Reader Comments
Post a new comment on this article
Post Your Discussion Comment
Please follow our guidelines for comments and review our competing interests policy. Comments that do not conform to our guidelines will be promptly removed and the user account disabled. The following must be avoided:
- Remarks that could be interpreted as allegations of misconduct
- Unsupported assertions or statements
- Inflammatory or insulting language
Thank You!
Thank you for taking the time to flag this posting; we review flagged postings on a regular basis.
closeReferee Comments: Referee 1
Posted by PLOS_ONE_Group on 20 May 2008 at 10:40 GMT
Referee 1's review:
**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication, the manuscript has been revised in light of these comments and to address other editorial requirements.
*********
The authors have responded thoroughly and satisfactorily concerning almost all points raised at the first review and have amended the manuscript accordingly.
My only remaining concern is with the analyses of CRP (p11 and Fig.2). It is now stated that 'notched boxplots' are displayed in Fig.2 and the symbols are explained.
The formula for calculation of the notches is given, but I believe there is a mistake: the minus-sign for the power '1/2' is superfluous. I understand that the notches should indicate a confidence interval for the median: if so, please state this so that the reader can interpret the information in the notches.
A number of 'notches' are reversed (in relation to the box quartiles) due to (a) the small sample sizes and (b) the skewness of the data. This makes the symbols difficult to grasp visually and may shed doubt upon the reliability of the notches as indicators of confidence intervals.
There is some disagreement between the statement in the text (p11) that CRP level was significantly lower on days 3,5,6 and 7 and the considerable overlap of the notches between placebo and verum groups on days 3 and 5. (Is the statement based on the 'post-hoc tests for isolated significances' referred to on p7? If so, is it not so that such post-hoc tests should only be carried out if the main 'global' test is significant? Please clarify.)
Therefore, I would recommend either removing the notches and employing simple boxplots or addressing these discrepancies.
A minor point: in Fig.5 the small n-values '3' and '1' appear to be erroneous.
Jeremy Miles's Comments (Reviewer #3) –e2159 Boutriau
**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication, the manuscript has been revised in light of these comments and to address other editorial requirements.
*********
I have only one, relatively minor comment.
I'd like to see a little more description of the random allocation procedure, for example, what covariates were used for blocking. The trial protocol mentions a standard SAS program, but doesn't tell us anything more about it.