Printer Friendly

Counseling outcomes from 1990 to 2008 for school-age youth with depression: a meta-analysis.

Depression is a mood state that involves frequent and intense feelings of sadness that cause an individual to lose interest in normal daily activities, such as work or school performance, social relationships, or normally pleasurable activities (American Psychiatric Association, 2000). Although depressive disorders are often thought of as affecting adults, it is not at all rare to encounter school-age youth with depression or depressive features (Imel, Malterer, McKay, & Wampold, 2008). Estimates of the prevalence of depression in school-age youth are approximately 2% to 3% for prepubertal children and 5% to 8% for adolescents (Birmaher et al., 1996; Costello et al., 1996; Costello, Erkanli, & Angold, 2006; Lewinsohn, Clarke, Seeley, & Rohde, 1994). By the age of 18, nearly 20% to 25% of youth will have met criteria for major depressive disorder at least once (Birmaher et al., 1996; Clarke et al., 2001; Lewinsohn et al., 1994), and Ryan (2005) estimated that clinically relevant symptoms can be observed in up to 30% of adolescents. Many youth with depression continue to experience depression as adults. Depression has been shown to affect social functioning, academic performance, family relationships, use of mental health services, drug use, suicide, and prevalence of other psychiatric disorders (Angold & Costello, 1993; Birmaher et al., 1996; Clarke, DeBar, & Lewinsohn, 2003; Gould et al., 1998; Le, Munoz, Ippen, & Stoddard, 2003; Stolberg, Clark, & Bongar, 2002).

In empirical studies, depression is usually operationalized quantitatively as an elevated score on a self-report or other-report depression scale (Nemeroff, 2007). Therefore, levels of depression can be measured with reasonable accuracy, and the effectiveness of treatments meant to reduce symptoms of depression can be assessed. Counseling/psychotherapy and antidepressant medications are the two most common treatments for depression in youth. Clinical trials and synthesized meta-analyses of the use of medication for the treatment of depression have yielded mixed results. Some studies indicated that medication had a small but significant advantage over psychotherapy in posttest and follow-up studies (Imel et al., 2008). Other studies concluded that neither approach yielded superior treatment outcomes over the other (De Maat, Dekker, Schoevers, & De Jonghe, 2006; Hazell, O'Connell, Heathcote, & Henry, 2002). What is clear is that concerns exist over the increased risks of suicide and suicidal ideation when using antidepressant medications with youth, and these concerns have resulted in the issuance of safety warnings (Costello et al., 2002; Vitiello & Swedo, 2004; Whittington et al., 2004). As a result, in the United Kingdom, the National Institute of Clinical Excellence recommended psychotherapy as preferable to medication in initial attempts to treat depression in youth (Ramchandani, 2004). A similar recommendation was made by the American Academy of Child and Adolescent Psychiatry (AACAP; 1998) for the treatment of depression in school-age youth in the United States.

If counseling or psychotherapy is the first avenue for treatment of depression in school-age youth, counselors and psychotherapists need answers to several important questions about the effectiveness of this broad therapeutic approach, and the following three questions form the purpose of this meta-analysis: (a) Is counseling/psychotherapy effective in reducing symptoms of depression in school-age youth when compared with various comparison conditions? (b) Do the effects of counseling/psychotherapy last? and (c) Is there a differential effectiveness when treatment of depression in youth is conducted in school settings as opposed to clinic/outpatient conditions?

* Is Counseling/Psychotherapy Effective for School-Age Youth With Depression?

The primary question of this meta-analysis involves whether counseling and psychotherapy are effective in reducing symptoms of depression in school-age youth when compared with various control or comparison conditions. Many clinical trials over the past several decades have led to mixed results. Some show no or only slight positive outcomes, whereas an impressive number of clinical trials have shown statistically significant results of counseling/psychotherapy in treating depression in youth. An advantage of using a meta-analysis is that effect size estimates are computed to compare the treatment condition with various comparison conditions (e.g., wait list, placebo, treatment as usual [TAU]), thereby helping to establish the average magnitude of differences across studies with similar designs.

More than a half dozen meta-analyses of the treatment of depression or related problems in children and adolescents have been conducted over the past dozen years, each focused on varying populations and study characteristics, and were of varying quality and comprehensiveness. Harrington, Whittaker, Shoebridge, and Campbell (1998) conducted a recta-analysis specifically aimed at assessing the effectiveness of cognitive behavior therapy (CBT) on depression in school-age youth and concluded that CBT is highly effective in this regard. Reinecke, Ryan, and DuBois (1998) synthesized six CBT adolescent studies and reported an average effect size of .97, well above most published clinical trial results of general youth mental health treatment. Lewinsohn and Clarke (1999) synthesized 12 controlled trials on treatment of adolescent depression, yielding an average effect size of 1.27. Finally, Michael and Crowley (2002) reviewed 14 adolescent controlled trials of diverse treatment approaches, not just CBT, and determined an average effect size of .72. Each of these studies used a fixed effects model, varied in degree of quality control, and reported average effect sizes equal to or larger than those traditionally reported for the general effectiveness of counseling and psychotherapy for the treatment of broad-based mental disorders.

Two more recent recta-analyses used a random effects model, tighter quality controls, and weighting of means by inverse variance procedures and synthesized larger numbers of studies. Weisz, McCarty, and Valeri (2006) synthesized the results of 35 studies of psychotherapy for depression in children and adolescents and reported a mean effect size of .34, which was a small to moderate effect, still significant, but much smaller than results reported by previously conducted meta-analyses. Weisz, McCarty, and Valeri reported no difference between cognitive and noncognitive therapeutic approaches, but also that psychotherapy was not superior to other treatment conditions, such as the use of medication. Watanabe, Hunot, Omori, Churchill, and Furukawa (2007) assessed and synthesized 27 studies and, unfortunately, reported relative risk of response (RR) data rather than classic effect size estimates. They observed a significant effect at the conclusion of treatment (RR = 1.18), especially for CBT and interpersonal therapy, and that psychotherapy seemed more effective for youth 12 to 18 years old with moderate to severe levels of depression.

Many clinical trials of depression have used wait-list control conditions, whereas many others have used a comparison with some other form of common treatment, usually referred to as a TAU comparison group. A few have used a placebo condition, which is meant to compare the active treatment with some alternative (active control) condition not designed to alter the participants' level of depression. As noted previously, in the past, comparison of treatment versus wait-list conditions usually showed large treatment effects. The few studies of treatment and placebo comparisons indicate much lower, sometimes even nonsignificant, effect sizes (Baskin, Tierney, Minami, & Wampold, 2003). The present recta-analysis explored the effectiveness of treatment methods for depression in school-age youth against each of these three comparison models: wait-list (inactive) control, placebo (active) control, and TAU (active) comparison groups. These studies each generated mean difference effect sizes for analysis (i.e., Cohen's d). A handful of studies also used no control or comparison conditions but did furnish pretest and posttest means and standard deviations. These single-group studies each generated mean gain effect sizes for analysis.

* Do the Effects of Counseling/ Psychotherapy Last for School-Age Youth With Depression?

A second important question involves whether the effects of counseling and psychotherapy with school-age youth last. Many clinical trials have assessed this question by conducting follow-up studies after the conclusion of the treatment phase to determine the staying power of the original intervention, and the results are mixed. Weisz, Jensen-Doss, and Hawley (2006) indicated no significant diminishment of treatment at 2 to 3 months, but both Reinecke et al. (1998) and Michael and Crowley (2002) indicated that the effects of treatment were substantially diminished on even short-term follow-up. The long-term effects of youth depression treatments in previous meta-analyses do not seem to last, because Watanabe et al. (2007) reported no significant effect of psychotherapy at 6 months after treatment terminated and Weisz, McCarty, and Valeri (2006) reported no significant lasting effect at a 1-year follow-up. Indeed, Weisz, McCarty, and Valeri demonstrated that gains made during treatment were gone 1 year later. The present meta-analysis again sought to assess the staying power of depression interventions on a more current set of clinical trials.

* Is There a Difference Between School-Based and Clinic-Based Results for School-Age Youth With Depression?

A third and final question was whether a differential effectiveness exists when treatment of depression in school-age youth is conducted in school settings as opposed to clinic or outpatient conditions. In a controlled experiment, laboratory conditions may produce significant effects that are more difficult to replicate in real-life venues, such as in schools, clinics, or outpatient settings. Nearly all youth with depression attend school, so the school is a natural environment for screening of youth depression, if not also for diagnosis and treatment. Most clinical trials use lab and clinic environments, but some studies have explored the effectiveness of depression interventions in schools, as indicated in a small meta-analysis conducted by Cuijpers, van Straten, Smits, and Smit (2006). Cuijpers et al. revealed an average effect size of .55 across eight studies aimed at early intervention with depression conducted in schools.

To answer each of these three main questions, we searched for published clinical trials of depression interventions that used quasi-experimental or true experimental designs. All selected studies assessed for depression using pretest and posttest instruments, and many assessed for follow-up effects either over the short term (a few months) or long term (6 months or up to 2 years). All the study samples included youth ages 6 to 17 years and used some variation of a depression rating scale, whether self-report, parent report, or clinician report, with established psychometric characteristics that were judged at least adequate for the purposes used. Effect sizes for mean difference scores were computed when a control or comparison condition was used (Chambers, 2004; M. W. Lipsey & Wilson, 2001), and averages for each type of comparison group (i.e., wait list, placebo, and TAU) were analyzed separately to avoid an apples-to-oranges comparison. Single-group pretest-posttest mean gain scores were computed when control conditions were not used (S. J. Lipsey, Lipsey, & Derzon, 2003; Netz, Wu, Becker, & Tenenbaum, 2005) and were analyzed separately. Tests of homogeneity (Cochran's Q) and inconsistency (/2) were conducted using a random effects model to determine the potential existence of moderator variables.

* Method

The definition of counseling or psychotherapy for depression used in this meta-analysis was any treatment or intervention aimed at the alleviation of depressive symptoms or disorders provided by a mental health professional or professional-in-training.

Inclusion and Exclusion Criteria

To be included in this meta-analysis, a study had to meet the following nine criteria. First, the study must use an intervention directly aimed at reducing symptoms of depression in participants diagnosed with, or identified because of, significant depressive symptoms. Thus, a study of some other childhood condition (e.g., anxiety, disruptive behavior, substance abuse) that coincidentally includes a depression outcome measure was excluded. Second, the participants' level of depression must be measured by a standardized instrument, so at least one standardized depression rating scale (self-report, parent report, clinician report) was used in all included studies. Third, the studies must provide sufficient output data to compute a mean difference effect size or mean gain effect size. Therefore, each study must provide pretest and posttest group means and standard deviations. We excluded studies that reported only t, F, or p statistics because the formulas used to convert these statistics to effect sizes lead to nonequivalent estimates when compared with mean gain and mean difference effect sizes. Fourth, the sample participants must receive some form of individual, group, or family counseling or psychotherapy. For example, studies that exclusively explored drug trials were eliminated, or, in the case where the study included a drug condition and psychotherapy condition, the effect sizes for the drug condition were not analyzed, unless that condition was described as the TAU comparison condition. Fifth, the studies must be composed of school-age participants 6 to 17 years old. Sixth, studies needed a minimum sample size of nine participants to help control for the effects of small sample size. Case study and single-subject research designs were excluded. Seventh, acceptable study designs included clinical trials with a single-group or some control or comparison group condition (e.g., wait list [inactive], placebo [active], or TAU [active comparison]). We eliminated nonexperimental or pre-experimental designs and analyzed only quasi-experimental or true experimental clinical trials. Eighth, the included articles needed to appear in print between 1990 and 2008. Finally, all studies needed to be published in the English language, although no restriction was placed on the sample's national or cultural origin. These criteria were implemented to establish a robust set of peer-reviewed studies of reasonably high quality. When separate articles were published using the same sample results, only one study was included in the analysis to maintain sample independence.

Search Strategies

Applying the aforementioned criteria for inclusion or exclusion, we examined the title and abstract of each article. Two authors served as judges, arriving at decisions independently, and the first author refereed disagreements in selection or elimination. For any study for which selection was in question, the full text was retrieved and consensus reached for inclusion and exclusion criteria among the two judges and the first author.

The studies included in this meta-analysis were obtained through three redundant search methods: computerized searches, reviewing reference lists, and hand searches of the most prominent contributing journals. Initially, computerized searches were conducted of PsycINFO and MEDLINE from 1990 to 2008. A beginning point of 1990 was chosen because nearly all electronic indexing of journals provides abstracts and full text (html or pdf) from 1990 forward, facilitating the selection of a comprehensive set of candidate studies from this time period. In addition, quality of clinical trials has become more standardized and of higher quality over the past 20 years.

We used keywords from full text related to the intervention of interest (i.e., counseling, psychotherapy) and condition (i.e., depression, major depressive, dysthymia). To the extent possible, we set search limitations to identify only studies with child and adolescent samples (ages 6 to 17 years), in the English language, using peer review, and treatment outcomes/ clinical trials. Second, we examined reference lists of review and synthesis articles and the clinical trials evaluated for this study. These included any meta-analyses or qualitative analyses on childhood depression previously published and any clinical trial study selected for inclusion in this study, among other sources. Finally, we hand searched journals with higher frequencies of already included articles from 1990 to 2008 (i.e., Journal of Consulting and Clinical Psychology, Journal of the American Academy of Child and Adolescent Psychiatry, Journal of Affective Disorders, Journal of the American Medical Association, European Child & Adolescent Psychiatry, and Archives of General Psychiatry).

We did not search dissertation abstracts, reasoning that high-quality research studies by a doctoral candidate would have been placed in consideration for publication in a peer-reviewed journal. Most previous depression-related meta-analyses (i.e., Cuijpers et al., 2006; Imel et al., 2008; Reinecke et al., 1998) did not include dissertations. In contrast, Weisz, McCarty, and Valeri (2006) selected 36 studies that included seven dissertations but reached very similar overall effect size estimates as those in the present study.

Coding Procedures

Each study was coded for multiple characteristics of participants, design, and method features for potential use in exploration of moderator or mediator variable in a subsequent phase of analysis should the sets of effect sizes underlying the hypotheses lack homogeneity. Participant characteristics included sample size; age, sex, and ethnicity of participants; country in which the study was conducted; and participant completion rate. Design characteristics included use of randomization, recruitment method, diagnosis method, setting of treatment, type of treatment, and type of control group. Method characteristics included use of therapists specially trained in the treatment, blind assessment, supervision, treatment manual, individual or group method, homework, number of sessions, duration of sessions, duration of study, completion rates, and therapist characteristics (i.e., degree, discipline, vocation). Missing data were coded as missing or "don't know" and analyzed for coder consistency; that is, if the data were missing, the coders should agree that they were missing. All characteristics were defined in a coding manual (Stock, 1994).

In total, 25 characteristics were coded for each article, in addition to effect size computations. At least two judges independently coded each study. With one exception, each judge was a graduate student who had completed course work in research, statistics, and assessment. All judges completed a training course and coded several practice studies under rigorous supervision of the first author. Discrepancies in coding were adjudicated by the first author, who refereed any disagreements by going to the full text article and using a consensus-building procedure. No formal assessment of study quality was conducted because the selection/inclusion process was rigorous and all articles were published, so the peer review process served as a proxy check on quality. Also, nearly all selected articles used randomized pretest--posttest control/comparison group designs.

Outcome Measures

Nearly all outcome measures used in the 42 selected studies were standardized self-report or parent report measures. The most commonly used measures were the Children's Depression Inventory (Kovacs, 1992; k = 17), the Beck Depression Inventory or the Beck Depression Inventory-II (Beck, Steer, & Brown, 1996; k = 13), one of the Hamilton scales (i.e., Hamilton Rating Scale for Depression [HAM-D or HRSD], Hamilton Depression Rating Scale [HDRS]; Hamilton, 1960; k = 12), the Center for Epidemiologic Studies Depression Scale (Radloff, 1977; k = 7), the Reynolds Adolescent Depression Scale-2 (Reynolds, 2002; k = 5), the Child Behavior Checklist (CBCL; Achenbach & Rescorla, 2001) Depression subscale (k = 3), the Edinburgh Depression Scale (Green & Murray, 1994; k = 2), and the Children's Global Assessment Scale (Shaffer et al., 1983; k = 2). A few informal measures of depression were also used in some studies, for example, a "depression severity rating" (Thompson et al., 2007) or "depressive symptoms" (Stice, Rohde, Seeley, & Gau, 2008). In order to be considered an outcome measure for inclusion in this meta-analysis, the variable needed to be a direct measure of depression. So a measure such as the Internalizing subscale from mother reports on the CBCL was not included in computation of effect sizes, whereas the Depression subscale score from the same instrument was included.

Statistical Methods

Because there are more than 40 formulas for computing effect size (Kirk, 1996) and because one should not combine the effect sizes derived from different effect size formulas, we excluded any study from which we could not derive either a mean difference or mean gain effect size. Furthermore, we combined only similar effect sizes (e.g., only mean difference effect sizes) on similar study designs (i.e., all wait-list, all placebo, all TAU, or all single-group designs, separately) as indicated by Erford,

Savin-Murphy, and Butler (2010). All effect sizes were independent; that is, (a) studies with multiple treatment conditions using the same comparison group advanced one effect size for only the most relevant treatment condition for further analysis and (b) multiple outcome measures of depression from a study were combined into a single effect size representing a single study prior to being averaged with independent effect sizes from other similar studies. Analysis of posttreatment effect sizes was conducted by combining effect sizes generated at the immediate conclusion of the treatment. Analysis of follow-up effects used the last follow-up effect size because this estimate tends to be the most conservative. Thus, if both 6-month and 12-month follow-ups were conducted, the 12-month follow-up effect size was advanced into the analysis.

M. W. Lipsey and Wilson (2001) suggested using Cohen's d to compute standardized mean difference effect sizes for wait-list, TAU, and placebo group samples by calculating the difference between the treatment and comparison group means and dividing by the pooled standard deviation. Directional designations were adjusted so that a positive d meant the treatment was effective. Computation of standardized mean gain effect sizes for single-group samples was conducted using the formula provided by M. W. Lipsey and Wilson (2001). A conservative reliability estimate of .65 was used when reliability data were not otherwise available.

Effect sizes were corrected for sample size bias using the formula d' = d{1 - [3/(4N- 9)]}. These unbiased effect sizes (d') were then corrected using an inverse weighting procedure described by Erford et al. (2010) and M. W. Lipsey and Wilson (2001), producing the corrected effect size d+ prior to being combined and averaged for hypothesis testing and subsequent assessment for homogeneity using a random effects model. The random effects model was chosen because it assumes that the selected studies were sampled from a larger set of studies and allows greater external generalizability (Hedges & Olkin, 1985).

Effect size standard errors and 95% confidence intervals (CIs) were computed using formulas reported by Erford et al. (2010) and M. W. Lipsey and Wilson (2001). The 95% CI was selected to allow easy evaluation of whether an effect size was greater than zero. For example, if a 95% CI for d+ = .25 is [+ or -] 20, a 95% CI results in a range of .05 to .45. Because the entire range for d+ is > 0, the null hypothesis of d+ = 0 can be rejected. On the other hand, if d+ = .10 and the 95% CI is [+ or -] 20, the resulting range would be -.10 to .30. Because a portion of that range is less than zero, the null cannot be rejected and the d+ cannot be considered to be significantly greater than zero.

Heterogeneity in effect sizes was estimated using Cochran's heterogeneity statistic (Q) and assessed using the chi-square distribution. A p < .05 for the Q statistic was sufficient to reject the null hypothesis of homogeneity, suggesting that heterogeneity of effect sizes across studies existed and allowing for exploration of possible mediator or moderator variables using the analysis of variance or regression analogs commonly used in meta-analysis (Hedges & Olkin, 1985; M. W Lipsey & Wilson, 2001). The degree of inconsistency ([I.sup.2]) was also computed (Higgins, Thompson, Decks, & Altman, 2003). When Q is less than the degrees of freedom, a negative [I.sup.2] results and is subsequently set to zero. When interpreting [I.sup.2], Higgins et al. (2003) suggested the following rule-of-thumb interpretations: 0% means no inconsistency (total homogeneity), 25% is low, 50% is moderate, 75% is high, and 100% means total inconsistency (total heterogeneity). Thus, [I.sup.2] > 50% ordinarily suggests substantial heterogeneity and that the presence of mediator or moderator variables should be explored.

Power is a substantial concern in the present meta-analysis because meta-analyses with k < 20 study results are potentially underpowered (Cornwell, 1993; Cornwell & Ladd, 1993), making them prone to Type II errors. In the current meta-analysis, wait-list (k = 18) and TAU (k = 18) studies approach the minimum of 20 sample results, so although not optimal, they may yield reasonably replicable results. The single-group (k = 5) and placebo (k = 2) condition results are probably underpowered and should be interpreted with caution.

Publication Bias

Four procedures were applied to assess and adjust for potential cautions related to publication bias. First, funnel plot analysis was conducted on each set of effect sizes to graphically detect publication bias, and several effect size estimates in the wait-list and TAU distributions were outside of the generally expected pattern. These potential outliers were all low or negative effect size estimates, rather than high, inflated estimates. Although it is possible that with a larger number of studies these possible outliers would conform to the expected funnel-shaped distribution, the decision was made to retain these potential outliers rather than trim or remove them. Because these potential low-end outliers were retained in the analyses, it is even more likely that concerns over publication bias can be mitigated and that the results can be viewed as conservative estimates when combined across studies. A second check for publication bias can be seen in the percentage of candidate studies with effect sizes greater than zero: 50% of placebo studies (one of two), 60% of single-group studies (three of five), 67% of TAU studies (12 of 18), and 89% of wait-list studies (16 of 18) were significantly higher than zero. Thus, only the wait-list distribution of effect size estimates seems unusual. As a third check, publication bias estimates for file drawer effects are presented in the Results section and were computed using Rosenthal's (1979) fail-safe N procedure. Finally, appropriate cautions should be given because of the fewer than optimal number of studies in each analysis (< 20, Cornwell, 1993; Cornwell & Ladd, 1993), leading to potentially underpowered results and an increase in Type II errors. Because no data were censored or effect size estimates eliminated, bias in the following analyses should be negligible.

* Results

Figure 1 provides the step-by-step decision-making process for article selection. We identified 420 relevant articles through computerized search procedures and an additional 17 through hand searches, for a total of 437 relevant articles. Initial review eliminated 383 articles for violation of at least one of the inclusion criteria, and an additional 12 articles were excluded after review of the full text. In total, a selection agreement rate of 97.2% between the two selectors was achieved ([kappa] = .93).


Study Characteristics

Of the final 42 articles advanced to the coding process, five were single-sample pretest--posttest designs, 35 used randomized samples, and we were unable to determine whether randomized assignment was used in two studies (Fine, Forth, Gilbert, & Haley, 1991; Nolan et al., 2002). Summary characteristics of these 42 studies are provided in Table 1, including purpose, sample size, mean age, percentage of males in the sample, percentage of White participants in the sample, control/comparison group type, and depression outcome measure(s) used. The total number of participants was 3,472, and 28 of the studies (67%) were conducted in the United States. Of the 42 studies, 16 (38%) were conducted in a school setting and 17 (40%) were conducted in an outpatient setting. Twelve of the studies (29%) used an individual treatment approach, 22 (52%) used a group approach, and the remaining eight studies (19%) used a mixed or family approach to treatment. The median number of discrete counseling/therapy sessions held per study was 12, and the range of number of sessions held was four to 30. The median length of sessions was 60 minutes, and the range was 35 minutes to 120 minutes.

Intercoder agreement across the 25 coded variables ranged from 71% to 100%, with a median percentage agreement of 97%. Related kappas ranged from .43 (whether blind assessment procedures were used) to 1.00, with a median kappa of .93. Landis and Koch (1977) indicated that kappas in the range of.41 to .60 were moderate and ordinarily sufficed for research purposes, whereas kappas of.61 to .80 were considered substantial, and kappas of.81 to 1.00 were almost perfect.

Is Counseling/Psychotherapy Effective for School-Age Youth With Depression?

The effectiveness of counseling or psychotherapy at termination (posttest) was evaluated using single-group studies (k = 5), and wait-list (k = 18), placebo (k = 2), and TAU (k = 18) comparison group studies. M. W. Lipsey and Wilson (1993) suggested that mean difference and mean gain effect sizes could be interpreted using the following rule-of-thumb designations: 0 means no effect, [less than or equal to] .30 is a small effect, .50 is a medium effect, and [greater than or equal to] .67 is a large effect. An alternative way of interpreting effect sizes involves transformation of the effect size, which conforms to a z-score distribution, to a percentile rank. Thus, a percentile rank of 50 indicates no effect of treatment, whereas percentile ranks greater than 50 indicate increasing degrees of effectiveness. Conversely, percentile ranks lower than 50 indicate ineffective or even harmful effects of treatment. For example, an effect size of.25, a small effect, means that the average person in the treatment group performed better than 60% of the comparison group participants, whereas an effect size of .50, a medium effect, means that the average person in the treatment group performed better than 69% of the comparison group participants.

Three of the four conditions (i.e., single group, wait list, and TAU) yielded average weighted effect sizes (d+) that were significantly higher than zero at termination, meaning that the treatments for depression were effective, and all tests of homogeneity (Cochran's Q and[I.sup.2]) indicated significant homogeneity and no effects of moderating or mediating variables. Only the placebo condition did not indicate a significant effect of treatment, but only two placebo controlled trials were located. Table 2 provides a summary of these effect size statistics.

Single-group studies. Five groups across four studies (Barrera, Chung, Greenberg, & Fleming, 2002; Kovacs et al., 2006; Miller, Gur, Shanok, & Weissman, 2008; Thompson et al., 2007; n = 62) combined to yield an average corrected effect size (d+) of .36 with a 95% CI range of .17 to .54. Because the entire 95% CI range is greater than zero, the null hypothesis of d+ = 0 can be rejected and one can conclude that d+ is greater than zero. Thus, counseling/psychotherapy at termination is significantly more effective than baseline measurement for school-age youth with depression in these five single-group studies. A d+ of .36 is a small to medium effect and means that the average participant at termination scored at the 64th percentile of the pretest score distribution. A d+ of .36 in a five-study analysis has a fail-safe N of 178, meaning that one would need to locate 178 single-group studies with an effect size of zero not included in this meta-analysis in order to mitigate this study's result and reduce the d+ of .36 down to a nonsignificant d+ of .01. In testing for homogeneity of the set of effect sizes from these five studies, we found that Cochran's Q was 4.68. With four degrees of freedom, this Q statistic value resulted in p > .05, meaning that the null hypothesis of homogeneity was retained and the set of five effect sizes is consistent and homogeneous. To corroborate this analysis, we computed [I.sup.2] to be 14%, also indicating homogeneity (i.e., [I.sup.2] < 50%). Because the distribution of effect sizes was homogeneous, there was no need to conduct moderator or mediator analyses.

Wait-list comparison groups. Eighteen studies (n = 1,520) reported wait-list comparisons for an average d+ = .55, 95% CI [.41, .69]. This d+ is greater than zero and has a fail-safe N of 9,882 studies, and so it is quite a robust result; that is, it is quite unlikely that almost 10,000 wait-list comparison group studies of youth depression demonstrating no effect of treatment (d' = .00) were not located. The homogeneity analysis indicated Q(17) = 16.03, p > .05, which was corroborated by [I.sup.2] = 0%. Thus, this set of wait-list effect sizes was homogeneous. A d+ of .55 is medium to large and indicates that the average treatment group participant was less depressed than were 71% of the wait-list control participants at termination.

Placebo comparison groups. Only two studies (Liddle & Spence, 1990; March et al., 2004; n = 122) reported using a placebo condition, yielding a weighted d+ of .01, 95% CI [-.37, .38], indicating no effect of treatment in comparison with the placebo group mean (percentile rank = 51). The two effect sizes were homogeneous, Q(1) = 0.61, p > .05, [I.sup.2] = 0%. It is important to note that only two placebo studies could be retrieved from the literature: Liddle and Spence (1990) with a d' of .36 for a sample size of 21, and March et al. (2004) with a d' of-.06 for a sample size of 223 in the Treatment for Adolescents With Depression Study (TADS). The Liddle and Spence study result was within the expected outcome range, given the wait-list and TAU results, whereas the TADS outcome indicated that the CBT treatment group actually performed worse than did the group in the placebo control condition.

TAU comparison groups. The mean difference comparisons for 18 studies reporting TAU comparisons (n = 1,358) combined for a d+ = .29, 95% CI [. 16, .43], which was significantly higher than zero and resulted in a fail-safe N of 529 studies. The average participant in the treatment condition, which in all but two instances was a cognitive behavior treatment, performed at the 62nd percentile of the TAU comparison group distribution, a small effect of treatment. The distribution of effect sizes was very homogeneous, Q(17) = 18.25, p > .05, [I.sup.2] = 14%.

Do the Effects of Counseling/Psychotherapy Last for School-Age Youth With Depression?

The effectiveness of counseling and psychotherapy at the most distant point of follow-up assessment was evaluated using single-group studies (k = 3) and wait-list (k = 9) and TAU (k = 9) comparison group studies. Each of these effect size groupings is analyzed separately to determine the likely staying power of therapeutic treatment effects on depression levels in the months or years after termination.

All three conditions yielded average weighted effect sizes (d+) that were significantly higher than zero at follow-up intervals, and all tests of homogeneity (Cochran's Q and [I.sup.2]) indicated significant homogeneity and no effects of moderating or mediating variables. Table 2 provides a summary of these follow-up effect size statistics.

Single-group studies. Three groups across three studies (Kovacs et al., 2006; Miller et al., 2008 [Group B]; Thompson et al., 2007; n = 33) combined to yield an average corrected effect size (d+) of .46 with a 95% CI range of .08 to .84, which indicates a treatment effect greater than zero. The amount of time lapsed between the termination of treatment and greatest follow-up measurement ranged from 5 to 12 months. This average d+ is a medium effect and indicates that the average participant at follow-up performed at the 68th percentile of the pretest distribution mean score. The fail-safe N for this analysis was 137.7 studies, meaning that 137 unpublished or unfound studies with d' = 0 would need to be located in order to reduce the d+ of .46 down to an insignificant effect size of .01, which is unlikely. The distribution of three effect sizes was homogeneous, Q(2) = 2.38, p > .05, [I.sup.2] = 16%; therefore, no moderator analysis was justified. It is interesting that the two studies reporting a 5- to 6-month follow-up (Kovacs et al., 2006; Miller et al., 2008) yielded a combined d+ of .32, whereas the two studies reporting a 9- to 12-month follow-up (Kovacs et al., 2006; Thompson et al., 2007) yielded a combined d+ of .65. Although these results must be viewed with caution because of the small sample sizes and numbers of studies, some evidence exists that treatment gains were maintained up to 1 year after termination; that is, the posttest d+ = .53 (k = 3), half-year follow-up d+ = .32 (k = 2), and approximately one-year follow-up d+ = .65 (k = 2) progression was observed for the three studies reporting follow-up effect sizes.

Wait-list comparison group studies. Effect of counseling and psychotherapy at longest follow-up for the nine wait-list studies providing follow-up data (Mdn = 6 months; range = 1- to 12-month follow-up; n = 923) when compared with wait-list controls resulted in a d+ = .29, 95% CI [.08, .49], which means that the follow-up mean d+ is greater than zero, and the fail-safe N was 257.4 studies, a robust result. A d+ of .29 is a small effect and means that the average participant in the treatment condition remained less depressed than did approximately 61% of the wait-list control participants after termination. This distribution of nine effect sizes was homogeneous, Q(8) = 13.22, p > .05, [I.sup.2] = 40%; therefore, no analysis for moderator or mediator variables was conducted. Thus, treatment of depression in school-age youth maintains an effective result after termination when compared with waitlist conditions. Not surprisingly, as the follow-up study time after termination and the initial follow-up increased, the effect sizes were diminished: d+ at termination was .42 (k = 9), d+ at 1 to 4 months was .71 (k = 4), d+ at 6 months was .23 (k= 3), and d+ at 9 to 12 months was .09 (k = 2). The correlation between time of follow-up and effect size was r = -.59 (p < .10, n = 9), but these results should be interpreted with caution because of the small number of studies involved.

TAU comparison group studies. Mean difference comparisons for the nine follow-up TAU studies resulted in a d+ of. 16, 95% CI [.08, .25], which was significantly higher than zero (percentile rank = 56) and a robust result with a fail-safe N of 144.9 studies. These nine studies had a total sample size of 743 participants, and the follow-up lapsed time periods ranged from 6 to 24 months, with a median length of 9 months. A recognizable pattern of effect sizes did not emerge as they did earlier because the average effect sizes from termination through the 24-month follow-up study time stayed relatively stable: d+ at termination was .22 (k = 18), d+ at 6 months was .12 (k = 4), d+ at 9 to 12 months was .18 (k = 3), and d+ at 24 months was .23 (k = 2), resulting in a correlation between time of follow-up and effect size of -.04 (p > .05, n = 9). As previously mentioned, these results should be interpreted with caution because of the small number of studies involved. Homogeneity of the TAU follow-up effect sizes distribution was negligible, Q(8) = 8.99, p > .05, [I.sup.2] = 11%, so there was no need for moderator analysis.

Is There a Difference Between School-Based and Clinic-Based Results for School-Age Youth With Depression?

The simple answer to this question is no. Because all the aforementioned analyses resulted in conclusions of homogeneity among effect size estimates, no moderator or mediator variables, including treatment setting, were in operation. However, because this was a primary research question of this rectaanalysis, it is instructive to present the mean effects across the various treatment conditions to underscore the point that there was no difference between treatments conducted in the school setting and treatments conducted in outpatient settings. Unfortunately, results are available only for the wait-list and TAU conditions because of the small number of studies in the analysis, but enough studies were present to yield preliminary estimates for both the posttest and follow-up time periods. More replication studies conducted in schools and clinics in the future will add power to future analyses.

Wait-list condition posttest results. Eight studies reported school-based results (n = 878) combining for a school-based d+ of .45, 95% CI [.25, .64], yielding a medium effect size (percentile rank = 67) with a fail-safe N of 356 studies. The outpatient sample was composed of seven studies (n = 529) and yielded a d+ of .64, 95% CI [.41, .87], a medium to large result (percentile rank = 74) with a fail-safe N of 449.4 studies. Although the magnitude of the outpatient d+ was a bit higher than the magnitude of the school-based d+, the difference was not significant, Q(1) = 1.61, p > .05.

TAU condition posttest results. Six TAU studies reported school-based results (n = 565) combining for a d+ = .41, 95% CI [.19, .63], meaning that the effect of school-based treatment was small to medium (percentile rank = 66) with a fail-safe N of 244.8 studies. The outpatient sample was composed of 12 studies (n = 793) and yielded a d+ of .23, 95% CI [.06, .39], a small effect (percentile rank = 59) with a fail-safe N of 270 studies. This time the magnitude of the school-based d+ was higher than the magnitude of the outpatient-based d+, although once again the difference was not significant, Q(1) = 1.69, p > .05.

Wait-list condition follow-up result. Six school-based waitlist studies (n = 686) conducted follow-up measurements of depression, which combined for a d+ of .27, 95% CI [.04, .51], a small effect (percentile rank = 61) with a robust fail-safe N of 139.8 studies. Only two outpatient wait-list studies (n = 115) were conducted resulting in a d+ of .33, 95% CI [-.15, .82], also a small effect (percentile rank = 63) with a fail-safe N of 66.6 studies. There was no significant difference between the school-based and outpatient-based wait-list follow-up results, Q(1) = 3.28, p > .05.

TAU condition follow-up result. Only four school-based TAU follow-up studies were located (n = 407) yielding a d+ of .15, 95% CI [.05, .26], which is a very small effect size (percentile rank = 56) with a fail-safe N of only 59.6 studies. Five outpatient TAU follow-up studies (n = 336) were included in an analysis resulting in a d+ of.19, 95% CI [.03, .35], another small effect size (percentile rank = 57) with a fail-safe N of 96 studies. There was no significant difference between the school-based and outpatient-based wait-list follow-up results, Q(1) = 0.83, p > .05.

* Discussion

The findings of this meta-analysis of 42 clinical trials suggest that counseling and psychotherapy are effective in the treatment depression in school-age children and adolescents at the termination of treatment. The effect sizes, derived using a random effects model on studies published between 1990 and 2008, were small to medium-sized. The results of this meta-analysis were consistent with the results of Weisz, McCarty, and Valeri's (2006) meta-analysis of 35 studies of psychotherapy for depression treatment in youth, which reported a d+ of .34 using the random effects model, also a small to moderate effect. Weisz, McCarty, and Valeri's study searched for clinical trials as far back as the 1860s and combined the effects of comparison studies if the comparison group consisted of "at least one untreated, waitlist, minimally treated, or active placebo control group" (p. 135), in effect, combining dissimilar comparison group effect sizes and lowering the overall general effect size. Weisz, McCarty, and Valeri's passive control comparison (equivalent to this study's wait-list posttest result) was a d+ of .41, close to the d+ of .55 reported in the present study. Weisz, McCarty, and Valeri also found that TAU comparisons resulted in a d+ of .24, also very close to the d+ of .29 reported in the present study. Thus, differences between the present meta-analysis and Weisz, McCarty, and Valeri's seem negligible and are likely due to differences in the studies selected for analysis. However, an essential difference between the two studies is that the current study actually sorted effect sizes according to specific comparison group conditions and analyzed each group of effect sizes separately, thus maintaining methodological and interpretive purity.

Other meta-analysts have reached basically similar conclusions about posttest effect sizes of depression studies. Sheard and Maguire (1999) identified 30 randomized trials of depression treatment in cancer patients and, using a random effects model, derived an effect size of .36, very consistent with the results of this study and Weisz, McCarty, and Valeri's (2006) meta-analysis. Furthermore, Watanabe et al. (2007) studied 35 comparisons across 27 randomized controlled studies using RR indices and concluded that psychotherapy was superior for wait-list and placebo comparison groups, but not TAU.

Effect sizes in the range of .25 to .55 for the present study and the Weisz, McCarty, and Valeri (2006) study are far lower than the effect sizes reported in previous meta-analyses of depression treatments in youth by Reinecke et al. (1998; effect size = .97), Lewinsohn and Clarke (1999; effect size = 1.27), and Michael and Crowley (2002; effect size = .72). However, each of these earlier meta-analyses used a fixed effects model to compute effect size averages and different meta-analytic statistical procedures. The earlier studies also varied in the degree of study quality, sample sizes, and number of studies used in the analysis. The effect sizes reported for the present meta-analysis are far more conservative than these earlier studies, were derived from more rigorous study selection and statistical methodology, and are much better estimates of the results clinicians in practice are likely to experience on average.

These more modest results require clinicians to rethink the overall effectiveness of depression treatments for school-age children and adolescents. No longer is it a given that the effectiveness of depression treatments exceed the general effectiveness of counseling and psychotherapy for other mental health issues (i.e., effect sizes higher than .70). Indeed, counseling-based treatment of depression may actually produce less effective results than experienced for other psychological conditions, such as anxiety, behavior disorders, and substance abuse. Of course, caution is warranted until direct comparisons of these conditions using similar methodology and time overlaps are conducted (Weisz, McCarty, & Valeri, 2006).

A second primary conclusion of the present meta-analysis is that the depression treatments seem to have significant staying power for perhaps up to 2 years. When combined for analysis of longest follow-up assessment, d+ for wait-list, TAU, and single-group studies were all significantly higher than zero. The long-term effects of counseling for depression require far more study because the number of studies reporting mid- to long-term follow-up results is small and these studies generally conclude that the further one proceeds after termination, the lower the effectiveness of treatment in minimizing depressive symptoms. Unfortunately, at this time, far too few follow-up studies exist to allow sufficient power to consistently detect long-term effects. Many of the average effect sizes in the Follow-Up Studies Continuum section in Table 2 are composed of the results of only two or three studies. In numerous instances, if even twice the number of follow-up studies of similar magnitude existed, the results would have been significantly higher than zero, rendering a tentative conclusion of long-term treatment effectiveness. Both this study and Weisz, McCarty, and Valeri's (2006) study reported correlation coefficients of approximately -.50 between effect size and length of time that passed after treatment termination.

In this study, as can be seen in Table 2, wait-list gains were maintained for only a few months and disappeared by 6 months after termination. Single-group gain studies, although not statistically significant at 6-month follow-up, were still substantial and significant (d+ = .65) at 12-month follow-up. TAU gains were not significant at 6-month follow-up but were significant at 1- and 2-year follow-up points. In their meta-analysis, Weisz, McCarty, and Valeri (2006) determined that therapeutic gains were sustained for several months, but virtually disappeared by 1 year after termination and were no longer significant. Watanabe et al. (2007) concluded that psychotherapy was no longer effective even 6 months after treatment ended. A primary obstacle to more definitive conclusions about the staying power of counseling interventions is the lack of a significant number of follow-up studies. It is difficult to reach reliable and valid conclusions when only nine of the original 42 studies provide a 6-month follow-up, only seven provide a 1-year follow-up, and only two extend the follow-up analysis to 2 years.

A third general conclusion from this meta-anatysis is that treatments implemented at school are just as effective as treatments conducted in an outpatient setting. Although this is the first study to directly test this hypothesis using meta-analytic procedures, a number of previous meta-analyses coded for treatment setting and did not identify setting as a moderating or mediating variable. School-based termination effect sizes were d+ = .45 (k = 8) for wait-list comparisons and d+ = .41 (k = 6) for TAU comparisons. Thus, counselors can be just as confident with the effectiveness of counseling procedures to treat students with depression at school as they can in clinics, private practices, and counseling laboratories. Like Weisz, McCarty, and Valeri's (2006) meta-analysis, the results of the present study also give practitioners confidence that treatment "procedures actually work with clinically referred youths, treated by clinical practitioners, in clinical practice settings" (p. 145), given that clinic-based termination effect sizes were d+ = .64 (k = 7) for wait-list comparisons and d+ = .23 (k = 6) for TAU comparisons. Follow-up comparisons for each group across each setting were less pronounced but were still significantly higher than zero.

The presence of homogeneity in all analyses was an unexpected but meaningful result. Homogeneous effect size distributions mean that moderator and mediator variables were not in operation; thus, the studies were robust across various treatments, client demographics, settings, and treatment characteristics. For example, one can conclude that the number of sessions (a measure of treatment potency) did not influence the results, indicating that brief treatments (e.g., four to eight sessions) were just as effective as longer term treatments (e.g., 20 to 30 sessions). Likewise, treatments seemed to be equally effective for children and adolescents, boys and girls, and White participants and participants of color. Group approaches were also as effective as individual approaches. Thus, although modest, these results are robust across diverse treatment, client, and study characteristics.

These three major conclusions are probably generalizable across relevant populations, treatment variations, outcome variables, and research designs. For example, most of the 42 studies included in this review used CBT approaches for the treatment of depression (k = 24, 57.1%). Of those CBT-based studies, 14 used wait-list comparison groups, two used placebo groups, and eight used a TAU comparison group. Interpersonal therapy was used in nine studies (21.4%), including two single-group studies, three wait-list studies, and four TAU comparison groups. The only other general treatment approach used more than once was family-based interventions, which occurred in three studies (7.1%). None of these three approaches were significantly more effective than the others. Similar to Weisz, McCarty, andValeri (2006), we found no significant differences among the various treatment approaches. However, the majority of these approaches emphasized changing cognitions, even though noncognitive approaches have been shown to be just as effective as cognitive approaches. There also were no significant differences in effect sizes between individual and group approaches.

Limitations of This Metastudy

This meta-analysis used the following rigorous methodological procedures: (a) extensive searches of published literature, (b) primarily standardized outcome measures, (c) nine-part inclusion criteria, (d) random effects model, (e) weighting of effect sizes for inverse variance to account for variability in sample sizes, (f) testing for homogeneity using both Cochran's test and F, and (g) assessment for publication bias using both funnel plots and computation of fail-safe Ns.

Still, as with any research design, a number of potential study limitations should be considered. Rigorous inclusion criteria may have eliminated viable studies of less methodological rigor, and using only published studies may limit generalizability of conclusions and lead to a publication effect. It is important to remember that the validity of results based on effect sizes is largely dependent on the quality of the studies included in the meta-analysis (Whiston, Tai, Rahardja, & Eder, 2011), but less rigorous or unpublished studies may result in somewhat different conclusions.

Perhaps the biggest limitation was the small number of studies in the single-group condition (k = 5) and the presence of only two placebo-controlled studies. Conclusions based on such small samples of studies always warrant caution, and numerous additional studies are needed to derive greater understanding of the effectiveness of depression treatment in school-age youth. Eighteen wait-list and an equal number of TAU condition studies were located, but more are still needed to boost power and avoid Type II errors (Cornwell, 1993; Cornwell & Ladd, 1993). One specific research design problem is worthy of special note: Less than half of the studies provided follow-up studies months after termination and even fewer for years after termination. Although some researchers believe that it is important to demonstrate only the effectiveness of the treatment under study, an even more important humanistic, societal, and professional issue is whether the intervention creates long-term change in clients, thus minimizing suffering and reducing costs of further treatments in the years after the intervention was terminated.

Several methodological issues in the individual clinical trials were problematic. Only half of the clinical studies selected for inclusion in this meta-analysis used a standardized treatment manual. Therefore, close to half of the studies probably provided insufficient information to replicate the treatment in another research study or clinical practice. Essential information on potential moderating variables was often missing from a study's description, and many authors did not report important basic information on design, sample, and treatment protocols. Methods used for diagnosis and inclusion of participants in a clinical trial varied widely, meaning that some potential participants would be included in one study but excluded from another; yet, all of these studies were combined in the same meta-analysis. Ultimately, diverse study characteristics can lead to less meaningful results when studies are combined. However, the homogeneity demonstrated in each analysis gives confidence that the diverse studies led to similar results. For example, the question of whether the effects of counseling were the same in school-based versus outpatient trials was answered with the conclusion of no differences.

Finally, there was some evidence of potential publication bias and outlier studies in this analysis. As mentioned earlier, these studies were retained rather than eliminated or trimmed, primarily because they yielded lower effect size estimates and therefore led to more conservative results.

Implications for Counseling Practice

Meta-analyses on the effects of counseling treatments for depression in school-age youth are consistently concluding that counseling is effective at termination, yielding at least a small to medium effect depending on the comparison groups under study (Watanabe et al., 2007; Weisz, McCarty, & Valeri, 2006). Thus, counseling and psychotherapy exist as an important treatment approach for depression, just as important as antidepressant medications, many of which have serious, even dangerous, side effects (March et al., 2004; Whittington et al., 2004). Counseling and psychotherapy can also be used is concert with medication, given that a combined approach seems most efficacious over the short term (Kennard et al., 2006; March et al., 2004).

Evidence of staying power at 6-month and 1-year follow-ups also exists, although a number of previous meta-analyses did not reach this same conclusion (e.g., Watanabe et al., 2007; Weisz, McCarty, & Valeri, 2006). One viable explanation for this inconsistent conclusion may simply be that only approximately half of clinical trials include a short-term follow-up (up to 6 months) and only a small percentage include a longer term follow-up of 1 year or more. Of the 42 studies in this meta-analysis, only 22 presented follow-up data, and of these, five presented follow-up data from 1 to 4 months after termination, nine from approximately six months after termination, seven from approximately a year after termination, and only two from 2 years or longer.

What is clear is that researchers should be focused on identifying or creating effective interventions with longer term staying power. Effective approaches may involve an intervention that maintains a more significant effect size over a longer period of time or postintervention booster sessions that help clients maintain treatment effects for longer periods of time. Clarke, Rohde, Lewinsohn, Hops, and Seeley (1999) and Weissman (1994) believed that treatment benefits could be expanded to greater lengths of time through booster sessions or treatment continuation, but few researchers have followed up to determine optimal efficiency for the long-term efficacy of these approaches.

Better answering the question of staying power will help clinicians determine whether booster or follow-up sessions may be necessary to maintain the significant gains demonstrated at the conclusion of the treatment phase. This issue penetrates to the core of the cost-effectiveness question of counseling and psychotherapy; if a client is willing to incur the expense to significantly and effectively reduce the degree of depressive symptoms experienced, is it not cost-effective to periodically assess and boost the initial treatment effect to maintain the demonstrated gains? Although the current analysis does not necessarily allow for a direct answer to this question, it may be helpful to frame the question this way. When a patient begins a treatment regimen of antidepressant medication, the patient usually takes the medication for at least months, and often for years. Studies exploring the cost-effectiveness of counseling and psychotherapy with posttreatment booster sessions versus long-term medication management may reveal important information regarding treatment efficacy, cost, and severity of side effects. For example, the TADS (Kennard et al., 2006; March et al., 2004) concluded that CBT plus fluoxetine produced a more effective long-term effect than did CBT alone. When long-term effects of medication and psychotherapy were tested separately, the TADS group indicated that neither treatment led to long-term improvement. Conversely, Imel et al. (2008) concluded that psychotherapy was superior to medication at follow-up after termination of both treatments. More clinical trials with long-term follow-up phases will help clarify these inconsistent findings in the current literature.

Most of the approaches reported in the clinical trials used in this meta-analysis were short term, with a median of 12 sessions. Westen and Morrison (2001) suggested that new approaches should be explored to find more effective approaches with longer term effects but questioned the efficacy of shorter term approaches when treating moderate to severe cases of depression and dysthymia. They concluded that "most treatments for depression in naturalistic samples unconstrained by managed care take roughly half a year for CBT and upward of 1 to 2 years for other forms of therapy ... [and that] therapeutic length doubles ... in the presence of comorbid conditions" (p. 887). Westen and Morrison went on to indicate that only approximately 50% of patients with depression improve with only 3 to 4 months of treatment. Likewise, Imel et al. (2008) suggested that the treatment of dysthymic disorder requires substantially more than a dozen hour-long sessions, concurring with McCullough (1991) that an average of 31 sessions is more likely needed to treat clients with dysthymia into remission.

Mental health practitioners should document any adverse effects of counseling and psychotherapy in the treatment of depression in school-age youth, such as suicidal thoughts and behaviors. With significant safety concerns regarding the use of antidepressant medications now in place (AACAP, 1998; Costello et al., 2002; Ramchandani, 2004; Vitiello & Swedo, 2004; Whittington et al., 2004), counseling and psychotherapy seem to have a substantial advantage when it comes to client safety and well-being. The fact that the effect sizes for counseling and psychotherapy are quite similar to antidepressant medication in head-to-head clinical trials both at termination and follow-up should give patients and clinicians confidence in terms of both effectiveness and safety. According to Imel et al. (2008), "psychotherapeutic treatment may often be preferable as it is relatively brief, appears to be accompanied by a protective effect not offered by medication, and does not carry the complications associated with medications" (p. 204).

Implications for Future Counseling Research

A great need exists for more clinical trials that measure the effectiveness of counseling and psychotherapy, especially upon long-term follow-up. We know that treatments for depression outperform wait-list comparison groups at termination but need more evidence that treatments are consistently superior to placebo and TAU conditions, otherwise we can have little confidence in the effectiveness of counseling or psychotherapy over and above that which can be gained through interpersonal support and sparkling, friendly conversation. Enough wait-list and TAU studies have been conducted to indicate that modest effect sizes likely exist upon termination, but these seem to be followed by somewhat diminished effects in 6 months and beyond, probably because the number of short- and long-term follow-up studies is so small that the longer term effects cannot be reliably estimated. Imel et al. (2008) suggested that although follow-up studies suffer from many threats to internal and external validity, they will be vital components of future research to answer continuing questions over the long-term effects of psychotherapy for not just depression treatments, but the wide range of mental and emotional disorders.

The need for better follow-up data over longer periods of time does raise potential ethical concerns, primarily that it is unethical to deprive participants of an active, effective treatment for long periods of time after termination of the initial clinical phase of the study. However, the TAU comparison condition is a solution to this ethical dilemma. As Weisz, McCarty, and Valeri (2006) pointed out, "no treatment and waitlist, currently the most commonly used control conditions, are not only the weakest experimentally but also the most difficult to sustain throughout the waitlist period, given ethical and humane concerns" (p. 146). Using a TAU comparison condition provides comparison participants with the active treatment they would have received even if they had not agreed to participate in the study.

Given the small to medium effect size average resulting from this analysis, three courses of further research activity seem prudent to enhance the effects of treatment: (a) strengthening the dose of current treatments, (b) packaging various effective treatments together to boost the overall effect, or (c) developing new approaches that are more effective and have more lasting effects (Weisz, McCarty, & Valeri, 2006). Out of all of the comparisons in this meta-analysis, four studies (Ackerson, Scogin, McKendree-Smith, & Lyman, 1998; Kahn, Kehle, Jenson, & Clark, 1990; Lewinsohn, Clarke, Hops, & Andrews, 1990; Young, Mufson, & Davies, 2006a) produced effect sizes of greater than 1.00 at termination and three studies (De Cuyper, Timbremont, Braet, De Backer, & Wullaert, 2004; Kahn et al., 1990; Thompson et al., 2007) produced follow-up effect sizes of greater than 1.00. Thus, treatments yielding large effect sizes do exist, and, perhaps, experimentation with a potent treatment or combinations of potent treatments may lead to better therapeutic results over the long term. Therefore, two important questions require further study: (a) How should clinicians best manage the treatment of depression over the long term? and (b) What should be the role of counseling/psychotherapy and medication in the short- and long-term treatment of depression?

The quality and description of studies that get published in journals is crucial, and currently many authors are not adequately describing sample characteristics such as sex, age, and race, let alone important characteristics of study design and procedure that would allow successful study replication or the use of the prescribed treatment in clinical practice. If a study shows a depression treatment to be effective, but practitioners reading the article have no idea how to replicate the treatment with their own clients, what real good is accomplished by publishing the article in the first place? Likewise, better "intent to treat" analyses are needed, because only approximately one third of the clinical trials in the present meta-analysis reported this information; thus, researchers do not know if the participants who drop out of a study do so because they perceived that they were not benefiting from the treatment. Removal of these attrition cases would lead to inflated effect size estimates because only those participants who are getting better would stay in the study and have their scores reflected in the group means. Journal editors and editorial board members should insist on inclusion of this essential information. Authors should use standardized treatment protocols and report specific characteristics of the study that will allow readers to understand the methodological rigor used and how to replicate the procedure with clients.

Regarding the use of outcome variables, it would be interesting to note whether there is a difference in self-report results (i.e., Does the youth think that he or she is improving?) versus parent report results (i.e., Does the parent think that the youth is improving?). Weisz, McCarty, and Valeri (2006) indicated that parent-report findings did not indicate significant results, whereas self-report outcome measures did indicate significant differences. This phenomenon has also been reported in individual studies (Angold & Costello, 1993; Capaldi & Stoolmiller, 1999).

Few clinical trials and no meta-analyses to date have studied the cost-effectiveness of either counseling/psychotherapy or medication in the treatment of depression in school-age youth. Mental health practitioners and researchers need to engage in much more in-depth study to determine the cost-effectiveness of counseling and psychotherapy interventions and how counseling and psychotherapy compare with medication interventions.

No clinical trials or outcome studies on childhood depression were identified for inclusion in the counseling literature, such as might be expected from the Journal of Counseling & Development, Professional School Counseling, or the Journal of Mental Health Counseling. Nearly all selected studies came from journals published by the American Psychological Association, psychiatric journals, or niche journal studying affective disorders. Counseling researchers need to focus more on counseling outcomes across a range of clinical concerns, including depression, and clinical trials of counseling effectiveness are significant studies that inform practice and bolster the scientific basis of the counseling profession.

Finally, it is striking that widespread implementation of cost-effective medical procedures regularly occurs with effect sizes of .30 and even far less (see McCarty & Weisz, 2007), but that calls for treatment of a condition such as depression go unheeded, even though as many as 30% of youth by age 18 years will present with significant symptoms of depression. Many, and perhaps most, will go untreated. The treatment effectiveness of counseling and psychotherapy at termination is at least small to medium and occurs without the safety concerns associated with antidepressant medication. Although much research needs to be conducted to further clarify these issues, counseling and psychotherapy have emerged as viable, effective choices for the treatment of depression in school-age youth.
Test to Earn CE Credit

Note: Correctly completing 3 of 3 test questions earns 1 continuing
education contact hour.

* Counseling Outcomes From 1990 to 2008 for School-Age Youth With
Depression: A Meta-Analysis (JCD, Volume 89, Number 4, Fall 2011)

* Learning Objectives

Reading this professional article will help you

1. Assess the level of effectiveness of counseling and
   psychotherapy with school-age youth with depression.

2. Analyze the lasting effectiveness of counseling and
   psychotherapy for youth with depression.

3. Examine needs for future clinical trials and meta-analyses.

* Examination Questions

1. What percentage of youth will have met the criteria
   for major depressive disorder at least once by age 18?

[] a. Less than 10%
[] b. 10% to 15%
[] c. 20% to 25%
[] d. 30% to 35%

2. Clinical trials to determine whether counseling and
   psychotherapy are effective with school-age youth
   over the past several decades have

[] a. Led to mixed results.
[] b. Shown positive outcomes.
[] c. Shown negative outcomes.
[] d. Been lacking in numbers sufficient to draw
      reasonable conclusions.

3. According to the authors, few clinical trials and no
   meta-analyses have studied the

[] a. Preparation of counselors to address depression
      in school-age youth.
[] b. Cost-effectiveness of either counseling/
      psychotherapy or medication in treatment of
      depression in school-age youth.
[] c. Management of depression in school-age
      youth on into their adult years.
[] d. All of the above

* Learning Assessments

On the basis of my reading of this article, I am able to

1. Comprehend the effectiveness of counseling and psychotherapy
   in the treatment of depression in school-age

   Strongly Agree          Agree         Strongly Disagree
         5           4       3      2            1

2. Study the various treatment approaches used in school
   and clinical settings.

   Strongly Agree          Agree         Strongly Disagree
         5           4       3      2            1

3. Assess the implications for future counseling research.

   Strongly Agree          Agree         Strongly Disagree
         5           4       3      2            1

Date: --

Signature: --

* Instructions

Online: Complete the test online at You will be able
to pay online and download your CE certificate immediately.

Mail: Complete the test and form above and mail (with check or
money order made payable to American Counseling Association) to:
ACA Accounting Department/JCD, American Counseling Association,
5999 Stevenson Ave., Alexandria, VA 22304. Allow 2-4 weeks for

For further assistance, please contact Debbie Beales at, or by phone at 800-347-6647, x306

Received 06/10/10

Revised 09/07/10

Accepted 11/23/10

* References

References marked with a single asterisk indicate studies included in the meta-analysis; references marked with a double asterisk indicate studies removed from the meta-analysis.

Achenbach, T. M., & Rescorla, L. A. (2001). Manual for the Achenbach System of Empirically Based Assessment (ASEBA). Burlington: University of Vermont, Research Center for Children, Youth, and Families.

* Ackerson, J., Scogin, F., McKendree-Smith, N., & Lyman, R. D. (1998). Cognitive bibliotherapy for mild and moderate adolescent depressive symptomatology. Journal of Consulting and Clinical Psychology, 66, 685-690. doi: 10.1037/0022-006X.66.4.685

American Academy of Child and Adolescent Psychiatry. (1998). Practice parameters for the assessment and treatment of children and adolescents with depressive disorders. Journal of the American Academy of Child and Adolescent Psychiatry, 37(Suppl.), 63S-83S.

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author.

Angold, A., & Costello, E. J. (1993). Depressive comorbidity in children and adolescents: Empirical, theoretical, and methodological issues. American Journal of Psychiatry, 150, 1779-1791.

* Asarnow, J. R., Scott, C. V., & Mintz, J. (2002). A combined cognitive-behavioral family education intervention for depression in children: A treatment development study. Cognitive Therapy and Research, 26, 221-229.

* Barrera, M., Chung, J. Y. Y., Greenberg, M., & Fleming, C. (2002). Preliminary investigation of a group intervention for siblings of pediatric cancer patients. Children's Health Care, 31, 131-142.

Baskin, T. W., Tierney, S. C., Minami, T., & Wampold, B. E. (2003). Establishing specificity in psychotherapy: A meta-analysis of structural equivalence of placebo controls. Journal of Consulting and Clinical Psychology, 71, 973-979.

Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Manual for the Beck Depression Inventory-Second Edition. San Antonio, TX: Psychological Corporation.

Birmaher, B., Ryan, N. D., Williamson, D. E., Brent, D. A., Kaufman, J., Dahl, R. E., ... Nelson, B. (1996). Childhood and adolescent depression: A review of the past 10 years, Part I. Journal of the American Academy of Child and Adolescent Psychiatry. 35, 1427-1439. doi: 10.1097/00004583-199611000-00011

* Bolton, P., Bass, J., Betancourt, T., Speelman, L., Onyango, G., Clougherty, K. F., ... Verdeli, H. (2007). Interventions for depression symptoms among adolescent survivors of war and displacement in Northern Uganda: A randomized controlled trial. Journal of the American Medical Association, 298, 519-527. doi:10.1001/jama.298.5.519

* Brent, D. A., Holder, D., Kolko, D., Birmaher, B., Baugher, M., Roth, C., ... Johnson, B. A. (1997). A clinical psychotherapy trial for adolescent depression comparing cognitive, family, and supportive therapy. Archives of General Psychiatry, 54, 877-885.

** Brent, D. A., Kolko, D. J., Birmaher, B., Baugher, M., & Bridge, J. (1999). A clinical trial for adolescent depression: Predictors of additional treatment in the acute and follow-up phases of the trial. Journal of the American Academy of Child and Adolescent Psychiatry, 38, 263-270.

** Butler, L., Miezitis, S., Friedman, R., & Cole, E. (1980). The effect of two school-based intervention programs on depressive symptoms in preadolescents. American Educational Research Journal, 17, 111-119.

Capaldi, D. M., & Stoolmiller, M. (1999). Co-occurrence of conduct problems and depressive symptoms in early adolescent boys: III. Prediction to young-adult adjustment. Development & Psychopathology, 11, 59-84.

Chambers, E. A. (2004). An introduction to met, a-analysis with articles from The Journal of Educational Research (1992-2002). The Journal of Educational Research, 98, 35-44.

Clarke, G. N., DeBar, L. L., & Lewinsohn, P. M. (2003). Cognitive-behavioral group treatment for adolescent depression. In A. E. Kazdin & J. R. Weisz (Eds.), Evidence-based psychotherapies for children and adolescents (pp. 120-134). New York, NY: Guilford Press.

* Clarke, G. N., Hawkins, W., Murphy, M., & Sheeber, L. B. (1995). Targeted prevention of unipolar depressive disorder in an at-risk sample of high school adolescents: A randomized trial of group cognitive intervention. Journal of the American Academy of Child and Adolescent Psychiatry, 34, 312-321.

* Clarke, G. N., Hornbrook, M., Lynch, F., Polen, M., Gale, J., Beardslee, W., ... Seeley, J. (2001). A randomized trial of a group cognitive intervention for preventing depression in adolescent offspring of depressed parents. Archives of General Psychiatry, 58, 1127-1134.

* Clarke, G. N., Hornbrook, M., Lynch, F., Polen, M., Gale, J., O'Connor, E., ... Debar, L. (2002). Group cognitive-behavioral treatment for depressed adolescent offspring of depressed parents in a health maintenance organization. Journal of the American Academy of Child and Adolescent Psychiatry, 41, 305-313.

* Clarke, G. N., Rohde, P., Lewinsohn, P. M., Hops, H., & Seeley, J. R. (1999). Cognitive-behavioral treatment of adolescent depression: Efficacy of acute group treatment and booster sessions. Journal of the American Academy of Child and Adolescent Psychiatry, 38, 272-279.

Cornwell, J. M. (1993). Monte Carlo comparisons of three tests for homogeneity of independent correlations. Educational and Psychological Measurement, 53, 605-618.

Cornwell, J. M., & Ladd, R. T. (1993). Power and accuracy of the Schmidt and Hunter meta-analytic procedures. Educational and Psychological Measurement, 53, 877-895.

Costello, E. J., Angold, A., Burns, B. J., Stangl, D. K., Tweed, D. L., Erkanli, A., & Wortham, C. M. (1996). The Great Smoky Mountains Study of youth: Goals, design, methods, and the prevalence of DSM-III-R disorders. Archives of General Psychiatry, 53, 1129-1136.

Costello, E. J., Erkanii, A., & Angold, A. (2006). Is there an epidemic of child or adolescent depression? Journal of Child Psychology & Psychiatry, 47, 1263-1271.

Costello, E. J., Pine, D. S., Hammen, C., March, J. S., Plotsky, P. M., Weissman, M. M., ... Leckman, J. F. (2002). Development and natural history of mood disorders. Biological Psychiatry, 52, 529-542. doi: 10.1016/S0006-3223(02)01372-0

Cuijpers, P., van Straten, A., Smits, N., & Smit, F. (2006). Screening and early psychological intervention for depression in schools: Systematic review and meta-analysis. European Adolescent Psychiatry, 15, 300-307. doi:l0.1007/s00787-006-0537-4

* De Cuyper, S., Timbremont, B., Bract, C., De Backer, V., & Wullaert, T. (2004). Treating depressive symptoms in schoolchildren: A pilot study. European Child and Adolescent Psychiatry, 13, 105-114. doi:10.1007/s00787-004-0366-2

De Maat, S., Dekker, J., Schoevers, R., & De Jonghe, F. (2006). Relative efficacy of psychotherapy and pharmacotherapy in the treatment of depression: A meta-analysis. Psychotherapy Research, 16, 566-578.

* Diamond, G., Siqueland, L., & Diamond, G. M. (2003). Attachment-based family therapy for depressed adolescents: Programmatic treatment development. Clinical Child and Family Psychology Review, 6, 107-127.

Erford, B. T., Savin-Murphy, J., & Butler, C. (2010). Conducting a meta-analysis of counseling outcome research: Twelve steps and practical procedures. Counseling Outcome Research and Evaluation, 1, 19-43.

* Fine, S., Forth, A., Gilbert, M., & Haley, G. (1991). Group therapy for adolescent depressive disorder: A comparison of social skills and therapeutic support. Journal of the American Academy of Child and Adolescent Psychiatry, 30, 79-85.

Gould, M. S., King, R., Greenwald, S., Fisher, P., Schwab-Stone, M., Kramer, R., ... Shaffer, D. (1998). Psychopathology associated with suicidal ideation and attempts among children and adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 37, 915-923.

Green, J. M., & Murray, D. (1994). The use of the Edinburgh Postnatal Depression Scale in research to explore relationship between antenatal and postnatal dysphoria. In J. Cox & J. Holden (Eds.), Perinatal psychiatry (pp. 180-198). London, England: Royal College of Psychiatry.

Hamilton, M. (1960). A rating scale for depression. Journal of Neurology, Neurosurgery, & Psychiatry, 12, 56-62.

Harrington, R., Whittaker, J., Shoebridge, P., & Campbell, F. (1998). Systematic review of efficacy of cognitive behavior therapies in childhood and adolescent depressive disorder. British Medical Journal, 316, 1559-1563.

Hazell, P., O'Connell, D., Heathcote, D., & Henry, D. (2002). Tricyclic drugs for depression in children and adolescents. Cochrane Database of Systematic Reviews 2002, Issue 2. Art. No.: CD002317. doi:10.1002/14651858.CD002317

Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis. Orlando, FL: Academic Press.

Higgins, J. P. T., Thompson, S. G., Decks, J. J., & Altman, D. G. (2003). Measuring inconsistency in meta-analysis. British Medical Journal, 327, 557-560. doi:10.1136/bmj.327.7414.557

* Horowitz, J. L., Garber, J., Ciesla, J. A., Young, J. F., & Mufson, L. (2007). Prevention of depressive symptoms in adolescents: A randomized trial of cognitive-behavioral and interpersonal prevention programs. Journal of Consulting and Clinical Psychology, 75, 693-706. doi:10.1037/0022-006x.75.5.693

* Hyun, M. S., Chung, H. I., & Lee, Y. J. (2005). The effect of cognitive- behavioral group therapy on the self-esteem, depression, and self-efficacy of runaway adolescents in a shelter in South Korea. Applied Nursing Research, 18, 160-166. doi:10.1016/j.apnr.2004.07.006

Imel, Z. E., Malterer, M. B., McKay, K. M., & Wampold, B. E. (2008). A meta-analysis of psychotherapy and medication in unipolar depression and dysthymia. Journal of Affective Disorders, 110, 197-206. doi:10.1016/j.jad.2008.03.018

** Jones, N. A., & Field, T. (1999). Massage and music therapies attenuate frontal EEG asymmetry in depressed adolescents. Adolescence, 34, 530-534.

* Kahn, J. S., Kehle, T. J., Jenson, W R., & Clark, E. (1990). Comparison of cognitive-behavioral, relaxation, and self-modeling interventions for depression among middle-school students. School Psychology Review, 19, 196-211.

* Kaufman, N. K., Rohde, P., Seeley, J. R., Clarke, G. N., & Stice, E. (2005). Potential mediators of cognitive-behavioral therapy for adolescents with comorbid major depression and conduct disorder. Journal of Consulting and Clinical Psychology, 73, 38-46. doi:10.1037/0022-006x.73.1.38

Kennard, B., Silva, S., Vitiello, B., Curry, J., Kratochvil, C., Simons, A., ... March, J. (2006). Remission and residual symptoms after short-term treatment in the Treatment for Adolescents With Depression Study (TADS). Journal of the American Academy of Child and Adolescent Psychiatry, 45, 1404-1411.

* Kerfoot, M., Harrington, R., Harrington, V., Rogers, J., & Verduyn, C. (2004). A step too far? Randomized trial of cognitive-behaviour therapy delivered by social workers to depressed adolescents. European Child & Adolescent Psychiatry, 13, 92-99. doi:10.1007/s00787-004-0362-6

Kirk, R. (1996). Practical significance: A concept whose time has come. Educational and Psychological Measurement, 56, 746-759.

Kovacs, M. (1992). Manual for the Children's Depression Inventory. North Tonawanda, NY: Multi-Health Systems.

* Kovacs, M., Sherrill, J., George, C. J., Pollock, M., Tumuluru, R. V., & Ho, V. (2006). Contextual emotion-regulation therapy for childhood depression: Description and pilot testing of a new intervention. Journal of the American Academy of Child and Adolescent Psychiatry, 45, 892-903. doi: 10.1097/01.chi.0000222878.74162.5a

** Lamb, J. M., Puskar, K. R., Sereika, S. M., & Corcoran, M. (1998). School-based intervention to promote coping in rural teens. The American Journal of Maternal Child Nursing, 23, 187-194.

Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for categorical data. Biometrics, 33, 159-174.

Le, H.-N., Munoz, R. F., Ippen, C. G., & Stoddard, J. L. (2003). Treatment is not enough: We must prevent major depression in women. Prevention & Treatment, 6, Article 10. doi: 10.1037/1522-3736.6.1.610a

Lewinsohn, P. M., & Clarke, G. N. (1999). Psychosocial treatments for adolescent depression. Clinical Psychology Review, 19, 329-342. doi: 10.1016/S0272-7358(98)00055-5

* Lewinsohn, P. M., Clarke, G. N., Hops, H., & Andrews, J. (1990). Cognitive-behavior treatment for depressed adolescents. Behavior Therapy, 21, 385-401.

Lewinsohn, P. M., Clarke, G. N., Seeley, J. R., & Rohde, P. (1994). Major depression in community adolescents: Age at onset, episode duration, and time to recurrence. Journal of the American Academy of Child and Adolescent Psychiatry, 33, 809-818.

* Liddle, B., & Spence, S. H. (1990). Cognitive-behaviour therapy with depressed primary school children: A cautionary note. Behavioural Psychotherapy, 18, 85-102.

Lipsey, M. W., & Wilson, D. B. (1993). The efficacy of psychological, educational, and behavioral treatment: Confirmation from meta-analysis. American Psychologist, 48, 1181-1209.

Lipsey, M. W., & Wilson, D. B. (2001). Practical recta-analysis. Thousand Oaks, CA: Sage.

Lipsey, S. J., Lipsey, M. W., & Derzon, J. H. (2003). The effects of school-based intervention on aggressive behavior: A meta-analysis. Journal of Consulting and Clinical Psychology, 71, 136-149.

** Maina, G., Fomer, F., & Bogetto, F. (2005). Randomized controlled trial comparing brief dynamic and supportive therapy with waiting list condition in minor depressive disorders. Psychotherapy and Psychosomatics, 74, 43-50.

* March, J., Silva, S., Petrycki, S., Curry, J., Wells, K., Fairbank, J., ... Severe, J. (2004). Fluoxetine, cognitive-behavioral therapy, and their combination for adolescents with depression. Journal of the American Medical Association, 292, 807-820.

** Marcotte, D., & Baron, P. (1993). The efficacy of a school-based rational-emotive intervention strategy with depressive adolescents. Canadian Journal of Counselling, 27, 77-92.

** Markowitz, J. C., Kocsis, J. H., Bleiberg, K. L., Christos, P. J., & Sacks, M. (2005). A comparative trial of psychotherapy and pharmacotherapy for "pure" dysthymic patients. Journal of Affective Disorders, 89, 167-175.

McCarty, C. A., & Weisz, J. R. (2007). Effects of psychotherapy for depression in children and adolescents: What we can (and can't) learn from meta-analysis and component profiling. Journal of the American Academy of Child and Adolescent Psychiatry, 46, 879-886.

McCullough, J. P. (1991). A naturalistic study often patients. Journal of Nervous and Mental Disorders, 179, 734-740.

* Melvin, G. A., Tonge, B. J., King, N. J., Heyne, D., Gordon, M. S., & Klimkeit, E. (2006). A comparison of cognitive-behavioral therapy, sertraline, and their combination for adolescent depression. Journal of the American Academy of Child and Adolescent Psychiatry, 45, 1151-1161. doi:10.1097/01.chi.0000233157.21925.71

* Mendlowitz, S. L., Manassis, K., Bradley, S., Scapillato, D., Miezitis, S., & Shaw, B. F. (1999). Cognitive-behavioral group treatments in childhood anxiety disorders: The role of parental involvement. Journal of the American Academy of Child and Adolescent Psychiatry, 38, 1223-1229.

Michael, K. D., & Crowley, S. L. (2002). How effective are treatments for child and adolescent depression? A meta-analytic review. Clinical Psychology Review, 22, 247-269.

* Miller, L., Gur, M., Shanok, A., & Weissman, M. (2008). Interpersonal psychotherapy with pregnant adolescents: Two pilot studies. Journal of Child Psychology and Psychiatry, 49, 733-742. doi:10.1111/j.1469-7610.2008.01890.x

* Mufson, L., Dorta, K. P., Wickramaratne, P., Nomura, Y., Olfson, M., & Weissman, M. M. (2004). A randomized effectiveness trial of interpersonal psychotherapy for depressed adolescents. Archives of General Psychiatry, 61, 577-584.

* Mufson, L., Weissman, M. M., Moreau, D., & Garfinkel, R. (1999). Efficacy of interpersonal psychotherapy for depressed adolescents. Archives of General Psychiatry, 56, 573-579.

Nemeroff, C. B. (2007). The burden of severe depression: A review of diagnostic challenges and treatment alternatives. Journal of Psychiatric Research, 41, 189-206. doi:10.1016/j.jpsychires.2006.05.008

Netz, Y., Wu, M., Becker, B. J., & Tenenbaum, G. (2005). Physical activity and psychological well-being in advanced age: A meta-analysis of intervention studies. Psychology and Aging, 20, 272-284.

* Nolan, M., Carr, A., Fitzpartrick, C., O'Flaherty, A., Keary, K., Turner, R., ... Tobin, G. (2002). A comparison of two programmes for victims of child sexual abuse: A treatment outcome study. Child Abuse Review, 11, 103-123. doi: 10.1002/car.727

* Puskar, K., Sereika, S., & Tusaie-Mumford, K. (2003). Effect of the Teaching Kids to Cope (TKC) program on outcomes of depression and coping among rural adolescents. Journal of Child and Adolescent Psychiatric Nursing, 16, 71-80.

Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385--401.

Ramchandani, P. (2004). Treatment of major depressive disorder in children and adolescents: Most selective serotonin reuptake inhibitors are no longer recommended. British Medical Journal, 328, 3--4. doi:10.1136/bmj.328.7430.3

Reinecke, M. A., Ryan, N. E., & DuBois, D. L. (1998). Meta-analysis of CBT for depression in adolescents: Dr. Reinecke et al. reply. Journal of the American Academy of Child and Adolescent Psychiatry, 37, 1006-1007. doi:10.1097/00004583-199810000-00005

Reynolds, W. M. (2002). Manual for the Reynolds Adolescent Depression Scale-Second Edition (RADS-2). Lutz, FL: Psychological Assessment Resources.

** Reynolds, W. M., & Coats, K. I. (1986). A comparison of cognitive-behavior therapy and relaxation training for the treatment of depressions in adolescents. Journal of Consulting and Clinical Psychology, 54, 653-660.

* Roberts, C., Kane, R., Thomson, H., Bishop, B., & Hart, B. (2003). The prevention of depressive symptoms in rural school children: A randomized controlled trial. Journal of Consulting and Clinical Psychology, 71, 622-628. doi:10.1037/0022-006x.71.3.622

* Rohde, P., Lewinsohn, P. M., & Seeley, J. R. (1994). Response of depressed adolescents to cognitive-behavioral treatment: Do differences in initial severity clarify the comparison of treatments? Journal of Consulting and Clinical Psychology, 62, 851-854.

Rosenthal, R. (1979). The file drawer problem and tolerance for null results. Psychological Bulletin, 86, 638-641.

* Rossello, J., & Bernal, G. (1999). The efficacy of cognitive-behavioral and interpersonal treatments for depression in Puerto Rican adolescents. Journal of Consulting and Clinical Psychology, 67, 734-745.

* Rossello, J., Bernal, G., & Rivera-Medina, C. (2008). Individual and group CBT and IPT for Puerto Rican adolescents with depressive symptoms. Cultural Diversity and Ethnic Minority Psychology, 14, 234-245. doi:10.1037/1099-9809.14.3.234

Ryan, N. D. (2005). Treatment of depression in children and adolescents. Lancet, 366, 933-940.

Shaffer, D., Gould, M. S., Brasic, J., Ambrosini, P., Fisher, P., Bird, H., & Aluwahlia, S. (1983). A Children's Global Assessment Scale (CGAS). Archives of General Psychiatry, 40, 1228-1231.

Sheard, T., & Maguire, P. (1999). The effect of psychological interventions on anxiety and depression in cancer patients: Results of two meta-analyses. British Journal of Cancer, 80, 1770-1780.

* Sheffield, J. K., Spence, S. H., Rapee, R. M., Kowalenko, N., Wignall, A., Davis, A., & McLoone, J. (2006). Evaluation of universal, indicated, and combined cognitive-behavioral approaches to the prevention of depression among adolescents. Journal of Consulting and Clinical Psychology, 74, 66-79. doi:10.1037/0022-006x.74.1.66

** Stark, K. D., Reynolds, W. M., & Kaslow, N. J. (1987). A comparison of relative efficacy of self-control therapy and a behavioral problem-solving therapy for depression in children. Journal of Abnormal Child Psychology, 15, 91-113.

* Stice, E., Burton, E., Bearman, S. K., & Rohde, P. (2006). Randomized trial of a brief depression prevention program: An elusive search for a psychosocial placebo control condition. Behaviour Research and Therapy, 45, 863-876. doi:10.1016/j.brat.2006.08.008

* Stice, E., Rohde, P., Seeley, J. R., & Gau, J. M. (2008). Brief cognitive-behavioral depression prevention program for high-risk adolescents outperforms two alternative interventions: A randomized efficacy trial. Journal of Consulting and Clinical Psychology, 76, 595-606. doi:10.1037/a0012645

Stock, W. A. (1994). Systematic coding for research synthesis. In H. Cooper & L. V. Hedges (Eds.), The handbook of research synthesis (pp. 125-138). New York, NY: Russell Sage Foundation.

Stolherg, R. A., Clark, D. C., & Bongar, B. (2002). Epidemiology, assessment, and management of suicide in depressed patients. In I. H. Gotlib & C. L. Hammen (Eds.), Handbook of depression (pp. 581-601). New York, NY: Guilford Press.

* Thompson, M. C., Pierre, C. B., Haber, F. M., Fogler, J. M., Groff, A. R., & Asarnow, J. R. (2007). Family-focused treatment for childhood-onset depressive disorders: Results of an open trial. Clinical Child Psychology and Psychiatry, 12, 403-420. doi:10.1177/1359104507078474

* Trowell, J., Joffe, I., Campbell, J., Clemente, C., Almqvist, F., Soininen, M., ... Tsiantis, J. (2007). Childhood depression: A place for psychotherapy. An outcome study comparing individual psychodynamic psychotherapy and family therapy. European Child & Adolescent Psychiatry, 16, 157-167. doi:10.1007/s00787-006-0584-x

** Verdeli, H., Ferro, T., Wickramaratne, P., Greenwald, S., Blanco, C., & Weissman, M. M. (2004). Treatment of depressed mothers of depressed children: Pilot study of feasibility. Depression and Anxiety, 19, 51-58.

Vitiello, B., & Swedo, S. (2004). Antidepressant medications in children. New England Journal of Medicine, 350, 1489-1491.

** Waslick, B. D., Walsh, B. T., Greenhill, L. L., Eilenberg, M., Capasso, L., & Lieber, D. (1999). Open trial of fluoxetine in children and adolescents with dysthymic disorder or double depression. Journal of Affective Disorders, 56, 227-236.

Watanabe, N., Hunot, V., Omori, I. M., Churchill, R., & Furukawa, T. A. (2007). Psychotherapy for depression among children and adolescents: A systematic review. Acta Psychiatrica Scandinavica, 116, 84-95. doi:10.1111/j.1600-0447.2007.01018.x

** Weersing, V. R., & Weisz, J. R. (2002). Community clinic treatment of depressed youth: Benchmarking usual care against CBT clinical trials. Journal of Consulting and Clinical Psychology, 70, 299-310.

Weissman, M. M. (1994). Psychotherapy in the maintenance treatment of depression. British Journal of Psychiatry, 165, 42-50.

Weisz, J. R., Jensen-Doss, A., & Hawley, K. M. (2006). Evidence-based youth psychotherapies versus usual clinical care: A meta-analysis of direct comparisons. American Psychologist, 61, 671-689.

Weisz, J. R., McCarty, C. A., & Valeri, S. M. (2006). Effects of psychotherapy for depression in children and adolescents: A meta-analysis. Psychological Bulletin, 132, 132-149. doi:10.1037/0033-2909.132.1.132

* Weisz, J. R., Thurber, C. A., Sweeney, L., Proffitt, V. D., & LeGagnoux, G. L. (1997). Brief treatment of mild-to-moderate child depression using primary and secondary control enhancement training. Journal of Consulting and Clinical Psychology, 65, 703-707.

Westen, D., & Morrison, K. (2001). A multidimensional meta-analysis of treatments for depression, panic, and generalized anxiety disorder: An empirical examination of the status of empirically supported therapies. Journal of Consulting and Clinical Psychology, 69, 875-899.

Whiston, S. C., Tai, W. L., Rahardja, D., & Eder, K. (2011). School counseling outcome: A meta-analytic examination of interventions. Journal of Counseling & Development, 89, 37-55.

Whittington, C. J., Kendall, T., Fonagy, P., Cottrell, D., Cotgrove, A., & Boddington, E. (2004). Selective serotonin reuptake inhibitors in childhood depression: Systematic review of published versus unpublished data. Lancet, 363, 1341-1345.

* Wood, A., Trainor, G., Rothwell, J., Moore, A., & Harrington, R. (2001). Randomized trial of group therapy for repeated deliberate self-harm in adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 40, 1246-1253. doi:10.1097/00004583-200111000-00003

* Young, J. F., Mufson, L., & Davies, M. (2006a). Efficacy of interpersonal psychotherapy-adolescent skills training: An indicated preventive intervention for depression. Journal of Child Psychology and Psychiatry, 47, 1254-1262. doi:10.1111/j.14697610.2006.01667.x

* Young, J. F., Mufson, L., & Davies, M. (2006b). Impact of comorbid anxiety in an effectiveness study of interpersonal psychotherapy for depressed adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 45, 904-912. doi: 10.1097/01. chi.0000222791.23927.5f

Bradley T. Erford, Gina Lattanzi, Janet Weller, Hallie Schein, Emily Wolf, Meredith Hughes, Jenna Darrow, Janet Savin- Murphy, and Elizabeth Peacock, Education Specialties Department, Loyola University Maryland; Breann M. Erford, Psychology Department, Drexel University. The authors would like to acknowledge the contributions of Sarah Short, Christina Daschille, Lauren Palmer, Joe DiBasilio, Megan Middleton, and Emily Pfiefer for their help on various facets of this study. Correspondence concerning this article should be addressed to Bradley T. Erford, Education Specialties Department, School of Education, Loyola University Maryland, Timonium Graduate Center, 2034 Greenspring Drive, Timonium, MD 21093 (e-mail:
Characteristics of Individual Studies Used in the Meta-Analysis

Study                                 Summary                  n

Ackerson et al.         Cognitive bibliotherapy for             22
(1998)                  adolescents with depression

Asarnow et al.          CB family education intervention        20
(2002)                  for depression in

Barrera et al. (2002)   Structured group intervention           13
                        for siblings of cancer patients

Bolton et al. (2007)    Group interpersonal psychotherapy      209
                        (IPT) with Ugandan
                        refugees to reduce depression
                        and anxiety

Brent et al. (1997)     Cognitive behavior therapy              97
                        (CBT), family and supportive
                        therapy for adolescents with

Clarke et al. (1995)    Group cognitive treatment              150
                        to prevent depression in

Clarke et al. (1999)    Use of individual CBT with             123
                        depressed adolescents and
                        CBT with parent adjunct

Clarke et al. (2001)    Cognitive therapy with adolescent       94
                        offspring of depressed

Clarke et al. (2002)    CBT treatment of depressed              88
                        offspring of depressed

De Cuyper et al.        CBT treatment of depressed              20
(2004)                  youth

Diamond et al.          Attachment-based family                 32
(2003)                  therapy for families of
                        adolescents with depression

Fine et al. (1991)      Short-term social skills group for      66
                        adolescents with depression

Horowitz et al.         CBT and IPT depression                 380
(2007)                  prevention programs in high
                        school students

Hyun et al. (2005)      CBT on depression in runaway            27
                        adolescents in S. Korea

Kahn et al. (1990)      CBT and relation therapy for            51
                        treatment of depression in
                        middle school students

Kaufman et al.          Group CBT for depressed                 93
(2005)                  adolescents

Kerfoot et al. (2004)   CBT for adolescents with                52

Kovacs et al. (2006)    Contextual emotion-regulation           20
                        therapy for children with

Lewinsohn et al.        CBT for depressed adolescents           59
(1990)                  and conjoint parent

Liddle & Spence         CBT (social competence training)        31
(1990)                  with depressed children

March et al. (2004)     CBT in the Treatment for               223
                        Adolescents With Depression

Melvin et al. (2006)    Treatment of adolescents with           42
                        diagnosis of a depressive

Mendlowitz et al.       Group CB interventions and              63
(1999)                  parental involvement for
                        depression/anxiety in children

Miller et al. (2008)    Group IPT for depression in             14
A (a)                   pregnant adolescents in
                        health class

Miller et al. (2008)    Group I PT for depression in            11
B (a)                   pregnant adolescents after

Mufson et al. (2004)    Use of IPT in adolescent                63

Mufson et al. (1999)    IPT for adolescents with                32

Nolan et al. (2002)     Treatment of sexually abused            30
                        children using individual and
                        combined (I/G) approaches

Puskar et al. (2003)    Group CB interventions for              89
                        adolescents with depression
                        from rural settings

Roberts et al.          Prevention of depression in            179
(2003)                  at-risk rural schoolchildren

Rohde et al. (1994)     Two cognitive treatment                 84
                        regimens for adolescents with
                        depression (Sample 2 only)

Rossello & Bernal       CBT and IPT in Puerto Rican             71
(1999)                  adolescents

Rossello et al.         Individual and group formats           112
(2008)                  of CBT and I PT for adolescents
                        with depression

Sheffield et al.        CBT approaches to prevention           248
(2006)                  of depression in adolescents
                        (indicated sample used)

Stice et al. (2006)     CBT depression prevention              117
                        program with high-risk

Slice et al. (2008)     CB prevention of depression            171
                        in high-risk adolescents

Thompson et al.         12-week family-focused treatment         9
(2007)                  for depressed school-age

Trowell et al. (2007)   Individual therapy and family           72
                        therapy for youth with mod
                        moderate to severe depression

Weisz et al. (1997)     Control enhancement training            48
                        with elementary students
                        with depression

Wood et al. (2001)      Group therapy for adolescents           63
                        who deliberately harmed

Young et al. (2006a)    IPTAST to reduce depression             41
                        in adolescents

Young et al. (2006b)    IPT for depressed adolescents           43

                          M Age      %      Control Group      %
Study                   in Years   Male         Type         White

Ackerson et al.           15.9      36        Wait list        68

Asarnow et al.             dk       35        Wait list        57

Barrera et al. (2002)     12.7      65      Single group       dk

Bolton et al. (2007)       dk       43        Wait list        0

Brent et al. (1997)       15.6      24     TAU-supportive      83

Clarke et al. (1995)      15.3      30     TAU-usual care      92

Clarke et al. (1999)      16.2      29        Wait list        dk

Clarke et al. (2001)      14.6      71     TAU-usual care      95

Clarke et al. (2002)      15.2      32     TAU-usual care      91

De Cuyper et al.          10.0      25        Wait list       100

Diamond et al.            15.0      22        Wait list        dk

Fine et al. (1991)        15.1      17     TAU-supportive      77

Horowitz et al.           14.4      46        Wait list        79

Hyun et al. (2005)        15.5      100       Wait list        0

Kahn et al. (1990)        12.5      48        Wait list        dk

Kaufman et al.            15.1      52     TAU-life skills     81
(2005)                                          group

Kerfoot et al. (2004)     13.9      54     TAU-usual care      dk

Kovacs et al. (2006)      10.4      65      Single group       90

Lewinsohn et al.          16.2      39        Wait list        dk

Liddle & Spence            dk       68      Wait list and      dk
(1990)                                         placebo

March et al. (2004)       14.6      46         Placebo         74

Melvin et al. (2006)      15.3      34     TAU-medication      dk

Mendlowitz et al.          9.8      43        Wait list        dk

Miller et al. (2008)      14.7       0      Single group       0
A (a)

Miller et al. (2008)      16.5       0      Single group       0
B (a)

Mufson et al. (2004)      15.1      16     TAU-counseling      dk

Mufson et al. (1999)      15.8      27      TAU-clinical       0

Nolan et al. (2002)       12.6      15      TAU-combined       dk

Puskar et al. (2003)      16.0      18           TAU           88

Roberts et al.            11.9      50       TAU-health        dk
(2003)                                        education

Rohde et al. (1994)       16.3      26        Wait list        99

Rossello & Bernal         14.5      46        Wait list        0

Rossello et al.           14.7      45         TAU-IPT         dk

Sheffield et al.          14.3      46        Wait list        dk

Stice et al. (2006)       18.4      30        Wait list        55

Slice et al. (2008)       15.6      44        Wait list        46

Thompson et al.           11.4      56      Single group       67

Trowell et al. (2007)     12.0      62       TAU-family        87

Weisz et al. (1997)        9.6      54        Wait list        62

Wood et al. (2001)        14.2      22     TAU-usual care      dk

Young et al. (2006a)      13.4      15       TAU-school        0

Young et al. (2006b)      15.9      16    TAU-school-based     dk

                               Depression Outcome
Study                               Measure

Ackerson et al.         Hamilton Rating Scale for
(1998)                  Depression (HRSD); Children's
                        Depression Inventory (CDI);
                        Child Behavior Checklist (CBCL)
                        Depression subscale

Asarnow et al.          CDI

Barrera et al. (2002)   CDI

Bolton et al. (2007)    Acholi Psychological Assessment

Brent et al. (1997)     Beck Depression Inventory
                        (BDI); Kiddie Schedule for
                        Affective Disorders and

Clarke et al. (1995)    Hamilton Depression Rating
                        Scale (HDRS); Center for
                        Epidemiologic Studies Depression
                        Scale (CES-D); Global
                        Assessment of Functioning
Clarke et al. (1999)    BDI; HDRS

Clarke et al. (2001)    CES-D; CBCL Depression
                        subscale; K-SADS; Hamilton
                        Depression Rating Scale

Clarke et al. (2002)    HAM-D; K-SADS; CES-D

De Cuyper et al.        CDI

Diamond et al.          BDI; HAM-D

Fine et al. (1991)      CDI; K-SADS

Horowitz et al.         CDI; CES-D

Hyun et al. (2005)      BDI

Kahn et al. (1990)      Reynolds Adolescent
                        Depression Scale (RADS);
                        CDI; Bellevue Index of

Kaufman et al.          Beck Depression Inventory-II;
(2005)                  HDRS

Kerfoot et al. (2004)   Mood and Feelings Questionnaire

Kovacs et al. (2006)    RADS

Lewinsohn et al.        CES-D; BDI

Liddle & Spence         CDI

March et al. (2004)     Children's Depression Rating
                        Scale; RADS

Melvin et al. (2006)    RADS
                        (Continued on next page)

Mendlowitz et al.       CDI

Miller et al. (2008)    BDI; Edinburgh Depression
A (a)                   Scale (EDS)

Miller et al. (2008)    BDI; HRSD; EDS
B (a)

Mufson et al. (2004)    HAM-D; BDI

Mufson et al. (1999)    HRSD; BDI

Nolan et al. (2002)     CDI

Puskar et al. (2003)    RADS

Roberts et al.          CDI

Rohde et al. (1994)     BDI; CES-D; HDRS

Rossello & Bernal       CD]

Rossello et al.         CDI

Sheffield et al.        CDI; CES-D

Stice et al. (2006)     BDI

Slice et al. (2008)     BDI; Depressive symptoms

Thompson et al.         Depression Outcome
(2007)                  Measure; Depression Self
                        Rating Scale

Trowell et al. (2007)   Children's Depression Scale

Weisz et al. (1997)     CDI; Children's Depression
                        Rating Scale-Revised

Wood et al. (2001)      Mood and Feelings Questionnaire

Young et al. (2006a)    CES-D; Children's Global
                        Assessment Scale

Young et al. (2006b)    HRSD; CGAS

Note. A study's sample size is the total number of participants
in the control and treatment groups used to compute the effect
size. dk= don't know; TAU = treatment as usual; S. = South;
CB = cognitive behavior; I/G = individual/group; AST = adolescent
skills training.

(a) The Miller et al. (2008) study contained two discrete samples
in two pilot studies and therefore was treated as two separate
studies (i.e., "A" and "B").

Summary of Posttest (PT) and Follow-Up Results

Condition                  k    n       d+         95% CI      Q

Termination (PT)
    Single group           5      62  .36 (a)    [.17, .54]   ns
    Wait list             18   1,520  .55 (a)    [.41, .69]   ns
    Placebo                2     122    .01     [-.37, .38]   ns
    Treatment as usual    18   1,358  .29 (a)    [.16, .43]   ns
Longest interval
  follow-up results
    Single group           3      33  .46 (a)    [.08, .84]   ns
    Wait list              9     923  .29 (a)    [.08 .49]    ns
    Placebo                1      31  .21 (a)   [-.17, .59]   ns
    Treatment as usual     9     743  .16 (a)    [.08, .25]   ns
Follow-up studies
    Single group
    Wait list
    Treatment as usual

                                     d+ (in Months)

Condition                   PT     k     1-4     k    -6    k

Termination (PT)
    Single group
    Wait list
    Treatment as usual
Longest interval
  follow-up results
    Single group
    Wait list
    Treatment as usual
Follow-up studies
    Single group         .53 (a)   3                 .32    2
    Wait list            .42 (a)   9   .71 (a)   4   .23    3
    Placebo              .34 (a)   1   .20       1
    Treatment as usual   .22 (a)   9                 .12    4

                                 d+ (in Months)

Condition                  -12     k      24     k

Termination (PT)
    Single group
    Wait list
    Treatment as usual
Longest interval
  follow-up results
    Single group
    Wait list
    Treatment as usual
Follow-up studies
    Single group         .65 (a)   2
    Wait list            .09       2
    Treatment as usual   .18 (a)   3   .23 (a)   2

(a) d+ > 0, p < .05.
COPYRIGHT 2011 American Counseling Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Erford, Bradley T.; Erford, Breann M.; Lattanzi, Gina; Weller, Janet; Schein, Hallie; Wolf, Emily; H
Publication:Journal of Counseling and Development
Article Type:Report
Geographic Code:1USA
Date:Sep 22, 2011
Previous Article:Self-efficacy, self-rated abilities, adjustment, and academic performance.
Next Article:An emergent theory of altruism and self-interest.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters