Lessons Learned About Barriers to Implementing School-Based Interventions for Adolescents: Ideas for Enhancing Future Research and Clinical Projects

Abstract

The majority of youth with mental health problems do not receive treatment, highlighting the critical need to transport evidence-based interventions into community settings, such as schools. Despite being able to reach a large number of adolescents and minority youth, the process of implementing evidence-based interventions to schools is challenging. This paper discusses some expected and unexpected challenges experienced during the implementation of an open trial and a pilot randomized controlled trial examining the acceptability and effectiveness of a school-based preventive intervention for adolescents at risk for internalizing disorders. First, we highlight key programs and findings on preventive interventions for adolescents at risk for depression and anxiety. Next, we provide a brief overview of the preventive intervention we implemented in schools. This provides a context for the section that describes implementation issues and highlights specific challenges and potential solutions for intervention implementation. Finally, the paper offers recommendations for researchers and clinicians interested in implementing school-based mental health services for adolescents.

By age 14, roughly half of all lifetime mental health disorders emerge (e.g., Kessler et al., 2007). Epidemiological surveys show only one fourth to one half of youth with mental health disorders receive professional services (Merikangas et al., 2011), with estimates of as few as 10% of school-age children receiving treatment (Ghandour, Kogan, Blumberg, Jones, & Perrin, 2012). These findings underscore a serious concern, as the majority of youth with mental health problems do not receive treatment. Multiple barriers prevent youths’ access to and utilization of services, including attitudinal (e.g., denial, stigma) and structural barriers (e.g., insufficient transportation, limited insurance coverage; Fontanella, Gupta, Hiance-Steelesmith, & Valentine, 2015; McLoone, Hudson, & Rapee, 2006; Owens et al., 2002). Further, the Centers for Disease Control and Prevention recognize mental health disorders as one of the most costly conditions to treat ($247 billion annually; Perou et al., 2013), highlighting the importance of preventing or treating youths’ mental health problems as early as possible.

For more than a decade, mental health professionals have promoted transporting and disseminating evidence-based interventions in community contexts where they may be accessed and utilized (Schoenwald & Hoagwood, 2001). Nevertheless, challenges remain in bridging the gap between positive outcomes achieved under highly controlled conditions and the typical clinical environment (Burns, Hoagwood, & Mrazek, 1999; Weersing & Weisz, 2002). These processes are complex and differentiation between transportability and dissemination involves consideration of many factors (see Schoenwald & Hoagwood, 2001, for more information). Generally, transportability typically occurs before dissemination and examines the movement of efficacious treatments into usual-care settings. Specifically, this explores which modifications to treatment protocols and practice settings are warranted so that effective treatments can be delivered in real-world settings. Dissemination involves examining the adoption of services as they were originally designed (without modification), but when utilized by the individuals themselves within that setting.

Schools are an important setting for implementing evidence-based services, as they provide access to nearly every child and adolescent (New Freedom Commission on Mental Health, 2003; Weist, Goldstein, Morris, & Bryant, 2003), thereby assisting youth who may not be reached through traditional mental health services. Indeed, of the small percentage of youth who receive services, the majority do so at school (Costello, He, Sampson, Kessler, & Merikangas, 2014). Although there may be variability in schools (i.e., public, private), many school resources are often limited and staff are often unable to meet many pressing mental health needs. Indeed, one fourth of schools nationwide do not have even a part-time counselor, less than two thirds of schools have a school psychologist, and less than half of schools nationwide have a social worker (Allensworth, 2014). For those that have school psychologists, these professionals may be responsible for several schools and manage significant caseloads that can slow the evaluation process and delay the start of services. In some cases, families seek evaluation outside of the school system to expedite the process and many schools have arrangements with community agencies or practitioners to provide services (Brener, Weist, Adelman, Taylor, & Vernon-Smiley, 2007). Furthermore, few services in schools are evidence based (Evans, Koch, Brady, Meszaros, & Sadler, 2012; Kelly et al., 2010).

School-based prevention and intervention strategies may be necessary for vulnerable populations, such as adolescents and minority youth. Specifically, many mental health disorders emerge in adolescence (e.g., Kessler et al., 2007), and minority youth are at increased risk for developing mental health difficulties, as research indicates they have less access to care (Alegria, Vallas, & Pumariega, 2010). Additionally, adolescents with internalizing problems experience difficulties that may be well suited for mental health professionals to address within the school setting. In particular, these adolescents often have challenges with peers, academic difficulties, and familial challenges (see Muroff & Ross, 2011, for a review). Furthermore, up to 30% of adolescents have symptoms that place them at risk for a major depressive episode (Cuijpers, Bolujt, & van Straten, 2008; Rushton, Forcier, & Schecktman, 2002), yet only 18–20% of those who screen positive for mental health problems seek treatment (Chisolm, Klima, Gardner, & Kelleher, 2009; Scott et al., 2009).

Even adolescents who have been identified as needing mental health services (e.g., those at risk for suicide) may not obtain needed services. One study showed that parents of adolescents identified as at risk for suicide reported they would take their adolescent for follow-up care, yet a third of parents did not follow through (Kataoka, Stein, Nadeem, & Wong, 2007). This same study found Latino adolescents were less likely to receive community-based mental health services than non-Latino youth—however, there were no ethnic/racial differences in adolescents’ rates of receiving school-based mental health services. Thus, addressing internalizing problems among adolescents within schools appears desirable, particularly for Latino adolescents.

Given the extent of unmet mental health needs among youth (especially adolescents and minorities) and the advantages of offering services in schools, implementing evidence-based interventions in school settings can have a sizeable public health impact. Nevertheless, this process can be challenging, as it involves identifying relevant interventions and collaborating with community partners to adapt interventions to fit the setting. The purpose of this article is to describe some of the lessons we learned from an open trial (OT) and a pilot randomized controlled trial (RCT) that involved implementing a school-based intervention¹ for adolescents at risk for internalizing disorders. Below, we briefly highlight programs and findings on preventive interventions for adolescents at risk for depression and anxiety. Next, we describe our school-based intervention project to provide a context for the challenges and barriers we experienced. We then turn to implementation issues, addressing specific challenges and potential solutions for future intervention implementation projects. Although our experiences involved transporting prevention interventions into schools, our description includes considerations that can be valuable for both researchers and clinicians involved in implementing, transporting, or disseminating school-based mental health services for adolescents.

School-Based Preventive Interventions for Adolescents at Risk for Internalizing Disorders

Meta-analyses offer support for the efficacy of depression prevention programs for children and adolescents, although effect sizes have been small (Sandler et al., 2014; Stice, Rhode, Gau, & Shaw, 2009). Recently, Brunwasser and Garber (2016) reviewed 37 studies to examine the efficacy, effectiveness, and readiness for dissemination of prevention programs for depression for youth. Consistent with the meta-analyses, the review showed that some prevention programs demonstrated efficacy (i.e., benefit under controlled circumstances)—however, no program demonstrated sufficient evidence of effectiveness within real-world settings to warrant widespread dissemination. Although discouraging, this review included studies that were diverse across a wide range of factors. For example, the majority of the programs were developed and studied internationally, included youth across a broad age range, and were implemented in a number of settings.

Within school settings across the United States, prevention programs for adolescents at risk for depression primarily have been cognitive-behavioral approaches or interpersonally focused. For example, the Coping With Stress course (Clarke & Lewinsohn, 1995) emphasizes monitoring mood, identifying activating events, changing behaviors, and changing beliefs. Another exemplary program, Interpersonal Psychotherapy—Adolescent Skills Training (IPT-AST; Young, Mufson, & Schueler, 2016), focuses on communication and the social skills necessary for reducing conflict and developing and maintaining positive relationships. Although both interventions have effectively lowered depressive symptomatology, these programs also have demonstrated success beyond efficacy. For example, adolescents who participated in the Coping With Stress course showed significant intervention effects even when implemented by school staff who were independent from the program developers (Clarke & Lewinsohn, 1995). In addition, adolescents receiving IPT-AST demonstrated significantly greater improvements in self-reported depression symptoms as well as evaluator-rated functioning relative to adolescents in rigorous control groups receiving group counseling delivered by school staff. Further, adolescents in IPT-AST continued to show improvements at 6-month follow-up (Young, Benas, et al., 2016). In sum, there is evidence to support that depression prevention programs can be effectively transported to school settings.

In terms of anxiety prevention, research has focused on a broad range of symptoms in children and early adolescents (e.g., Barrett, Farrell, Ollendick, & Dadds, 2006; Essau, Conradt, Sasagawa, & Ollendick, 2012). However, meta-analytic findings reveal only modest improvements in anxiety-specific symptoms following universal prevention programs (Ahlen, Lenhard, & Ghaderi, 2015). In terms of social anxiety in particular, no evidence-based interventions are available for adolescents ages 14–18 years. In fact, only one prevention study focused on social anxiety, and this was conducted with children and early adolescents (11–14 years; Ahlen et al., 2015; Aune & Stiles, 2009). Thus, especially in the area of social anxiety prevention, no effective programs for older adolescents have been identified yet.

In terms of the joint consideration of internalizing problems, a systematic review of 28 studies examined school-based prevention programs for depression and anxiety in adolescence (Corrieri et al., 2013). Results showed the majority of interventions were effective—however, the overall mean effect sizes were small, even for studies that included youth with a clinical disorder. The findings fit with a recent review of the joint efficacy of universal, selective, and indicated prevention programs for anxiety and depression combined, which only obtained evidence for short-term intervention effects (Stockings et al., 2016).

Findings such as these suggest that preventive interventions can be transported to school settings to reduce adolescents’ internalizing symptomatology, yet such programs are still limited. Researchers and clinicians who wish to address this gap often must overcome incredible challenges in program implementation. Across studies, the primary barriers reported include lack of support from school administration, staff, and parents (e.g., Masia Warner & Fox, 2014); limited recruitment period due to the academic calendar; school hours that may prevent parents from participating due to competing work demands; and competing demands of adolescents’ after-school schedules (e.g., work, home responsibilities; Mufson, 2010).

The existing literature describes these challenges broadly, but lacks detailed explanations of how these barriers manifest on a regular basis. Thus, informing the research and practice communities of the challenges that arise may help those trying to implement or transport evidence-based interventions to adolescents in schools. This paper describes the key barriers we encountered and potential solutions for future studies that were derived from an OT and a pilot RCT. In those studies, we implemented an evidence-based preventive intervention (PI) for high school students (La Greca, Ehrenreich-May, Mufson, & Chan, 2016a; La Greca, Mufson, Ehrenreich-May, Girio-Herrera, Ehrlich, & Cicila, 2016b; La Greca, Mufson, Ehrenreich-May, & Ehrlich, 2018) that occurred within the school context, and focused on adolescents from predominantly underserved, low- to middle income Latino backgrounds.

Description of the Intervention Project and School Setting

Our goal was to develop a PI for high school students at risk for social anxiety or depression who also experienced high levels of interpersonal peer victimization (IPV); we evaluated the intervention’s feasibility, acceptability, and preliminary benefit. Our goal was to assist adolescents who were the targets of peer victimization, a key risk factor for social anxiety and depression (e.g., La Greca & Harrison, 2005). Further, few antibullying programs target high school youth, and no social anxiety prevention programs specifically address this age group. Given the shared interpersonal nature of peer victimization, social anxiety, and depression, we based our PI on IPT-AST (Young, Mufson, & Gallop, 2010) for preventing adolescent depression, which was modified to additionally target social anxiety and peer relationship problems (see La Greca et al., 2016a, 2016b, 2018).

Study Design

First, we conducted a small OT in two schools (seven adolescents at each school) over the course of a spring semester, using a baseline to postintervention study design. Then, the following spring, we conducted a pilot RCT in four schools with a total of 49 adolescents randomly assigned (within school) to either the PI or an active education/support (ES) group. We used an active comparison condition to make the intervention project more acceptable to schools. RCT participants were evaluated at baseline, postintervention, and 6-month follow-up (i.e., the fall of the following school year). OT and RCT procedures were very similar and are described below.

Engaging Schools and Planning Logistics

The four schools participating in the OT or RCT were large (i.e., range of 1,700–3,500 students enrolled) urban/suburban high schools from a large public school district in the Southeast. The schools enrolled youth from mainly low-income backgrounds (70–90% qualified for free or reduced cost lunch at 3 schools; 17% at 1 school) with a high representation of minority youth (91–96% across the 4 schools) from predominantly Latino/Hispanic backgrounds (about 90%). The schools indicated that they had limited mental health services available for students. Many schools in the district previously employed social workers to assist with students’ mental health needs—however, funding cuts eliminated these positions prior to our project. Additionally, school counselors reported that the majority of their time was devoted to academic advising and scheduling.

Typically, collaboration with schools involves partnership and communication that begins well before the start of an intervention. Although our recruitment began in early fall, we gauged interest from school administrators in the academic year and summer before the project began. In July and August, we met with key administrators (e.g., principal, head of counseling) in each school to explain the project, provide written materials about the intervention and issues surrounding confidentiality, and enlist their participation. Each school designated a school liaison (typically a counselor or assistant principal interested in the project and knowledgeable about school policies and procedures) to serve as our contact person for the logistics of planning and implementing the intervention. Correspondingly, we assigned a staff member to serve as our school coordinator (one per school). We cultivated relationships with the school liaisons by strengthening interpersonal relationships and occasionally providing small tokens of appreciation, like food, plants, and/or personalized notes.

The first task coordinated by the liaisons and coordinators were focus groups that occurred the first few weeks of the school year (details below). We gathered information about available resources and potential challenges, as well as how to present the program to adolescents within these schools and recruit their participation. Using this information as a guide, the liaison and coordinator together developed a school-specific plan for informing teachers and students about the intervention and implementing the project logistics (e.g., when to send letters and consent forms home; when and where to schedule screenings, individual assessments, and intervention sessions). The school liaisons preferred to communicate directly with teachers about project details; our school coordinators were available, as needed, to assist in this process. The remainder of the fall semester was devoted to recruiting adolescents, screening potential participants, and determining initial eligibility; the spring semester focused on delivering the intervention, as well as conducting the baseline and postintervention assessments.

Focus Groups

We conducted focus groups across two schools (two comprising school principals, counselors, and teachers; two comprising adolescents). In general, we were interested in learning about available resources and potential challenges, as well as how to present the program to adolescents. Specifically, we solicited ideas for a name for the intervention, feedback on flyers and recruitment materials, strategies for stigma reduction and encouragement of participation, and logistics of program scheduling. Adolescents strongly emphasized that our program should have a positive tone and highlighted the importance of approaching parents with sensitivity. Many thought some parents might be wary of mental health interventions and might not support their adolescent attending such a group. These concerns helped to shape the way we communicated with families, school personnel, and participating adolescents.

Recruitment, Screening, and Initial Inclusion Exclusion Criteria

Because the intervention targeted “at-risk” youth, adolescents were invited to participate in a screening to determine eligibility. Flyers were distributed in participating schools that described the project positively as a “way to improve your social life and peer relationships.” Recruitment materials emphasized that the program was designed to enhance social skills.

After obtaining active parental consent (via letters and consent forms that adolescents brought home and returned to school) and adolescent written assent (obtained in person) from interested youth, we screened 108 adolescents for the OT and 283 for the RCT using standardized questionnaires that assessed social anxiety, depressive symptoms, and IPV. Adolescents in grades 9–11 were initially eligible if they had (a) clinical elevations in either social anxiety or depressive symptoms, and (b) elevations in IPV as reflected in high scores for either relational or reputational victimization (see La Greca et al., 2016a, 2016b, 2018), for details). We also evaluated whether adolescents reported elevated peer aggression or overt/physical victimization, as these were exclusionary criteria. We did not want to include aggressive youth (i.e., potential bullies) in the intervention, and we referred youth who were overtly victimized to the school counselor so they would receive immediate help (consistent with school policies on bullying).

Rollout and Consenting Procedures

Adolescents who were eligible based on screening participated in a diagnostic interview to rule out the presence of a clinical disorder or suicidal risk (exclusionary criteria), and completed questionnaires to establish a baseline. The interview and questionnaires were readministered at postintervention and at follow-up (for the RCT).

Informed parental consent and adolescent assent were required prior to the baseline assessment and intervention participation. Most adolescents who were initially eligible (based on screening) agreed to be evaluated; a few declined, saying they were no longer interested. Across the OT and RCT, about half of those who completed baseline remained eligible and enrolled in the intervention project. The remaining youth were excluded due to the presence of a clinical disorder; their parents were provided with referral information for clinical services (see La Greca et al., 2016a, 2018, for details).

In all, 14 adolescents enrolled in the OT and received the PI; 49 adolescents enrolled in the RCT and were randomly assigned to the PI or ES group. Across the OT and RCT, 73% of the adolescents were girls and about 89% were from Latino backgrounds; they ranged in age from 14 to 18 years, with a mean age of about 15 years.

Intervention

Both the PI and ES groups involved 2 initial and 1 midgroup individual sessions (45 minutes each), and 10 weekly 90-minute group sessions (5–8 adolescents each, with 2 group leaders/clinicians). Based on the strong preferences of adolescents and school personnel, the intervention groups took place during the regular school day. The specific content of the PI and ES groups has been described in detail elsewhere (La Greca et al., 2018).

Outcome Measures

Across the OT and RCT, the primary outcomes included (a) independent evaluator (IE) ratings of clinical severity (for subclinical depression or social anxiety) on the Anxiety Disorders Interview Schedule for DSM-IV—Child Version (ADIS-C-IV; Silverman & Albano, 1996; Wood, Piacentini, Bergman, McCracken, & Barrios, 2002), (b) IE ratings of severity on the Clinical Global Impressions Scale (Guy, 1976), and (c) adolescents’ reports of relational and reputational peer victimization on the Revised Peer Experiences Questionnaire (R-PEQ; De Los Reyes & Prinstein, 2004). Secondary outcomes included adolescents’ reports of (a) social anxiety on the Social Anxiety Scale for Adolescents (La Greca & Lopez, 1998), (b) depressive symptoms on the Center for Epidemiological Studies Depression Scale (Radloff, 1977), and (c) social support on the Friends subscale of the Perceived Social Support Scale (Procidano & Heller, 1983; see La Greca et al., 2016a, 2018), for descriptions of the measures). Finally, because we were interested in the intervention’s acceptability, participating adolescents also completed a Feedback Survey at postintervention, similar to the Treatment Satisfaction Questionnaire developed for the IPT-AST (Young, Mufson, et al., 2016). Five items assessed the quality, helpfulness, satisfaction, and effectiveness of the intervention rated on a 4- or 5-point scale (e.g., very satisfied, mostly satisfied, indifferent, mildly dissatisfied, quite dissatisfied).

Clinicians and IEs

Across the OT and RCT, we trained 12 different clinicians (2 postdoctoral fellows and 10 advanced Ph.D. students in clinical psychology). Each group had two clinicians; at least one had a master’s or doctoral degree. Clinicians received a full day of training, plus weekly group supervision (from one of the project leaders; see La Greca et al., 2018, for details). The IEs were graduate students in clinical psychology who were trained in the administration of the ADIS-C-IV. The IEs administered approximately four interviews at any given intervention phase (baseline, postintervention, follow-up); different IEs were available across the intervention phases.

Barriers to Treatment and Other Qualitative Data

Because this was an intervention development project, we also gathered data on “barriers to treatment” from nine adolescents in the RCT who dropped out before completing the intervention (see Table 1). All who dropped out did so due to scheduling concerns; none were concerned about potential reactions from school staff, friends, or other peers regarding participation in the program. Additionally, three adolescents dropped out prior to completing the OT and reported informally to team members that they were reluctant to miss class. Also, at postintervention, we interviewed our four school liaisons regarding their perceptions of program acceptability, feasibility, and sustainability. Last, at postintervention, our clinicians provided informal input on the intervention and its implementation through an anonymous survey. In our discussion of key implementation issues (next section), we provide quotes from these interviews.

Table 1.

Barriers to Program Participation for Adolescents Who Dropped Out From the RCT (n = 9)

Top barriers (50% or more endorsed)	% endorsed
I found it hard to participate in the program because …	Yes or a little
I did not want to miss class.	100.0%
I was concerned that attending the program would affect my grades.	77.8%
I lost interest in participating in the program.	77.7%
I did not think it would be helpful to me.	66.6%
I was involved in other activities (e.g., football, gym, band).	55.5%
I did not think I would be comfortable with other students in the group.	55.5%
The meetings lasted too long (90 minutes).	55.5%
Potential concerns that were not barriers (less than 25% endorsed)
I found it hard to participate because …
Other students were not supportive or kind when they saw I was in the program.	0%
I did not feel comfortable with program staff.	0%
My friends were not supportive of me participating in the program.	0%
I did not want other students to know I was participating in the program.	0%
I did not want school staff to know I was participating in the program.	0%
I thought I might be made fun of for participating in the program.	0%
I did not think I would be comfortable with the group leaders.	11.1%
The program involved discussing topics that I do not want to discuss.	22.2%

Note. RCT = randomized controlled trial.

Logistic Challenges That Affect Implementation

A. Scheduling of Sessions

What We Did

Individual sessions were conducted during the school day based on clinician and adolescent schedules. Weekly group sessions occurred at a fixed day and time each week for all but one school (where sessions alternated days of the week, consistent with the school’s policy of alternating classes every other day). Overall, our scheduling was effective, in that most of the enrolled adolescents attended the groups and completed the intervention. For example, mean attendance for the OT was 11.8 out of 13 sessions (90.0%) and 10.0 out of 13 sessions (74.8%) for the RCT. However, the group sessions were challenging to schedule (see Table 2 for a summary of this and other challenges, as well as possible solutions).

Table 2.

Summary of Challenges and Potential Approaches

Challenge	What we did	Advantages	Limitations	Recommendations
Scheduling group or individual meetings	Scheduled sessions during school day (based on feedback from schools and students)	• More inclusive and accessible • No transportation barriers • Does not conflict with after-school commitments • Does not require supervision from school staff after hours	• Missing class is major barrier for students and teachers • May conflict with exams and state-mandated testing • May be more difficult to find an available room	• Schedule during elective classes • Consider scheduling during lunch (and providing food) • May need to schedule on different days of the week accommodate testing or minimize specific classes missed • Limit the number of sessions missed for a particular class
Limited space available	Scouted rooms on weekly basis	• Makes it possible to meet during school day, even when a consistent room is not available	• Causes confusion about location for students • Clinicians may have to track down students individually	• Use a sign in an agreed-upon location to designate the meeting room • For schools where space is barrier, consider scheduling after school instead
Communication with parents	Sent detailed information to parents about the study; sent information on assessment results	• Informs parents of the study procedures • Alerts parents that child reported elevated levels of distress • Provides referrals for additional intervention • Meets ethical standards	• Some parents may be upset to learn that their child may be distressed • Some parents may be dismissive of concerns or suspicious of intervention	• Use multiple communications to prepare parents for receiving feedback about their child • Use neutral and nonstigmatizing language when referring to child’s distress • Use a liaison to address cultural barriers and improve communication
Balancing confidentiality and safety concerns	Prepared staff for issues of confidentiality; did not share information about child’s functioning with school, unless a safety concern arose	• Protects child’s privacy • Meets ethical and legal requirements • Allows child/family to be more comfortable with sharing personal information with clinicians	• School staff may not understand why such information cannot be shared • School staff may feel ill prepared to handle issues that arise unexpectedly	• Periodically discuss and reinforce with school staff ethical/legal issues of confidentiality • Develop a clear policy about confidentiality/student safety issues and share with all staff

Rationale and Advantages for Using Our Approach

We decided to implement our program during the school day, based on focus group feedback from adolescents and school personnel. The intervention would be more inclusive if it could fit within the regular school day, as transportation issues and after-school commitments would preclude many students from participating. However, daytime scheduling presented a challenge for group sessions as we needed to accommodate clinician schedules along with multiple students with different class schedules. Thus, we developed several guidelines for scheduling sessions.

Our first priority was to minimize the amount of academic instruction adolescents would miss; we did this by scheduling groups during elective classes (rather than core academic subjects), to the extent possible. Prioritizing this goal often required negotiation between school personnel, teachers, adolescents, and clinicians.

Next, we adjusted our scheduling around state-mandated achievement testing and test preparation. One counselor alerted us that “students are almost always testing, so it is a challenge.” We worked with school counselors and liaisons to prevent adolescents from missing standardized tests. However, near testing time, some teachers were reluctant to allow students to attend sessions due to concerns about how missing class would affect exam performance. Overall, testing schedules affected some group sessions, whereas individual meetings were more flexibly completed.

We also considered scheduling group sessions during lunchtime. However, in some schools, lunch periods occurred throughout the late morning and early afternoon—thus, finding a mutual lunchtime for all group members also proved challenging. When groups were scheduled during lunch, we provided food for the students. Although this offered a solution to conflicting schedules, it added the cost and coordination of ordering, transporting, and providing food.

Alternative Approaches and Recommendations

One possible solution to scheduling barriers would be to conduct an intervention before or after school. All participants from the RCT who dropped out from intervention reported concerns about missing class (see Table 1). Further, in a feedback survey following intervention, 78% of clinicians reported it would preferable to conduct the intervention after school. Although potentially desirable, an after-school option would also be challenging, as schools require a teacher or other school personnel to be present during after-school activities (which also might require overtime pay for extra school personnel). Interventions conducted outside of school hours also present transportation issues for participants, or interfere with other after-school activities or work responsibilities. If interventions are scheduled after school, it may be worth inquiring about whether schools offer a late bus for those with after-school activities. Alternatively, funded projects might budget for student transportation or school personnel overtime, thus bypassing potential obstacles to after-school scheduling. It is possible that a combination of during and after-school scheduling could address some of the challenges described above.

Another potential scheduling solution may be to fit the intervention into a standard class period. Clinical interventions that do not require randomization of participants to conditions naturally offer more flexibility for scheduling, especially for group sessions. For research, investigators might consider a quasi-experimental approach, with individuals randomized by school (rather than within schools), which would enhance scheduling flexibility. Although these recommendations apply most directly to interventions containing a group component, any intervention offered on a regular basis is likely to encounter some scheduling challenges.

In general, we recommend conducting focus groups and working closely with school liaisons, administrators, teachers, and adolescents to identify the best overall option for a particular school. Importantly, this is a decision that involves coordinating with teachers, both early in the planning process and throughout implementation, and facilitating their awareness of relevant events that could present obstacles along the way.

B. Space for Sessions

Interventions require space for sessions to be held in privacy. In our experience, the ease of finding an available, acceptable room to conduct confidential intervention sessions was highly variable across schools. For some schools, we negotiated in advance with school personnel to secure a room that remained available throughout the intervention program. In other schools, we had to identify an intervention room on a week-by-week basis, due to space constraints. This was a common challenge for our team (see Table 2), as well as for school staff. As one counselor sympathetically observed, we “have to give standardized exams in hallways or outside at picnic tables or in the fields because there is no space [in this school].”

What We Did

When consistent space was not available, we developed creative solutions. For example, some clinicians scouted out available rooms on a weekly basis and worked with school staff and security guards to get rooms unlocked prior to the intervention’s start. A complication, however, was that the intervention participants did not know in advance where the sessions would be held, and as a result, clinicians had to “physically track down most kids to bring them to each session.” This became a weekly challenge for the clinicians, especially because the schools were large and students freely moved around the building during the day. On a few occasions, the time involved in identifying space and tracking down students reduced the time available for intervention.

Space sometimes was an issue even at schools that had a consistent intervention room available. At one school, the intervention sessions took place in a department lunchroom. While this provided a consistent, private location, it displaced school personnel during their lunch breaks on a weekly basis. To ensure the privacy of the sessions, we put signs on the door requesting privacy, which decreased the number of interruptions. However, on occasion, there were brief interruptions when personnel opened the door before noticing the sign indicating that a group was in progress. Even with the challenges and constraints noted above, we were successful at maintaining privacy for the intervention. In our clinician survey, 89% of clinicians reported that sessions were never interrupted by someone outside of the group.

Alternative Approaches and Recommendations

Solutions for obtaining an appropriate space varied widely by school and by when the intervention was scheduled. It is important to consider the above recommendations regarding scheduling to solve space issues. For example, clinicians will likely find ample, private, and uninterrupted space for group interventions when they are scheduled after school. Also, grouping students within a school based on their schedule may also ease the challenges of securing appropriate space, although this may not be possible for research-based interventions. For schools unable to provide a consistent meeting room during the school day, another possibility would be to use a sign at an agreed-upon location to designate the meeting room each week. In our experience, this would have been challenging in the large schools involved in our program, as there were many school buildings and it took a great deal of time to travel around the school. However, this approach may work well in a smaller school with an accessible, centralized main office. The success of this approach also may depend on whether students are responsible and able to remember to check the meeting room sign on a regular basis. Overall, our team members engaged in flexible and creative problem solving to determine the best solution within the context of their specific school, given the limitations of an RCT.

C. Communication With Parents

What We Did

An important project activity that also posed challenges involved parent communication (see Table 2). At the project outset, we sent information and consent forms to parents and made our team available to answer any questions. Because of our predominantly Latino population, all written materials were provided in English and Spanish and several bilingual staff were available to communicate with parents in their preferred language. Following our screening for intervention eligibility, we also communicated with parents by phone or in writing about their child’s continued eligibility, and directly contacted parents if their child reported experiencing symptoms of social anxiety or depression, suicidality, or met criteria for a clinical diagnosis.

Research Considerations

Although some of the challenges we encountered were similar to those experienced by clinicians who treat adolescents, others were related to conducting research within schools. For example, some parents did not fully complete informed consents (e.g., did not sign forms completely or left critical information blank). This resulted in considerable follow-up work by our team, including considerations surrounding the Family Educational Rights and Privacy Act of 1974 (FERPA), a federal law that protects the privacy of student education records. Under this law, school staff are afforded more flexibility with one another in their communication about students. However, clinicians who are part of a research team or who are employed by entities other than that school are not privy to the same flexible communication allowed by school personnel. Therefore, we had to rely upon school personnel to approach the adolescents and seek their permission for us to speak to them directly, to secure corrected and completed consents.

As another example, some parents were “caught off guard” by our letters informing them that their child reported symptoms of social anxiety or depression. Consistent with ethical procedures for screening at-risk students, and as explained in our recruitment letters and consent forms, we told parents we would notify them in writing if their child reported elevated levels of social anxiety or depression. Although this notification procedure generally went very well, a few parents were surprised and contacted the school directly (rather than the research team) for additional information. School staff were not privy to this information and thus were unable to assist parents. In addition, we were unable to reach a few parents (by letter or phone) as they had relocated or changed their phone service. When we did reach parents, most were understandably concerned to learn their child had reported elevated levels of social anxiety or depression. However, a few parents were wary of psychological assessment, did not believe that their child reported distress symptoms, or acknowledged but minimized their child’s difficulties.

Clinical Considerations

Although some of the parent communication challenges are unique to a research study, some issues are pertinent even in nonresearch clinical situations. In particular, parents’ hesitations to seek help or continue to seek help was a theme our staff witnessed when working in the schools. In particular, the role of culture is an important consideration in terms of communicating with parents and understanding their receptivity to psychological intervention for their child. Evidence suggests that individuals of Latino backgrounds are less likely to utilize mental health services than their non-Latino counterparts (e.g., Alegria et al., 2007). Thus, it may not be surprising that even after receiving a detailed letter and engaging in a phone conversation about their child’s symptoms of anxiety and depression, some parents insisted their teen’s emotional or behavioral functioning did not seem unusual or warrant services. Although the majority of parents approved of their adolescent’s involvement, about 20% were no longer interested after being identified as eligible.

Recommendations

Interventionists in schools must carefully navigate the expectations of confidentiality between participants and parents. This is especially true during the assessment process, when participants may report distress or suicidal thoughts. Our experience highlighted the importance of alerting parents and clearly informing them about the nature of the written or verbal communication they might receive. Although we informed parents that we would contact them if their child was distressed, we believe that our procedures could be improved and that others might benefit from suggested adaptations moving forward.

Specifically, additional efforts might be useful to ensure that parents were well-informed of procedures ahead of time and would not be surprised by receiving personal feedback about their child. For example, one might send a “thank-you” note to parents of enrolled participants expressing gratitude for their participation and remind parents that they will be recontacted if the adolescent appears to need more support. Another potential solution is to include a bolded, bulleted list within the consent form to visually highlight important points regarding results from intervention-related assessments. Alternatively, a “summary sheet” could be sent to parents along with the consent form that reinforces the timeline of events and potential future communications. It is important to note that for researchers, most Internal Review Boards have very strict guidelines pertaining to the information included in a consent form, but we urge researchers to explore these options, especially as communication to parents about study-related assessments concerns the welfare of the research participants.

Further, it may be beneficial to use very neutral language to inform parents of the results of assessments. For example, parents may respond more favorably to news that their adolescent reported “elevated levels of distress” than that they “may be experiencing symptoms related to social anxiety or depression.” Of course, the referral information may be the same, but parents may be more likely to seek additional diagnostic information on their own if language is neutral and nonstigmatizing. Last, it would be valuable to have a parent liaison/consultant from within the school or community advise researchers or clinicians working in schools as to how to communicate with parents. This person could help to establish trusting relationships and address cultural barriers specific to the community, and may even help to facilitate communication between researchers, parents, and school staff. Additionally, it may reduce barriers and facilitate services among those families who may have concerns related to legal/immigration status. Another suggestion would be to remind families of the intervention project and goals during a school-based “parent night” or school-based open-house events.

Finally, it is worth noting that even if researchers or clinicians modify procedures to improve communication and engage parents, a small proportion of parents will likely continue to experience barriers that hinder their adolescent’s ability to access services. The literature suggests that parent barriers can be practical (e.g., transportation issues) or involve unfavorable perceptions about mental health problems or services (Owens et al., 2002). A parent’s ability to recognize or accept that a problem exists is predictive of service use (Teagle, 2002), and this was evident in our experience wherein parent discomfort or lack of understanding about mental health led to limited (or lack of) participation, even when their adolescent desired help. Given that a majority of parents will experience at least one barrier to service use (Girio-Herrera, Owens, & Langberg, 2013), it is important to be aware of the role of barriers, to anticipate them, and to be open to reducing as many as is feasible and malleable.

D. Balancing Confidentiality and Safety Concerns

Beyond informing parents when their child is distressed, balancing confidentiality and safety concerns is particularly challenging to manage within the school context (see Table 2). Clinicians must carefully navigate the expectations of confidentiality between participants and parents, especially during the assessment process, when participants may report distress or suicidal thoughts. It is even more challenging to balance privacy and confidentiality (on the one hand) with the school’s interest in knowing how students are doing (on the other hand), especially when the intervention is conducted within a research context.

Both research and practice require communicating limitations surrounding confidentiality and how information will be shared. Yet, especially in a research context (and possibly in other clinical contexts), the dual role of protecting adolescent confidentiality (i.e., not sharing adolescents’ at-risk status with teachers and other school personnel) and ensuring adolescent safety (i.e., getting immediate help in situations of abuse or self-harm) often were at odds with the responsibility felt by school staff to know how their students were doing.

Specifically, the limits of confidentiality in a research context meant we could not provide specific information to school staff about adolescents’ psychological functioning. (This contrasts with services provided by school staff, where school personnel may communicate with one another about students’ functioning.) In some cases, school administrators and staff did not understand why we were unable to share such information (i.e., FERPA constraints), despite multiple discussions about this issue.

Moreover, confidentiality was further complicated by the importance of maintaining adolescent safety, as in the case of potential abuse or threat of serious harm. When we identified a safety issue, our protocol required that we communicate the issue to appropriate school personnel. Determining “with whom” and “how much” information could be shared was complicated and required balancing safety with protecting adolescent confidentiality. On one occasion, we became aware of a reportable event related to a parent’s treatment of their adolescent. In this case, a mandated abuse report was made, but school staff were not informed to protect adolescent confidentiality and because the report was not relevant to safety at school. However, upon learning of the reportable event, the adolescent’s parent contacted the school to inquire about the report. In processing the event, the school personnel understandably shared that they felt ill prepared to respond to the parent’s call.

Recommendations

First, we recommend that individuals going into schools to implement interventions explicitly have conversations about FERPA. School staff may not realize the implications of FERPA, or may be aware but later forget as they begin to work collaboratively with researchers or clinicians employed outside of the school. Thus, initial conversations about FERPA with ongoing reminders are often warranted.

Similarly, it is important to discuss and reinforce with school personnel that, although student safety is paramount, ethical protections dictate that clinicians must also prioritize confidentiality. This should be explained clearly as issues of confidentiality may not be as salient, pivotal, or as central to the work of the personnel in school settings, relative to that of outside clinicians and research staff. Second, we recommend that school personnel be made aware of exceptions to this general policy, as when a student’s safety or the safety of others in the school are at imminent risk. Third, we recommend emphasizing and reminding school staff that feedback will be provided only to the parent, except in the case of certain safety issues.

Finally, in regard to mandated abuse reporting, it was valuable that we prioritized confidentiality. However, school personnel could have been informed generally that a report had been made and to direct specific questions about it to our team, should something arise. This would have allowed the school to be prepared to respond to potential calls from parents, without being fully aware of specifics. Last, we recommend parent feedback be provided as soon as possible after participants’ assessment and to be explicit about the time lapse between assessment and parent feedback, so that parents know when to expect news.

Discussion

Mental health problems are common among adolescents, yet few with such problems seek or receive professional help (Merikangas et al., 2011), and this is especially true of youth from ethnic-minority backgrounds (Alegria et al., 2010). Mental health interventions conducted in schools provide an important avenue for expanding the accessibility of evidence-based psychological interventions for affected youth, and for reducing the significant cost burden of mental health problems (Perou et al., 2013). Despite the appeal of school-based mental health services for youth, implementation can be challenging. In this paper, we reviewed some of the practical issues that occur in implementing a school-based intervention with adolescents from ethnic-minority backgrounds, and offered suggestions for addressing these issues in future research and practice. It is important to disseminate information about the challenges of working in this environment and propose solutions to improve the feasibility of conducting interventions in schools.

We emphasize that it is possible to provide effective interventions in schools, but doing so requires flexibility, creativity, and problem solving, as well as strong, communicative alliances among mental health care providers (and/or clinical research staff), parents, and school personnel, including administrators, counselors, and teachers. As school policies and educational environments evolve, clinicians and researchers must continue to evaluate how to adapt interventions to be accessible to students in varying circumstances.

In addition to the specific suggestions offered throughout the text, here we offer general recommendations for clinicians and researchers. First, establish a communication plan from the outset for dealing with schools and with parents and reevaluate it periodically. Second, establish and review a “safety plan” with specific algorithms for how to handle challenging situations. Third, both through phone contacts and in writing, inform parents, teachers, and school personnel of any limits of confidentiality and “remind” them of the limits periodically. Fourth, if conducting a selective preventive intervention, develop a detailed plan for what to do with youth who are excluded (but who still may be distressed) that includes linking such youth with community services and developing strategies for handling parents who do not believe their child might have a problem or be distressed. Fifth, consult with other resources as needed. For example, involving a parent from the community to review procedures and materials, or consultation with others who have expertise in school-based ethical issues can be invaluable.

In conclusion, reaching youth in schools and other primary care settings is critical, but requires careful thought and planning. Nevertheless, such solutions are critical to reducing the high incidence and burden of untreated mental health problems in youth.

Highlights.

• School-based, preventive intervention for adolescents at risk for depression and anxiety

• School-based prevention intervention processes involved in implementation

• Specific challenges and potential solutions given for issues involved in implementation

• Recommendations for implementing school-based mental health services for adolescents