IN A CERTAIN SENSE, Merck & Co.'s decision to cease selling Vioxx, its best-selling painkiller, demonstrates how the peculiarly American combination of government regulation and private-sector competition can, serendipitously, sometimes work well. Vioxx was approved after trials held under the auspices of the Food and Drug Administration showed it to be effective (which it was). After some initial evidence that the drug might be linked to cardiac disease, the FDA required Merck to attach an additional warning (which was justified). Merck, which knew Vioxx was in competition with similar drugs, then conducted further trials, largely to find out if Vioxx could be legitimately recommended for other conditions. Through this process, Merck discovered that Vioxx did indeed carry a small but unacceptable risk of cardiac problems when used for longer than 18 months. Competition combined with regulation forced the drug off the market. And if courts determine that Merck was negligent, the company will pay a heavy price in compensation.

On its own, this incident does not make a good argument for more stringent FDA approval procedures. Extensive and lengthy pre-approval testing of the sort that would have kept Vioxx -- and all other drugs -- under examination for years ultimately would have harmed more patients than it would have helped. As hundreds of patients' groups can attest, there are high costs when the FDA does not approve drugs quickly, even drugs with serious side effects. People die every day waiting for new treatments.

But the Vioxx withdrawal should inspire some sharper thought about what happens to drugs once they have been approved, and whether the FDA should have the authority to mandate more thorough, ongoing, long-term testing. This is a safety issue, but also an economic one. Vioxx is a new-generation painkiller, one of several that are (or were) prescribed instead of aspirin or other over-the-counter drugs. Vioxx was easier on the stomach than aspirin, but it cost more. Did the 20 million Americans who used the drug since its launch in 1999 really have to spend that extra money and, as it turns out, incur a slight extra risk? Or were they persuaded to do so by Merck's advertising?

According to studies carried out by the Kaiser Family Foundation and others, the pharmaceutical industry's advertising encourages some people to consult doctors, which is a good thing, particularly for those who may not have known they have, say, high cholesterol or blood pressure. But ads also persuade people to spend money on unnecessary drugs, which is a bad thing for their health and for insurance premiums. The solution is not a total ban on advertising, as some suggest, but rather more clinical trials of drugs, aimed specifically at determining effectiveness as well as long-term safety. It is in the insurance industry's interest, the FDA's interest and the federal government's interest -- because the federal government is a major provider of health insurance -- either to require drug companies to conduct such comparative tests or to set up a neutral agency to do so. If there is a lesson to be learned from the Vioxx withdrawal, surely it is the need for renewed attention to evidence-based medicine and evidence-based drug prescription as well.