The broad objectives of the review were stated, but there was no clear definition of eligible participants, comparisons or outcomes. Relevant sources were searched for studies. It was unclear whether the search was restricted by language or publication status. If so, it was possible that some studies were missed. The potential for publication bias was not discussed. It was unclear whether steps were taken to minimise risks of reviewer bias and error by having more than one reviewer independently select studies and extract data. It appeared that study validity was not assessed.
The review findings were difficult to interpret, partly because there was wide variation in the outcomes measured, types of data reported (within-group change, between-group change) and the samples analysed (completers only, responders only, study subgroups) and also because no confidence intervals were provided. Population characteristics, interventions and comparators varied. Most studies reported a large number of outcomes and it was not always clear which was the primary outcome. Sample sizes were very small in most cases and drop-out rates, where reported, ranged up to more than 40%.
Limitations in the review, which included failure to assess study quality, differences between the studies and a lack of predefined outcome measures in the review, mean that the authors’ conclusions require cautious interpretation.