Many people struggle with anxiety disorders such as panic disorder, social phobia, and Posttraumatic Stress Disorder (PTSD). Moreover, about 30% of us have an anxiety disorder at some point in our life (Kessler et al 2005) but most people never receive treatment.

Self-help interventions may provide a solution when people are unable or unwilling to participate in face-to-face treatment: waiting lists don’t have to be an issue, there is no need to travel to appointments, and costs can be low.

But do these interventions work? And if so, how well do they work compared to face-to-face treatment?

A team of Scandinavian researchers set out to:

• Describe the effectiveness of self-help for anxiety disorders compared to wait-list/placebo and face-to-face treatment, and
• See whether specific person- or treatment characteristics play a role in the effects. Their method consisted of a meta-analysis and a meta-regression, respectively.

They selected 56 randomized controlled trials with adults who struggled with a variety of anxiety disorders (mostly panic disorder/agoraphobia, social phobia, and PTSD). The self-help treatment had to be either printed on paper or made available as computer software or via the internet. Therapist contact had to be non-existent (“pure self-help”) or only facilitative (e.g. a 10 min phone call per week; “guided self-help”). The total sample consisted of 4713 individuals.

The researchers found that self-help treatment groups did significantly better than the wait-list or placebo control groups. The mean effect size (Hedge’s g) post-treatment was 0.78 (CI 95 0.67-0.90). In the multiple regression, self-referred samples did better than clinical samples. The authors see this as an indication that people with lower symptom levels or without comorbidity may benefit more from self-help than clients with many symptoms and/or comorbidity. There was no difference between pure self-help and guided self-help. The conclusion: engaging in self-help treatment is indeed (much) better than doing nothing.

When self-help was compared to face-to-face treatment, the latter was superior (the mean effect size was -.20; CI 95 -.037 – -.02). However, when the researchers looked at ‘treatment as usual’ and cognitive behavioural therapy (CBT) separately, something interesting happened. CBT outperformed the self-management interventions strongly, whereas treatment as usual was not significantly different from self-help. It suggests that as a patient, if you cannot get face-to-face CBT treatment, you can choose what you prefer; non-specialized psychological assistance or a self-management program. From a public health perspective, it makes sense to make self-help treatment widely available since face-to-face resources are very limited.

In the future we may be able to differentiate better according to the type of anxiety disorder involved. For example, self-help programs targeting PTSD or specific phobia yielded the smallest effect sizes and the most variability. Closer examination suggested that  written programs were less effective for PTSD than computerized self-help. This result also came up in one of the meta-regressions for the whole set of studies. It may have to do with computerized options enabling the tailoring of treatment.

In my view, this study provides hopeful insights. We may well be able to further ameliorate and extend self-help programs, particularly those online. Combined with more insight in how to increase uptake and completion of these interventions, there are opportunities to help far more people with anxiety disorders than we have done previously.

Links

Haug T, Nordgreen T, Ost LG, Havik OE. Self-help treatment of anxiety disorders: A meta-analysis and meta-regression of effects and potential moderators. Clin Psychol Rev. 2012 Jul;32(5):425-45. Epub 2012 Apr 21. [PubMed abstract]

Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005 Jun;62(6):593-602. [PubMed abstract]