There is a small but growing literature discussing the validity and reliability of the peer review process.1 There has been substantial effort to address the specific selection bias (usually referred to as ‘ publication bias’) where decision to publish is based on the research results.2 However, there has been little discussion in the literature about perhaps the largest publication selection bias; editorial licence. The aim of this editorial is to make explicit the critical issues that underlie Injury Prevention ’s editorial selection process and invite the journal’s readers to provide further comment.

There were 250 original research articles submitted to Injury Prevention in 2017. Over that same 12-month period, 25% made their way successfully through all the many editorial filters to print. That is, 25% of only those injury prevention papers that were submitted to the journal and for which the journal’s editorial team had direct responsibility. Right from the point of choosing their research questions, authors themselves make a series of editorial decisions about what to research, what manuscripts to write and whether (or not) to send their manuscripts to Injury Prevention. In the context of a world of possible manuscripts relevant to the prevention of injury, the manuscripts published each year in Injury Prevention are a highly select group.

Journal editors do not publish a random sample of the papers submitted to their journal. Indeed they have an obligation to do otherwise. While it may be self-evident that an editor’s role is to select the best manuscripts from the ones submitted—efforts to define ‘best’ reveals the nuanced nature, and potential impact, of an editor’s influence. Among the responsibilities of editors of medical journals enumerated by the World Association for Medical Editors are four of particular interest here.

Editors are responsible to readers, and should learn about their needs and interests. …… Editors are responsible for the editorial content of the journal; that is, the subject matter and types of articles and the actual content of the articles that are published in the journal. …… Editors should work to improve not only the quality of manuscripts but also the quality of research in the field. …… Editors should plan for the future of their journals.3

All this for an editor whose main tool of trade is manuscript selection! So how is does it work? First, editors select from the manuscripts submitted to the journal those they believe are worthy of further review. About half of the manuscripts submitted to Injury Prevention are rejected at this stage of the process. As noted in the journal’s letter to these authors, decisions at this stage are made on the basis of quality of science, relevance to the audience, usefulness for prevention and balance of topics across the field. Second, for those manuscripts being referred for external peer review, the editor seeks reviewer judgement not only on the scientific strengths of the manuscript but also on the article’s importance, its usefulness, its originality, its contribution to knowledge, its clarity and its internal coherence.4 Third, in the instances where manuscripts are judged to meet most of these criteria but are not yet deemed to be publishable as they stand, the editor invites manuscript authors to consider reviewer concerns and resubmit a revised version of the manuscript for further consideration. Throughout each of these steps the editor uses subjective judgement to improve the journal’s responsiveness, relevance, quality and impact.

While editors have a responsibility to be selective, they also have an obligation to be transparent about the selection process and hold themselves accountable for the results. This transparency is usually managed in a qualitative manner. Editorial teams are comprised of a diverse range of respected research leaders. Journal policies are explained at national and international meetings and appropriate checks and balances are built into operational editorial processes. Feedback from readers features in the letters and commentaries pages of the journal.

Transparency could be managed perhaps more optimally by editors publishing the explicit criteria they use to ensure consistency of their decision making. This is not usual practice. The most rigorous approach, which to the best of my knowledge has not yet been adopted by any journal, is to publish the journal’s publication metadata in such a way as to make the editor’s decisions a quantifiable and contestable characteristic of the journal. The effect of a journal’s editor’s decisions can be monitored by flow charts published annually, detailing submission rates, the points at which rejection occurs, the numbers rejected at each point and the characteristics of the papers lost. An annual content analysis by topic, study population and methods of the manuscripts that do reach publication papers would be informative, as would a quantitative summary of the use of these papers by the journal’s readership.

For a journal such as Injury Prevention, which has a unique place in its coverage of the public health approach to injury,5 detailed content analysis of publications in print, combined with a bibliometric analysis of the content and authorship of injury prevention articles published in the range of other scientific journals, would provide global insights about the nature and extent of the entire injury prevention field.