The review addressed a clear question with respect to participants, intervention and outcome. Inclusion criteria regarding study design was not explicitly stated. Relevant databases were searched for studies in any language and efforts were made to retrieve additional articles by searching references lists. The search was somewhat limited as there were no attempts to retrieve unpublished studies, increasing the possibility of publication bias.
The quality assessment suggested that the included studies may have been of poor quality, all of which were open-label studies with small number of patients. Statistical heterogeneity was not assessed, but the authors considered clinical heterogeneity which was found to be present. Given the heterogeneity of the data and a lack of description of statistical methods, combining the studies in a meta-analysis may not have been appropriate.
The inclusion of small studies with questionable quality and the variability among studies raises uncertainties about the reliability of the results, so the authors' conclusions may have been optimistic and are unlikely to be reliable.