Cognitive behavioural therapy for depression, panic disorder and generalized anxiety disorder: a meta-regression of factors that may predict outcome
- HABY Michelle M., et al
- Journal article citation:
- Australian and New Zealand Journal of Psychiatry, 40(1), January 2006, pp.9-19.
- Blackwell Publishing
Cognitive behavioural therapy (CBT) is known to be an effective treatment for depression, panic disorder and generalised anxiety disorder (GAD). This study examines the factors related to clinical practice or to the conduct of trials which impact on the effect size of this treatment. Meta-analysis was used to determine the overall effect size and meta-regression to determine the factors that impact on this effect size in randomised controlled trials (RCTs). In 52 comparisons extracted from 33 RCTs of CBT for depression, panic disorder or GAD the effect size was 0.68 (95% CI=0.51-0.84). The type of disorder was not a significant predictor of the effect size, and the only factors that were so predictive were the type of control group and the inclusion of severely affected patients. CBT is significantly less effective for severely affected patients, and trials that compared CBT to a wait-list control group found significantly larger effect sizes than those comparing CBT to an attention placebo, but not a pill placebo. Further research is suggested to determine whether CBT is effective when given by those other than psychologists.
- Extended abstract:
HABY Michelle M.; et al.;
Cognitive behavioural therapy for depression, panic disorder and generalized anxiety disorder: a meta-regression of factors that may predict outcome.
Journal citation/publication details
Australian and New Zealand Journal of Psychiatry, 40(1), January 2006, pp.9-19.
This meta-analysis of 33 randomised controlled trials (RCTs) of cognitive behavioural therapy (CBT) for depression, panic disorder and generalised anxiety disorder (GAD) showed an effect size of 0.68 (95% CI=0.51-0.84). The only factors that were significant predictors of the effect size were the type of control group (in particular wait-list control group and pill-placebo control groups) and the inclusion of severely affected patients. Further research is needed to determine whether CBT is effective when provided by those other than psychologists.
Previous studies have shown significant variability in the effect size of CBT for depression, panic disorder and GAD. This study aims to determine which factors are significant predictors of effect size, and which explain heterogeneity in results.
What sources were used?
Existing meta-analyses of CBT for depression, panic disorder and GAD were used to identify suitable studies. These were supplemented by additional searches of Medline and the Cochrane Collaboration Controlled Trials Register up to November 2002.
What search terms/strategies were used?
Search strategies are not given.
What criteria were used to decide on which studies to include?
Eligible studies wereRCTs with one of the following control groups: wait list (or no treatment); pill placebo; or attention/psychological placebo. Study participants had to be 18 years and over, and have depression, panic disorder or GAD based on standardised DSM diagnostic criteria. Trials had to be studies of CBT, or the behavioural component alone or cognitive restructuring alone, with outcomes given as symptoms, functioning or health-related quality of life measures. Studies were excluded if means and standard deviations were not reported (this making it impossible to calculate effect sizes).
Who decided on their relevance and quality?
Two of the reviewers (not named) decided on relevance and quality.
How many studies were included and where were they from?
A total of 64 studies were initially retrieved, of which 31 were excluded, generally because of inadequate diagnosis of depression. The remaining 33 were published between 1984 and 2000 and only three were in a language other than English.. Forty comparisons were carried out in the USA , seven in the UK and five in other countries.
How were the study findings combined?
Mean results from each treatment and control condition were extracted to calculate effect sizes. Factors that might impact on treatment efficacy were also extracted including disorder, treatment type, duration and intensity of treatment, mode of therapy, type of therapists and their training, and whether severely affected patients were included. Where appropriate, a Beck Depression Inventory score was extracted, as were other factors relating to the conduct of trials.
Data were combined statistically to derive an effect size for each study and an overall effect size across all studies (standardised mean difference calculated using Hedges' g). Meta-regression was then used to determine the factors influencing the efficacy of CBT.
Findings of the review
The 33 papers reported 52 treatment comparisons, of which 30 involved in-patients with panic disorder, 17 in-patients with depression and five patients with GAD. The pooled effect size for all comparisons of CBT with any type of control group was 0.68 (95% CI=0.51-0.84). Variability in effect sizes was fully explained by: type of treatment; duration of therapy; inclusion of severely affected patients; year of study; country of study; control group; language of study; and number of dropouts from the control group.
The only factors that were significant predictors of effect size were the type of control group (in particular wait-list control group and pill-placebo control groups) and the inclusion of severely affected patients.
A number of factors have been identified which explain the heterogeneity of effect sizes in previous RCTs of CBT to treat depression, panic disorder and GAD. However, the results also show that CBT is significantly less effective for severely affected patients, while trials that compared CBT to a wait-list control group produced significantly larger effect sizes than those comparing CBT to attention (but not pill) placebo.
Further research is needed to determine whether CBT is effective when provided by those other than psychologists, or for non-English speaking patient groups.
Implications for policy or practice
None are discussed.
- Subject terms:
- randomised controlled trials, anxiety, behaviour therapy, cognitive behavioural therapy, depression;
- Content type:
- research review
- Journal home page
- ISSN print: