“This is a very confusing situation,” a top Food and Drug Administration scientist, Sandra Kweder, said of the controversy over Celebrex and other popular pain relievers. If the FDA is confused, imagine the feelings of doctors and patients caught in the backlash.

The story so far: Celebrex, along with Vioxx and Bextra, were a new class of drugs that held great promise to relieve suffering from arthritis without causing the digestive discomfort and bleeding associated with older medications. They were heavily marketed, and sales reached billions of dollars.

Three months ago, a study confirmed fears that Vioxx increased the risk of heart attack and stroke. Now Vioxx is off the market, pulled by its manufacturer, and last Thursday the FDA urged doctors to limit prescriptions for Celebrex and Bextra.

The Vioxx debacle exposed a much larger danger: The FDA’s monitoring of new drugs is inept. Even after Vioxx was implicated in up to 40,000 deaths, the agency seemed to react only after one of its scientists, David Graham, turned whistleblower and told Congress that the FDA’s procedures guarantee that unsafe drugs remain on the market.

As we’ve noted previously, that problem needs to be addressed urgently. Drugs that show Vioxx-like dangers should be removed, or their use should be sharply restricted. But another danger is emerging: the risk that overreaction will make helpful drugs unavailable to seriously ill patients for whom significant risk is acceptable.

The consumer group Public Citizen, for instance, is urging that 181 drugs be taken off the market. Doing so, no doubt, would save some lives. It also would relegate many others to pain, disability and premature death.

Nearly half of Americans take at least one prescription drug, and rarely are the safety issues clear-cut.

Consider Celebrex. A study released earlier this month found that patients who received several times the typical dose had an increased risk of heart attack and stroke. Previous studies showed the opposite. Another new study found that Celebrex caused fewer heart problems than naproxen (sold over the counter as Aleve).

If Celebrex were pulled from the market, millions who’ve found relief from painful, disabling arthritis would have to switch to other medicines such as ibuprofen or naproxen. But those drugs cause severe gastrointestinal problems in some people, resulting in more than 16,000 deaths a year. Patients and doctors have to decide whether the potential cardiac risk outweighs the drug’s benefit.

Certainty about drugs is impossible because clinical trials before FDA approval involve only a few thousand patients. New risks, or benefits, can emerge once the drug is prescribed to millions. Vioxx, for example, was showing promise for preventing precancerous colon polyps.

Overreaction can cause harm by:

Delaying new drug approvals. Some patient-advocacy groups worry that the FDA is slowing approval of new drugs for cancer and other deadly diseases. Marqibo, a pill to treat an aggressive cancer that kills at least half of all patients with it, was turned down for accelerated approval earlier this month. That caution may be appropriate, but there is another side: “The number of people who die because of delayed approval far exceeds the number hurt by unsafe drugs,” says Steven Walker of Abigail Alliance, a group that wants quicker access to new drugs.

Distorting the level of risk. Clinical studies often contradict each other, and wide-ranging conclusions are drawn from small samples. In the recently reported link between naproxen and cardiac risk, fewer than 3% of the 2,500 elderly patients suffered heart attacks or strokes over three years, a relatively small number for that age group. “We’re being given flash results based on single trials that haven’t been properly analyzed,” says Michael Weber, associate dean of New York’s Downstate College of Medicine. “There may or may not be a safety problem. At this point, we don’t know.”

Jeopardizing health. In October, the FDA ordered antidepressants to carry tough labels warning of increased suicidal behavior in children. Banning the drugs would have threatened the health of a million kids getting relief from the agony of depression. The suicide rate in 15 nations has dropped by one-third during the past 15 years since these drugs were introduced. The FDA was right to urge greater monitoring instead.

Make no mistake: The FDA is badly off-track. It’s not watching closely enough to see that the drugs it approves are safe as well as effective. But most often, the right solution will be to restrict a drug’s use, not to ban it.