This review answered a clearly defined review question. The authors stated that they tried to identify both unpublished and published studies. However, relevant data may have been missed as only one database was searched and only English language studies were eligible for inclusion, which increased the risk of language bias. Although two reviewers extracted the study data, similar precautions were not reported for the selection of studies and assessment of trial quality, so the risk of reviewer error and bias was unclear.
Relevant criteria were used to assess the methodological quality of the included trials, but a number of the trials appeared to have methodological flaws. Statistical heterogeneity was assessed and only found to be significant for one pooled analysis. However, there appeared to be differences between the trials with in the definition of outcomes, trial methods and measurements, which suggested that the findings may not be reliable (as acknowledged by the authors).
The evidence supported the authors' findings, but the methodological shortcomings of the data and the variability between trials suggest that the conclusions should be interpreted with caution.