Australian school-based prevention and early intervention programs for anxiety and depression: a systematic review

Prevention and early intervention programs are normally classified into four types: universal, selective, indicated, and treatment programs.4 Although definitions differ somewhat, universal programs are presented to all students regardless of symptoms; selective programs target children and adolescents who are at risk of developing a disorder by virtue of particular risk factors, such as being children of a depressed parent; indicated programs are delivered to students with early or mild symptoms of a disorder; and treatment programs are provided for those diagnosed with the disorder.4 We use this classification system in this review.

The Cochrane, PsychInfo and PubMed databases were searched in June 2006, with the key search terms “school-based OR school*”, “depression OR anxiety”, and “Australia*”. Programs were included if they addressed symptoms of anxiety or depression in a school context, or increased student resilience through the development of positive coping skills. Only programs developed in Australia and overseas programs that had been trialled in Australia were included in the review. No other restrictions were placed on program inclusion. In the light of recent COAG reforms in mental health, our explicit aim was to review tested programs with established efficacy or effectiveness in the Australian school environment.

A search of the Primary Mental Health Care Australian Resource Centre database was also undertaken using the terms “school setting” and “youth”. This search was conducted to identify unpublished studies. Once programs were identified, additional materials were obtained from program websites, reference lists and authors. The authors of each program were contacted via email and invited to provide unpublished outcome data to be included in the review.

Data abstraction was completed independently and the process was not blinded.

As poor quality intervention studies overestimate the size of intervention effects,5 study quality was assessed using a validated measure that assesses quality against three key criteria: randomisation, double-blinding, and withdrawals and dropouts.6 Quality ratings can range from 0 to 5, although intervention trials of mental health programs within schools, which often cannot achieve double-blind conditions or full randomisation, rarely receive scores above 3.

Trials were considered efficacious or effective if significant differences were apparent between the control and treatment groups’ mean levels of anxiety and depression at post-test or follow-up. For trials without a control group, efficacy was based on significant differences being present from pre-test to post-test and follow-up.

Where data were available and extractable, effect size was estimated using Cohen’s d,7 which was calculated by subtracting the mean intervention score from the mean control score, and dividing the result by the pooled standard deviation.

A formal meta-analysis was not conducted because of the heterogeneity of programs, small number of trials with follow-up times, and available data quality.

A large proportion of Australian-based programs reported positive outcomes either immediately or at follow-up. The effect sizes for controlled trials varied from small (0.18) to moderate (0.83). About 80% of indicated anxiety programs and 50% of depression programs were associated with reductions in, respectively, anxiety and depression symptoms. Results for universal programs were similar, with improvements associated with 60% of programs for anxiety and 58% of those for depression. Effectiveness persisted for at least 6 months for some programs. Programs such as FRIENDS, which included booster sessions, reported positive outcomes at 12 months. Although the effect size and trial quality were variable, our findings provide strong support for mental health prevention and early intervention programs. Both indicated and universal approaches appear to produce short- to mid-term small to moderate reductions in anxiety and depression in schools.

Our conclusion differs somewhat from the more cautious support for prevention programs voiced by some researchers, based on Australian and international trials. These researchers have highlighted the need to establish effectiveness in routine practice.41 However, our conclusion is consistent with that in a review of international research,42 in which the findings from depression prevention programs were encouraging and similar for both targeted and universal prevention programs.

Prevention and early intervention programs need to be assessed and evaluated against the standards issued by the Society for Prevention Research (SPR).43 According to these standards, programs must fulfil particular requirements before they can be considered efficacious, effective or ready for broad dissemination.

The FRIENDS program appears to stand up well against the SPR standards. Seven Australian school-based trials of this program consistently showed efficacy and effectiveness across a range of samples, with successful implementation by both program developers and classroom teachers. The quality of the trials has also been very high, with five of the seven trials being RCTs. Psychometrically sound outcome measures were used in these trials,18,23 as well as parent and clinician ratings.16,18,23 The FRIENDS program is further supported by its successful implementation with migrant students21,22 and maintenance of treatment effects for up to 36 months.24 It also offers extensive training and a variety of program manuals and evaluation tools.22

Some other programs approach the standards, but require more evaluation. The RAP has had several positive efficacy and effectiveness trials, with students’ symptoms of depression successfully reduced. Long-term outcomes up to 12 months have been published. However, most trials have not been RCTs, so the efficacy of the program requires consolidation.

A number of the studies in our review were underpowered, with the consequence that effects may have been missed. Some of these non-significant interventions were associated with small to medium effect sizes, supporting this interpretation and indicating that outcomes may be more positive than the Box suggests. Many of the trials collected follow-up measurements at one time point only, so the longer-term effectiveness of these programs is unclear. In some cases, positive effects were evident at 6 months but, as no further data were collected, it is unclear how long these effects persisted.

We have grouped programs under the titles provided by the researchers, but the content and quality of these programs may have changed during their development. We were also unable to evaluate the effectiveness of newer “whole of school” approaches, as data from these programs have not yet been published.

Australian researchers have developed various school-based prevention programs. A few highly developed programs, such as MindMatters and the beyondblue schools program, have yet to be evaluated for their effects on anxiety or depression. Nevertheless, programs exist on which to build interventions that are scalable and capable of wide dissemination. Trials of senior school-based programs are particularly needed, and must include adequate attention placebo comparisons. Indeed, there may be an opportunity to dovetail research on school-based alcohol and drug programs with the research on mental health programs, with each serving as an attention placebo control condition for the other. A further area of investigation is the use of educational content to improve mental health. Although CBT programs constitute the bulk of the interventions, some psychoeducational materials might also be preventive.42