A Scoping Review of School-Based Efforts to Support Students Who Have Experienced Trauma

The current review sought to describe the published literature relative to addressing trauma in schools. Through a systematic review of peer-reviewed publications as well as gray literature, we identified a total of 91 publications that were coded for study rigor as well as a number of intervention characteristics. Publications included in the review mentioned a variety of intervention components, most notably an emphasis on counseling services, skill development, psychoeducation related to trauma, and parent engagement. We identified a relative lack of empirical evaluation of whole-school approaches and interventions intended to be delivered by non-clinical staff. We also found that less rigorous publications were more likely to highlight the needs of particularly vulnerable groups of youth and to emphasize cultural competence and community engagement in efforts to address trauma in schools. We call for more rigorous evaluation of practices and policies that take a whole-school approach and can be implemented by non-clinical staff. In particular, we highlight the need to evaluate professional development strategies that can help school staff acquire knowledge and skills that can translate into improved outcomes for students—especially students from historically marginalized groups. We also emphasize the importance of ensuring that high-quality research be made accessible to policymakers and school staff to ensure that clear, evidence-based guidance is available to avoid programs, practices, and policies that may inadvertently traumatize students or exacerbate symptoms among students who have already experienced trauma.


Introduction
Over the past decade, increasingly sophisticated research has confirmed that trauma-the long-lasting, adverse response to a physically or emotionally harmful or life-threatening event, series of events, or circumstances-can significantly alter the brain (Anda et al., 2006;Stark et al., 2015). Exposure to traumatic stress in childhood can contribute to mental, emotional, and behavioral challenges including mood disorders and difficulties with self-regulation that can lead to poor academic performance (Terrasi & de Galarce, 2017). At the same time, many jurisdictions across the country have adopted punitive approaches to discipline that do not align with the current research on the effects of trauma and are often disproportionately applied to students of color and students with disabilities (Skiba & Peterson, 2000;Curran, 2016).
As awareness about trauma has grown, researchers, policymakers, and practitioners are paying more attention to understanding the impacts of trauma and increasing the capacity of service systems to adequately address trauma in recent years (Donisch, Bray, & Gewirtz, 2016;Ko et al., 2008). In response to this increasing awareness, nearly a dozen states have passed legislation encouraging or requiring school staff training on the impacts of trauma on students (Chriqui et al., 2019). Yet, despite this growing consensus that school staff play a critical role in supporting students that have had traumatic experiences, there remains a need to identify the most effective strategies for increasing schools' capacity in these efforts.
Much of the existing work around trauma is informed by the US Substance Abuse and Mental Health Services Administration (SAMHSA, 2014). SAMHSA defines individual trauma as "an event, series of events, or set of circumstances that is experienced by an individual as physically or emotionally harmful or life threatening and that has lasting adverse effects on the individual's functioning and mental, physical, social, emotional, or spiritual well-being" (p. 7). Accordingly, SAMHSA describes a trauma-informed program, organization, or system as one that "(1) realizes the widespread impact of trauma and understands potential paths for recovery; (2) recognizes the signs and symptoms of trauma in clients, families, staff, and others involved in the system; (3) responds by fully integrating knowledge about trauma into policies, procedures, and practices; and (4) seeks to actively resist re-traumatization" (p. 9).
The concept of trauma-informed care emerged within health care settings-including physical and behavioral health systems-in the early 2000s (Harris & Fallot, 2001). As the concept gained prominence, other systems including child welfare and juvenile justice began to seek effective ways to integrate trauma-informed approaches into their delivery models (Ko et al., 2008). Education settings are now seeking ways to integrate trauma-informed approaches into their schools and classrooms. However, education settings differ in scope and structure from medical, child welfare, and juvenile justice settings. For example, public schools are intended to educate all children, whereas the other systems typically serve a selected group of youth with identified symptoms or risk factors. Additionally, many medical, child welfare, and juvenile justice professionals receive intensive training in behavioral health topics and typically work to address the needs of individual youth and their families. The majority of school staff, on the other hand, receive limited training in behavioral health topics and are typically charged with addressing the varying needs of a classroom of students (Chafouleas, Johnson, Overstreet, & Santos, 2016). Thus, it is critical to identify and address the gaps in knowledge with respect to effective school-based interventions to address trauma.
Previous reviews of school-based trauma interventions highlight two important gaps in knowledge. First, there is a lack of rigorous evaluation of trauma-informed interventions that are delivered by teachers-even fewer evaluate the effectiveness of training for school personnel (Rolfsnes & Idsoe, 2011;Zakszeski, Ventresco, & Jaffe, 2017). This dearth is particularly striking given evidence suggesting that interventions delivered by teachers can be effective in improving student behavioral health outcomes. For example, a review of 49 studies of mental and behavioral health interventions found that 40% included some involvement of classroom teachers (Franklin, Kim, Ryan, Kelly, & Montgomery, 2012). Notably, the interventions delivered by classroom teachers were just as effective as those delivered by mental health clinicians. Durlak, Weissberg, Dymnicki, Taylor, and Schellinger (2011) conducted a meta-analysis of over 213 school-based, universal social and emotional learning (SEL) programs and found that interventions led by teachers were the most effective for a range of outcomes including behavior, emotional distress, and academic performance. These reviews suggest that non-clinical school staff can also be effective in supporting students experiencing trauma.
Second, many schools have embraced the Multi-Tiered Systems of Support (MTSS) approach to addressing students' needs (Sugai & Horner, 2019). However, efforts to rigorously evaluate trauma-focused interventions across different tiers of support have been uneven and have left schools with little evidence of what works with respect to Universal (Tier 1) interventions. The reviews published by Rolfsnes and Idsoe (2011) and Zakszeski et al. (2017) almost exclusively focused on intensive interventions that would be classified as Targeted (Tier 2) or Intensive (Tier 3) approaches. Somewhat unique to trauma-related interventions, a Targeted or Intensive intervention could be classified as Universal in a post-traumatic event situation. For example, in a school where students have experienced a natural disaster, providing a high-intensity intervention to all students could be classified as Universal implementation, even though the intervention itself would generally be considered Tier 2 or 3. Zakszeski et al. (2017) noted that most of the 15 studies they reviewed that fell into the category of Universal interventions were actually high-intensity interventions being delivered broadly in response to a particular incident. Neither of the previously published reviews we identified included Universal interventions beyond those used to screen students to identify youth in need of more intensive services.
Given the lack of rigorously evaluated studies of schoolbased, trauma-focused interventions that address the role of non-clinical staff, a broader review of relevant publications is warranted. Such a review can serve to characterize the literature from which schools draw to inform their efforts to address trauma and highlight future research priorities.

Objectives of Review
The current review is intended to describe the published literature relative to interventions used to address trauma in schools. Our definition of intervention includes policies (i.e., a set of guidelines intended to address trauma, such as requiring training on trauma), programs (i.e., a structured set of activities to address trauma, such as a staff training on trauma or a set of lessons targeting coping skills), and practices (i.e., a behavior or set of behaviors intended to address trauma, such as avoiding using loud noises or turning off lights to get student attention). Specifically, we sought to answer two questions: (1) What types of school-based, trauma-focused interventions are described in the literature and (2) how does the literature vary across different levels of scientific rigor?
Recognizing that schools access information from a variety of sources, the current review includes peer-reviewed studies as well as publications from reputable sources such as reports from government agencies, research centers, and professional associations (i.e., gray literature). Given the variety of scientific rigor reflected across the literature, we classified publications into five categories based on type and rigor. Publications that reported on an evaluation of the effectiveness of a particular intervention were classified as experimental if they randomized participants to an intervention or control condition and quasi-experimental if they did not randomly assign participants but did make comparisons either between intervention and comparison groups or within the same group of participants before and after exposure to the intervention. The publications classified as systematic reviews met the following criteria: (1) summarized existing research; (2) described their search strategy; (3) searched multiple databases; and (4) screened search results against pre-established inclusion/exclusion criteria. Publications that summarized existing research but did not meet the other three criteria were classified as literature reviews. Finally, we classified publications that focused primarily on describing how an intervention was-or should be-delivered without presenting an evaluation of its effectiveness as descriptive. Publications that described implementation but also reported the effects of the intervention were classified as experimental or quasi-experimental depending on the rigor of the study design.

Article Identification
This paper follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Moher, Liberati, Tetzlaff, & Altman, 2009). See Fig. 1 for the flow diagram. We took a dual approach to identifying relevant peer-reviewed and gray literature publications. Both approaches used similar search strategies, although most gray literature publications did not include an abstract and therefore required a full-text review. Below, we outline search strategies for each approach. Full-text reviews were conducted in an identical manner for both approaches.
To identify peer-reviewed publications, we searched the following bibliographic databases: PubMed, ERIC (EBSCOhost), PsycINFO (ProQuest), PAIS (ProQuest), and Web of Science Core Collection. Searches were conducted for the literature published from January 2010 through July 2018 in order to focus on interventions that reflect current understanding of the neurobiological effects of trauma (Shonkoff, 2016;Thomason & Marusak, 2017). The search strategies for each database included both controlled vocabulary terms and keywords that were based around the three conceptual domains of school-based, trauma, and interventions. In order to make comparisons across different contexts without arbitrarily excluding non-western countries, we limited search results to only countries classified as low risk by the Organization for Economic Cooperation and Development (OECD) including the USA, Canada, European countries, Israel, Japan, South Korea, Chile, Australia, and New Zealand. See [Supplementary File A1] for complete search strategies. We also included peer-reviewed publications that were submitted through a Call for Submissions which was distributed through Child Trends' listserv of nearly 40,000 researchers, practitioners, and policymakers with an interest in research that focuses on the health and well-being of children and youth. Our research team screened the abstracts to ensure they mentioned trauma and included some indication that the publication addressed a school-based intervention. To achieve reliability across reviewers, the team of reviewers initially reviewed the same 30 abstracts and discussed any differences of opinion until all reviewers came to consensus. Subsequently, reviewers flagged any abstract for which they were uncertain for a second review. In the case of a disagreement, a senior member of the team was consulted to make the final decision.
To identify gray literature publications, the research team developed a list of relevant government agencies, research centers, and professional associations, which was reviewed by senior researchers with expertise in trauma-informed interventions. Research assistants then conducted searches of the associated Web sites using the same search terms as described above. Because most publications identified through the Web search did not have abstracts, they were included in the full-text review. We also included gray literature publications that were submitted through a Call for Submissions.
During the full-text review of the 221 publications identified through the database and Web searches described previously, publications were excluded if (a) they did not describe a specific intervention or set of interventions; (b) there was no evidence that the intervention targeted trauma-related outcomes (e.g., psychosocial functioning, staff ability to recognize trauma and make appropriate referrals, or promotion of physical and emotional safety); (c) the intervention targeted preventing a driver of trauma (e.g., bullying, suicide, 1 3 child maltreatment), but not responding to the trauma response; or (d) the target of the intervention did not include students or adults working in a K-12 education setting (e.g., early childhood, out of school, or higher education).

Data Extraction
To describe the school-based, trauma-focused interventions in the literature, we imported all 91 publications into Dedoose, a qualitative software for review (Dedoose, 2018). We also compiled a summary spreadsheet as a double-entry data extraction method. We extracted information for the following characteristics: intervention characteristics, intervention components, facilitator characteristics, study rigor, and outcomes. These data were compiled into a summary table.

Data Coding
To describe the interventions and identify differences across the varying levels of scientific rigor, including a bias assessment for experimental and quasi-experimental studies (Higgins & Green, 2011), the research team developed a standard set of codes and definitions for each code (Table 1). Coded variables were grouped into several categories to describe the contexts within which interventions were delivered, the components of the interventions, the expected outcomes, and the scientific rigor of the studies. Training on coding was conducted by the first author and included a discussion of the codes to be used and definitions of those codes. Codes were developed by the first author in collaboration with experts in childhood trauma and schoolbased wellness interventions. After training, coders each coded the same publication and met to discuss their interpretation of the codes to establish agreement and address any lingering questions about definitions. Coders met twice during the full-text review to discuss coding and resolve any challenges they were having with applying the codes. Coders also conferred with the coding coordinator, who consulted with the first author to resolve any disagreements or address uncertainties in how to apply codes. Additionally, coders all coded the same publication in order to assess the percent of agreement. We calculated an 82% agreement across the coders; 80% and above is commonly considered an acceptable degree of agreement (Belur, Tompson, Thornton, and Simon (2018;p. 7).
In addition to the training and double-coding, the team also employed a double-entry system for coding intervention components whereby coders (a) coded relevant text using Dedoose and (b) completed a spreadsheet where they indicated whether information on a particular variable was present in a given publication. This double-entry system was used to conduct quality control of the data coding. A member of the research team constructed a coding matrix using Dedoose to determine which codes were, or were not, used in coding each publication. The matrix created by Dedoose was then compared to the spreadsheet, and any discrepancies were addressed by a second coder. The second coder conferred with the first author to resolve uncertainties. Variables related to study design, study sample, dosage, setting, and level of rigor were not entered into the spreadsheet but were all double-coded to ensure agreement.

Results
A total of 5841 publications were retrieved from the electronic databases, 181 publications were identified through a search of relevant Web sites, and 24 were identified through a call for submissions. After removing duplicate entries, 4125 publications remained. A total of 3904 records were excluded based on screening criteria, resulting in a total of 221 publications that were maintained for full-text review. During the full-text review, another 130 publications were excluded due to not meeting the study criteria. Ultimately, 91 publications were selected for data extraction. Table 2 provides a brief description of each publication included in the review.
Below, we summarize key findings from these publications. The findings of this review have been organized around publication rigor rather than some other characteristic-such as MTSS tier or intervention setting-in order to highlight the gaps in evidence with respect to programs, policies, and practices that are implemented in schools to address trauma. This strategy allows policymakers and practitioners to better understand the strength of evidence behind these interventions. It also allows researchers to identify interventions that lack rigorous evaluation in order to prioritize studies that will fill identified gaps in evidence when it comes to what is currently being implemented in schools.

Rigor
We classified the scientific rigor of each publication into five categories: experimental studies, quasi-experimental studies, systematic reviews, literature reviews, and descriptive publications. Descriptive studies-the least rigorous of the five categories-were the most common (44%) followed by literature reviews (23%) and quasi-experimental studies (21%); experimental studies (6.5%) and systematic reviews (5.5%) were the least common. Many of the literature and systematic reviews included experimental and quasi-experimental studies that were also included in our review-most often evaluations of Cognitive Behavioral Interventions for Trauma in Schools (CBITS). A structured set of activities to address trauma, such as a staff training on trauma or a set of lessons targeting coping skills Practice A behavior or set of behaviors intended to address trauma, such as avoiding using loud noises or turning off lights to get student attention Intervention Setting Whole school Intervention is delivered throughout the school, such as an assembly Classroom Intervention is delivered in a classroom Non-classroom wellness Intervention is delivered in a non-classroom setting dedicated to wellness promotion, such as a health suite or counselor's office Non-classroom other Intervention is not "whole school" and is not delivered in a classroom or wellness space, such as an intervention delivered at recess on the playground Target Individual Seeks to change student outcomes and is delivered one-on-one Small group Seeks to change student outcomes and is delivered in small groups (generally not in the classroom) Whole classroom Seeks to change student outcomes and is delivered to all students in a classroom together Whole school Seeks to change student outcomes and is delivered to all students in a school together Staff Staff receive the intervention. This could be staff wellness-focused or a training that focuses on staff behaviors Objective Identify The intervention seeks to identify students with exposure to traumatic events or exhibiting symptoms (impaired functioning) Referral The intervention seeks to connect students with unmet clinical needs to school-or community-based clinical services Coping skills The intervention seeks to increase students' positive coping skills School climate The intervention seeks to avoid exposing students to trauma or re-traumatizing students Dosage Number of sessions How many times are students are exposed to the intervention, such as the number of lessons for a program and the number of trainings for staff. For some interventions it may be challenging to assess Frequency How often are students exposed to the intervention. This may be once or annually for some programs, or daily for some practices (such as daily meditation) Approach Universal The intervention is intended to benefit all students Selected The intervention is intended to benefit students at risk for impaired functioning (such as those exposed to trauma or exhibiting some symptoms) Targeted The intervention is intended to benefit students exhibiting impaired functioning related to trauma exposure Sequenced The intervention occurs over time and is intended to meet different needs at different developmental stages Intervention components Assessment The intervention seeks to assess whether a student meets criteria for mental health condition related to trauma (e.g., PTSD, acute stress, depression, anxiety, etc.) Code of conduct The intervention outlines behavioral expectations related to trauma or addresses elements of school discipline Community engagement The intervention involves collaboration or partnership with community members/ organizations, such as community providers of mental health services or other youth/family programs Counseling The intervention provides clinical support in an individual or small group format to allow students to process trauma and gain coping skills Crisis response The intervention includes immediate and short-term responses to take a potentially traumatic event within a school or community crisis (e.g., natural disaster, school shooting) The intervention addresses mindfulness, including things like guided imagery or breath exercises Ongoing implementation support The intervention addresses structural supports such as policy changes, establishment of intervention support teams, or other methods that are intended to achieve sustainable improvements in the way the school supports students that have experienced trauma Parent engagement The intervention seeks to engage parents, such as through parent meetings or homework that students complete with parents Peer support The intervention facilitates youth-led opportunities for youth to support one another. If directed by an adult with mental health training, this would be counseling and not peer support Physical activity The intervention includes some movement component, such as yoga or some physical activity Skills development The intervention provides opportunities to practice skills, not just talking about them Psychoeducation The intervention seeks to increase knowledge related to trauma Referral The intervention seeks to link students with unmet mental health needs to schoolor community-based services Safety enforcement The intervention addresses enforcement of policies to support physical and emotional safety Screening The intervention seeks to identify students with exposure to traumatic events or who may be experiencing trauma symptoms (or both) SEL The intervention focuses on socio-emotional learning: self-management, selfawareness, social awareness, relationship skills, decision-making Staff self-care The intervention addresses staff wellness in the context of supporting students who have experienced trauma, including recognizing and addressing the effects of vicarious trauma Staff training The intervention seeks to increase staff knowledge and skills with respect to recognizing and addressing trauma or taking steps to avoid retraumatization of students that have experienced trauma Technology The intervention either uses technology or addresses it (such as discussing social media and trauma/coping skills) Wellness policy The intervention is reflected in a school wellness policy Facilitator components Role Administrator Intervention is appropriate for implementation by school principal, assistant principal, or other school administrator General staff Intervention is appropriate for implementation by any school staff member Mental health clinician Intervention is appropriate for implementation by school psychologist, social worker, counselor, or other mental health clinician School nurse Intervention is appropriate for implementation by school nurse Teachers Intervention is appropriate for implementation by teachers Other staff Intervention is appropriate for implementation by staff in non-clinical, noninstructional, or non-administrative roles (e.g., coaches, community school coordinator, school resource officer) Training Yes Training is explicitly required to implement this intervention No Training is not explicitly required to implement this intervention Demographics A majority of the experimental studies focused on elementary schools serving mostly students of color, while quasi-experimental studies spanned all grade levels but focused on middle and high schools serving mostly students of color. Review publications of both types (systematic or literature reviews) and descriptive publications either explicitly addressed K-12 or did not specify any particular grade level; they rarely described the racial or ethnic identity of intervention participants. Nearly all mentions of historically marginalized and underserved groups of students, including youth in foster care or experiencing homelessness, children with special health care needs, and youth who identify as LGBTQ, came from descriptive publications with the exception of one literature review that mentioned LGBTQ students (Biegel & Kuehl, 2010) and one that mentioned youth in foster care (Berardi & Morton, 2017).

Intervention Characteristics
Overall, the strength of evidence across intervention characteristics was uneven, favoring Tier 2 programs delivered by mental health professionals in health and wellness settings. Practices and policies-especially those that target whole-school or classroom approaches delivered by teachers or other non-clinical school staff-were most often described in descriptive publications and literature reviews. Both programs and practices were mentioned frequently, although rigorous studies tended to focus on programs while descriptive publications focused more on practices. All three MTSS tiers were also mentioned frequently. Tier 1 interventions were mostly discussed in descriptive publications, while Tier 2 interventions were more likely to be the focus of experimental and quasi-experimental studies. Wholeschool approaches were the most frequently mentioned Bounce Back produced differential treatment effects including improved PTSD symptoms and coping skills but did not produce significant changes in depression or anxiety symptoms or in teacher-reported classroom behavior school staff This report presents a systematic review of interventions that target traumatic stress among children exposed to trauma other than maltreatment or family violence.
Among school-based interventions, the authors found mixed results, although CBT interventions were generally effective This article reviews the literature with respect to opportunities and challenges posed by universal school-based mental health screening. The authors conclude that universal screening in schools holds promise, but that many critical questions remain unanswered  The authors concluded that the screening procedures were successfully implemented with young children using modified administration of instruments This article reports on a mental health screening approach that considers student strengths, in addition to symptoms of distress.
The authors conclude that dualfactor screening can identify strengths as well as unmet needs and may help to reduce stigma This is a tip sheet that provides guidance for educators in how to speak to students about the interplay of race and trauma. The tip sheet reviews trauma, historical trauma, racial trauma, and effects of racial trauma by age-group and outlines practical recommendations for educators setting, but similar to the MTSS tiers, most of those mentions occurred in descriptive publications with fewer than one in five experimental or quasi-experimental studies focusing on whole-school approaches. The most rigorous studies focused almost exclusively on interventions delivered in health and wellness settings such as in a counselor's office. Mental health clinicians and classroom teachers were each described as primary implementers in approximately half of the publications. Notably, mental health clinicians were mentioned across all levels of rigor in this review, while classroom teachers were primarily mentioned in descriptive publications and literature and systematic reviews, but rarely mentioned in experimental or quasi-experimental studies. Aligned with the findings for implementers, training requirements were explicitly mentioned in the majority of experimental and quasi-experimental studies and were not explicitly mentioned in about three quarters of literature and systematic reviews as well as the descriptive publications.
It is important to note that more than half of the review and descriptive publications were classified as "unclear" because the descriptions provided were insufficient to determine training requirements.

Intervention Components
Given the large number of intervention components we examined, we have placed each component into one of five categories based on the consistency and frequency of mentions of that component across our five levels of rigor: (1) common, (2) emerging, (3) bubbling up, (4) research/practice gap, and (5) rarely mentioned. Notably, efforts to train staff and to ensure that trauma-informed approaches are implemented in a way that respects the unique needs of different groups of students-especially those from historically marginalized and underserved communities such as youth who identify as LGBTQ or who are system-involved-were classified as bubbling up because they were prevalent in best practice guides and other descriptive publications but were absent from many of the most rigorous studies. Below, we describe each category in more detail: (1) Common: frequent mentions across all levels of rigor. Components that were classified as common were generally aligned with aspects of well-established clinical interventions for children and youth including counseling; psychoeducation; skill development; and parent engagement. (2) Emerging: consistent mentions across all levels of rigor but with varying frequencies. Emerging components tended to fall into three broad categories: techniques to enhance self-regulation and emotion management; strategies to identify students with unmet behavioral health needs related to trauma; and systematic implementation supports to promote high-quality, sustainable interventions. (3) Bubbling up: frequently mentioned in less rigorous publications with some Components that were rarely mentioned consisted of four broad categories: policies to support students and staff (as opposed to programs or practices, which were more common); leveraging supportive relationships among peers or through mentoring; strategies to support students through physical activities; and use of technology.

Outcomes
We examined mentions of the following outcomes to assess the proportion of publications that provided some level of evidence with respect to them. We focused on the following outcomes of interest to schools and mental health professionals: academics, aggression, appropriate referrals, attendance, coping skills, identification of unmet need, knowledge of trauma, knowledge of trauma-related resources, mental health symptoms, supportive parenting behaviors, participation in clinical services, school climate, supportive staff behaviors, and trauma symptoms. We classified experimental and quasi-experimental studies as mentioning an outcome of interest if they reported it in their results section. Other publications, including descriptive publications, were classified as mentioning one of our outcomes of interest if it either provided direct evidence or included citations to published research relative to that outcome. Our findings are not intended to assess the quality of evidence for these outcomes, but rather to describe the frequency with which these outcomes are mentioned across various levels of rigor within the literature that we reviewed. There was only one outcome that was mentioned in more than half of the publications we reviewed: changes in trauma symptoms. Quasi-experimental and experimental studies were the most likely to mention trauma symptoms, while only one in five descriptive publications mentioned trauma symptoms as an outcome. There were three other outcomes that were mentioned in at least one third of all the publications we reviewed: mental health symptoms, aggressive behaviors, and coping skills. All three of these outcomes were primarily mentioned in experimental and quasi-experimental studies. The most commonly mentioned outcome among descriptive publications was identification of unmet needs which was mentioned in one-quarter of descriptive publications but was not mentioned as an outcome in any of the other publications. Notably, only 13% of the experimental or quasi-experimental studies included academics as an outcome. The least mentioned outcomes included knowledge of the impacts of trauma, appropriate referrals for services, supportive parent behaviors, and knowledge of trauma-related resources, all of which were mentioned as outcomes in fewer than 10% of publications.

Discussion
This review synthesizes 91 different publications describing school-based, trauma-focused interventions from peerreviewed journals as well as reports and guidance documents from government agencies, research centers, and professional associations. Slightly fewer than half of the publications were classified as descriptive with the other half evenly split across empirical evaluations (i.e., experimental and quasi-experimental studies) and reviews (i.e., systematic reviews and literature reviews). The publications we reviewed described interventions that took a school-wide approach or were implemented in classrooms or health and wellness offices and were implemented by mental health professionals, classroom teachers, and other school staff. Publications mentioned a variety of intervention components, most notably an emphasis on counseling services, skill development, psychoeducation related to trauma, and parent engagement. We identified some alignment across the various levels of rigor as well as several gaps, including a lack of empirical evaluation of whole-school approaches and interventions intended to be delivered by non-clinical staff. We also found that less rigorous publications were more likely to highlight the needs of particularly vulnerable groups of youth and to emphasize cultural competence and community engagement in efforts to address trauma in schools. Below, we discuss some of the gaps and opportunities in more detail.

Common Practices
We identified several areas of alignment across all levels of research rigor. For example, our review highlighted consistent and frequent references to several well-supported strategies including the importance of high-quality mental health treatment (Fitzgerald & Cohen, 2012;Overstreet & Mathews, 2011;Plumb, Bush, & Kersevich, 2016), an emphasis on skill development (Dariotis, Mendelson, & Blanchard, 2010;Guarino & Chagnon, 2018b;Ijadi-Maghsoodi et al., 2017), psychoeducation to raise awareness of the effects of trauma (Chafouleas et al., 2016;Jaycox, Kataoka, Stein, Langley, & Wong, 2012;National Child Traumatic Stress Network, 2018), and the value of engaging parents and other caregivers to effectively meet the needs of students (Beehler, Birman, & Campbell, 2012;Langley, Santiago, Rodríguez, & Zelaya, 2013;Santiago, Fuller, Lennon, & Kataoka, 2016). We also found frequent mentions of the potential benefits of integrating trauma-informed approaches into existing MTSS (Evers, n.d.;National Association of School Psychologists, 2017;NCTSN 2017b;Reinbergs & Fefer, 2018), suggesting that schools could leverage processes they currently have in place to identify students in need of additional supports to intentionally address the needs of students who have experienced trauma.
We also identified a handful of trauma-focused programs-including Cognitive Behavioral Interventions for Trauma in Schools (CBITS)  and Trauma-Focused Cognitive Behavioral Therapy (TF-CBT) (Fitzgerald & Cohen, 2012)-that are well supported by a set of rigorous evaluations across multiple studies with diverse groups of students. These interventions were frequently mentioned among publications across all levels of rigor, suggesting that these evidence-based programs have achieved some success at scaling up in schools throughout the nation. In fact, we also identified evidence of the scaling out of CBITS (Aarons, Sklar, Mustanski, Benbow, & Brown, 2017). Scaling out is a process by which well-supported, evidence-based interventions are systematically adapted to new populations or settings. In the case of CBITS, promising adaptations for younger students (i.e., Bounce Back) and for delivery by classroom teachers (Support for Students Exposed to Trauma) are currently being evaluated to ensure greater reach of the program (Langley, Gonzalez, Sugar, Solis, & Jaycox, 2015;Schultz et al., 2010). Of note, many of the evaluations of these programs have been conducted with diverse groups of students, a critical aspect of ensuring evidence-based programs work equally well for all students (Allison & Ferreira, 2017;Goodkind, LaNoue, & Milford, 2010;Schultz et al., 2010).

Gaps in the Literature
While there is much to celebrate, our review highlights the uneven attention given to key aspects of trauma-informed approaches in schools across publications of varying levels of rigor. In several instances, we found that descriptive publications were pushing the field forward, challenging schools to be more intentional in supporting their own staff and ensuring that their efforts are culturally aligned with the needs of the communities they serve through an emphasis on meaningful community engagement (Guarino & Chagnon, 2018c;Langley et al., 2013;Love & Cobb, 2012;NCTSN 2017a;Pickens & Tschopp, 2017). In contrast, few rigorous publications mentioned these topics. Descriptive publications were also much more likely than empirical publications to highlight the unique needs of particularly vulnerable groups of youth. For example, several less rigorous publications highlighted the needs of youth in the foster care system, experiencing homelessness, or who identify as LGBTQ, while none of the empirical publications referenced these groups. Given inequities with respect to exposure to trauma and access to high-quality mental health services-many of which are related to well-documented systemic oppression-it is critical that the research community help schools establish evidence of effective practices and policies in addition to programs so that schools can effectively allocate their resources (López et al., 2017;Marrast, Himmelstein, & Woolhandler, 2016). In addition, in focus groups conducted with school staff after a workshop, Blitz, Anderson, and Saastamoinen (2016) noted that participants rarely mentioned race when discussing trauma experienced by students. Those who did talk about race tended to discount it as contributing to trauma experiences with some white focus group participants noting that they found workshop discussions during a school district training on cultural responsiveness to be offensive. While these findings come from a small study conducted in one elementary school, the themes suggest that schools and school districts would benefit from more evidence-based guidance on how to address the intersection of racism and trauma in schools.
Our review also highlights the imbalance among the empirical publications with respect to the public health approach, a sharp contrast to the literature reviews and descriptive publications we reviewed. For example, while rigorous publications often referenced MTSS, they focused almost exclusively on Tier 2 or Tier 3 interventions, generally only addressing the role of Tier 1 (universal) interventions with respect to strategies for screening students to make appropriate referrals for more intensive treatment. Literature reviews and descriptive publications, on the other hand, often referenced all levels of support (Chafouleas et al., 2016;Evers, n.d.;National Child Traumatic Stress Network, 2017b;Reinbergs & Fefer, 2018). Given that the public health model is reliant on a strong set of universal interventions to reduce the need for more intensive interventions and-especially in the case of schools-to ensure that the benefits of high-quality Tier 2 and Tier 3 interventions are not inadvertently undone by sending students into settings that are ill equipped to provide appropriate supports, it is critical that the research community pay more attention to establishing evidence for effective interventions across all tiers of support. That requires extending the body of research to include more rigorous evaluations of wholeschool approaches and interventions that are intended to be delivered by a range of school staff, including investigations of how to best train non-clinical school staff to implement trauma-informed interventions. In fact, a recent study of an effort to infuse trauma-informed practices within an MTSS approach in a large urban school district highlights several implementation challenges that deserve more rigorous investigation (von der Embse, Rutherford, Mankin, & Jenkins, 2018). It also means expanding the outcomes that are measured to include metrics commonly used by schools including academic performance and attendance as well as referrals to and use of health and mental health services.
While many of the gaps we identified could be classified as examples where the research community needs to catch up with the practice community, there were also a few examples of potential opportunities being missed across all categories of publications in our review. In particular, few publications addressed the role of policies in supporting the types of programs and practices that make a difference for students and staff. This is particularly important given the recent increase in legislative attention to trauma in schools (Chriqui et al., 2019). Another policy lever that was rarely mentioned in descriptive publications and completely absent in empirical publications we reviewed is the role of school wellness policies. As with MTSS, more attention should be paid to ensuring that existing policies are effectively leveraged to support students and staff.

Recommendations for Future Research and Practice
The findings of this scoping review point to a need to better align research and practice. To date, empirical research related to school-based interventions to address trauma have focused almost exclusively on evaluating and scaling up programs, many of which are intended to be implemented by mental health professionals. While it is useful to identify evidence-based programs that can be effectively implemented in schools, it is also critical to examine what works when it comes to practices and policies that can have widereaching effects across an entire school and, in some cases, a school system. In particular, future research should pay more attention to universal supports that can be implemented in schools serving diverse communities with often limited mental health resources. Further, researchers must follow the lead of education professionals and begin to conduct research in a way that allows for comparisons across multiple groups of students to ensure that the benefits of effective interventions accrue to all students. An important step in that process is making sure that marginalized communities have a seat at the table when interventions are being developed and that research studies are being designed to evaluate those interventions.
As we noted in Discussion, more attention should be paid to identifying and disseminating the most effective strategies for identifying students with unmet needs as a result of experiencing trauma, especially when that trauma is ongoing. Several of the publications we reviewed advised caution when considering a large-scale or universal screening to identify students who may be experiencing trauma. With increasing pressures from legislators, as well as mounting evidence that unaddressed trauma can have negative consequences for classroom engagement and learning, schools are looking for answers and the research community should ensure that the limitations of current screening options are widely shared while also partnering with schools and communities to identify valid, reliable, and cost-effective strategies for identifying students in need.
More attention should also be paid to the role that codes of conduct can play in addressing trauma. Codes of conduct were mentioned in several descriptive publications but were entirely absent from the empirical publications we reviewed. It is critical that schools have access to high-quality evidence with respect to ensuring that codes of conduct are enforced in a way that supports physical and emotional safety for all students and avoids retraumatization of students who are currently experiencing trauma. This gap in the literature is particularly concerning given the potential for punitive and exclusionary discipline to exacerbate trauma symptoms among students experiencing trauma. Well-documented inequities in discipline for "acts of defiance" combined with unequal use of exclusionary discipline suggest that schools would benefit from research-based guidance that takes into account the role of codes of conduct in addressing trauma effectively (Okonofua, Walton, & Eberhardt, 2016;Skiba, Michael, Nardo, & Peterson, 2002).
We also found it striking that only one of the publications in our review explicitly mentioned the use of technology. There are a number of ways in which technology could be brought to bear upon the problem of how to address trauma in schools. Virtual reality simulations such as one that was recently released by Kognito (2018) hold promise because they can allow teachers to practice critical skills in a safe and realistic environment. Technology could be developed to help researchers better capture the true effects of professional development for teachers related to addressing trauma. There are many school districts around the country that are spending professional development resources on training teachers about trauma with little evidence to demonstrate whether those trainings actually translate into changed behaviors in the classroom and improved outcomes for students. Technology also holds promise to help schools and communities better assess the level of unmet need. Big data and smart phone applications offer opportunities to collect large amounts of data that can be used to help identify which schools or neighborhoods might be experiencing the most unmet need, allowing policymakers to target resources to the schools that most need them, but also to help them be more intentional about coordinating efforts across various government and nonprofit service providers. There are, of course, serious privacy and ethics concerns when it comes to harnessing these sources of data that must be addressed (Vayena, Blasimme, & Cohen, 2018).
Schools, government agencies, and nonprofit service providers should also begin to identify ways to leverage existing data and partnerships to better understand the needs in their community and to more efficiently monitor the effects of their efforts. Given the nascent stage of the field of schoolbased interventions to address trauma, schools cannot wait for a list of evidence-based interventions that have been proven effective in a population just like theirs. Rather, they must build upon what already exists in intentional ways, leveraging data that they already collect to monitor progress and make course corrections when necessary. This is challenging work that requires trust and collaboration. Researchers must continue to play a role in this process as well. The development of the CBITS program is an example of a university-community partnership that has resulted in an intervention that has ultimately been adapted to meet the needs of many communities beyond the original participating schools.

Limitations
Our review had many strengths, including a more expansive search strategy compared to previously published reviews that included both peer-reviewed and gray literature publications. However, there are also important limitations to note. First, we were necessarily limited by our search criteria. Although we endeavored to include search terms that reflected core concepts in the field, if authors did not use the terms we identified to describe a school-based trauma intervention, their papers would not have been identified. Also, we excluded dissertations and books. For this reason, though many of the publications we identified have also been identified in previously published review articles, it is possible that we missed some potentially relevant publications. Similarly, due to the fact that gray literature is not systematically organized and indexed as are scholarly bibliographic databases, our Web search was necessarily restricted. In order to maximize the comprehensiveness of this search, we identified a number of government agencies, national professional associations that work in education and mental health, and a few wellestablished academic research centers through a process that included review and comment from several experts, including researchers involved with the National Child Traumatic Stress Network. Additionally, while our search was thus restricted, several publications were cross-published across multiple Web sites, suggesting that we have identified many publications that are likely to be familiar to school and mental health stakeholder groups. Another limitation is the nature of the review; due to the fact that we identified 91 relevant publications, we had to find ways to restrict our analyses. It may be that selecting different constructs would have resulted in a different summary of the literature. However, we believe that the constructs we identified present a useful and cohesive review that is well aligned to the needs of researchers, policymakers, and educators. In the end, because we maintained a large number of constructs, we had to rely on a team of reviewers to complete the review in an efficient and timely manner. Having multiple reviewers can introduce the potential for inconsistencies across coders, and it is possible that a different team of researchers could code the data in slightly different ways. However, we believe that our double-coding of a portion of the resources combined with regular check-ins and final decisions being made by the first author allowed us to benefit from the perspectives of researchers from various sectors including education, mental health, and public health as well as library science resulting in a rigorous and practical review of the data.

Conclusion
Schools are increasingly undertaking efforts to better meet the needs of students who are experiencing trauma. There is a large and growing body of evidence with respect to the effects of trauma. There is also a growing list of clinical interventions with proven effects on reducing trauma, including a handful of well-supported school-based interventions. More and more of this research is also exploring the effectiveness of these interventions for diverse groups of youth. Despite these important advances, schools continue to face significant challenges when it comes to identifying and implementing effective programs, policies, and practices across all levels of the public health model. Researchers must pay more attention to practices and policies that support students and staff, including identification of professional development strategies that can help school staff acquire knowledge and skills that can translate into improved outcomes for students. High-quality research must also be made accessible to policymakers and school staff to ensure that clear, evidence-based guidance is available to avoid programs, practices, and policies that may inadvertently traumatize students or exacerbate symptoms among students who have already experienced trauma.