Talk about unnerving: In this month's New York Review of Books, Marcia Angell argues that "it is simply no longer possible to believe much of the clinical research that is published, or to rely on the judgment of trusted physicians or authoritative medical guidelines" when it comes to drugs or medical devices. Big Pharma, she argues, has corrupted the clinical trial process too thoroughly. (Interestingly, the essay focuses primarily on drugs, although the approval processes for medical devices or, say, new surgical techniques are often even less stringent—surely big financial interests have a lot of sway there, too.)

Angell's full essay is worth taking for a spin, but note that the problems she discusses are especially acute when it comes to off-label uses of drugs. Once the FDA approves a drug for a specific purpose, doctors can legally prescribe it for any other disease they feel like. So drug companies design and bankroll often-dubious trials to "test" those other uses, pay some esteemed researcher a fat consulting fee so that he'll slap his name on the results (unless the results are unfavorable, in which case into the dustbin they go), and then lobby doctors—with gifts, if necessary—to start prescribing the drug for the new use. Oh yeah, step four: Profit!

But if all this subtle—and often not-so-subtle—corruption is hard to eliminate, as Angell says it is, why not create stricter FDA regulations on off-label use? Angell estimates that roughly half of all current prescriptions are off-label, and no doubt many of those are valid, but we've also seen, for instance, a 40-fold increase in the diagnosis of bipolar disorder among children between 1994 and 2003, with many being treated with powerful psychoactive drugs off-label—side-effects and all. It makes little sense that a drug company has to spend years and millions of dollars getting FDA approval for one use of a drug, but then all other uses are automatically kosher, no matter how little reliable data exists. Or is this off-label "loophole" useful and effective in some way I'm not grasping?

On a related note, a few weeks ago, The New York Times had a great piece on how psychiatrists were gearing up to put together the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders, the "bible" of psychiatry. Edward Shorter, a historian in the field, was quoted arguing that psychiatry is not like, say, cardiology—no one really knows the causes of various mental disorders, so "political, social, and financial" factors often drive the classifications. Financial, indeed: As Angell notes, of the 170 contributors to the DSM-IV, 95 had financial ties to drug companies.

Now, does that help explain why, as Christopher Lane reports in his new book,
Shyness: How Normal Behavior Became a Sickness, ordinary shyness went from being listed as a rare "social phobia" in 1980s DSM-III to a full-fledged "social anxiety disorder" in 1994's DSM-IV? Maybe, maybe not, though GlaxoSmithKline certainly found it profitable to push the idea of social-anxiety disorder on the public ("Imagine being allergic to people...") and marketing its antidepressant, Paxil, as a treatment.

That's not to say the disorder's fake, or that Paxil can't help people; just that there's a set of murky judgments at work here not grounded in rigorous scientific evidence, and a little more transparency would be nice. But alas, as the Times reported, the contributors to DSM-V have all signed non-disclosure agreements. No peeking behind the curtain. True, the participants have agreed to limit their income from drug companies to "only" $10,000 a year while working on the DSM-V, though, given that drug companies potentially have billions at stake on what does and does not go in the manual, it's not hard to imagine that they might try to, ah, push the envelope a wee bit.