The review question and supporting criteria were well defined. Multiple sources were searched to identify relevant evidence in any language, reducing the possibility of language bias; it was unclear whether unpublished studies were sought, so some studies may have been missed. Study selection, data extraction and quality assessment were performed in duplicate, which reduced the risk of error and bias in the review.
Quality assessment of the included trials was undertaken, but limited details were provided; most trials appeared to be of poor quality and contained small sample sizes. The chosen method of synthesis appeared to be appropriate given the absence of statistical heterogeneity. All included trials were conducted in North America, so the generalisability of the findings to other setting is uncertain, as acknowledged by the authors.
This was a generally well-conducted review, but the poor quality of included trials and concerns regarding generalisability should be borne in mind when interpreting the results.