Is ENAR really useful? Leonardo Costa, Professor from the Master's in Physiotherapy Program University City of São Paulo 5 November 2010 I would like firstly to congratulate the authors for their trial. However it seems that neither the authors and the reviewers paid attention to a number limitations of the study that changes the conclusions of the study completely. In addition to the lack of intention to treat analysis and a loss of 20% of follow up (30/24) the biggest problem in this paper is that there is a clear imbalance of the baseline values from all outcomes observed (as can be easily seen in figures 2, 3, 4 and 5 and tables 4, 5 and 6), being the patients allocated to the ENAR intervention with higher levels of pain and disability. This issue could be explained due to the per protocol analysis, small sample size, randomization corruption or a combination of both. This imbalance is crucial to determine the between-group differences and therefore the effects observed by the authors cannot be attributed to the intervention. This is important for readers to know before they start applying this new intervention in their patients. I suggest to readers to be careful while interpreting this study and high quality RCTs are needed to test this new intervention. Competing interests There is no competing interests.