Journal Conflicts: Drilling Down Farther

When I was writing HOOKED, one of the most frustrating subtopics that I encountered was the financial aspect of medical journal publication. I found it nearly impossible to gather data about the degree to which most journals were entangled with the drug industry. Andreas Lundh and his colleagues at the Nordic Cochrane Center were able to do a better job than I at uncovering some key data but still came up partly empty-handed:
http://www.plosmedicine.org/article/info%3Adoi%2F10.1371%2Fjournal.pmed.1000354

In this somewhat older article (hat tip to Primary Care Medical Abstracts and Drs. Bukata and Hoffman for pointing it out) the authors tried to gather data on how many studies in six of the top medical journals in English are industry-sponsored, what impact these articles have on the impact factor of the journal; and how much money the journals make from industry ads and from sales of reprints to drug companies. The impact factor is key; journals are in cutthroat competition over this measure of how widely cited their articles are in the rest of the medical literature. Impact factor plays a role in determining where important studies are submitted, how many libraries subscribe to the journal, and how easily the journal can attract advertising.

When Lundh and colleagues went looking for the financial data, they adopted the excellent initial strategy of asking. Of the six journals they were studying, BMJ and Lancet replied, while JAMA, Archives of Internal Medicine, New England Journal, and Annals of Internal Medicine refused to supply any data. When the authors then obtained tax reports and tried to calculate some of the relevant numbers, and sent their calculations to the publishing organizations for confirmation, American College of Physicians (publishers of Annals) responded but the others again refused to divulge any financial information.

Bottom line: most medical journals keep their financial numbers very close to the vest--even journals that have editorial policies that encourage full disclosure of financial conflicts among other parties.

So with what they could get their hands on, Lundh and colleagues noted that the percentage of clinical trials supported solely by industry varied from a low of 3% in BMJ to 32% in New England Journal. Industry-funded trials were more widely cited than other studies (a phenomenon noted in a number of previous surveys; being sure to write more articles that cite a successful study, and then placing those articles in turn in higher-impact journals, is standard industry marketing practice). Therefore as one would expect, publishing more industry-sponsored studies has an effect on the journal's impact factor. Had those studies not been published the impact factor would have dropped only by 1% at BMJ, the low end of the scale, but by 15% for NEJM. In other words, journals have a significant financial interest in publishing industry-sponsored studies on the grounds of impact factor alone even before we get to ad and reprint sales.

For the only two journals for which they could raise the data, Lundh et al. found that journal reprints made up only 3% of revenue for BMJ but all of 41% for Lancet. Tax returns indicate that the AMA, publisher of JAMA and Archives, derives 12% of revenue from reprint sales and a whopping 53% from ads.

The authors end with the following sensible recommendation: "We suggest that journals abide by the same standards related to conflicts of interest, which they rightly require from their authors, and that the sources and the amount of income are disclosed to improve transparency."