Just one example on the matter of drug approvals.

A case in point is Replimune’s melanoma treatment, which the FDA rejected last month. About a third of patients who hadn’t responded to prior immunotherapy showed a strong response to Replimune’s in a clinical trial.
Tumors shrank in nearly all patients, and responses proved durable over three years. Serious side effects were rare. Oncologists who treated patients in the trial hailed the results.

These are responses in absolute terms. The drug was safe, and it worked.

The FDA blocked its release into the market though:

[T]he FDA said the trial was “not considered to be an adequate and well-controlled clinical investigation that provides substantial evidence of effectiveness.”
Its quibble is that the trial lacked a control group.

This is a demand for a relative outcome—whether the drug worked better or worse, and whether it was safer or less so, than the status quo. The status quo is what a control group presents.

The answer, though, is not to stop “quibbling” about control groups when assessing drug trial efficacy. Instead, it’s necessary for the FDA to get out of the business of requiring, as a condition of approval, that a drug work. FDA’s role should hold out only for assessing a drug’s safety. The market, formed by patients and their doctors, will do a perfectly fine job of assessing the drug’s effectiveness, with no more exceptions than are extant in any other market. That Replimune’s drug was shown to work in absolute terms is a happy additional outcome and should not represent even this much of an acceptance criterion.

This is where FDA Commissioner and medical doctor Marty Makary can—and should—make the changes to the FDA’s approval processes. A doctor’s primary injunction is “first, do no harm.” So it should be with the FDA. A doctor continues, with his patient, actively to treat the medical problem. The FDA, on the other hand, should stop at the do no harm part. Let the practicing doctors and their patients do the rest.