The review addressed a clear question and inclusion criteria were defined. The literature search was appropriate for published studies, but restrictions of the review to published data meant that there was a possibility of publication bias. Appropriate steps were taken to minimise bias and errors when extracting data and assessing inclusion, but it was unclear whether such steps were also taken when assessing quality.
Trial quality was assessed using appropriate criteria; the results were clearly presented. Methods used to pool data were appropriate. Results were clearly presented with the help of forest plots.
The authors' conclusions were supported by the data, but should be interpreted with some caution due to the possibility of publication bias and (as the authors acknowledged) the methodological limitations of the included trials.