Having electronic preprints is logical

EDITOR— Delamothe asks what the policy of journals should be on publication on the internet of electronic preprints (“eprints”).1He points out that some journals, including the New England Journal of Medicine and, currently, the BMJ, reject papers whose substance has already appeared as an eprint because they regard this manoeuvre as prior publication. This policy is illogical and impossible to police. What is the difference between the published abstract of a study presented at a conference and an eprint? The only differences are that the internet version is more accessible and likely to be a fuller account of the study. To enforce the restrictive policy journals must even now be searching the internet for evidence of prior publication of every paper they receive; we do not envy them this nightmare.

We can imagine a scenario in which journals offer a home for eprints. When authors have ready a draft of a paper they could ask for the eprint to appear on a journal's website. Those papers that survive the in house selection procedure (about a third in The Lancet's case) would appear on the website within a few days of submission, accompanied by a note of the expected date of submission of the “definitive” draft and, as Delamothe suggests, a request not to quote the study. Once the paper had gone through the formal peer review process the eprint would be removed from the web and its fate (including, for an accepted paper, a link to the definitive published version) recorded in the eprint archive.

And what of papers that have appeared as eprints on non-journal websites? Editors who support the eprint policy could not logically refuse to consider these papers on their own merits when formally submitted. The disadvantage of this system for authors, as Bingham has pointed out, is that they would not be able to submit rejected papers to journals that opposed the eprint policy.2

The system outlined above should work for the bulk of medical research that interests only other medical researchers. An issue that requires careful consideration, however, is whether journals should host the eprints of studies whose results have the potential to cause real public concern. If the answer to this is yes, will the eprint be viewable by everyone or will there be controlled access? Perhaps the authors of such contentious studies can be encouraged to work with editors, reviewers, and public health agencies at the earliest stage of manuscript preparation to ensure that publication is swift and quality assured.