The Food and Drug Administration is being buffeted by a kind of perfect storm. First there were claims that the labeling of certain antidepressants failed to warn doctors that the drugs caused some adolescents to commit suicide. Then the agency was blindsided by contamination that made half the nation's flu vaccine supply unavailable. Thereafter came revelations about apparent side effects of several widely prescribed anti-inflammatory analgesic drugs.
Taking a candidate drug from discovery in the lab to the marketplace requires 12 to 15 years and more than $800 million in direct and indirect costs. On average, it involves more than 60 clinical trials with more than 4,000 patients, but even this extensive testing may not elicit all possible reactions to a new drug. Thus, regulators must always make decisions on the basis of data that are, in a sense, incomplete. Infrequent side effects, or adverse events, might not show up until hundreds of thousands, or even millions, of patients are exposed to the drug in normal use.
Drug companies must report adverse events and injuries caused by their products to the FDA, but they depend on practicing physicians to provide these data. Because doctors are unlikely to notice side effects that occur in only a small percentage of their patient populations and are under no legal obligation to communicate adverse events at all, underreporting is common.
The "safety" of a drug is a relative thing. Safety and efficacy, the two criteria required for marketing approval of a drug, are inextricably linked; regulators' judgments require a global and often difficult calculation of risk and benefit. We tolerate greater uncertainty and more severe side effects for a potential cure for pancreatic cancer or AIDS, for example, than for a drug for treating heartburn. When the FDA grants marketing approval for a given product, that drug is deemed to be safe and effective for the conditions on the label.
The efficient detection of side effects is essential, and the United States needs to improve pharmacovigilance—the monitoring of the safety of approved drugs. But the newly created FDA Drug Safety Oversight Board is not the answer. Rather than more bureaucrats, we must have better data.
We need to encourage physicians' reporting of adverse events (perhaps by rewarding them with the Continuing Medical Education credits they need to retain licensure), to contract with organizations that treat large patient populations to monitor and report adverse events, and to share data with foreign regulators. We might also consider some variation of the United Kingdom's "yellow card" system, in which doctors, dentists, and pharmacists report adverse events to federal regulators (on a small yellow card, of course).
But we will not get any of these improvements—or for that matter, desperately needed reforms to redress the FDA's chronic risk aversion and overregulation of drug testing—without strong leadership. Lester Crawford, the president's nominee for commissioner, hardly seems the person for the job, but let us hope he will rise to the occasion and provide it.