It Is Time to Fix the Peer-Review System
Henry R. Black, MD, MACP, 15 West 81st Street, New York, NY 10024
Peer review is the essence of scientific exchange. Those of us who produce what we think is publishable science depend on our peers to evaluate our work and rely on the editors of appropriate journals to publish it in a timely fashion.
But with the advent of seemingly scores of new journals, medical and traditional periodicals with stories penned by undereducated reporters, webcasters, bloggers, and the like, getting good science presented promptly to colleagues has a higher degree of urgency. This situation has put added stress on editors, investigators, and the public. Editors are pressured to reduce the turnaround time from submission to acceptance and publication; investigators need to choose more carefully as to where to send their results; reviewers are pressured to submit their evaluations rapidly; and the public, both the scientific and lay audiences, are hungry for news, which is often carelessly and incorrectly described when the source is not privy to the full details of the study or capable of correctly interpreting the data.
Can This Situation Be Improved?
The solution may lie in borrowing from the internship matching program, which, though criticized by a few as unfair, has stood us in good stead for decades. As those of us who participated remember, the candidate, after investigating which programs he or she hoped to be matched to, ranked them in order of preference. The hospitals, after reviewing applications and interviewing the finalists, also ranked applicants in the order they preferred to make up their incoming class. After several weeks, as the algorithm ground through these data, each candidate was notified on a prespecified date as to which program he or she was matched. Hospitals also got their lists on a prespecified date. Those candidates participating in the matching program were committed to accepting their assignments and the hospitals were committed to accepting the candidates. A few applicants opted out, but almost all candidates for certified internship programs participated. (Some fellowship programs have also adopted this system, although it hasn’t worked as well.)
How Would This System Work for Journals and for Reviewers?
- •Journals willing to participate would need to form a consortium that would agree to abide by the rules they develop.
- •Investigators would submit their manuscript to the consortium with a list of participating journals in the order in which they would prefer to have their paper published. To limit abuse of the program (eg, hundreds of journals being listed), the investigators could list only a predetermined number of potential journals, perhaps five. They could always resubmit the paper with a second or third list if none of the desired journals accepted the paper.
- •Journals would agree to share reviewers, but they would still have the independence to accept, accept with revisions, or reject a paper, just as they do now.
- •Reviewers would agree to participate in the created consortium. This would be a great advantage to reviewers, who are now often asked to re-review the same paper that they had seen when it was submitted to a different journal days or weeks earlier. The list of topics he or she would review can become more specific, because the number of potential reviewers is likely to be larger than any one journal could recruit. Reviewers are the underappreciated links in this chain. The peer-review process requires a considerable amount of time and effort to properly evaluate a paper, and there is little, if any, reward for this work.
How Would This Process Improve Our Current System?
- • The time between the initial submission and final acceptance would be dramatically reduced for all but those papers accepted by the first journal to which it was sent.
- • Investigators would no longer have to send the manuscript out multiple times until a journal agrees to publish it. A paper submitted to the consortium would effectively be submitted to all participating journals. Investigators would no longer be in the unfortunate position of submitting a paper to a journal that wasn’t the one(s) they hoped would publish it, as the system exists now. The pressure to get a paper published promptly (and avoid multiple submissions and all the time and effort that takes) sometimes leads authors to send the paper to a journal with less prestige and impact than it might have deserved.
- • Each journal would not have to assemble its own panel of qualified reviewers. This would be the responsibility of the consortium.
- • Reviewers would not or should not be asked to review papers that are not specifically in their comfort zone. Once they have agreed to be part of the panel of reviewers, arrangements could be made to limit the number of papers a single individual is asked to review in a given period.
If a system such as this can be implemented, investigators will no longer have to guess whether their paper is acceptable to the journal(s) in which they most wish it to be published. The matching system will assure them that the journal that is both highest on their list and is interested will be the one that sends them an acceptance or provisional acceptance. If not, the paper would then go to the next journal on the list that has expressed interest. The time from submission to publication will surely be reduced. The most sought-after journals will have to triage more manuscripts than they currently do, but how much of a problem this will be remains to be determined.
A system such as proposed here will streamline the peer-review process, disseminate good science to all audiences more rapidly, and reduce the burden on investigators, editors, and reviewers. Our colleagues and the public deserve a more efficient and effective system than the one that is currently in place.