Unaccepted Medline

Last week, I discussed how to get free research papers online without having to hack into any publisher’s database.

Now, my Google Alert for Journal of Biological Chemistry has raised an interesting issue on the MEDLIB-L list from Thomas B. Craig, Assistant Director of Library Services at The University of Texas Health Center at Tyler.

Craig points out that he gets a lot of eprint requests from users after JBC papers that simply don’t show up in his library system as yet being available.

As Sciencebase readers know, JBC puts its “Papers in Press” free online. There can be a few dozen of over a thousand available at its website at any one time. The journal indexing system Medline, automatically indexes these “in press” papers and so as Craig explained some time ago on MEDLIB-L, “we get Docline requests for them when a library does not recognize a citation as coming from JBC Papers in Press.”

The requests contain an ID number but no page or issue number, just an annotation to say: “e-pub ahead of print”.

The JBC papers are essentially preprints, they go live on the site even before they’re edited or accepted officially for publication.

Craig thus asked: “Does this type of paper belong in MEDLINE, and what is its usefulness when it has yet to be accepted and is subject to change?”

Early awareness of research in a particular area would be one reason, but this issue makes a mockery of the out-dated and pointless system of artificial media embargoes that journals such as JACS, Nature, PNAS, Science, and others slap on their papers to help them coordinate the press response to a particular paper’s publication. Moreover, given that a paper published ahead of acceptance and seen by more peers might lead to flaws being spotted that were missed by the official referees, before it goes to press, might help the review process and prevent dubious papers ever making it to the Table of Contents.

The recent suspected research fraud being investigated at the University of Missouri is just the most recent example in a string of problems associated with fabricated and massaged data stretching back decades.