SCIENCE POLICY:

SUMMARY: After requiring independent statistical analysis in academia prior to publication, the Journal of the American Medical Association has been publishing fewer studies financed by industry.

The Journal of the American Medical Association (JAMA) has required all manuscripts financed by industry to be independently checked by statisticians in academia prior to publication, starting in mid-2005. This commendable policy has unfortunately not been adopted in similar form elsewhere within the medical publishing field, and has (predictably) faced criticism from some quarters.

Has JAMA's scientific policy resulted in the journal publishing fewer industry-sponsored studies? Elizabeth Wager (Sideview, United Kingdom) and coworkers show that it has, although we can't tell from their analysis whether industry is shunning the journal, as opposed to whether such research is no longer surviving peer review.

Data analysis.

Two scientists studied all randomized controlled trials published in the JAMA from July 2002 through June 2008. Such articles often clearly listed their funding sources.

The articles were listed as either industry funded (direct funding, provision of materials, or joint commercial and industrial funding), non-commercial, or unclear/no information provided. A similar procedure was performed for two other medical journals, The Lancet and the New England Journal of Medicine (NEJM), which do not have the statistical requirements of JAMA.

Industry vs JAMA.

The percentage of all randomized controlled trials published in these three journals changed little (from a decrease of 5% in JAMA to an increase of 4% in NEJM) after JAMA introduced their statistical requirements. The situation regarding industry-funded studies was quite different.

Such studies funded by industry decreased by 26% in JAMA. In contrast, they increased by 12% in NEJM, and by 10% in The Lancet.

Implications.

Is industry shunning JAMA, or is their funded research no longer surviving the peer review process as well as it had been previously? In other words, is JAMA's statistical requirement decreasing manuscript submission rates, or is it decreasing manuscript acceptance rates?

Unfortunately, there's no way to answer this question. JAMA says that such information is not available.

The scientists didn't attempt to measure article quality, with a specific emphasis on statistical methods. In my opinion, this question should be pursued, to determine whether current JAMA statistical policies have resulted in increased statistical quality in their articles.

Has JAMA's statistical policy been successful? We don't know this either.

If shoddy research is being funneled to other journals instead of JAMA, then the purpose has been defeated. This question needs to be answered if we are to know whether the visionary policy adopted by JAMA is having a positive effect on the technical medical literature.

NOTE: The scientists report no sources of funding for their research. Possible competing interests, which they explicitly declare in their article, include the fact that one of them advises pharmaceutical companies and technical journal employees, and another has received financial support from public sources as well as industry.

Wager, E., Mhaskar, R., Warburton, S., & Djulbegovic, B. (2010). JAMA Published Fewer Industry-Funded Studies after Introducing a Requirement for Independent Statistical Analysis PLoS ONE, 5 (10) DOI: 10.1371/journal.pone.0013591