M Shih, YH Yang, and M Koo.
Review published: 2009.
The authors concluded that hypnotic intervention appeared to be a viable non-pharmacological intervention to treat the symptoms of depression. In light of the poor quality of available trials, the presence of heterogeneity and the absence of information of comparator conditions, the reliability of the authors' conclusions is unclear.
To evaluate the effectiveness of hypnosis in the treatment of depressive symptoms.
PubMed, the Cochrane Library, PsiTri, PsycLIT, EMBASE, the Cochrane Depression, Anxiety and Neurosis Review Group (CCDAN) and Google Scholar were searched for articles in any language. Search terms were reported.
Randomised controlled trials (RCT) comparing a hypnotic intervention with no treatment or standard treatment for the reduction of depressive symptoms were eligible for inclusion. Included RCTs had to use a standardised assessment tool to evaluate depressive symptoms.
Included RCTs were of hypnotic interventions delivered by a therapist, audiotape or through self-hypnosis. The length of hypnotic intervention, where reported, ranged from 30 to 60 minutes, and the frequency ranged from daily to bi-weekly. The duration of intervention ranged from 10 days to six months. Participants in the included trials were cancer sufferers, first-time mothers or adult volunteers with ages ranging from 18 to 81 years. The majority of participants were female. Severity of depressive symptoms at baseline ranged from non-depressed to severely depressed. Depressive symptoms were measured using the Edinburgh Post-Natal Depression Scale, the Self-Rating Depression Scale, the Beck Depression Inventory and the Hospital Anxiety and Depression Scale. The trials were conducted in Japan, China, South Africa, the USA and the UK.
The authors did not state how the studies were selected for the review, or how many reviewers performed the study selection.
Assessment of study quality
The methodological quality of the included trials was assessed using the Jadad, a three-item scale assessing randomisation, blinding and withdrawals and drop-outs, with a maximum score of 5 points.
Two researchers independently conducted the validity assessment, with disagreements resolved by consensus.
Means and standard deviations of post-treatment scores for intervention and control groups were extracted and used to calculate effect sizes for each trial. Where means and standard deviations were not available, the t or p values were extracted and used to calculate effect sizes.
Two researchers independently conducted the data extraction, with disagreements resolved by consensus
Methods of synthesis
Pooled effect sizes, with 95% confidence intervals (CI), were calculated using a random-effects model. Statistical heterogeneity was assessed using the I2 statistic. The fail safe N was calculated.
Results of the review
Six RCTs were included in the review (n=258 participants). All trials were categorised as poor quality according to the Jadad scale.
Hypnotic intervention was associated with a moderate reduction in depressive symptoms (effect size 0.57, 95% CI 0.32 to 0.81). There was evidence of moderate statistical heterogeneity (I2=47.4%). The number of trials with an effect size of zero needed to reduce the pooled effect size to zero was 25 (fail safe N 25).
Hypnosis appeared to be a viable non-pharmacologic intervention for depression.
The review addressed a clear question, with well-defined inclusion criteria for intervention, study design and outcomes. Inclusion criteria for participants were not stated. Characteristics of control conditions were not reported, which made it difficult to determine what hypnosis was being evaluated against. Several relevant databases were searched for articles in any language, minimising the risks of language bias. Attempts did not appear to have been made to identify unpublished data, but publication bias was assessed and was deemed unlikely to have influenced the results. Appropriate steps were taken in the validity assessment and data extraction processes to minimise reviewer error and bias. However, it was unclear whether similar steps were taken in the study selection process, so reviewer error and bias could not be definitively ruled out.
A suitable validity assessment was conducted. The quality of included trials was low, partly due to the difficulty in double-blinding hypnotic interventions. Appropriate methods were used to combine the trials. The authors conclusions represent the evidence presented. However, the presence of statistical and clinical heterogeneity between trials limited any conclusions that could be drawn.
In light of the poor quality of available trials, the presence of heterogeneity and the absence of information of comparator conditions, the reliability of the authors' conclusions is unclear.