguistic differences working with families. In order to optimize schools’ ability to sustainably absorb an intervention like SAFETY-A, targeted strategies are needed to address these implementation barriers.
Introduction
Suicide is the third leading cause of death among youth ages 15–24 in the United States (Center of Disease Control, 2024). High rates of suicidal behavior have been observed among youth of color, with alarming recent increases in suicidal thoughts, attempts, and deaths among Latinx, Black, Asian American/Pacific Islander, and Indigenous youth (e.g., between 2019 and 2021 attempts increased by 16%, 17%, and 44% among Latina, Black, and Native girls, respectively; Gaylor et al., 2023; Keum et al., 2024; Meza & Bath, 2021; Sheftall et al., 2022). Furthermore, there are racial disparities in access to care among suicidal youth, with youth of color underutilizing mental health services (Nestor et al., 2016). Schools are well positioned to intervene with suicidal youth from underserved groups (Hall & Nielsen, 2020). State mandates that require schools to respond to suicidal crises (O’Neill et al., 2021) can enhance access to evidence-based interventions (EBIs) and reach youth of color who are more likely to receive care in schools (Cummings et al., 2010; Kataoka et al., 2007). Yet, risk assessment, intervention, and linkage to services present challenges in schools, and racially minoritized students are less likely to receive care following risk detection (Kodish et al., 2020; O’Neill et al., 2021; Yu et al., 2023).
Studies of school-based suicide risk assessment practices suggest common reliance on emergency services (e.g., psychiatric mobile response teams [PMRT], police transports, hospitalization), even when not clinically necessary (O’Neill et al., 2021). Numerous challenges contribute to this practice, including extreme pressures on school personnel to discern which students are at imminent risk in the context of variable mental health training, competing job demands, limited resources and time, difficulty contacting parents, and worries about assuming liability (Brown et al., 2018; Pearlman et al., 2018; Pisani et al., 2011; Yu et al., 2023). Minoritized youth and caregivers often experience school-based suicide risk assessment as intrusive and coercive, feeling they exclude families from decision-making (Kodish et al., 2020). Emergency department (ED) visits and hospitalization for suicidal youth can be traumatic and sometimes ineffective, while incurring high financial costs to families (Huey et al., 2004; Hughes et al., 2017; Lear & Pepper, 2018). These encounters can exacerbate distress (Asarnow et al., 2022), erode family trust of schools and mental health services (Haussman-Stabile et al., 2018; O’Neill et al., 2021), and worsen disparities in follow-up care (O’Neill et al., 2021).
One approach to improve school capacity to respond to suicide risk is to employ an EBI that can help providers (A) assess risk, (B) build hope, (C) connect youth to supports, and (D) develop a plan to maintain safety in the school and home environments, potentially reducing unnecessary emergency service deployments. Safe Alternatives for Teens and Youth-Acute (SAFETY-A, originally known as the Family Intervention for Suicide Prevention) is an example of a developmentally and trauma-informed, strengths-based EBI designed to provide a behavioral assessment of risk, safety planning, and linkage to needed follow-up care among suicidal youth presenting in the ED (Asarnow et al., 2009, 2020, 2022), with adolescents of color showing the greatest benefit (Kodish et al., 2023). SAFETY-A emphasizes work with the caregiver, youth, and family, identifying and coaching caring adults who can provide protective support and monitoring (for detailed description, see Asarnow et al., 2009, 2020, 2022; www.asapnctsn.org). Behavioral risk assessment strategies involve collaborating with the youth and caregiver to: (1) identify strengths in the youth and family; (2) identify patterns of emotional escalation using a feeling thermometer, (3) develop a safety plan; (4) practice and assess commitment to using the safety plan; (5) support the caregiver in protective monitoring and lethal means restriction; and (6) assess and bolster motivation to obtain follow-up care. Components of SAFETY-A are well-suited to address the determinants of disparities in follow-up care, including mistrust and stigma.
SAFETY-A has been deployed in diverse community settings (Esposito-Smythers et al., 2021) including primary care (Pinciotti et al., 2025) and medical units (Hutcherson et al., 2021). To extend implementation to school systems serving minoritized families, an assessment of implementation determinants is needed to promote EBI adoption and sustainment (Bauer et al., 2015; O’Neill et al., 2021). Implementation determinants can occur at multiple levels, including the inner (e.g., school district policies, practices, capacity; school provider characteristics; student and family needs) and outer context (e.g., state mandates, youth serving systems of care, social determinants, and community characteristics; Moullin et al., 2019; Powell & Beidas, 2016).
The current study draws data from the initial phase of a feasibility trial to adapt SAFETY-A for implementation in schools to reduce racial inequities in student service use outcomes following a suicidal crisis. The trial is characterized by three aims: (1) theater testing with stakeholders to develop preliminary intervention adaptations and implementation strategies thought to enhance fit for the context; (2) refining these pre-implementation adaptations and strategies based on an open trial case series with a small number of providers, and (3) conducting a feasibility trial to further iterate on adaptations and strategies across four school districts who received training and implementation support (see Vargas et al., under review for more details). The current manuscript focuses on activities related to Aim 1 (pre-implementation exploration phase; Moullin et al., 2019), in which we assessed anticipated barriers by eliciting participants’ perspectives through individual interviews.
Method
Participants
Participants were recruited through purposive sampling from five Southern California school district partners who primarily served racially minoritized communities (see Table 1). Study participants included school district staff with experience conducting student suicide risk assessments (providers), as well as students and caregivers (see Table 2). Providers were nominated for participation by school district leaders. District leaders and/or providers referred students and caregivers who could serve as informants for our study. Students (11–19 years old) were eligible if they had used school mental health services or were involved in mental health advocacy in their schools. Caregivers were eligible if their child was eligible.
School District Demographics.
Participant Demographics.
Procedures
Participants who provided (parental) consent participated in individual, semi-structured interviews via videoconference to provide their perspectives on the fit of SAFETY-A with their school communities. All participants were presented an animated video storyboard overview of the SAFETY-A process, and videos using actors to demonstrate selected SAFETY-A intervention components (e.g., identifying strengths, lethal means restriction, protective monitoring, safety planning). Providers, caregivers, and students were shown different video clips corresponding to their role. Interview questions were developed by the research team to elicit perceptions about the intervention and general barriers to implementation (e.g., “Meeting with school counselors about student mental health and safety concerns can sometimes feel scary for students and their parents. What could be changed to make this less scary?”). Interviews were conducted by Clinical Psychology graduate students and a Clinical Psychologist postdoctoral trainee who were women of color, and a White male faculty School Psychologist. Members of the interview team shared their relevant background information with participants (e.g., professional background, role in the research project). Interviews lasted approximately 1 h and were transcribed. Participants were compensated with a $25 gift card. Study procedures were approved by the UCLA Institutional Review Board.
Qualitative Coding and Analysis
Interviews were coded and analyzed using an immersion crystallization approach, an inductive and iterative qualitative analytic method of identifying themes and patterns that emerged within each interview and then across all interviews (Borkan, 2022). Immediately following each interview, interviewers wrote synthesis notes to highlight the main points discussed. The coding team was comprised of four coders who were women of color (three of whom conducted interviews); a postdoctoral fellow in Clinical Psychology with specialization in adolescent suicide prevention, two Clinical Psychology PhD students, and a postbaccalaureate Project Coordinator. Coders independently reviewed all interview notes to identify initial codes. Codes were discussed and finalized using an iterative process of reviewing interviewer notes along with interview transcripts. Thematic analyses based on emerging patterns in codes across all interviews were identified through team consensus. Coders frequently shared reflections on their potential biases in interpreting data, including how their personal experience and professional training and experience may impact their interpretations. Given their varying professional stages, coders explicitly discussed power dynamics within the team and frequently engaged in practices to promote participation from all coders (e.g., encouraging more junior members of the team to state their opinions or coding decisions first). This process of independent coding followed by consensus discussion was established to enhance internal validity and rigor.
Coding and analysis followed six stages: initial engagement (during study design for the larger project and the current study), iterative cycles of immersion (beginning with data collection and continued through repeated review of interview transcripts), crystallization (during review of transcripts and interviewer notes to identify patterns and group consensus discussions about themes), creative synthesis (through group consensus discussions to support categorization of findings), corroboration and search for alternative interpretations (through group discussion of codes and generation of themes), and final interpretations (during group consensus discussions of findings until reaching saturation).
The themes were identified through an inductive process. However, the themes below are organized using the Health Equity Implementation Framework (HEIF) to reflect our interest in multi-level implementation determinants: inner context, recipient factors, characteristics of the innovation, and clinical encounters (Woodward et al., 2019, 2021). The HEIF integrates well-established implementation frameworks (e.g., i-PARIHS; Harvey, Gill & Kitson, 2015) with health equity principles to identify implementation determinants that may impact inequities. Equity concerns are proposed to arise across three domains: (1) culturally relevant factors; (2) the clinical encounter (i.e., the school-based suicide risk assessment), where transactions between consumers (i.e., caregivers and students) and providers may influence whether an innovation is accepted by a consumer; and (3) the larger societal context which permeates across levels and is influenced by physical structures, economies, and social and political forces.
Results
Demographics
The study sample included 27 participants, of whom 63% were providers and 37% were school community members (22% caregivers, 15% students). Participant demographics are described in Table 2. Student and caregiver demographics broadly reflect the communities served in the participating school districts.
Qualitative Findings
Participants were asked about anticipated barriers to SAFETY-A implementation based on their initial perception of the intervention. Ten themes were identified and organized based on the HEIF. Themes encompass determinants related to SAFETY-A and routine practices in school-based suicide risk assessment that could influence implementation (see Figure 1). Provider (P), caregiver (C), and student (S) perspectives are denoted.
Themes organized using the Health Equity Implementation Framework.
Outer Context
Providers and caregivers also pointed to
Inner Context
In addition, providers stated that implementing SAFETY-A would place greater responsibility and liability on school staff. They expressed concern that
School Community/Recipients
Providers and students raised concerns about ways in which SAFETY-A may not fit with logistical factors in working with underserved families in schools. Providers anticipated
Characteristics of the Innovation
Nature of the Clinical Encounter Into Which SAFETY-A Would be Integrated
Perhaps because of this perception, caregivers described that
Relatedly, students emphasized that SAFETY-A would require
Discussion
The current study explored providers’, caregivers’, and students’ initial perceptions about potential barriers associated with implementing a trauma-informed, family-based EBI for suicide prevention in under-resourced schools serving racially minoritized families. Our analysis showed that determinants appeared thematically across multiple levels of the HEIF (Woodward et al., 2019), illustrating how structural barriers cascade into organizational practices and ultimately influence clinical encounters where risk assessments take place. This can be observed across three central challenges for the implementation of SAFETY-A: (1) resource constraints in school-based settings; (2) systems’ historical over-involvement with communities of color, leading to (l)earned mistrust; and (3) cultural and linguistic differences across families. These findings align with domains outlined in the HEIF that perpetuate inequities. Although our data collection and analysis were not informed by a pre-existing implementation framework, our themes are organized by the HEIF due to the multi-level nature of our results. Specifically, our results outline how these disparity determinants which are maintained across the inner and outer context likely reflect challenges to implementing any trauma-informed suicide prevention EBI in schools—not just SAFETY-A. Implementation strategies should be responsive to these initial perceptions to promote equity in suicide prevention efforts for underserved youth and their families.
Impact of Resource Constraints in School-Based Settings
Providers described that schools often rely on overburdened systems for crisis response, including PMRT, police, or other emergency services to evaluate, triage, and transport students. Given that emergency services involve significant time delays, providers described a need for quick risk screening to mobilize emergency services given long response times. They expressed confusion about how a strengths-based intervention like SAFETY-A would be integrated into this practice and preferred to deploy emergency services due to concerns about incurring the liability of making risk determination (O’Neill et al., 2021; Yu et al., 2023). While many providers appeared receptive to a therapeutic strengths-based approach, they described disincentives to adopt SAFETY-A in the larger context of the service landscape, the affordances and constraints of schools, and the demands already placed on school personnel. These concerns elaborate on previously documented barriers among school personnel, including limited time and resources, training and concerns about liability (Brown et al., 2018; Pearlman et al., 2018; Pisani et al., 2011; Yu et al., 2023). Tailored implementation strategies are needed to address the outer context factors that may deter school districts from adoption, with attention to ways in which the integration of SAFETY-A in routine risk assessment protocols may require de-implementation of current practices in deploying emergency services. At a minimum, this would require committed leadership support, policy clarification/revision, and organizational messaging about the adoption of a strengths-based therapeutic risk assessment to manage liability.
Conducting Risk Assessments in the Context of Systems’ Historical Over-Involvement With Communities of Color
Participants explained how the impacts of limited resources are compounded by a history of social service systems over-involvement with communities of color, resulting in (l)earned mistrust and racial trauma among families. Providers and caregivers described a broader sociopolitical context in which negative experiences with systems, including law enforcement, immigration, Child Protective Services, and even mental health services, perpetuate mistrust towards systems among families. Communities of color experience disproportionate surveillance and mistreatment by social service systems (Edwards, 2016; Fagan et al., 2016; Merkel-Holguin et al., 2022; Narayan et al., 2024; Sewell et al., 2016) and receive poorer quality mental health care (Cook et al., 2017), due in part to structural racism (Merkel-Holguin et al., 2022; Narayan et al., 2024; Shim, 2021). Participants detailed how these past experiences impact relationships with school providers, who may be viewed as extensions of a coercive system. Furthermore, they described practices that may unintentionally reinforce or perpetuate these perceptions. For example, participants described a practice of schools preferring to use emergency services – including police and school resource officers – to transport students to hospitals, rather than intervening with students on-site (Kodish et al., 2020; Yu et al., 2023). Similarly, they described attempts to compel caregivers to heed the recommendations by suggesting that child protective services may be called.
Yet, ambivalence about obtaining mental health services, doubt and difficulty accepting a youth's suicidal risk are common experiences among caregivers (Reardon et al., 2017). This may be particularly the case among families of color or from immigrant backgrounds with differing cultural explanatory models for emotional problems (Yeh et al., 2005, 2016). While SAFETY-A does not explicitly address cultural factors, the intervention aims to meet families where they are to develop collaboration. This approach may counter ambivalence and promote trust with families. Implementation strategies should consider integrating tailored training to support provider skill development in identifying and responding to ambivalence and practicing cultural humility to yield shared understanding and effective communication during risk assessments.
Providers and caregivers described challenges that showcase how domains of the HEIF may interact, highlighting the clinical encounter as a setting where the broader societal context of marginalization influences provider and consumer interactions (Woodward et al., 2019, 2020). In school-based risk assessments, (l)earned mistrust of systems can make families reluctant to speak openly with school providers. Providers may feel constrained by the lack of time, resources, and training to be able to engage in the necessary relationship-building to earn trust and respond to ambivalence. In response, caregivers may retreat or disengage when they feel confronted or dictated to. Consequently, providers may engage in coercive or re-traumatizing practices such as moving quickly to involuntary hospitalization, ultimately reinforcing mistrust (Kodish et al., 2020). Implementation strategies should support the de-implementation of routine care practices that erode trust and reduce the likelihood of families connecting with mental health services, such as through leadership messaging and tailored didactics about trauma-informed practices within the broader context of racism in systems’ over-involvement with families of color.
Challenges With Conducting Family-Based Risk Assessments in the Context of Cultural and Linguistic Differences
Providers, caregivers, and students expressed ambivalence about the joint family meeting, raising a range of concerns about schools’ preparation to address communication about suicidal risk and enhance emotional support within families. First, providers raised concerns about serving non-English speaking families, indicating how limited interpreter resources in schools contributed to a deferral of care to emergency services which may or may not be better resourced to intervene. Additionally, some participants viewed open communication about youth emotional distress as incompatible with cultural norms in immigrant families. While SAFETY-A does promote family communication, the intervention is not intended to be family therapy. Providers felt unprepared to broach family issues and bridge cultural differences involved in SAFETY-A. Implementation strategies may involve intervention adaptations to guide the provider in navigating acculturation gaps within families in service of increasing protective support. Additionally, implementation strategies could include leadership messaging reinforcing the alignment between SAFETY-A practices and state mandates and district policies to involve parents in suicide risk assessments.
Findings from this study were organized according to the HEIF (Woodward et al., 2019) to shed light on implementation determinants driving disparities. While the HEIF was developed for use in healthcare settings, our study showcases how the model can be extended for use in non-specialty service sectors. Our analysis highlights how school-based risk assessments occur within a broader societal context of structural racism, overtaxed service systems, and pervasive mental health stigma. EBIs like SAFETY-A are being transported to community settings shaped by this larger context, so implementation efforts should be explicitly responsive to these contextual factors to enhance health equity. For example, adaptations to raise awareness of racial trauma and the need to promote trust may increase the chances that providers adopt a strengths-based and trauma-informed approach, to increase the likelihood that immigrant families connect to care (Kodish et al., 2023). Youth and caregivers who experience a positive interaction during a suicide risk assessment may be more likely to seek help. Providers who adopt an EBI that reduces coercive interactions may benefit from greater self-efficacy and job satisfaction.
The current manuscript describes the first part of a larger study informed by aspects of human-centered design (HCD; Lyon et al., 2019). Consistent with HCD, the broader study integrated end-user feedback using theatre testing, rapid qualitative analysis, sequential interviews, focus groups, a case series design, and a feasibility trial to continuously iterate on our SAFETY-A implementation plan. HCD often relies on co-design through power sharing with community partners or participants. In our study, we only sometimes worked with school leaders to respond to specific problems. The implementation team (researchers, clinical trainers, and a school consultant) largely made decisions about how to respond and iterate based on participant feedback.
There were several study limitations. First, our study based in the Greater Los Angeles area may not generalize to school districts in other areas. Second, while efforts were made to recruit students and caregivers, few school districts were able to successfully encourage engagement of school community members in interviews. Low caregiver and student participation likely reflect stigma about discussing mental health topics and negative lived experiences with suicide risk assessments. Third, participants’ views may not be generalizable. Caregiver and student participants were all female and monoracial Black participants were notably absent. While our participants reflected the racial demographic composition of the school districts we surveyed, our results may not generalize to Black participants. As in all interviews, the interviewers’ perceived backgrounds, including professional credentials or experience, may have influenced participant responses. The use of selected video demonstrations may have shaped their responses, given their limited exposure to the intervention. Due to manuscript length, our analysis focused specifically on barriers, but participants were also asked about facilitators. Participants were queried about SAFETY-A and yet raised many issues pertaining to suicide risk assessment practices as usual within schools. Barriers raised were largely applicable to a range of trauma-informed, family centered practices that are not specific to SAFETY-A. Further, some of the barriers described may apply to other care settings. Prior quality improvement studies integrating SAFETY-A into EDs, consultation-liaison teams, and county behavioral health systems have reported provider time as a barrier to implementation (Esposito-Smythers et al., 2021; Giles et al., 2021; Hutcherson et al., 2021), as well as challenges related to organizational rules, and concerns about involving parents due to logistics, language barriers, or reactivity (Esposito-Smythers et al., 2021).
Conclusion
School-based providers, caregivers, and students identified an array of equity determinants that could pose barriers to implementing SAFETY-A in under-resourced schools serving minoritized families. Through the HEIF, determinants at multiple levels were identified which largely converged on three equity domains. Our analysis shows how multi-level determinants contribute to issues present within the clinical encounter where the EBI would be delivered. Notably, several components of SAFETY-A, including the trauma-informed approach, strengths-based family focus, may directly address determinants contributing to inequity in schools (i.e., stigma and mistrust). However, equity-focused implementation strategies are needed to address these barriers and support EBI uptake.
Footnotes
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the National Institute of Mental Health (R34MH126670; PI: Lau). SMV was supported by the National Institute of Mental Health (T32MH073517, PI: McCracken), Ford Foundation Postdoctoral Fellowship, and UCLA Chancellor's Postdoctoral Fellowship Program. SHY was supported by the National Institute of Mental Health (T32MH018261, PI: Pfiffner) at the time this manuscript was accepted. PM was supported by the National Institute of Mental Health (T32MH073517, PI: McCracken).
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.