Open peer review: a randomised controlled trial

Background Most scientific journals practise anonymous peer review. There i s no evidence, however, that this is any better than an open system. Aims To evaluate the feasibility of an open peer review system. Method Reviewers for the British Journal of Psychiatry were asked whether t hey would agree to have their name revealed to the authors whose papers the y review; 408 manuscripts assigned to reviewers who agreed were randomised to signed or unsigned groups. We measured review quality, tone, recommendat ion for publication and time taken to complete each review. Results A total of 245 reviewers (76%) agreed to sign. Signed reviews were of higher quality, were more courteous and took longer to complete than uns igned reviews. Reviewers who signed were more likely to recommend publicati on. Conclusions This study supports the feasibility of an open peer review syst em and identifies such a system's potential drawbacks. Declaration of interest G.W. is the Editor, L.A. an Assistant Editor and E. W. and M.R. Trainee Editors of the British Journal of Psychiatry.