Patients and their families may see hope in the news that the Food and Drug Administration has approved a drug to slow the pace of Alzheimer’s disease. Aducanumab is the first medicine authorized to treat what might be an underlying cause of the disease: amyloid beta clumps that accumulate in the brain. Yet the evidence from clinical trials that this drug reduces Alzheimer’s is incomplete; it may be only minimally effective, if it works at all. The FDA’s own expert panel strongly objected to the drug’s approval.

All of which makes the FDA’s announcement both perplexing and wrong. It threatens to mislead millions of Alzheimer’s patients. At the same time, because Biogen plans to price the drug at $56,000 a year for the average patient, it also stands to cost Medicare and other insurers a bundle. This underlines the need for two kinds of reform. First, the FDA needs to rethink its processes for approving drugs. Second, the rules controlling how Medicare chooses and pays for medicines need a thorough overhaul.

The FDA has failed to fully explain why it overruled the almost unanimous advice of its advisory panel on aducanumab. The agency didn’t deny that the clinical-trial evidence was poor. It simply ignored that problem and used different reasoning to grant the drug “accelerated approval.” To the general public, this term suggests urgent approval of an especially valuable treatment. In fact, it’s a more limited endorsement than ordinary approval, indicating that the drug has not proven a “real effect on how a patient survives, feels or functions,” in the FDA’s words, but nevertheless acts on an underlying disease process in a way that might predict such a benefit.

The FDA noted that patients who received aducanumab in the drug’s two truncated trials experienced a significant reduction of amyloid plaque, while those in the control arms did not. If the amyloid hypothesis is correct, the loss of plaque might cause a lessening of Alzheimer’s disease. A problem with this reasoning is that other experimental drugs that have targeted amyloid plaques have failed their clinical trials, and now many experts question whether plaque is a good predictor of Alzheimer’s disease progression. Yet the FDA settled on accelerated approval without seeking comment from outside experts or fully explaining its thinking on the plaque question.

The agency has asked the manufacturer, Biogen, to conduct another trial within eight years – and says that if no real benefit is seen, it may withdraw approval. But that won’t be easy. Now that the FDA seems to have said aducanumab works, it will be difficult to recruit subjects for a trial in which half the patients are given placebos. For this reason alone, the essential first step is for the FDA to help patients understand its announcements. Accelerated approval doesn’t mean the drug is any good.

Despite that, Biogen will sell aducanumab – under the brand name Aduhelm – at an exorbitant cost. (The nonprofit Institute for Clinical and Economic Review has said that even if the drug works as well as Biogen suggests it can, it would be cost-effective at a price of $2,600 to $8,300 a year.) An estimated 1.2 million to 1.4 million Americans with Alzheimer’s stand to be given the medicine, because they have mild cognitive impairment and plentiful amyloid deposits.

It would be absurd to spend so much on a drug that remains essentially unproven. Yet that is what Medicare is poised to do. It’s forbidden to negotiate prices, and it is expected to pay for medicines the FDA approves unless a special exception is made. Both of those strictures should be changed. Legislation championed by House Speaker Nancy Pelosi would give Medicare negotiating power it now lacks. And if, in addition, Medicare were generally under no obligation to buy drugs before full clinical trials were done, drugmakers would have a powerful incentive to conduct those trials as quickly as possible. In effect, the current arrangement allows high prices to be charged and collected until a merely promising drug is proven to be of no use.

Strange as it seems, the aducanumab story highlights systemic failures in U.S. health care. Some FDA employees worry that the agency has been reacting to public pressure by too readily endorsing questionable drugs, including the failed COVID-19 treatment hydroxychloroquine last spring. Maybe so, but the problem is more deeply entrenched than this suggests. The U.S. needs to take a good, hard look at its system for approving and pricing medicines. From now on, the FDA should make intelligible announcements. And Medicare should be freed to exercise intelligent discretion and deploy all the bargaining power at its command.