The Instructions for Authors at JAMA includes a form that that must be signed by authors that includes the following:

“2. (check at least 1 of 2 below)
[ ] drafting of the manuscript
[ ] critical revision of the manuscript for important intellectual content”

This is, clearly, a restatement of the ICMJE rule.

Similarly, the NEJM’s current policy statement about who should be an author: “The specific requirements for authorship promulgated by the International Committee of Medical Journal Editors (ICMJE), which have been adopted by many biomedical journals, are posted on its Web site at http://www.icmje.org... As in the past, our policy is that all persons listed as authors must meet the ICMJE requirements for authorship. ”

There’s never even a clue as to who actually wrote the paper, and that’s considered to be more or less irrelevant.

That is the problem that this policy is specifically designed to address.

Very different meaning if you leave out random words.

Very different meaning if you leave out random words.

Two words. Graduate students.
In theory this would save a lot of time. In practice, not so much.

No, I don’t think anyone follows that rule. I’d never heard of any rule like that. And I work in this field.

If you contributed significantly to the research, you become an author. Look at any article detailing the results of a clinical trial. You’ll probably see a dozen or two dozen authors, representing a dozen different institutions. These are all the doctors who organized the trials at those institutions, plus a couple statisticians and experimental technicians. There’s never even a clue as to who actually wrote the paper, and that’s considered to be more or less irrelevant. “Author” represents who did the work.

Now, some journals require the submitters to detail which of the authors contributed what to the paper. So it’ll say “JMM, HUY and DXX designed the experiments. DXX, SOS and POS carried out the experiments. BFF performed statistical analysis. JMM and DXX wrote the manuscript.”

Ghostwriting and guest authorship are all too common in academic publishing and, naturally, raise concern over the integrity of the underlying research. However, undisclosed authorship, per se, is not the same as reporting of fraudulent data. The suppression or ommission of negative results (as in the case of Vioxx) is a larger threat to the public than the practice of borrowing political support from the scientific and medical community to report true results.

Again, it was once acceptable for drug companies to attach conditions on publication to research funds given to universities. It isn’t acceptable any longer.

The manufacturers get what they want by paying academics to consult. Carrots, not sticks.

Two words. Graduate students.

This was accepted practice in academic medicine a generation ago. Now, according to the Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication of the International Committee of Medical Journal Editors, it is necessary, but not sufficient, that your contribution includes “drafting the article or revising it critically for important intellectual content”. That is, some actual writing. I’m sure this norm is still abused, but I hope we have made progress.

More worrisome to me than this sort of venality is the practice of corporate right of first refusal on publication. In this way they can commission 10 articles, trash the 9 that they don’t like, and publish the 1 that says what they like. I think this means that even if the 1 that is published passes muster for its methodology, statistical analysis, etc., its epistemic import is artificially strengthened through lack of context. Though I think more and more journals are requiring disclosure of such arrangements, I don’t think this disclosure practice is yet universal.

I’m not an expert in this sort of applied epistemology, though, so if I’m off base here, I’d be happy to be corrected.