| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ann Thorac Surg 2004;77:10-11
© 2004 The Society of Thoracic Surgeons
a Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania, USA
* Address reprint requests to Dr Edmunds, Annals of Thoracic Surgery, Hospital of the University of Pennsylvania, 3400 Spruce St, 5000 Ravdin Court, Philadelphia, PA 19104, USA
e-mail: hank.edmunds{at}uphs.upenn.edu
Peer review is the process by which grants are awarded and scientific journal articles are published; it is a surrogate for quality control [1]. The process has many critics and not a few deficiencies. For journal articles, peer review is expensive, delays publication and is vulnerable to unfairness, censorship, bias, and conflicts of interest [1]. Peer review does not validate methodology, detect duplicate publication and cannot certify truthfulness and ethical compliance [2]. Selection of reviewers is unsystematic, secretive and subjective. The process may invite inexpert reviewers, the authors' competitors, conflicted persons and individuals who are unwilling or unable to critically evaluate the paper. To be fair, reviewers are rarely provided guidelines for critiquing a paper or writing a review; the process is "ad hoc" from start to finish. "Quality control" often amounts to advising the editor to publish or not; the author receives no useful feedback. Consensus is rare and the benefits of the process have never been convincingly validated [3]. Yet there is no practical alternative [1, 4].
In the last two decades attempts have been made to assess and improve the process [1, 5]. Blinding reviewers of authors, revealing (unmasking) reviewers to authors and checklists do not noticeably improve the quality of editorial peer review [2–4]. Peer review does improve the quality and readability of the manuscript [3] but fails to improve the quality of the reported research and to detect fraud, plagiarism, conflicts and duplicate publishing [3, 4, 6]. Consequently, editors, educators and scientists have recently looked to the internet for solutions. Physicists now solicit colleagues’ input prior to publication and during the past decade have established preprint, open review of submitted articles by the entire physics community through the Los Alamos National Laboratory in New Mexico. Open peer review, as this process is called, works well for physics. Variations of open review have been tried by biomedical journals with inconclusive and mixed results [7, 8].
So where does The Annals of Thoracic Surgery stand? In this environment of change, it is wise to examine our protocols and perhaps initiate some "experiments" to improve our peer review process. To begin we must affirm the objectives of our journal, recognize the realities of our profession, and tailor our "experiments" accordingly. Our mission statement, "To provide a place for ALL thoracic surgeons to relate experiences which will help other practicing thoracic surgeons give better patient care," defines The Annals' objectives. No need to change direction in our 40th year. Our authors and readers are nearly all practicing cardiac or thoracic surgeons who spend most of their time in operating rooms and intensive care and patient units. Our reviewers are selected from these peers and all have varying fields of interest and expertise, but do not have unlimited hours to review manuscripts. Reviewing takes time and not all surgeons are willing to donate their time.
Several immediate experiments come to mind: improving selection of reviewers, increasing the number of reviewers for original scientific articles, decentralizing some of the editors' tasks and providing guidelines for the reviewers.
With the advent of electronic peer review, The Annals developed a keyword list with the idea that authors could define the content of their articles by choosing up to five "words" from this list and that reviewers could choose keywords from the same list to describe their interests and expertise. HighWire Press made it possible for the editor to link 2–5 keywords to search for reviewers that matched all of the linked words. Linking makes it possible to identify very specific expertise and content that is not covered by a single keyword. For example, the editor received a manuscript on DeJeune's syndrome and after looking up the term keyed in "chest wall" and "pediatrics." Nine potential reviewers appeared from which four were selected. Clearly this list and the linking feature improved this editor's ability to match reviewer expertise and manuscript content. Thus far, 2,391 of 3,487 registrants have filled out their personal keyword list. The Annals plans to revise the keyword list each year to keep pace with new developments in our fast moving specialties. Revised lists will appear in each January issue. The Annals hopes that all registrants will annually update their personal expertise list; 20–40 words usually describe most surgeons' focus of interest and expertise, but do not describe all they know. Too many keywords dilute our ability to match expert reviewers with manuscript content and may cast some reviewers into the marginal domains of their expertise.
Electronic peer review has enabled The Annals to increase the number of reviewers for each manuscript. For most biomedical journals, original scientific articles are reviewed by two individuals. The editor, who is rarely able to read a manuscript critically, resolves conflicting recommendations. Since January 2003, The Annals has invited four or more reviewers, one of whom is an editorial board member, to review each original article and two or three individuals to review each case report and "How To Do It" manuscript. Reviewers who decline are generally replaced; reviewers who accept but do not deliver within 6 weeks account for the shortfalls. More reviewers for each manuscript have been shown to detect more mistakes in manuscripts [8] and does add extra work for the editorial office. However, if it improves our journal, the editorial staff and editor will manage.
In the first 8 months since the inauguration of our website, www.atseditorialoffice.org, The Annals received a total of 760 case reports, "How To Do It" articles, and images—feature articles—as compared with 531 the previous year. We also received 112 more non-supplement, original scientific articles than in the first 8 months of 2002. Page space is limited for feature articles; the original scientific article remains the heart and soul of The Annals. Electronic peer review provides an opportunity to decentralize the selection of reviewers and decision making for feature articles to Feature Editors. This new category of editorial board members was initially recruited from the University of Pennsylvania for logistic reasons. With electronic peer review working well, Feature Editors will now be recruited worldwide. At the end of 2003, HighWire Press completed new, customized software that enables decentralization of the reviewing process for feature articles. This software and new operational protocols are being tested and prospective Feature Editors are being identified. The Annals hopes to implement decentralization of the review process for feature articles in early 2004. The editor will still approve all accepted articles, but will have a diminished role in the review process of these manuscripts.
Guidelines for reviewers raise the issue of what I, as editor, expect from the reviewers. With due respect to those with more experience, this editor expects reviewers to provide two insights. The first, which is the lesser of the two, is a recommendation supported by specific, objective comments that identify strengths and weaknesses of the article. The second, most important input is to instruct the author as to how the study could be or could have been improved to meet the modern standards of an original scientific article in the field of cardiothoracic surgery. Many of our authors write from all parts of the world, and often in their second and weaker language. This editor values reviewers' constructive suggestions to authors: how they can make this paper or their future papers better by pointing out fixable and unfixable flaws. Regardless of the decision to publish or not, this latter critique is "the gift from Anonymous," which is generously bestowed on authors by each reviewer. Multiplied, this gift raises the quality of care (and science) of our entire profession and indirectly rewards all of us and our patients.
The editor has added "Guidelines for Reviewers" alongside "Information for Authors" in this issue. A draft of this document was e-mailed to editorial board members and was modified by their responses. It is still a "work in progress," but is a start. The document will be further defined after discussion during the January editorial board meeting. Hopefully these "Guidelines" will prompt responses from reviewers and authors throughout the year and will be further improved by them. Like any initiative, these Guidelines are an "experiment" until we, all of us, can get it "right" for The Annals.
The Annals cannot address all of the deficiencies of the peer review system for ensuring quality control. Quality control starts with the integrity and ethics of each member of our profession. There are those among us whose ambitions entice them to "game the system" by plagiarism, duplicate publication, "editing raw data" and worse, but these misguided individuals are very rare. These miscreants can still ply their arts, but eventually will be discovered and cast out from the ongoing dialogue of our vigorous, dynamic profession.
The Annals of Thoracic Surgery needs peer review and in the editor's opinion the profession needs it also. Peer review protects patients from applications of "untried ideas," "one-time miracles" and "poorly conceived procedures." Written critiques are powerful vehicles of education; few authors fail to read their reviews carefully. The review process does improve the reporting of science, although it cannot improve or verify the science [3]. Realistically, no alternatives are available. Hopefully some of these "experiments" will work and raise The Annals to the "next level." Meanwhile the door of the editorial office is always open to your thoughts and new ideas.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
