I edit a journal called Ophthalmology that competes with the American Journal of Ophthalmology (AJO). We are all in this together, and I am flattered to be asked to offer editorial commentary on a article written by the Editor-in-Chief, Thomas J. Liesegang of the American Journal of Ophthalmology (AJO) along with associates, Shaikh and Crook.1 They ask an important question, and journal readers will want to know the answer to their question: What happens to papers sent to a leading journal that are not accepted? The authors determined if the paper apparently is not published (based on no match after an electronic literature search), or if they do find a match, they tell us if the paper appears in a better journal or a lesser journal, based on the journal impact factor as the scoring criterion.
First, I hope all journal readers understand that editors do not know everything, as any author of a rejected manuscript quickly will agree. Dr Liesegang is an expert in some areas, for example, corneal disease and I am an expert in aspects of retinal disease. Editors consider themes and features of papers important to them and that they believe are important to readers; for example, Is the writing succinct and clear? and Is the paper organized in a thoughtful manner? For clinical projects, Is the sample size meaningful? and Are the conclusions reasonable and not overreaching? We need the advice of reviewers for areas in which we are not experts, and frankly, we need them in areas in which we are experts. Readers should be reassured that they do not simply see papers selected by Dr Liesegang or by Dr Schachat, but rather, by an editor who presumably has already seen favorable comments from experts reviewers.
Before I comment on the findings, I will mention some important caveats. The lack of a match does not mean the rejected or withdrawn manuscript will not published. It simply means that Liesegang and associates did not find a match using their methods, which included an electronic search in PubMed using the same title or the main components of a title. They also search all manuscripts by the first author and by the corresponding author. They correctly note that papers may be published in journals not indexed by PubMed or MEDLINE, but I also note that when papers are resubmitted, titles may change, perhaps based on the peer reviewer suggestions, and authorship may change as well. Perhaps a reviewer has asked for more statistical analysis and a biostatistician is added to the author list. After considering the reviewer or editor comments or additional data collection or analysis, conclusions may change, and with that, the title may change. Suffice to say that things change, and no match does not reliably mean that the findings have not appeared in another journal. I accept that most of the time no match does not mean (yet) printed.
The impact factor, which is the number of times a journal was cited in the previous two years divided by the number of articles in that journal over the same period, is used as a marker for a more prominent or less prominent journal. Prominence is variably defined and the website at www.dictionary.com offers “standing out so as to be seen easily; conspicuous; particularly noticeable.” To me, because the impact factor measures how often there are citations, based on this definition, it is a reasonable measure of prominence. It need not be a measure of good, better, or best. Other aspects of journals need to be considered. Journals need to be publish in a timely manner, be affordable, and in this day and age need to have good web sites. Impact factor alone does not mean the journal is a good one. The impact factor is a reasonable surrogate marker for being good, but it is not perfect.
And what did the authors find? Approximately 50% of original manuscripts and brief reports are rejected or withdrawn from the American Journal of Ophthalmology (AJO) were published elsewhere. Approximately 90% of the time, publication was in a lower-impact journal. There is a little systematic bias here. There are only a few clinical ophthalmology journals with a higher impact factor than the AJO’s, so if a paper is published subsequently, there is a statistical tendency for it to be in a lower-impact journal. But the point nevertheless is valid-if your paper is not accepted in this journal and it is printed somewhere, it is likely to be in a less prominent journal.
I would like to highlight a few take-home messages. I believe the research summarized in the report by Liesegang and associates helps to validate the importance of peer-review. The authors in their Methods section do not explain how the editors decide to accept or reject a paper, although their procedures are explained on the journal website, AJO.com. I am making the assumption that the AJO editors make decisions using an approach that is similar to my own. The acceptance rate of the journal I edit is similar to that of the AJO, as is the number of manuscripts received per year. Our impact factors are similar. We use an overlapping pool of manuscript reviewers. If I receive two favorable reviews, I generally read the article, add my comments and suggestions for revision, and usually receive a revised manuscript that eventually is accepted. When there are two negative reviews, the paper is almost always rejected. I spend most of the time on papers with disparate reviews. Depending on the topic, sample size, likelihood of generalizability, my assessment of readership interest, our manuscript backlog at the time, and other factors, I make a decision about whether to reject the paper or to request and consider a revision. If my assumption is correct and the decisions at the AJO are made in a fashion similar to that of decisions at Ophthalmology, then the authors essentially have shown us that the papers that they are not accepting, based on the large part of the advice of their reviewers, appear in lower-impact (surrogate for less good) journals or do not appear at all. This information helps to validate the rationale for peer review. The AJO prints better papers; rejected papers are not printed or if printed, usually appear in lower-impact journals.
A second point to remember is that there are a lot of eye journals and eventually, I believe most papers are accepted somewhere if the authors persevere. With web searches and electronic systems, most can be found, and every day a patient comes into my office carrying an article printed in a lower-quality journal. Because poor-quality reports are printed, ophthalmologists need to have a firm basis in clinical research to assist them in interpreting the literature. Many editorialists have reminded readers what to look for: prospective study design, contemporaneous randomized comparison groups, large sample sizes, long follow-up, an a priori selected primary outcome, and other markers of top quality clinical work. The highest certainty recommendations are graced with supportive evidence from two or more high-quality randomized trials.
Finally, I make mistakes, and so does Dr Liesegang. We accept some papers we should not. When valid letters arrive correcting us, we accept them. We accept papers that prove us wrong and set the record straight. We also erroneously reject papers that are publishable. The remedy here is considering an appeal from the authors. Alternatively, there are other high-impact journals, and if a paper is a good one, it will be accepted elsewhere. I congratulate the authors on their excellent report.
FULL TEXT
