The FDA wants less research. Doctors want more.
On February 18, 2026, the FDA announced they would lower the bar from two to one well-designed trials before approving a drug. The shift began in 1997, when Congress allowed single-trial approvals at the agency’s discretion. At first, they rarely used this option, but single-trial approvals rose in the 2010s through new, accelerated pathways. The new policy makes single-trial approvals the default option.
FDA commissioner Marty Makary and Vinay Prasad announced the decision in the New England Journal of Medicine, and now that journal has published criticisms of the policy.
The Problems
Adding bad meds, missing good ones. The lower standard will allow ineffective medications to slip through, but will also cause some potentially useful ones to be denied. While Makary and Prasad argue that the change will lead to more treatment options, a study of 21 recent approvals found 13 of them were approved when one of their two trials failed. Without the second trial, those approvals may not have made it.
Where’s the data? Makary and Prasad didn’t cite any data, relying on theory as they argued for less empiricism.
Learning from history. With the recent spike in single-trial approvals, more drugs have to be withdrawn as evidence emerges that they are unsafe or do not work. 18 drugs were withdrawn in 2021-2022 alone, compared to a rate of one withdrawal every two years over the previous three decades. In the five years before 2021, single-trial approvals nearly tripled, from 20 to 59 per year. Trial sizes shrank, from a median of 106 patients to just 59, which means less safety data. Follow-up trials meant to confirm a drug actually works dropped from 75% to 42%.
The case for replication. When something matters, we double check. Requiring two trials protects against random error, bias, and hidden confounding factors that can fool even careful researchers. Nearly a third of drugs approved from 2001 to 2010 were later tied to serious safety problems, and a second trial does a better job at catching these.
When Single Trials are Appropriate
Single-trial approval still makes sense for rare diseases or drugs with an overwhelming effect
Stand-ins for Health
The FDA says they’ll include other sources of evidence besides the single trial, but they don’t have a good history of upholding that, and the other sources are problematic.
Many medications look good in the lab but fail to make a difference in people. These biological markers are a starting place, meant to inspire clinical research, but in the last 10 years the FDA has allowed approvals based on these stand-ins alone. Drugs that fail to sharpen cognition are approved for dementia because they lower amyloid plaques. Drugs that improve symptoms or lifespan in cancer are approved based on imaging scans.
SOURCE: Letters in response to a NEJM article, by Carl Peck (a former FDA director under Bush Sr); Isabella Xu, Irene Ulrich, and Peter Lurie (Center for Science in the Public Interest), and Mahavir Singh and Yashodhara Singh (University of Louisville, KY).
—Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report
