 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
| |
|
||||||||||||||||
| |
|
||||||||||||||||
|
Volume 22, No. 2
|
Book Review Peer Review in Health SciencesLondon (England): BMJ Publishing Group, 1999;
The main message of Peer Review in Health Sciences is based on the observation that despite its critical importance to science, the peer review process is flawed and has remained "remarkably untouched by the rigours of science". The book is a call to action to improve peer review, consider alternatives and increase research into the review process. The two editors, Fiona Godlee, a scientific editor from the British Medical Journal, and Tom Jefferson, from the Surrey and Cochrane Centre in the United Kingdom, have brought together data and the views of many editors and researchers interested in peer review to give a description of the current state of peer review and reflections on its future directions. In the first section, Drummond Rennie, a scientific editor at the Journal of the American Medical Association, co-director of a Cochrane Centre and organizer of three international congresses on peer review, masterfully identifies the main problems with peer review. It is unreliable, unfair and fails to validate or authenticate; it is unstandardized and open to bias; blinded peer review invites malice, either from ad hominem attacks on the author or by facilitating plagiarism; it stifles innovation; it lends spurious authority to reviewers; reviewers knowledgeable enough to review a study are often competitors, and therefore have a conflict of interest; and it causes unnecessary delays in publication. To this sobering list is added, in a chapter on the peer review of grant applications, the observation that "the most important question to be asked ... is whether or not [peer review] assists scientists in making important discoveries that stand the test of time. We do not know." Some may wonder - in light of the depth and breadth of these difficulties - why there is peer review at all. A chapter on the state of the evidence for journal peer review concludes that "It is the only [viable] system we currently have" and as other authors point out, the alternatives, either a free-for-all or audits, look worse. Some optimists may be tempted to say "Peer review may not be perfect. So what?" noting that while these problems exist, they do not appear to be rampant, so in general the system continues to work. The answer to this is in the chapter on bias, subjectivity, chance and conflict of interest: if the studies that get funding and the studies that are published are subject to a flawed review system, this leads to publication bias. They then note: "publication bias is perhaps one of the more important practical and ethical issues currently facing biomedical journals." In a much shorter section on the future of peer review, the editors agree that peer review is here to stay, but could be improved. Clinical trial registries would help to stem the tendency to publish only positive trials, but this will not improve the review process. Electronic communications can, and likely will, affect future peer review. In some disciplines, such as physics, authors post electronic drafts of their manuscripts, called "eprints", for comment prior to formal publication. The Medical Journal of Australia experimented with electronic posting and the authors were given the opportunity to make revisions prior to the print publication. This journal's editors have also tested online peer review and note that one of its advantages is that it allows interactivity among reviewers, authors and the editor. A central issue in peer review is exactly how the evaluation is carried out. This has remained largely discretionary. Although most journals and granting agencies use checklists, it is the reviewer comments that carry much more weight. Teaching the peer review process is a surprisingly new concept. In a tongue-in-cheek interview near the end of the book, "Socrates" is engaged in a discussion on the problems of peer review, then poses this rhetorical question regarding some research results on the subject: "You asked untrained people to do what you concede is a difficult job. And then you went to the trouble of carrying out a study which showed that they weren't very good at it?" In an attempt to address this problem, the book also includes a "how to" section. This includes a researchers' guide to the peer review of grant applications, an authors' guide to editorial peer review, a reviewers' guide to peer review, and an editors' guide to setting up a peer review system. Although this is a useful section, the changing target audience for each chapter, as well as the contrast in the target audience for this section (mostly neophytes) with respect to the rest of the book, which targets seasoned editors and researchers, may seem somewhat disjointed to readers who read the book from cover to cover. Peer Review in Health Sciences is a credible, fairly expensive, paperback reference text that identifies the major issues in peer review today and summarizes the problems and dearth of evidence to support this critical practice. Its biggest contribution is in highlighting the humbling fact that the peer review process is far from perfect and that continued efforts are needed to improve it.
Pat Huston
[Previous] [Table of Contents] [Next]
|
 |
|||
|
Last Updated: 2002-09-30 |  |
|
