F E A T U R E Alexandra Hilt-P Panahon Lee Kern Anuja Divatia Project REACH, Lehigh University Frank Gresham Project REACH, Louisiana State University Key words: depression; anxiety disorders; internalizing disorders; school-based intervention Introduction Depression is increasingly recognized as a significant problem for school-aged children. When left untreated, childhood depression is associated with several negative outcomes, including lowered self-esteem, social withdrawal, poor academic performance in school, and, in severe cases, even suicide (Rawson & Tabb, 1993). The existing literature suggests that as many as A B S T School-based Interventions for Students with or at Risk for Depression: A Review of the Literature 2.8% of children and up to eight percent of adolescents in the United States experience depression, making it one of the most prevalent childhood mental health disorders (Collins et al, 2004). Considering the high prevalence, effective intervention options are imperative. Despite the large number of children and adolescents affected by depression and depressive symptoms, the majority do not access intervention. Collins et al (2004) identified several barriers at the individual, provider, and systemic levels that influence whether an individual seeks and/or receives services. Variables at the individual level include willingness to disclose R A C T Internalizing disorders are increasingly recognized as mented in school settings to reduce children’s depres- a significant problem for school-aged children. sive symptoms. A variety of variables related to inter- Students with depression may experience lowered vention implementation and effectiveness were self-esteem, withdrawal, lack of concentration, and considered. Cognitive behavioral therapies emerged poor academic performance. Given these negative as the intervention with the strongest evidence base outcomes, as well as growing support for school- for reducing depressive symptoms, showing moderate based mental health services, it is critical to examine to large effect sizes. In addition, relaxation training the evidence supporting school-based interventions was identified as a promising practice, particularly for students with or at risk for depression. This paper for children with co-morbid symptoms of anxiety. provides a review of research on interventions imple- Implications for both research and practice are discussed. 32 Advances in School Mental Health Promotion INAUGURAL ISSUE - October 2007 © The Clifford Beers Foundation & University of Maryland F problems, fear of stigma and embarrassment, and demographic factors (such as age). Provider variables include knowledge of mental health problems, skill level in assessing problems, and willingness to diagnose and treat mental health issues. Finally, systemic variables consist of factors such as awareness of the range of effective treatment options, availability of mental health providers, and integration of mental health services into primary care. While these variables affect individuals at all ages, additional barriers faced by children and adolescents, such as reliance on parent/caretakers to access services and the lack of family health insurance, result in this population being largely under-served. Although the majority of children with mental health challenges do not receive the services they need, among those who do, schools are the primary provider (Burns et al, 1995). In fact, research indicates that schools provide 70–80% of the mental health services that school-age children receive. The Individuals with Disabilities Education Act (IDEA, 2004) holds schools responsible for providing support services to children identified as having emotional and behavioral disorders. Unfortunately, there has been little understanding of what types of school-based practices and programs are effective and what makes them effective (Kutash et al, 2006; Weist & Evans, 2005). The poor outcomes among children identified as having an emotional or behavioral disorder underscore the urgency to identify evidence-based practices (Bradley et al, 2004). Thus, given the central role of schools in providing mental health services, coupled with marginal student outcomes, it is critical to scrutinize carefully the effectiveness of school-based intervention practices. Existing reviews In the recent past, several reviews have examined the evidence to support a variety of school-based mental health programs and practices (Rones & Hoagwood, 2000; Foster et al, 2005). These are summarized in a comprehensive monograph by Kutash, Duchnowski, and Lynn (2006). The reviews greatly advanced our understanding of the empirical data underlying the immense number of programs used in America’s schools. For the most part, however, the programs and practices evaluated in the reviews were broad in scope, in that they were delivered to all or most of the school student body. In addition, they usually focused on prevention and frequently targeted general behavior Advances in School Mental Health Promotion E A T U R E problems (for example emotional and behavioral problems, bullying) or skill development (such as social skills, academics). As Rones and Hoagwood (2000) reported, few studies have targeted particular clinical syndromes (such as depression). Nor did the reviews evaluate specific program features that would make them applicable and feasible in school settings. Thus, a reasonable next step is to examine closely effective school-based interventions for targeted populations and psychiatric problems. A relatively recent literature review by Curry (2001) examined the effectiveness of psychosocial interventions for childhood and adolescent depression. A total of 15 studies were reviewed, and the findings indicated that psychosocial intervention was indeed effective for reducing depressive symptoms into the normative range. Specifically, applying Chambless & Hollon’s (1998) criteria, cognitive behavioral therapy (CBT) was both efficacious and superior to no intervention or other types of intervention (including family therapy and relaxation training). These findings were encouraging; however, the majority of studies (n=10) were conducted in clinical settings. Considering that the vast majority of children receive mental health intervention at school, it is imperative to examine intervention effectiveness when implemented in school settings (Kahn et al, 1990). The review The purpose of the present review was to examine the literature on school-based interventions for depression and depressive symptoms. The literature was reviewed to determine: „ what interventions have been implemented in the school setting „ whether interventions were effective in the school setting „ who implemented the interventions. In addition, once efficacious interventions were identified, we further evaluated issues related to implementation in school settings by typical school personnel. Method Two methods were used to identify studies for inclusion in this review. First, computer searches were conducted using Psycinfo, Medline, and ERIC for articles published between 1982 and 2006. The following search terms INAUGURAL ISSUE - October 2007 © The Clifford Beers Foundation & University of Maryland 33 F E A T U R E were used to identify articles: depression, depressive symptoms, social withdrawal, school, intervention, school-based intervention, children, and adolescents. Through this initial search 1473 articles were identified. Abstracts were reviewed for relevance, and those that met inclusion criteria (described below) were retrieved and the full article reviewed. In addition to the computer search, an ancestral search of each reviewed article was conducted by examining the reference section to locate additional articles not identified by the computer search. Articles were identified for inclusion using the following criteria. Only articles appearing in peer-reviewed journals were included (dissertations were excluded). The study also had to describe a prevention or intervention program designed to reduce depressive symptoms. Intervention had to be implemented with children, aged 6 to 17, enrolled in grades K-12. Only studies that described interventions implemented in a school setting (public or private) in the United States were included. a doctorate. Both coders had previous experience of reviewing and coding research articles. Before the start of coding, all categories were defined operationally, and examples and non-examples were discussed by the two reviewers. Inter-observer agreement When all studies had been coded, each coded category was reviewed to assess inter-observer agreement. Agreement was determined by dividing total category agreements by agreements plus disagreements. Agreement was 100% across all categories, with the exception of type of intervention. Reviewers disagreed on the nature of the intervention (CBT & relaxation training vs. CBT only) for one study, making overall agreement for this category 93%. Coders subsequently reviewed the study and established consensus for analysis purposes. Results Areas of evaluation Each identified article was coded along the following dimensions. Average age, age range and/or grade of all participants were coded. In addition, ethnicity, when described, was coded. The type of school (private or public; elementary, middle, or high) in which the intervention was conducted was also noted. The procedures for participant enrollment were coded as one of the following: a) open, in which enrollment in the study was open to all students in the class or school; b) inclusion contingent on screening for depressive symptoms, using cut-off scores on measures of depression or c) enrollment based on non-specific symptoms suggestive of depression. The type of intervention (for example CBT, relaxation therapy) was coded along with the duration of the intervention, features of intervention delivery (such as group size), and the individual who conducted the intervention (researcher, teacher, psychologist, for example). The research design and dependent variables were coded for each study as well. Studies were also coded for presence of treatment fidelity. The effectiveness of each intervention was coded by determining effect size. Lastly, the presence of follow-up data was noted. Coding procedures Each article was independently coded by two individuals. One coder was a doctoral student and the other had 34 Advances in School Mental Health Promotion Fifteen studies met the inclusion criteria. Table 1, opposite, provides detailed descriptions of the individual studies. There were a total of 2652 participants across all studies. Participants ranged in age from 6 to 17 years. Intervention programs were implemented with students from 1st to 12th grade. The ethnicity of participants was reported in 10 of the 15 studies. In these studies, 48% of participants were Caucasian, 34% African American, 16% Hispanic, and less than 1% other. It is important to note that some minority populations, such as Asian and Native American, were underrepresented in these investigations. Studies were conducted in a range of school settings. Sixty-six percent of the studies were conducted in public school settings, and the remaining third in either private school or residential settings. Criteria for participation in intervention None of the 15 studies required an existing psychiatric diagnosis of depression for participation in the intervention. Instead, the majority of the studies (eight) conducted multi-tiered screening for depressive symptoms. This consisted of initial screening of all potential participants using a standardized measure of depressive symptoms (such as the Child Depression Inventory (CDI) or Beck’s Depression Inventory (BDI)). Those who met a pre-determined cut-off score were further assessed using additional standardized measures, INAUGURAL ISSUE - October 2007 © The Clifford Beers Foundation & University of Maryland F TABLE 1 E A T U R E Description of Reviewed Studies Article Participants/Diagnosis Asarnow, Scott, & Mintz 23 students; 15 female, 8 male (2002) Grade= 4-6 57% White,17% Hispanic,13% Asian, 13% African American No diagnosis Cardemil, Reivch, & Seligman (2002) Clarke, Hawkins, Murphy, & Sheeber (1993) 152 students; 81 female, 71 male Grade= 5-8 Latino: 23 intervention, 26 control African American: 47 intervention, 56 control No diagnosis Study 1: 622 students; 361 intervention, 261 control Grade=9-10 Study 2: 380 students; 190 intervention; 190 control Grade=9-10 No diagnosis Intervention Dependent measures and Effect Sizes CBT CDI= .31 CBT CDI (Latino)= 1.01 CDI (African American) = .16 Depression Education CES-D = .016 Clarke, Hawkins, Murphy, Sheeber, Lewinsohon & Seeley (1995) 150 students; 90 female, 60 male 76 prevention group 74 usual care control Mean age = 15.3 years 92.5% non-Hispanic White No diagnosis CBT CES-D = .35 HDRS = .30 Hains & Szyjakowski (1990) 21 males Age=16-17 years No diagnosis CBT BDI = .18 Hains (1992) 25 males Age=15-16 years No diagnosis CBT AM/RT RADS = .93 RADS = 1.14 Hains (1994) 10 intervention; 7 females, 3 males 9 control; 7 females, 2 males Grade= 11th 2 African American, 1 Hispanic 16 Caucasian No diagnosis CBT RADS = .78 Kahn, Kehle, Jenson, & Clark (1990) 68 students; 35 females, 33 males, Grade 6-8 No diagnosis CBT RADS = 1.9 CDI = CNC RADS = 1.30 CDI = 1.01 RADS = 1.11 CDI = 1.11 AM/RT Self Modeling Kellam, Rebok, Mayer, Ialongo, & Kalodner (1994) 685 students; 350 female, 335 male Mean age=6.3 years 56.9% African America, 19.2% Caucasian, 1% other ethnic groups, 22.8% unspecified No diagnosis Jaycox, Reivich, Gillham, 143 students & Seligman (1994) 69 treatment (34 female, 35 male) 74 control (32 female, 42 male) Age=10-13 years 83% White, 11% African American No diagnosis Miller & Cole (1998) 1 male student Age= 14 years Emotional behavioral disorder and previous diagnosis of depression Advances in School Mental Health Promotion Academic Intervention CBT Social skills CDI (males) = -.22 CDI (female) =.18 CDI = .27 RCDS = .38 Effect size could not be calculated. Scores on measure of depression reduced to sub-clinical range post intervention INAUGURAL ISSUE - October 2007 © The Clifford Beers Foundation & University of Maryland 35 F E A TABLE 1 T U R E Description of Reviewed Studies (continued) Article Rawson & Tabb (1993) Participants/Diagnosis 99 students; 9 female, 90 male Age=8-12 years Behavior Disorder Reynolds & Coats (1986) 30 students Mean age=15.65 years No diagnosis Intervention Reinforcement RCDS = .36 CBT BDI = 1.64 BID = 2.22 RADS = 1.32 BDI= 1.67 BID = 2.45 RADS = 1.31 AM/RT Stark, Reynolds, & Kaslow 28 students; 12 female, 16 male (1987) Mean age=11.17 years No diagnosis Dependent measures and Effect Sizes CBT CDI = 1.25 CDS = .82 CDRS-R = .98 Problem solving CDI = 1.04 CDS = .41 CDRS-R = .65 CBT with Relaxation CDI = .63 training CDRS-R = .39 Weisz, Thurber, Sweeney, 48 students; 22 female, 26 male Proffitt, & LeGagnoux Mean age=9.6 years (1997) 30 Caucasian, 18 ethnic minorities No diagnosis Notes CDI =Child Depression Inventory (CDI), CES-D = Center for Epidemiologic Studies - Depression Scale HDRS = Hamilton Depression Rating Scale BDI = Beck's Depression Inventory RADS = Reynolds Adolescent Depression Scale RCDS = Reynolds Child Depression Scale CDRS-R (Children's Depression Rating Scale - Revised) (13%) BID = Bellevue Index of Depression, CDS = Child Depression Scale. structured interviews, or a combination of both. In six studies, participation was open to all students in the school, district, or classroom. In one study, participants had to meet a list of inclusion criteria suggestive of symptoms of depression (for example learning and adjustment problems, classroom behavior problems). ty of studies (n=8). Cognitive restructuring involves teaching children to challenge distorted and negative cognitions about themselves and their environment and to replace those cognitions with more realistic ones. This technique is based on the assumption that children are depressed due to a maladaptive style of information processing (interpreting events as negative). If cognitions are more realistic (and potentially more positive), the individual should experience less depression. Problem solving, also considered a CBT approach, was the second most frequently implemented technique, occurring in six studies. Problem solving involves teaching children to evaluate stress-provoking situations by gathering relevant information, thinking about alternative responses, and choosing the best response. Pleasant activity scheduling was implemented in five studies. This intervention entails systematic planning of children’s or adolescents’ daily activities to incorporate pleasant and desirable events. For instance, children generate a list of preferred activities, which are incorporated into their daily routine to increase pleasant experiences. Self-change (making self-evaluations and changing behavior as a result), attribution retraining (teaching children to make more realistic and adaptive attributions), and activities to link thoughts, feelings, and behavior (teaching children how all three are linked and influence Type of intervention Several types of intervention were used in the 15 studies reviewed, and many of the studies compared multiple interventions, so the number of interventions exceeds 15. Cognitive behavior therapy (CBT) was implemented most often. This intervention was used in 73% (n=11) of the studies examined. Of the eleven studies, three compared CBT with anxiety management/relaxation training and one evaluated CBT and anxiety management/relaxation training in combination. All other interventions were implemented in only one study. These were reinforcement, academic intervention, education about depression, and social skills training. The term ‘cognitive behavior therapy’ describes techniques that incorporate cognitive and behavioral models of behavior change. A number of different techniques and combinations of techniques were used as CBT interventions. Cognitive restructuring was used in the majori- 36 Advances in School Mental Health Promotion INAUGURAL ISSUE - October 2007 © The Clifford Beers Foundation & University of Maryland F each other) were used in three intervention packages. Self-instruction was employed in two studies while selfmodeling was used in one. Duration of intervention The duration of intervention implementation was reported in terms of the number of weeks, number of sessions, and length of each session in 13 studies. The interventions lasted from five to twelve weeks, and required between two and sixteen sessions, lasting from twenty to ninety minutes. Two studies did not specify intervention duration; one reported that the intervention was delivered through the entire day, and the second indicated that the intervention was delivered throughout the year. E A T U R E was lacking or teachers had no intervention experience. The amount of training provided to the intervention agent varied as well. Forty-seven percent of the studies reported providing specific training, ranging from two to forty hours. A detailed intervention/training manual was used in 40% (n= 6) of the studies. Research design The most common design implemented was some form of pre-test/post-test with random assignment to conditions (73%, n=11). Other designs included nonrandom assignment of participants (20%, n=3) and a multiple baseline across behaviors in the one singlesubject research study. Twelve of the fourteen group design studies included a control condition, most often in the form of a wait-list control. Intervention delivery With the exception of one single subject study, all interventions were implemented, at least in part, in a group format. For CBT interventions, 73% (n=8) were conducted exclusively in group sessions and 27% (n=3) used a combination of group and individual sessions. All anxiety management/ relaxation training interventions were delivered in a group format. Other types of intervention used group formats, with the exception of social skills instruction. One study implemented family therapy in addition to group intervention for the participant. Detailed information regarding group size for interventions identified as effective is provided in Table 2, below. TABLE 2 Summary of Evidence-B Based Interventions Intervention CBT AM/RT alone (N=11) (N=3) School Characteristics Public Private Elementary Middle High 73% 27% 0% 55% 45% 67% 33% 0% 33% 67% 0% 73% 27% 0% 100% 0% 27% 36% 36% 33% 33% 33% 55% 27% 9% 0% 9% 67% 0% 0% 0% 33% 45% 36% 36% 27% 33% 67% 33% 0% Intervention setting Individual Group Combination Group size Intervention agent Interventions were conducted by school staff (teachers, school psychologists, etc) in 33% (n=5) of the studies reviewed. In the remaining studies, interventions were conducted by graduate students not affiliated with the school (27%; n=4), trained personnel not affiliated with the school (20%, n=3), graduate students placed in the school (13%, n=2), or the researcher (7%, n=1). Intervention agent training The level of experience of intervention agents varied. Interventionists had psychological training in 73% (n=11) of studies. Of those individuals, 36% (n=4) had previous experience of implementing similar interventions. In the remaining studies, information Advances in School Mental Health Promotion 2-5 students 6-12 students Not specified Intervention Agent Graduate student Trained outside agents Researcher School staff School staff & Researcher Intervention Agent Training Intervention manual Intervention training Previous intervention experience No information Intervention Duration Number of sessions Minutes per session Total minutes Number of weeks 8-15 30-90 260-1080 5-12 9-12 30-50 270-600 5-8 INAUGURAL ISSUE - October 2007 © The Clifford Beers Foundation & University of Maryland 37 F E A T U R E Dependent variables All the studies reviewed used one or more standardized measure of depression as a dependent variable. Measures used were: „ „ „ „ „ „ „ „ „ Child Depression Inventory (CDI) (47%) Reynolds Adolescent Depression Scale (RADS) (33%) Reynolds Child Depression Scale (RCDS) (13%) Children’s Depression Rating Scale – Revised (CDRS-R ) (13%) Beck’s Depression Inventory (BDI) (13%) Center for Epidemiologic Studies – Depression Scale (CES-D) (13%) Child Depression Scale (CDS) (7%) Bellevue Index of Depression (BID) (7%) Hamilton Depression Rating Scale (HDRS) (7%). Additional dependent variables included selfesteem (33%), anxiety (27%), anger (20%), and explanatory style (13%). One study also assessed problem behavior, classroom behavior, reading achievement, daily and major life stressors, and negative cognitions. Intervention fidelity Studies varied in the assessment and reporting of intervention fidelity. In four of the fifteen studies, fidelity was not assessed. For an additional five studies, there was mention of measures to ensure treatment fidelity, such as training, supervision, and comprehensive treatment manuals, but fidelity data were not reported. For the remaining investigations (40%, n=6), fidelity was assessed and reported. In all six studies, intervention implementation was assessed by trained observers, who reviewed and scored audiotapes of intervention sessions using various measures, such as Likert scale ratings, treatment adherence rating scales, and point by point scoring of intervention components. Reported fidelity was high for these studies, averaging 93% (range, 83%-100%). Intervention effectiveness The effectiveness of each intervention was assessed by calculating effect sizes. Effect size was computed for each dependent measure by subtracting the treatment mean from the baseline mean and dividing by the pooled standard deviation (Cohen’s d) for each study with a group design (93%). In the 11 CBT studies, the 38 Advances in School Mental Health Promotion range of effect sizes was 0.16–2.22 (note: does not include ‘low emotional arousal’ students who did not have depressive symptoms in Hains, 1994). The effect sizes across the majority of CBT studies were moderate to large. Low effect sizes were found in two studies. In the first study (Hains & Szyjakowski, 1990) an effect size of 0.18 was obtained for depressive symptoms. The authors hypothesized that limited effects were due to reduction in depressive symptoms of both experimental and control groups, most probably related to low levels of initial symptoms. The low effect size for the second study (Cardemil et al, 2002) was attributable to a sub-group of participants. The same intervention program was provided to two ethnically diverse groups of students at two schools. While results were positive for the Latino children (ES =1.01), little positive effects were noted for the African American participants (ES=.16). The authors provided several possible explanations for these results, including regression to the mean, differential expression of symptoms across ethnic groups, and ethnic variation in the response to different intervention components. The effect size was large in the study that combined CBT with relaxation training (.63). Three studies evaluated relaxation training in isolation. Effect sizes were large, ranging from 1.14 to 2.45. Other interventions, such as academic interventions (ES=-.22 for males, .18 for females), education about depression (ES=.016), and non-contingent rewards (ES=.36) had small to moderate effect sizes. The study implementing a social skills intervention was a singlesubject study with one participant. Although an effect size could not be calculated, results indicated that depression was reduced to non-clinical levels following intervention. Follow up Maintenance of intervention effects was assessed in 53% (n=8) of the studies at 1–12 months post intervention. Among these studies, treatment gains were maintained for all CBT and relaxation training interventions. In the study using depression education, follow-up data revealed that the small post-intervention decreases in depressive symptoms were not maintained at 12 weeks. Likewise, maintenance of effects for self-modeling alone did not occur. Evidence-based interventions Chambless and Hollon (1998) developed criteria to determine empirically supported efficacious interventions. INAUGURAL ISSUE - October 2007 © The Clifford Beers Foundation & University of Maryland F The criteria include: „ a randomized control trial with results superior to no treatment or equivalent or superior to another treatment of known efficacy „ use of a treatment manual, a specified population, valid and reliable outcome measures „ use of appropriate statistical analyses. The criteria must have been met in at least two different research settings. Applying these criteria, CBT is considered an evidence-based intervention when implemented in school settings. Although Anxiety Management/Relaxation Training (AM/RT) does not meet the definition of an evidence-based practice, the only criterion that was not met was the use of a treatment manual, absent in two of the three studies. Thus we consider it a promising practice. Table 2 provides a detailed summary of the interventions that emerged as evidence-based or promising practices. Both CBT and AM/RT have been implemented in public and private school settings, but neither intervention has been implemented with elementary age children. The interventions have been implemented almost exclusively in a group setting, generally by graduate students as part of a research study. The total intervention duration for CBT ranged from 4 hours, 20 minutes to 18 hours, while the AM/RT intervention duration ranged from 4 hours, 30 minutes to 10 hours. Discussion The results of this review indicate that, among the school-based interventions for depressive symptoms, CBT is the most researched, and benefits from the strongest empirical support with respect to effect size and maintenance of intervention gains. The CBT interventions used in the studies reviewed included variations, using different types of cognitive and behavioral components, all of which proved to be effective. The most frequently used components were cognitive restructuring, pleasant activity scheduling, and problem solving. Self-change skills, attribution retraining, and linking feelings, thoughts and behaviors were also relatively common components of school-based CBT. Anxiety management/relaxation training appears to be a promising intervention, particularly for children with co-morbid symptoms of depression and anxiety. Given the limited number of studies, however, it is difficult to make conclusive remarks regarding this Advances in School Mental Health Promotion E A T U R E intervention. The absence of an intervention manual also limits the supportive evidence. Despite the positive findings described above, there are several issues that need to be addressed related to implementation of interventions in the school setting. First, although CBT and AM/RT decreased depressive symptoms in middle and high school aged students, there were no studies that included participants with a diagnosis of depression. As mentioned earlier, the majority of studies conducted multi-tiered screenings for depressive symptoms, but did not require a diagnosis of depression for inclusion. There are two reasons why this might have occurred. First, there may have been ethical considerations about assigning children with a clinical diagnosis to a control group. In fact, in one study, potential participants who were identified as having major depressive or dysthymic disorder were excluded from participation and referred for community services (Clarke et al, 1995). Second, children with internalizing disorders often go undiagnosed. Past research indicates that up to 30% of adolescents will exceed clinical cut-offs on self-report measures of depression (Hammen & Rudolph, 1996), although only a fraction of these children will ever receive a diagnosis. By including children with depressive symptoms, rather than just those with a diagnosis of depression, it is most likely that a larger and more representative sample was identified. Thus, although the effectiveness of schoolbased intervention with a clinical population per se was not evaluated, the results with symptomatic populations were positive. Whether interventions for depression can be implemented by typical school personnel remains unanswered. Of the five studies with intervention implemented by teachers, three had low to moderate effect sizes, indicating that these interventions were moderately efficacious at best. It is important to note, however, that the implemented interventions had limited or no empirical support for the treatment of depression (depression education, academic intervention, reinforcement). The lack of effects, therefore, may have been due to the intervention itself, rather than the interventionists. The studies provided little to no training for teachers, and did not assess intervention integrity. It is possible that, with adequate training and feedback on fidelity, positive effects might have been observed, as was the case in the fourth study that used school personnel (Miller & Cole, 1998). Intervention was implemented by a well-trained interventionist, and INAUGURAL ISSUE - October 2007 © The Clifford Beers Foundation & University of Maryland 39 F E A T U R E high levels of fidelity were documented, which may have contributed to the reduction of depressive symptoms to non-clinical levels after the implementation of a social skills intervention with the study’s single participant. Considering just the interventions identified as evidence-based or promising, the feasibility of implementing CBT and AM/RT by typical school personnel also remains unclear (Adelman & Taylor, 1998). CBT and AM/RT were implemented by school staff in only one study. Specifically, in Kahn et al (1990), the school counselor implemented CBT, relaxation training, and self-modeling for some intervention groups after training from the researcher. On-going supervision and frequent integrity checks were conducted throughout the intervention. Results indicated that the interventions were effective in reducing depressive symptoms, with effect sizes above 1.0 for all three interventions. There were no discernable differences in effect sizes or implementation fidelity between this study and other CBT and AM/RT studies implemented by outside intervention agents. These results strongly suggest that, with the proper training, school personnel can implement effective interventions. Future research should document the amount of training to school-based personnel and level of fidelity necessary to ensure positive student outcomes. The findings of the current review indicate that CBT and AM/RT required an average of approximately 10 hours of intervention. Given the high risk of later negative sequelae without intervention, including lower educational attainment, substance abuse, and poor work history (Weersing & Brent, 2006), this seems to be a minimal investment on the part of schools. In addition, from a cost–benefit analysis, without intervention, the need for academic remediation and social/emotional/behavioral interventions is almost certain. Many students will require costly tertiary interventions, including placement in more restrictive educational settings or hospitalization. In the long run, considering student outcomes and cost-effectiveness, the reasonableness of 10 hours of intervention seems indisputable. It is important to note also that CBT and AM/RT were demonstrated to be effective in a group format, thus minimizing the time and resources required. Future research should identify barriers to implementation. Another issue important to address in future research is determining which CBT components are necessary and/or sufficient for positive outcomes. Given the numerous variations described in the literature, it is 40 Advances in School Mental Health Promotion important to ascertain which components, or combinations of components, are most efficacious. By identifying the components that have the quickest, largest, and longest-lasting effects, intervention can be streamlined. A promising intervention identified through this review was anxiety management/relaxation training. It appears that this intervention may be effective because, if anxiety is reduced, an individual will have less cause to be depressed and consequently depressive symptoms should decline. AM/RT appears to be somewhat easier to implement than CBT and requires less training and expertise. Additional research should be conducted to determine whether this promising approach can be considered evidence-based and to evaluate ease of implementation. Ultimately, what is most important is that children and adolescents with depressive symptoms receive evidence-based interventions before negative outcomes occur. Considering data on the provision of children’s mental health services, this will happen only if interventions occur in school settings. Preliminary research suggests that the interventions will be effective when implemented by school personnel. It is imperative that schools identify staff members who can receive training and begin to implement evidence-based interventions. Simultaneously, research must continue to refine effective practices, evaluate promising practices, and develop models of collaboration that allow school personnel and mental health experts to work together to ensure that schools provide effective interventions to all children in need. Address for correspondence Alexandra Hilt-Panahon, College of Education, Lehigh University, 111 Research Drive, Bethlehem, PA 18015 References Adelman HS & Taylor L (1998) Reframing mental health in schools and expanding school reform. Educational Psychologist 33 135–53. Asarnow JR, Scott CV & Mints J (2002) A combined cognitive-behavioral family education intervention for depression in children: a treatment development study. Cognitive Therapy and Research 26 221–9. Bradley R, Henderson K & Monfore DA (2004) A national perspective on children with emotional disorders. Behavioral Disorders 29 211–23. Burns BJ, Costello EJ, Angold A, Tweed D, Stangle D & INAUGURAL ISSUE - October 2007 © The Clifford Beers Foundation & University of Maryland F Farmer EMZ (1995) Children’s mental health service use across service sectors. Health Affairs 14 148–59. Cardemil EV, Reivich KJ & Seligman EP (2002) The prevention of depressive symptoms in low-income minority middle school students. Prevention and Treatment 5 8–38. Chambless DL & Hollon SD (1998) Defining empirically supported therapies. Journal of Consulting and Clinical Psychology 66 7–18. Clarke GN, Hawkin W, Murphy M & Sheeber L (1993) School-based primary prevention of depressive symptomology in adolescents: findings from two studies. Journal of Adolescent Research 8 (2) 183–204. Clarke GN, Hawkin W, Murphy M, Sheeber LB, Lewinsohn PM & Seeley JR (1995) Targeted prevention of unipolar depressive disorder in an at-risk sample of high school adolescents: a randomized trial of a group cognitive intervention. Journal of American Academy of Child Adolescent Psychiatry 34 312–21. Collins KA, Westra HA, Dozois DJA & Burns DD (2004) Gaps in accessing treatment for anxiety and depression: challenges for the delivery of care. Clinical Psychology Review 24 583–616. Curry JF (2001) Specific psychotherapies for childhood and adolescent depression. Biological Psychiatry 49 1091–0. Foster S, Rollefson M, Doksum T, Noonan D, Robinson G & Teich J (2005) School Mental Health Services in the United States, 2002-2003. DHHS Pub. No. (SMA) 05-4068. Rockville, MD, USA: Center for Mental Health Services, Substance Abuse and Mental Health Services Administration. Hains AA (1992) Comparison of cognitive-behavioral stress management techniques with adolescent boys. Journal of Counseling and Development 70 600–5. Hains AA (1994) The effectiveness of a school-based, cognitive behavioral stress management program with adolescents reporting high and low levels of emotional arousal. The School Counselor 42 115–24. Hains AA & Szyjakowski M (1990) A cognitive stressreduction intervention program for adolescents. Journal of Counseling Psychology 37 79–84. Hammen C & Rudolph RD (1996) Childhood depression. In: EJ Mash & RA Barkley (Eds) Child Psychopathology. New York: Guildford Press. Individuals with Disabilities Education Improvement Act Amendments of 2004. 20 U.S.C. § 1462(h) et seq. Jaycoz LH, Reivich KJ, Hillham J & Seligman MEP (1994) Prevention of depressive symptoms in school Advances in School Mental Health Promotion E A T U R E children. Behavior Research Therapy 32 801–16. Kahn JS, Kehle TJ, Jenson WR & Clark E (1990) Comparison of cognitive-behavioral, relaxation, and self-modeling interventions for depression among middle-school students. School Psychology Review 19 196–211. Kellam SG, Rebok GW, Mayer LS, Ialongo N & Kalodner CR (1994) Depressive symptoms over first grade and their response to a developmental epidemiologically based preventive trial aimed at improving achievement. Development and Psychopathology 6 463–81. Kutash K, Duchnowski AJ & Lynn N (2006) SchoolBased Mental Health: An empirical guide for decisionmakers. Tampa, FL, USA: University of South Florida, Louis de la Parte Florida Mental Health Institute, Department of Child and Family Studies, Research and Training Center for Children’s Mental Health. Miller DN & Cole CL (1998) Effects of social skills training on an adolescent with comorbid conduct disorder and depression. Child and Family Behavior Therapy 20 35–53. Rawson HE & Tabb LC (1993) Effects of therapeutic intervention on childhood depression. Child and Adolescent Social Work Journal 10 39–53. Reynolds WM & Coats KI (1986) A comparison of cognitive-behavioral therapy and relaxation training for the treatment of depression in adolescents. Journal of Consulting and Clinical Psychology 54 653–60. Rones M & Hoagwood K (2000) School-based mental health services: a research review. Clinical Child and Family Psychology Review 3 223–41. Stark KD, Reynolds WM & Kaslow NJ (1987) A comparison of the relative efficacy of self-control therapy and a behavioral problem-solving therapy for depression in children. Journal of Abnormal Child Psychology 15 91–113. Weersing VR & Brent DA (2006) Cognitive behavioral therapy for depression in youth. Child and Adolescent Psychiatric Clinics of North America 15 939–57. Weist MD & Evans SW (2005) Expanded school mental health: challenges and opportunities in an emerging field. Journal of Youth and Adolescence 34 3–6. Weisz JR, Thurber CA, Proffitt VD, Sweeneny L & LeGagnoux GL.(1997) Brief treatment of mild-tomoderate child depression using primary and secondary control enhancement training. Journal of Counseling and Clinical Psychology 65 (4) 703–7. INAUGURAL ISSUE - October 2007 © The Clifford Beers Foundation & University of Maryland 41