I found this article in The American Scholar via Arts and Letters Daily, entitled “Flacking for Big Pharma”. As you might have possibly guessed from the title, it’s a broadside against the advertising practices of the drug industry, and particularly against its interactions with physicians and the medical journals.
And I’ll say up front that the piece is not, in fact, completely wrong. It’s probably not even mostly wrong. There really are big problems in these areas, such as too-aggressive promotion, minimization of side effects, too many payments to “key opinion leaders”, too many studies that don’t see the light of day, and so on. And these things really do lower the respect that people have for the drug industry – assuming, by this point, that there’s much respect left. But overall, this article is sort of a summary version of Marcia Angell’s book, for people who would like to hate the drug industry but find themselves pressed for time. And as such, it manages to get some important things wrong in the process of getting some things right.
For example, it makes much of subgroup analysis of clinical trials, but as a way for drug companies to pull the wool over readers’ eyes. I wonder how much this really happens, though, since overzealous data mining of a trial that wasn’t powered to generate such conclusion is (you’d think) a well-known pitfall by now. Perhaps not, though. But the example given in the article is BiDil:
BiDil proponents published studies that supported their claim of a racially mediated genetic anomaly that was addressed by BiDil, making it an ideal drug for blacks but not for whites.. . .
NitroMed won FDA approval of a new trial that included only 1,050 black subjects, with no white subjects to provide comparison data. Furthermore, BiDil was not tested alone, but only in concert with heart medications that are already known to work, such as diuretics, beta-blockers, and angiotensin-converting enzyme (or ACE) inhibitors. The published results of the trial were heralded as a success when subjects taking the drug combinations that included BiDil enjoyed 43 percent fewer heart-failure deaths.
. . .excluding whites was a medically illogical but financially strategic move because it eliminated the possibility that the drug would test well in whites, thereby robbing NitroMed of its already thin rationale for calling BiDil a black drug. The “black” label was crucial, because BiDil’s patent covering use in all ethnic groups expired in 2007, but the patent for blacks only allows NitroMed to profit from it until 2020. BiDil is a case study in research methodology “flaws” that mask strategies calculated to make a dodgy drug look good on paper, for profit.
But this doesn’t appear to be correct. First off, as the article itself mentioned earlier, the BiDil combination was originally tested (twice) in racially mixed (in fact, I believe, mostly white) trial groups. Secondly, the 1,050-patient trial in black patients was done with other therapies because to do otherwise would be unethical (see below). And what you wouldn’t realize by reading all this is the BiDil, in fact, was a failure. No one’s making piles of profits on BiDil until 2020, especially not NitroMed. You wouldn’t even know that NitroMed itself gave up trying to sell BiDil three years ago, and that the company itself was acquired (for a whopping 80 cents a share) in 2009.
Now, about those placebo-controlled trials. This article makes much of a British Medical Journal satire from 2003 on how to make a drug look good. But it’s confused:
A placebo, such as a sham or “sugar” pill, has no active ingredient, and, although placebos may evoke some poorly understood medical benefits, called the “placebo effect,” they are weak: medications tend to outperform placebos. Placebo studies are not ethical when a treatment already exists for a disorder, because it means that some in the study go untreated. However, if you care only that your new drug shines in print, testing against placebo is the way to go.
Well, which is it? We can’t, in fact, run placebo-controlled trials just to “shine in print” when there’s a standard of care, you know. You can only do that when there’s no standard of care at all. And in those cases, what exactly should we use as a comparison? Using nothing at all (no pills, nothing) would, in fact, make our drugs look even better than they are, because of that placebo effect. This is a specious objection.
And when there’s a standard of care that a new drug will be added to (as was the case with BiDil), then you actually do have to run it with those therapies in place, at least when you get to Phase III. The FDA (and the medical community) want to know how your drug is going to perform in the real world, and if patients out in that real world are taking other medications, well, you can’t pretend that they aren’t.
In another section, the article makes much of the Merck/Elsevier affair, where Elsevier’s “Excerpta Medica” division set up some not-really-journals in Australia (blogged about here). That was, in fact, disgraceful (as I said at the time), but disgraceful apparently isn’t enough:
. . .Elsevier, the Dutch publisher of both The Lancet and Gray’s Anatomy, sullied its pristine reputation by publishing an entire sham medical journal devoted solely to promoting Merck products. Elsevier publishes 2,000 scientific journals and 20,000 book-length works, but its Australasian Journal of Bone and Joint Medicine, which looks just like a medical journal, and was described as such, was not a peer-reviewed medical journal but rather a collection of reprinted articles that Merck paid Elsevier to publish. At least some of the articles were ghostwritten, and all lavished unalloyed praise on Merck drugs, such as its troubled painkiller Vioxx. There was no disclosure of Merck’s sponsorship. Librarian and analyst Jonathan Rochkind found five similar mock journals, also paid for by Merck and touted as genuine. The ersatz journals are still being printed and circulated, according to Rochkind, and 50 more Elsevier journals appear to be Big Pharma advertisements passed off as medical publications. Rochkind’s forensic librarianship has exposed the all-but-inaccessible queen of medical publishing as a high-priced call girl.
Fifty journals? Really? As far as I can tell, that figure comes from this analysis at the time, and seems to be mostly nonce publications, one-off conference proceedings, and the like. There is a whole list of “Australasian Journal of So-and-Sos”, which would be the same reprint advertorials as the other Excerpta Medica stuff, but do these still exist? (Did all of them on the list, in fact, ever actually publish anything?)
You’d get the impression that Elsevier is (or was, until Big Pharma came along) an absolute shining pinnacle of the medical establishment – but, with apologies to the people I know who work there, that is unfortunately not the case. They’re big, and they’re very far from the worst scientific publishers out there, but some of their titles are, in fact, not adding much to the total of human knowledge. Nor has the conduct of their marketing department always been above reproach. But no, this has to be built up to look even worse than it is.
The irritating thing is that there’s plenty to criticize about this industry without misrepresenting reality. But does that sell?