| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
University of Pennsylvania Philadelphia, PA 19104-4283
Peer review has been a topic of much debate in the biomedical literature in the past decade [1]. The double-blinded peer review process, in which reviewers are unaware of author identity and institutional affiliation, has been implemented by many journal editors to safeguard against subtle forms of publication bias. Despite recent espousal of the double-blinded review process and the potential advantages of this approach, anecdotal experience of reviewers suggests that unblinding is rampant [2]. Unblinding of reviewers may be caused by author inclusion of identifying information in various components of the submitted journal manuscript. The editors and reviewers of journals that use a double-blinded review process must ensure the integrity of the double-blinded review because the readers are unaware of any unblinding that occurs during the review process.
As a reviewer for several journals with double-blinded peer review policies, I have encountered numerous and repeated occurrences in which authorship is easily ascertained from the purportedly masked manuscript. Current author instructions for these journals request the omission of authors' names or initials and institutional affiliations from the manuscript or figures. Unfortunately, such blinding rules are frequently violated, and current author requirements fail to delineate or account for all potential sources of unblinding [2]. When I have encountered such unblinding, I have contacted the editor and relinquished my responsibilities as a reviewer for the manuscript, yet I remain unaware of any action to blind subsequent reviewers.
The editorial decision for a journal to adopt a double-blinded review process is elective because many publications do not use such an approach [3]. Journals implementing a double-blinded review process, however, have an ethical obligation to provide quality assurance of confidential peer review. Author instructions for the submission of manuscripts to such journals should detail all potential sources of unblinding. The editorial office should review the quality of blinding for each manuscript with a detailed checklist of potential sources of bias before dissemination for review. The development of a quality checklist for blinding of submitted journal manuscripts may parallel the evolution of the CONSORT statement for the quality of reporting randomized controlled trials [4]. Until such guidelines can be established and implemented, the double-blinded review process remains a myth.
References
Winthrop-University Hospital Mineola, NY 11501
Radiology Richmond, VA 23219
RadioGraphics Bethesda, MD 20814
Dr. Liebeskind is to be complimented for his appropriateness in relinquishing his responsibilities as a reviewer when he believes he can identify the authors of a manuscript as a result of unblinding and for suggesting the development of a checklist of potential sources of bias. However, although unblinding by authors is relatively common in our recent experience [1], the frequency with which the reviewers will see this unblinding is substantially diminished because the editorial offices of Radiology and RadioGraphics remove from all manuscripts, including from the figures, any mention of the authors' initials or names, as well as of their institutions, before submission to the reviewers. These sources of potential unblinding accounted for 57% (214/374) of the individual violations by authors with regard to unblinding or potential unblinding of the 880 manuscripts inventoried in our study [1]. The number of instances in which the reviewers would know the identities of the authors or of their institutions, solely on the basis of the manuscripts that they receive during the review process, is therefore substantially less than if the manuscripts were sent to the reviewer without blinding by the editorial offices.
The development of a checklist to routinely inventory all submitted manuscripts is an interesting concept. However, we believe authors should also be involved in ensuring that their submitted manuscripts cannot be unblinded. Additionally, Liebeskind does not mention the other component of double-blinding, in which the authors do not know the identities and institutions of all reviewers [1, 2]. This policy of concealing the identity of the reviewers from the authors of all manuscripts is strictly enforced by our editorial offices. Although as we stated in our initial article, we realize that it is impossible to completely blind all reviewers (because even if authors make every effort to avoid unblinding, many reviewers will be aware of which researchers are working in specific niche subspecialty areas [1] or of which researchers have presented abstracts at certain meetings from which the current manuscripts under review originate), we agree with Dr. Rogers that it is the information contained in a manuscript, not the source of the information, that should make the difference in a reviewer's assessment of that manuscript and that knowing the source of a manuscript introduces a potential bias that is best avoided [2].
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
