The review addressed a clear question with well-defined inclusion criteria for study design, intervention and outcomes. Inclusion criteria for participants were not stated. Limited details were available for participants, making it difficult to assess the homogeneity of the studies and generalisability of the findings. Several relevant databases were searched for articles in any language, minimising the risk of language bias. There did not appear to have been any attempts to identify unpublished data. Publication bias was not assessed, so it could not be ruled out. It was unclear whether appropriate steps were taken in the review process to minimise the risk of reviewer error and bias.
A validity assessment did not appear to have been conducted, so it was not possible to ascertain the quality of the included trials. Despite the fact that only RCTs were eligible for inclusion, data only appear to have been extracted from the treatment arms. Therefore the trials were treated as uncontrolled trials, losing the benefits of RCTs. Also, the results appeared to have been summed across trials, rather than using appropriate methods for pooling data.
Given the potential for publication bias and bias in the review process, the unclear quality of the included trials and uncertainty regarding the analyses, the authors' conclusions are unlikely to be reliable.