Anxiety disorders are common. They interfere with normal living, and they tend not to go away without treatment. Effective treatments are available, including cognitive behavioural therapy. These are known to work when delivered in person, but many people cannot access face-to-face treatment. This review examined 101 clinical trials of self-help and statistically analysed 92 of them. In these trials, 8403 people received self-help or were assigned to a control condition.
Overall, self-help with some support from a professional appears to be more effective than no treatment. Only half of the people who used self-help were better at the end of treatment, but self-help may still be considered effective because people with anxiety do not tend to get better without treatment. Self-help may be less effective than normal face-to-face therapy. Some results were difficult to interpret because the effects of treatments varied and the risk of overestimating the results was serious because of limitations in the study methods. We conclude that self-help is probably better than no treatment, but many people with an anxiety disorder would get better results from treatment provided by a skilled psychologist. Furthermore, most of the self-help materials used in these studies are intended for research and are not available to the public, so the results reported here may not apply to commercially available products.
Self-help may be useful for people who are not able or are not willing to use other services for people with anxiety disorders; for people who can access it, face-to-face cognitive behavioural therapy is probably clinically superior. Economic analyses were beyond the scope of this review.
Important heterogeneity was noted across trials. Recent interventions for specific problems that incorporate clinician support may be more effective than transdiagnostic interventions (i.e. interventions for multiple disorders) provided with no guidance, but these issues are confounded in the available trials.
Although many small trials have been conducted, the generalisability of their findings is limited. Most interventions tested are not available to consumers. Self-help has been recommended as the first step in the treatment of some anxiety disorders, but the short-term and long-term effectiveness of media-delivered interventions has not been established. Large, pragmatic trials are needed to evaluate and to maximise the benefits of self-help interventions.
Anxiety disorders are the most common mental health problems. They are chronic and unremitting. Effective treatments are available, but access to services is limited. Media-delivered behavioural and cognitive behavioural interventions (self-help) aim to deliver treatment with less input from professionals compared with traditional therapies.
To assess the effects of media-delivered behavioural and cognitive behavioural therapies for anxiety disorders in adults.
Published and unpublished studies were considered without restriction by language or date. The Cochrane Depression, Anxiety and Neurosis Review Group's Specialized Register (CCDANCTR) was searched all years to 1 January 2013. The CCDANCTR includes relevant randomised controlled trials from the following bibliographic databases: The Cochrane Library (all years), EMBASE (1974 to date), MEDLINE (1950 to date) and PsycINFO (1967 to date). Complementary searches were carried out on Ovid MEDLINE (1950 to 23 February 2013) and PsycINFO (1987 to February, Week 2, 2013), together with International trial registries (the trials portal of the World Health Organization (ICTRP) and ClinicalTrials.gov). Reference lists from previous meta-analyses and reports of randomised controlled trials were checked, and authors were contacted for unpublished data.
Randomised controlled trials of media-delivered behavioural or cognitive behavioural therapy in adults with anxiety disorders (other than post-traumatic stress disorder) compared with no intervention (including attention/relaxation controls) or compared with face-to-face therapy.
Both review authors independently screened titles and abstracts. Study characteristics and outcomes were extracted in duplicate. Outcomes were combined using random-effects models, and tests for heterogeneity and for small study bias were conducted. We examined subgroup differences by type of disorder, type of intervention provided, type of media, and recruitment methods used.
One hundred and one studies with 8403 participants were included; 92 studies were included in the quantitative synthesis. These trials compared several types of media-delivered interventions (with varying levels of support) with no treatment and with face-to-face interventions. Inconsistency and risk of bias reduced our confidence in the overall results. For the primary outcome of symptoms of anxiety, moderate-quality evidence showed medium effects compared with no intervention (standardised mean difference (SMD) 0.67, 95% confidence interval (CI) 0.55 to 0.80; 72 studies, 4537 participants), and low-quality evidence of small effects favoured face-to-face therapy (SMD -0.23, 95% CI -0.36 to -0.09; 24 studies, 1360 participants). The intervention was associated with greater response than was seen with no treatment (risk ratio (RR) 2.34, 95% CI 1.81 to 3.03; 21 studies, 1547 participants) and was not significantly inferior to face-to-face therapy in these studies (RR 0.78, 95 % CI 0.56 to 1.09; 10 studies, 575 participants), but the latter comparison included versions of therapies that were not as comprehensive as those provided in routine clinical practice. Evidence suggested benefit for secondary outcome measures (depression, mental-health related disability, quality of life and dropout), but this evidence was of low to moderate quality. Evidence regarding harm was lacking.