News
2018; Lippincott Williams & Wilkins; Volume: 40; Issue: 12 Linguagem: Inglês
10.1097/01.eem.0000550378.47421.79
ISSN1552-3624
Autores Tópico(s)Radiology practices and education
Resumopeer review: peer reviewFigureAll emergency physicians, whether author, reader, or educator, should be aware of and contribute to the discussion of the changing nature of peer review. Why? Let's start with the sheer number of papers published each year. Peer review can thin the herd in an age of information overload. Emergency medicine journal rejection rates are high (71% in my last 93 reviews for two leading emergency medicine journals), but thousands of suspect articles are published anyway. One 2010 estimate said more than 400,000 papers had been published in questionable journals without true peer review. (BMC Med 2015;13:230; http://bit.ly/2yHwbeW.) Peer review also exerts quality control over the research on which we rely. Is it worth reading if it's not in PubMed? Editors performing peer review are held accountable, and potential conflicts of interest are addressed. The journals that use peer review also have recall and correction mechanisms in place. The value of peer review changes with one's perspective. Authors generally think that editors improve their papers. Nothing focuses one's writing like rejection. Readers find that confidence in peer review contributes to the lazy luxury of forming opinions based on an abstract. (Evid Based Med 2013;18[2]:48; http://bit.ly/2Ade0Qc.) Academics question whether they are up to date and credible, asking which evidence they should use to answer questions, evaluate research ideas, and resolve disparities. And malpractice attorneys use peer review to determine standard of care or an expert's qualifications. Criticisms of traditional peer review abound, however. One article with the provocative title, “Why Most Published Research Findings Are False,” raises concerns that that might be true. (PLoS Medicine 2005;2[8]:e124.) Another study found that “the small proportion of results chosen for publication are unrepresentative of scientists' repeated samplings of the real world.” (PLoS Med 2008;5[10]:e201; http://bit.ly/2yKEYN4.) There's also a fair amount of criticism that says peer review is a poor method for evaluation. “If the only thing Amazon ever published were reviews of the first three people who bought a product, then we'd have a very ineffective system for knowing what was good and bad,” Michael Eisen, a Howard Hughes Medical Institute investigator at the University of California, Berkeley, told NPR. (Feb. 24, 2018; https://n.pr/2NEPhYJ.) And an article in TheEconomist said “a rising number of journals that claim to review submissions in this way do not bother to do so.” (June 23, 2018; https://econ.st/2AdI7qU.) Around 8,000 predatory journals published, for a fee, an estimated 420,000 articles in 2014. (BMC Med 2015;13:230; http://bit.ly/2yHwbeW.) Multiple egregious examples demonstrate the absence of “peer review,” including one paper that was accepted for publication despite repeating the same seven words over and over (The Guardian Nov 24, 2014; http://bit.ly/2yGLTHf) and another in which students included references from the show “Rick and Morty.” That was accepted by three journals without question and five in exchange for payment. (Ozy Oct. 4, 2018; http://bit.ly/2Ep3JVl.) Another downside of peer review? Waiting for review and revision is often excruciating. Open Science Publishing The proliferating forms of peer review include traditional prepublication peer review in professional society-sponsored print journals, predatory publishers, and online educational resources such as FOAMed—Free Open Access Medical Education. Fewer than 15 percent of the 100 top emergency medicine and critical care open educational resources (OERs) describe any editorial process, and only 18 sites listed academic affiliations. (AEM Educ Train 2018;2[3]:204.) Open science publishing, defined as literature that is free on the internet to read, download, copy, distribute, and link, is a challenge to traditional publishing. Multiple variations allow all to follow the process and collaborate. (F1000 Research; http://bit.ly/2En9TW8; Figshare; http://bit.ly/2AeAel8.) Different terms are used when talking about OERs. Open access generally means freely available on the internet with or without peer review. Some funders, such as the National Institutes of Health and some charitable foundations, require open access publication. Open data means making data available to and reusable by anyone for further analysis, like the human genome project. Open peer review means posting all reviewer names and reviews online before publication, as BMJ has done since 1999. (NPR. Feb. 24, 2018; https://n.pr/2NEPhYJ.) Open review employs crowdsourcing. (Surowiecki J. The Wisdom of the Crowds. New York: Doubleday; 2004.) An unlimited number of reviewers (the public) discuss an already published manuscript, as Wikipedia and WikEM.org do. (Adv Physiol Educ 2007;31[2]:145.) Megajournals, like PLoS One, is a successful example. Publications are checked for scientific rigor and validity, but the significance of the results is left to the readers to decide. (Tohatoha Oct. 23, 2013; http://bit.ly/2PDoPjO.) Then there's post-publication peer review where preprints are published without review (F1000 Research; http://bit.ly/2En9TW8) and then followed by various review formats, including no review (like the distribution service bioRxiv where authors post unpublished preprints and readers provide feedback before submission to journals), formally invited reviewers (articles that pass are indexed in scholarly databases), volunteer reviewers with variable impact on the status of the published article, and journal-led reviews (journals host online post-publication discussion of their own articles). (J Am Coll Radiol 2018;15[1 Part B]:173; http://bit.ly/2ypyHXX.) Review also results from uninvited comments on blogs, podcasts, Twitter, and third-party sites. (CJEM 2018;20[1]:3.) Other sites such as F1000 Research charge a fee for immediate open access publication, transparent post-publication peer review by invited referees, and full data deposition and sharing. All scientifically sound articles are accepted and indexed in PubMed. Cureus is free to the author and reader, offers rapid publication, and relies on professional in-house editing, a community of scholars on editorial boards, and pre- and post-publication peer review. (Gaslighting the Medical Literature. Medscape; https://wb.md/2COAKsn.) Forces Affecting Change As always, we have to follow the money. Paying to receive or deliver science has hidden costs to science and to society. But how long can the current model of mailing monthly journals to dues-paying members last? Millennials will lead. It is important to distinguish among the various functions of peer review and the forms they are taking: determining if a paper is worth publishing, if it contributes to knowledge, and if it is scientifically valid; improving the work through editing; and disseminating and translating knowledge. The nature of peer review remains the author's choice, and peer review is better than no peer review. More than other specialties, we emergency physicians continually modify our practice with information from sources outside our own literature. Sorting good science from bad is a valid role for one's professional organization and peers. Knowledge translation is slow enough when the content has been reviewed, but speedy dissemination of bad information helps no one. Can we really be comfortable with the wisdom of the crowd?
Referência(s)