Lifting the veil on the editorial process

Abstract

Recently a large group of medical editors from around the world met in Prague to discuss the current state of peer review and the future of the editorial process in biomedical journals. A common theme running through the International Congress on Biomedical Peer Review and Global Communications was that the traditional secrecy and lack of transparency of the editorial process had to cease. Powerful arguments were made to open up the process of peer review to easier scrutiny, to make editors more accountable for their decisions and to ensure the system of reviewing and accepting papers for publication is fair, equitable and reproducible. The body of scientific knowledge being built on research into peer review is in its infancy. Few of the processes involved in peer review have been studied and fewer still validated. The development of instruments to measure the quality of papers, the quality of peer review, and the reliability of reviewer assessments is as yet poor. The evidence on whether reviewers give better reports when blinded to the identity of authors is now conflicting, despite initial work from Journal of General Internal Medicine1 suggesting this was so. Indeed a variety of journals showed that attempted blinding of reviewers to the identity of authors was successful as little as 50% of the time2,3,4, especially in smaller specialties. The question whether reviewers' reports would be better if authors knew their identity has not been investigated. Despite the lack of evidence that better quality papers result, there is a strong feeling amongst medical editors that biomedical journals must move towards opening up their processes and lifting the veil on the peer review process. Anonymous peer review is an example of privilege and power disconnected from reviewers' responsibilities to fellow scientists. Justice is poorly served by secrecy. I am convinced that revealing the identity of reviewers to authors is essential. Providing a report in the knowledge that authors will not know the reviewers' identity can result in negative rather than positive criticism. While most reviewers refrain from making disparaging remarks under the current system, knowing that authors will be aware of the reviewers' identity can only encourage accountability for comments. Some editors went further and suggested that perhaps the reviewers' identities should be published with the article. Indeed the Medical Journal of Australia has pioneered an even more innovative idea, opening up the peer review process to comment on the Internet, effectively increasing the pool of reviewers for a given paper. The MJA has posted some of its papers accepted for publication on the Internet, inviting comment from interested readers. A selection of these (edited) comments has been fed back to authors where the editors felt the comments were pertinent. Craig Bingham of the MJA showed that around 30% of papers were modified following additional comments received on the Net. Richard Horton, editor of the Lancet, reported that in July 1996 his journal established an ombudsman to deal with complaints related to delays in handling manuscripts, editorial discourtesy, and failure to take reasonable account of authors' concerns5. While Emergency Medicine does not yet have the resources for similar initiatives, there is little doubt that we should be moving to more transparency and accountability in our peer review processes. To this end, the Editorial Board recently approved a proposal to reveal the identities of reviewers of a paper to the authors. I believe we should not do this immediately though; rather we should now begin to survey authors on their satisfaction with the reviewing process, and then in 12 months time begin advising authors of the identity of the reviewers. We will then be able to compare authors' satisfaction with the peer review process before and after the change. We will also be measuring the quality of the reviews produced to assess whether this intervention improves quality, but given the lack of reliable measures of quality this is likely to be unproductive. This study should at least show us whether contributing authors feel they are getting a better deal from a more open review system than the present one. Although readers buy journals and hence determine its viability, without authors there would be no journals. We must encourage researchers in our field to submit their work to our journal, in the knowledge that they will be dealt with fairly and openly.