The Food and Drug Administration found itself backed into a corner the other week.

After Merck pulled its arthritis pain reliever, Vioxx, from the market two-and-a-half years ago, the FDA endured relentless and even vitriolic attacks from academia, Congress, and the press for paying too little attention to drug safety.

In the meantime, well beneath the public radar screen, the New England Journal of Medicine and other elite advocates implored the FDA to raise the bar for new drugs — a dangerous move.

Instead of just requiring new drugs to be safe and effective, these advocates want new drugs to be either safer or more effective than existing drugs. The idea is to force the pharmaceutical research industry to be more innovative and to provide patients with new drugs that have even fewer risks. But, instead, the opposite would happen if this became a new requirement that had to be enforced by the FDA.

The beleaguered FDA and one of its advisory committees faced a new drug application from Merck for Arcoxia, which works like Vioxx. Both are Cox-2 inhibitors, which in turn are part of the larger class of nonsteroidal anti-inflammatory drugs, or NSAIDs.

NSAIDs are probably the most-prescribed drugs in the world. They include such familiar over the counter medicines as ibuprofen, often sold as Advil, and naproxen, or Alleve, and a raft of prescription versions including less well-known drugs such as diclofenac. The only Cox-2 drug on the market in America is Celebrex, but Arcoxia is available in more than 60 nations including virtually all of Europe.

The Cox-2 inhibitors selectively suppress the Cox-2 enzyme, which causes pain. Traditional NSAIDs suppress the Cox-2 enzyme, too, but they also suppress the Cox-1 enzyme, which protects the upper gastrointestinal system from lesions and ulcers. Bleeding ulcers caused by NSAIDs create great distress in millions of patients and often lead to death, perhaps several thousand per year. The Cox-2 inhibitors are easier on the stomach and esophagus although the relief is far from complete.

It turns out the Cox-2 inhibitors also cause extra heart attacks when compared to a sugar pill or placebo. That’s why Vioxx was pulled from the market. But arthritis patients in pain usually take NSAIDs, not sugar pills.

The post-Vioxx drug safety investigations generated two remarkable discoveries. First, all the Cox-2 inhibitors tend to cause heart attacks, although not very often. Vioxx is a little worse than the others in that respect, but it is better in preventing ulcers. Second — this is the real news although the press has pretty much missed it — all the NSAIDs, traditional and Cox-2s alike, increase the risk of heart attacks. The only NSAID that is noticeably safer than the others is naproxen.

This put the FDA and its advisory committee in a bind. In clinical trials comparing it to traditional NSAIDs, Arcoxia was consistently better in terms of intestinal damage, especially compared to naproxen. It was comparable in terms of heart safety except, of course, relative to naproxen.

Having been approved by nearly all other advanced nations, Arcoxia approval here should have been a slam dunk. But if the FDA committee voted to approve Arcoxia, it would invite criticism for tolerating a drug known to cause heart attacks. So the FDA offered a way out: simply raise the standard so that only superior drugs are approved. That is essentially what the committee did by a vote of 20 to 1 when it recommended against Arcoxia approval because comparable drugs were already on the market.

What neither the FDA nor its committee seems to have considered is what its superiority standard will do to patients and health care costs. Clear superiority for the average patient can be demonstrated only through mammoth clinical trials at great expense. But some drugs are only superior for subgroups of patients. The only way to demonstrate that kind of superiority is to run even larger trials; Arcoxia’s huge 34,000-patient trial was insufficient. Under the old approval standard, patients can find out which works best; under the new standard, they will have to wait years just to get a choice.

Even worse is what happens when two or more pharmaceutical firms are in a race to bring new breakthrough drugs to market. No one knows who will be first, but everyone knows the second-place finisher either will have to start all over with a much bigger trial to meet a new standard, or simply give up. Either way, research and development costs go up. And competition goes down. The first approved drug would have years of monopoly leverage while its competitors compile superiority data or again, simply give up.

The FDA’s recent move was a giant step toward a calamitous new drug approval policy.

Mr. Calfee is a resident scholar at the American Enterprise Institute, which receives a small portion of its funding from pharmaceutical firms.