The journal conducts double-anonymous reviews of all manuscripts submitted. As part of the decision process, reviewers routinely receive a copy of the decision letter, which includes reviewers’ comments. In the transition to a new editorial staff, a change to the email template inadvertently meant that the full letter was sent out, including the corresponding author’s name. Before this was discovered, decision letters for multiple manuscripts had been shared with reviewers. Four were sent to the same reviewers to review the resubmitted article. No reviewers contacted the editors to let them know if they were aware of the authors’ identities.
Decisions were made based on submitted reviews for two of these (one reject, one minor revision) and two await reviews. Several manuscripts could still be submitted as revisions. The editorial team is now debating whether it is ethical to send revisions of the manuscripts to reviewers we should assume know the identity of the corresponding author. Finding new reviewers familiar with the subject area will delay the process. New reviewers may make additional requests of the authors, adding to their burden and further delaying the process. We would prefer to allow the original reviewers to remain in the process, as they were carefully selected for their expertise.
Questions for the Forum
- Is it ethical to send revisions of the manuscripts to reviewers we should assume know the identity of the corresponding author?
- Can we make an exception to our double-anonymous process if we inform both authors and reviewers that we are doing so and get their approval? Reviewers would be asked if they have a conflict of interest, knowing the corresponding author’s identity, and could opt out of the process. Essentially, we would be moving to a single-anonymous review process only for these manuscripts.
Once a mistake like this has occurred, the journal should implement as transparent a policy as possible over their actions and mitigations. The journal should send the affected authors an apology, explaining what has happened and asking if they would like their work to be sent to an additional reviewer (which will take more time) or whether they are happy proceeding with a single anonymised review process. These actions would ensure the integrity of the review process was maintained, especially since many journals use single anonymised review processes as a matter of course. It was not felt to be advisable to disclose the identity of the reviewer to the authors since this would remove any anonymity in the process. Instead, all reviewers should be asked to confirm that they have no conflicts of interest in reviewing the article. Any paused articles with a decision should be treated in the same way. When the papers are published the journal should consider including a notice stating that they have been through a single anonymised peer review.
The journal followed COPE’s advice and informed all authors and reviewers of affected manuscripts of the situation. All authors were asked if they wished to request new reviewers, continue with the existing reviewers in a single-anonymous review, or withdraw their manuscript. The single manuscript with a rejection decision made prior to discovering the error was invited to resubmit it to a new slate of reviewers. The author declined this offer, having submitted the manuscript elsewhere. In nearly all cases, authors opted to continue with a single anonymous review. Before proceeding, all reviewers were asked to confirm they could be unbiased in a review of the manuscript, knowing the corresponding author. Authors were informed that any accepted manuscript would be published with a notification which stated that a technical error had resulted in part of the review process being single-anonymous rather than double-anonymous. All authors and reviewers agreed to proceed and no conflicts of interest were reported by the reviewers concerned.
Authors and reviewers seemed to appreciate the journal’s transparency in this process.