The review was based on defined inclusion criteria and was underpinned by a search of a small range of resources. It was possible that studies were missed as the review only considered articles published in English. Unpublished studies were not eligible for the review, so there was the possibility of publication bias which could not be fully assessed because of the low number of studies. It was unclear if more than one reviewer was involved in all processes of the review to minimise reviewer bias and error.
Study quality was assessed. Statistical pooling may not have been appropriate given the diverse study designs and varying steroid regimens.
The authors' conclusions on ineffectiveness should be approached with caution. However, given the size and limitations of the included evidence, the authors' recommendation for further research appears to be appropriate.