Will David Graham’s Avandia study fall victim to the Ingelfinger Rule?
Last week, an unpublished study by researchers including longtime Avandia critic and FDA reviewer David Graham — who was a well-known Vioxx critic before that — made a bit of a splash after it was leaked to Pharmalot’s Ed Silverman. The study found, Ed reported:
48,000 excess events attributable to Avandia among patients 65 years or older between 1999 and June 2009. And since 62 percent of Avandia use has been among people younger than 65, they estimate the national impact is probably 100,000 or more.
I’ll leave summarizing the findings to Ed, who deserves the kind of traffic this scoop should get. You can also read the manuscript, which the authors were planning to submit to JAMA. According to emails leaked along with the manuscript, they were concerned that the FDA would try to block their submission.
Larry Husten, at Cardiobrief, picked up on the news, and ended his post with this:
It is theoretically possible that the early release of the manuscript will jeopardize or complicate the acceptance of the study by JAMA or another medical journal.
Larry’s suggestion is what prompted this post. In short, he’s questioning whether JAMA will invoke the Ingelfinger Rule and reject the paper. As I noted in a previous post, the rule, named for the New England Journal of Medicine editor who codified it, says that for a journal to publish a study, its results can’t have appeared elsewhere, with the exception of limited conference presentations.
It’s important to David Graham that the paper is published, because that will mean that it has been vetted by peer review. At this point, it’s too easy for Avandia maker GlaxoSmithKline to dismiss its findings, as they did in a comment to Pharmalot, because the study isn’t yet peer-reviewed.
Peer review is certainly not without its flaws; good papers can be rejected thanks to withering criticism from competitors, and weak papers can make it through superficial reviews. But to paraphrase Churchill’s views on democracy, it’s the worst way to vet scientific findings, except for all those other forms that have been tried from time to time.
Many people, including me, have concerns about reporting on studies that aren’t yet peer-reviewed, although I’d be quick to say I have no problem with Ed and others giving this one prominent play, given David Graham’s track record of getting things right. Avandia has already been the subject of criticism — as Larry notes, this study may just tighten the noose around the drug’s neck — but there’s still a public good argument here, if nothing else.
It’s unclear what will happen. We don’t know, for example, who leaked Ed the paper. Does the Ingelfinger Rule apply if the author didn’t seek any attention himself?
The Wall Street Journal, picking up Ed’s scoop, reported that a JAMA spokeswoman “said the journal doesn’t ever confirm or deny receipt or acceptance of any manuscript until publication.” A journal spokesperson told my colleagues at Reuters the same thing.
In the absence of any indication, it’s worth noting what JAMA editor in chief Catherine DeAngelis told New York Times medical correspondent Lawrence Altman in 2000 in a story on JAMA‘s tightening of the Ingelfinger Rule:
Many scientists who submit manuscripts believe ”they are on to something that is the most important thing since sliced bread, and they want it out there,” Dr. DeAngelis said.
”Fine,” she added, ”put it out there, but don’t submit it to JAMA.”