This review addressed a clear question. The inclusion criteria seem to have been appropriate, but it was unclear whether the included study designs were all defined in advance. The search covered a range of relevant databases, but its restriction to peer-reviewed studies, published in English, means that some relevant studies could have been missed. The methods for study selection, data extraction and quality assessment were not reported, which means that the risk of reviewer error or bias is unclear.
Standard methods were used for quality assessment, and the results were applied in the synthesis. Some relevant details of the included studies were presented, but there was little information about the study participants. A narrative synthesis was appropriate, but the vote counting approach weighted all studies equally, which could have produced misleading results. Differences between studies were investigated, but no general conclusions could be drawn.
Despite these limitations, the authors' conclusions broadly reflect the evidence presented and are likely to be reliable. A lack of detail on which types of intervention were most effective, limits the utility of the review and supports the authors' recommendation for further high-quality research.