At the root of the conflict is the fundamental principle that the FDA uses to evaluate drugs. Decisions are based almost entirely on what is known as a randomized controlled trial, in which patients are randomly assigned to receive a treatment or not. Other types of studies have, again and again, failed to deliver accurate information about medicines’ benefits and risks, and are used sparingly in making medical decisions. Three randomized studies have now shown no benefit for hydroxychloroquine in hospitalized patients.

The study that sparked the latest controversy was anything but randomized. Not only was it not randomized, outside experts noted, but patients who received hydroxychloroquine were also more likely to get steroids, which appear to help very sick patients with Covid-19. That is likely to have influenced the central finding of the Henry Ford study: that death rates were 50% lower among patients in hospitals treated with hydroxychloroquine…

Nissen and Borio say observational studies simply cannot be used to determine whether a medicine is effective. Again and again they have been wrong. In one famous example, estrogen replacement therapy after menopause was thought to have benefits in preventing heart and other problems; large studies showed this was not the case. In another, a knee surgery for arthritis was shown to have no benefits over medical care.