The abrupt withdrawal last week of the best-selling painkiller Vioxx is an event rich in ironies and lessons that may ultimately lead to a rethinking of the way drug safety is evaluated in the United States.

Drugs whose stated advantage is that they are safer than their competitors -- which was Vioxx's claim -- are likely to get much closer scrutiny. The Food and Drug Administration in the future may force companies to run new, definitive clinical studies specifically to address safety issues even after a drug is on the market and in widespread use.

Vioxx, taken by 1.3 million Americans, was removed from sale worldwide on Thursday because it raised the risk of cardiovascular disease, the nation's leading cause of death, but proof of that did not come until five years after the drug was licensed and three years after the first hints of a problem.

"We will want more data on [the safety of] long-term use for both drugs in the pipeline and those already approved," Steven K. Galson, acting director of the FDA's Center for Drug Evaluation and Research, said last week, two days after the agency learned that Merck & Co. planned to withdraw its blockbuster drug.

Vioxx's downfall is likely to spur other changes as well, experts said.

It may lead doctors and patients to take a fresh look at the consequences of direct-to-consumer advertising, the root of much of Vioxx's enormous popularity in a field full of competing pain relievers. It may also dampen some of the enthusiasm for expensive, targeted, "smart" drugs, which sometimes turn out to be no better -- or in Vioxx's case, actually worse -- than their cheaper, blunter, dumber ancestors.

"I think there are important issues here that we would be foolish not to reflect upon," said David Wofsy, a physician at the University of California at San Francisco and president of the American College of Rheumatology.

Vioxx, whose generic name is rofecoxib, is a "selective COX-2 inhibitor." When it and others in the same family were in development, their main appeal was that they were expected to cause fewer stomach ulcers than many existing painkillers.

COX-2 inhibitors block the action of an enzyme used throughout the body to make a large and diverse family of compounds called prostanoids that are involved in regulating myriad physiological actions, including inflammation, blood clotting and protection of the stomach lining from the destructive effects of digestive acids.

The fact that Vioxx acts only on COX-2 distinguishes it from aspirin and the large family of nonsteroidal anti-inflammatory drugs (NSAIDs) that includes Motrin (ibuprofen) and Aleve (naproxen), which block not only COX-2 but also a related enzyme called COX-1. Blocking COX-1 is responsible for the stomach-irritating effects those drugs can cause.

Selective COX-2 inhibitors are meant to be the pharmacological equivalents of fine surgical dissection, doing their work on COX-2 and its painful inflammatory products while leaving COX-1 alone. They do live up to their claim of producing fewer ulcers and other stomach irritation. People taking them have only one-quarter to one-half as many of those problems as people taking aspirin or NSAIDs.

The patients deemed most likely to benefit were arthritis sufferers, 10 to 25 percent of whom have ulcers (often small and not painful) if they take NSAIDs regularly. Arthritis physicians thought elderly people and other patients taking blood-thinners such as warfarin, in whom bleeding ulcers can be fatal, would especially be helped.

But when the first COX-2 inhibitors arrived in January 1999, they were marketed to a far broader group -- basically anyone who did not have effective pain relief with what they were using.

The drugs were alluring on many fronts.

They were new, which alone holds great sway with American medical consumers and physicians. They were supposed to be safer, another natural appeal. Perhaps most important, they had the cachet of being an expensive, prescription medicine in a world of mundane, over-the-counter generics.

"One of the things that drove their overuse was the widespread perception, even among many doctors, that they were somehow better pain relievers, and they never really were," said Jerry Avorn, a physician at Brigham and Women's Hospital in Boston, who studies patterns of drug use.

It is likely that Vioxx and other similar drugs did work better for some people, as individuals vary widely in their response to painkillers. The novelty of the COX-2s, however, may have produced something analogous to a placebo effect that enhanced their real effect. In clinical trials, patients who are unknowingly taking an inert placebo typically get about one-third the pain relief of those taking NSAIDs, according to a study published in the New England Journal of Medicine in 2001.

Drug companies boosted COX-2 inhibitors' popularity by promoting them heavily through direct-to-consumer marketing. That strategy accounted for $3.2 billion out of $25 billion spent on pharmaceutical promotion in 2003, and the figure has been rising steadily since 1996, according to the market research company IMS Health.

The amount that Merck spent marketing Vioxx directly to patients is not readily available, but an IMS survey estimated that Merck spent $500 million in 2003 promoting the drug to physicians through advertisements in medical journals, free samples and salesmen's visits to doctors' offices.

The payoff was evident in a survey of 2,300 physicians that IMS conducted in 2001. It found that 52 percent of doctors were willing to write prescriptions for COX-2 inhibitors if patients requested them and needed painkillers.

"When a patient comes in and wants something, there is a desire to serve them," said Wofsy, of the American College of Rheumatology. "There is a desire on the part of physicians, as there is on anyone else who provides service, to keep the customer happy."

The COX-2 drugs caught fire. Current annual sales in the United States total $5.6 billion. In 2003, Vioxx alone represented 11 percent of the revenue for Merck, the fourth-largest drug company by sales in the United States.

What is particularly ironic about this past week's events is that they came as aspirin, the hoary ancestor of both the nonsteroidal and the COX-2 drugs, continues to be reborn as something close to a miracle drug.

Synthesized in 1897 by the German chemist Felix Hoffmann as a pain-reliever for his arthritic father, aspirin since then has been found to lower the risk of heart attack in people with heart disease, and possibly to lower the risk of colon, prostate and breast cancer, Hodgkin's disease and Alzheimer's.

Although the reason Vioxx caused heart problems is not certain, researchers believe its selective action on the COX-2 enzyme, but not the COX-1, throws the prostanoids involved in clotting out of balance.

The result is a tendency for blood to clot -- and clots are the first step of heart attacks and most strokes. If that theory is correct, it means that Vioxx's cleverness was also its Achilles' heel.

The experience of "targeted" drugs -- the goal of most pharmaceutical research today -- has been mixed. While some have been huge successes -- the statins that lower cholesterol by inhibiting a single enzyme -- others have been disappointing. In particular, numerous new "biological modifier" anti-cancer drugs (such as those that try to stop the growth of blood vessels in tumors) have failed to live up to their promise. They are clever, but not as clever as the body's ability to find a way around them.

In the end, the Vioxx debacle suggests that figuring the true worth of new drugs may be a more difficult calculation than it seems.

Patients and doctors are left wondering if a drug is worth trying simply because it is new and supposedly a little bit better than what they are already using.

Companies such as Merck may wonder whether it is wise not to address directly early hints that there are problems with their golden goose products.

American society may also wonder whether expensive new drugs such as Vioxx -- part pharmaceutical, part fetish object -- are worth the money.

Avorn, who in August published a book, "Powerful Medicine: The Benefits, Risks, and Costs of Prescription Drugs," thinks the answer generally is no.

The use of Vioxx by people who did not need the modest amount of stomach protection it offered "has cost billions of dollars that could have been better used for other purposes in our health care system," he said. "There is this newer-is-better mentality, and this is why we can't afford health care."