Attacking the FDA – Pseudoscience masking itself as patient advocacy

Speaking of libertarians, reading the JCI this week I came across this wonderful review of Richard Epstein’s new book, “Overdose: How excessive government regulation stifles pharmaceutical innovation”.

We’ve discussed Epstein, and his ilk before. The libertarians that routinely attack the FDA as some kind of bogeyman, killing kids, eating babies, blah blah blah, when the market could be making all these drug decisions for us. David Ross, writing for JCI, sees through the nonsense.

Although Epstein terms Overdose a study, it’s really a legal polemic that could be subtitled “What’s good for pharma is good for America,” his title for a 2006 newspaper opinion piece (1). Overdose raises questions but fails to persuade, largely because Epstein consistently fails to mention evidence that might inconvenience his argument.

First, he ignores the reality that patients and health care providers cannot acquire on their own the information they need to make rational choices about drugs. Without complete information, drug market failures, in the form of poor patient outcomes, are inevitable. Epstein dismisses this problem, proposing, for instance, surfing the Internet as a replacement for FDA-required information on drug efficacy and safety.

Second, Epstein seems unconcerned about the effects of market failure. He notes the Elixir Sulfanilamide disaster of 1937, which cost over a hundred lives, without explaining how his proposals would prevent such catastrophes. He posits the effects of safety issues on companies’ reputations as a deterrent to the knowing marketing of unsafe drugs, ignoring repeated examples of manufacturers doing just that. He concedes that the average patient does not have the ability to tell whether a drug is contaminated yet assumes that judging clinical trial results is a simple matter.

Finally, Epstein repeatedly distorts clinical trial science. The most peculiar section of Overdose is Epstein’s attack on randomized controlled trials (RCTs). Epstein seizes on the heterogeneity of responses to any therapeutic intervention as evidence that reliance on RCTs is flawed. He contends that drugs that are equivalent to (or even inferior to) placebo should still be approved, arguing that patient responses in the experimental arm that are greater than the placebo mean imply efficacy for the drug in those patients — committing the blunder of rejecting the null hypothesis for a post-hoc subgroup. Epstein appears unaware, or indifferent, to the role of variations in natural history, drug response, and assessment, and the role of type 1 error. Under his scheme, patent medicine nostrums would be legal again.

Epstein’s hidden agenda appears to be opposing more FDA-mandated information on drug efficacy and safety that might level the playing field between drug manufacturers and consumers.

I don’t have much hope that we’ll be able to keep these arguments out of the debate over the FDA forever. People have such a poor understanding of history (and everyone from that period is dead or soon will be – the drugs didn’t help), that they may actually be convinced of this nonsense. But I can think of few suggestions more profoundly stupid and intellectually dishonest than to suggest that we can create this ideal market system in which omniscient consumers navigate through the vagaries of pharmacologic science and RCTs to find the best drug for their condition, or that the market will magically provide information to consumers when the evidence is that companies have already shown they are disinclined to do so.

The argument over whether or not we should have validation of safety and efficacy of drugs by non-partial government scientists is over. The only way one can argue for a return to a pre-FDA market regime is to argue dishonestly.