• DOI:10.1017/S0033291707000062
• Corpus ID: 12980225

What makes self-help interventions effective in the management of depressive symptoms? Meta-analysis and meta-regression

@article{Gellatly2007WhatMS,
  title={What makes self-help interventions effective in the management of depressive symptoms? Meta-analysis and meta-regression},
  author={Judith Gellatly and Peter Bower and Sue Hennessy and David A. Richards and Simon Gilbody and Karina Lovell},
  journal={Psychological Medicine},
  year={2007},
  volume={37},
  pages={1217 - 1228}
}

• J. Gellatly, P. Bower, +3 authors K. Lovell
• Published 19 February 2007
• Psychology, Medicine
• Psychological Medicine

ABSTRACT Background Although self-help interventions are effective in treating depression, less is known about the factors that determine effectiveness (i.e. moderators of effect). This study sought to determine whether the content of self-help interventions, the study populations or aspects of study design were the most important moderators. Method Randomized trials of the effectiveness of self-help interventions versus controls in the treatment of depressive symptoms were identified using previous reviews and electronic database searches. Data on moderators (i.e. patient populations, study design, intervention content) and outcomes were extracted and analysed using meta-regression. Results Thirty-four studies were identified with 39 comparisons. Study design factors associated with greater effectiveness were unclear allocation concealment, observer-rated outcome measures and waiting-list control groups. Greater effectiveness was also associated with recruitment in non-clinical settings, patients with existing depression (rather than those ‘at risk’), contact with a therapist (i.e. guided self-help) and the use of cognitive behavioural therapy (CBT) techniques. However, only guided self-help remained significant in the multivariate analysis [regression coefficient 0·36, 95% confidence interval (CI) 0·05–0·68, p=0·03]. In the subset of guided studies, there were no significant associations between outcomes and the session length, content, delivery mode or therapist background. Conclusions The results provide some insights into moderators of self-help interventions, which might assist in the design of future interventions. However, the present study did not provide a comprehensive description, and other research methods might be required to identify factors associated with the effectiveness of self-help.