Going Overboard; The FDA
Moving into high gear, the Food and Drug Administration last month boldly removed Pfizer Inc.'s Bextra from the marketplace and ordered stringent black-box warnings to be placed on just about the entire set of nonsteroidal anti-inflammatories that include not only the much maligned Celebrex but other familiar household drugs such as Mobic, Motrin, Naprosyn and perhaps a dozen other similar drugs. Aspirin and Tylenol (acetaminophen) are covered by the current orders, but who knows where the FDA will move next. The FDA decisions are based on its preliminary assessment of increased cardiovascular, gastrointestinal and deadly skin conditions associated with their use. As is its wont, the FDA insists that this action-which, atypically, went beyond the recommendations of its own advisory panel-will "protect" the public, even as it limits their options.
Whether these maneuvers protect the public is an open proposition. But the one party that surely benefits from these decisions is the FDA itself, which has been under a relentless assault from members of Congress and consumer groups on all sides of the political spectrum. Senator Chuck Grassley, R-Iowa, vies with Dr. Sydney Wolfe of Public Citizen in denouncing the lackluster performance of the FDA, which they think is in the pocket of the various pharmaceutical companies whose conduct is now under the microscope.
In today's overheated atmosphere, one often-touted proposal seeks to clip the wings of the FDA by creating a new independent board whose task would be to oversee the safety and efficacy-the two touchstones of FDA policy-of approved drugs already in the marketplace. This board would have extensive power to boost warnings and remove established drugs from the marketplace on the strength of ongoing studies of the performances of these drugs.
The FDA stoutly resists this proposal, and it is right, but for the wrong reason. Make no mistake about it: Any new board convened today could survive politically only by claiming the scalp of some drug already on the marketplace. The targets are there for the asking. The problem here is not that FDA-ordered clinical trials (which continue to get ever longer) are too lax. Rather, it is that they are, in the nature of things, finite.
To see what the problem is, note that the critics of the drug industry often urge two propositions simultaneously. The first holds that no individual should be made the guinea pig for untested therapy. The second insists that all drugs released into the marketplace should have received extensive clinical trials before they are made available to the general public. Unfortunately, these two desirable objectives are flatly inconsistent with each other. All the FDA, or any omniscient substitute, could do is to ask how many trials should be taken before a drug is released on the market.
The facile answer is the more trials the merrier. But that response ignores one of the inherent conflicts of interest that dog the FDA. When drugs misfire after they are released into the marketplace, the FDA draws flak from all sides. But when safe drugs are mired in trials, there are huge, hidden social benefits that are forgone. Typically, therefore, the critiques of the FDA's performance are directed largely to visible losses, to the exclusion of the benefits forgone, which are entitled to equal weight in the social calculus. Stated bluntly, I don't think that the FDA has done anyone a public service by taking drugs off the market, and placing endless detailed warnings of tiny risks on various products will all too often just be treated by physicians and ordinary folk as hyperbole, not information.
Worse still, we can be pretty sure that any independent board will always find grist for the mill once new blockbuster drugs hit the market. Suppose that an exhaustive clinical trial has 10,000 individuals, divided into two groups of 5,000 each. Any measure of effectiveness can usually be found in a small sample because we are looking for dramatic effects. Thus if the new drug is 200% better than no treatment at all, we might find improvement in 400 people in the placebo group, and 1,200 in the treated group. It doesn't take any sophisticated test to know that some good is in the works. But safety is a different matter altogether. The increments in question could be on the order of 1% heightened risk, which is much harder to detect in a small population, especially since clinical trials cannot cover the full range of dosages in patients whose responses may differ by age, sex, ethnicity and prior medical condition.
Once a successful launch is made, the drug is no longer used by thousands. Millions prescribe. At this point, small differences in safety level can be identified, which makes it fair game for the FDA or its independent substitute to remove it from the market. But the question is: to whose benefit? These public discussions often feature the use of absolute numbers of individuals who are said to be injured or killed by the new drug. But if 65 people are claimed to have died from liver involvement from a given drug, it is worth asking whether that comes from a user population of 10,000 or 2,000,000.
Why? Because the rate gives us some clue about the benefit side of the equation. If those large boosts in effectiveness hold for general populations, then tens of thousands of individuals could receive some improvement from the drug in question. Why are Chuck Grassley and Sydney Wolfe so confident that the FDA can do a better job in managing risk for a typical patient than he and his doctors? What the FDA grandly calls protection looks more like misguided paternalism.
Richard A. Epstein, an NLJ columnist, is a professor of law at the University of Chicago Law School and a senior fellow at the Hoove