The paper reviewed here is ‘Self-help Therapy for Insomnia: A Meta-Analysis’ by van Straten and Cuijpers. The authors note the use of pharmaceutical approaches to insomnia as well as psychotherapeutic interventions. However they are interested in those approaches that the person with insomnia can use by themselves with some supervision if necessary. For this purpose the authors have selected randomised controlled trials looking at self-help for insomnia in comparison with a control group. They describe the process for selection of papers – they searched through several databases – ‘Pubmed, PsycInfo, Embase and digital dissertations’. Here they inform the reader that papers are selected up to January 2007 although they do not stipulate the start year in each database. They also use a secondary method for identifying papers of searching through the references of identified papers. They excluded papers that had greater than 50% loss to follow-up but otherwise papers under the general heading of ‘self-help’ were included meaning that a patient could use this approach ‘independently’ although they go on to add that they did include papers in which there was time-limited supervision for patients in learning to use these approaches. Since sleep diaries were used, the authors were able to use total sleep time, sleep onset latency, sleep efficiency, wake after sleep-onset and the secondary outcomes of anxiety and depression. The authors explain their analysis methods very clearly – they subtract means scores of the control group from the experimental group and divide by the ‘pooled standard deviations of both groups’ to produce an effect size. What I found interesting here is that there is a specialised computer package for use in meta-analysis – Comprehensive Meta-Analysis which the authors used. There were a variety of approaches used in calculating the effect sizes in the different groups and they used the package to look for any publication bias with a funnel plot followed by a trim and fill procedure*. They identified 14 randomised studies with varying conditions for the control groups (e.g. waiting list and active treatment). 7 of the studies included participants aged 18 or over although other studies included other age groupings. The impression I got from reading through the results section was that there was great variation amongst studies in a number of areas including inclusion and exclusion criteria, recruitment strategies, support and self-help approaches. With a random effects model they found variation in the effectiveness of interventions on sleep-outcomes whereas with a fixed effects model they found benefits on all sleep outcome measures with the exception of total sleep duration. There weren’t found to be any significant benefits in the analysis of studies that had follow-up – with the follow-up period of 4 months chosen for the analysis. The authors found that smaller studies more positive results and after using the Trim and Fill method the effect size wasn’t statistically significant. There weren’t any significant findings from an analysis of subgroups in the study. In their discussion, the authors note that they couldn’t identify any significant improvement in depression or anxiety scores. I didn’t see any mention of what the authors thought about the difference in the results obtained using random versus fixed effect models in the discussion however. Initially I couldn’t see any benefit to including such a heterogenous group of studies in the meta-analysis but on further reflection thought it provided a useful starting point to asking more specific questions.
* Duval and Tweedies trim and fill procedure
Steps To Treatment = 4
Steps To Treatment (STT)
STT = Steps To Treatment. An estimate of the number of steps between the results and translation into treatment. This is an opinion.
Annemieke van Straten and Pim Cuijpers. Self-help Therapy for Insomnia: A Meta-Analysis. Sleep Medicine Reviews. 2009. 13. 61-71.
You can listen to this post on Odiogo by clicking on this link (there may be a small delay between publishing of the blog article and the availability of the podcast).
If you have any comments, you can leave them below or alternatively e-mail email@example.com
The comments made here represent the opinions of the author and do not represent the profession or any body/organisation. The comments made here are not meant as a source of medical advice and those seeking medical advice are advised to consult with their own doctor. The author is not responsible for the contents of any external sites that are linked to in this blog.