Hook
Every Approval Is a Promise Made to Strangers
There is a particular kind of silence that follows an approval: the moment after a stranger trusts the stamp enough to swallow the pill, and before anyone knows whether that trust was earned. An approval cannot promise that nothing will go wrong. It can only promise that someone looked.
In September 1960, an application for a sleeping pill reached the Food and Drug Administration in Washington. It landed on the desk of a physician who had been with the agency for about a month, one of the first assignments she was given. By her own later account, her superiors expected an easy one (Bren, 2001).1
The drug was thalidomide. Germany had sold it over the counter since 1957, and by 1960 it sat on shelves across Europe, in Canada, and in South America. Its makers described it as safe even for pregnant women. The Cincinnati firm seeking an American license, Richardson-Merrell, expected a routine yes.
The physician was Frances Oldham Kelsey, forty-six years old, and she did not give it. She did not say never. She said, over and over, show me.
Historical Case
A Law Written After the Funerals
To see what she was standing against, it helps to know what the law of 1960 assumed. A manufacturer filed an application, and if the agency did not act within a set period, the application took effect on its own (Meadows, 2006).2 A reviewer who judged an application incomplete could send it back, but each resubmission started the sixty days again (Bren, 2001). The default, in other words, was yes, and a no had to be renewed every two months.
The law itself had come out of a funeral procession. In 1937 a Tennessee company dissolved a new antibacterial drug in a sweet, cherry-flavored solvent and sold it as a syrup children would take willingly. The solvent, diethylene glycol, is chemically related to antifreeze, and the elixir was linked to 107 deaths, many of them children. Kelsey, then a graduate student at the University of Chicago, helped run the animal tests that identified the solvent as the poison, and in 1938 Congress required manufacturers to show that a drug was safe before it was sold (Bren, 2001; Meadows, 2006).
She had watched, as a student, a law get written after the bodies. Now she was asked to apply it to a pill most of the world had already decided was harmless. In 1960 she was one of only seven full-time physicians reviewing drugs for the agency, and the company’s representative could telephone her directly (Bren, 2001).
In the FDA’s account, built largely on Kelsey’s own recollections, the pressure was ordinary, which is what made it hard to resist. The company’s representative telephoned, visited, and complained to her superiors that she was unreasonable and delaying a drug already sold abroad. She remembered the firm saying outright that it wanted the pill on the market before Christmas, when sleeping pills sell best, and she recalled having always expected industry to push (Bren, 2001).
For the firm, the deadline was a date on a sales calendar. For the people who would take the pill, whatever it did could not be undone.
Last week we followed Wilbur Tennant, a West Virginia farmer who drove a stack of his own videotapes to a lawyer because no agency would test his creek. Kelsey’s desk is the other half of that picture: an official who did insist on tests, and who paid for it in telephone calls and complaints to her superiors.
Biblical Lens
Three Verses for a Room That Has Already Said Yes
Scripture records a room like the one Kelsey walked into. In First Kings, King Ahab of Israel gathers about four hundred prophets and asks whether he should go to war at Ramoth-gilead, and they answer in one voice that he should go up and prosper. One more prophet is available, Micaiah, whom the king dislikes because he never prophesies anything pleasant. The messenger sent to fetch him offers advice on the way: the others are unanimous, so be like them.
Behold now, the words of the prophets declare good unto the king with one mouth…
1 Kings 22:13 (KJV)
Micaiah’s reply is shorter than the advice.
As the LORD liveth, what the LORD saith unto me, that will I speak.
1 Kings 22:14 (KJV)
The comparison needs its limits stated first. Micaiah claims to carry a word he has received. Kelsey claimed no vision, and when she began she did not know that thalidomide harmed unborn children, only that the dossier did not yet support what was claimed for it. Her chorus, too, was larger than Micaiah’s and was not made of flatterers: it was physicians and regulators in country after country who had judged the drug fit for sale. Their shared ground is not office or authority; it is that neither let the agreement of the room stand in for the standard the room was meant to meet.
Her no was smaller and more modest than a prophet’s. It said the case had not yet been made.
The second verse comes from a law code, and its setting is a courtroom.
Thou shalt not follow a multitude to do evil…
Exodus 23:2 (KJV)
The verse goes on to forbid tilting testimony toward the majority in a lawsuit. The logic cuts both ways: a crowd’s size is no argument for a claim, and a lone dissenter’s isolation is no argument either. What the verse asks about is what holds the claim up.
The third is a proverb that could have been written about a dossier of testimonials.
The simple believeth every word: but the prudent man looketh well to his going.
Proverbs 14:15 (KJV)
The prudent person in the proverb is not a cynic. She walks the road, and she looks at where her feet will land. Kelsey’s requests were of that kind: specific and answerable, naming what was missing rather than doubting everything.
As far as I can find, Kelsey never described her work in scriptural terms, and I will not put those words in her mouth.
The Bible nowhere teaches that every disaster is a punishment sent from above. It records something slower: a society’s repeated choices set into a character, and in time the character shows, including which evidence it declined to read and which voices it learned to wave away.
Pattern Insight
Eighteen Months Against the Clock
The arithmetic of her refusal is simple. The application arrived on September 8, 1960, and the company withdrew it on March 8, 1962 (National Women’s History Museum).3 Across those eighteen months, each request for more data restarted the sixty-day clock, so that on paper every one of her refusals was a request.
The requests were not vague. By the FDA’s account, she and her colleagues found the long-term toxicity studies too short, the data on absorption and excretion thin, the manufacturing controls weak, and the clinical reports closer to testimonials than to designed studies (Bren, 2001). A testimonial records that a patient slept well. A designed study asks what else the drug did to her.
These were questions a better dossier could have answered. A delay that names what is missing, and what would satisfy it, is a different thing from a delay that merely waits.
In December 1960 the British Medical Journal published a physician’s report of nerve damage in patients who took the drug for long periods, and when Kelsey came across it she wrote to the company at once. Her thinking reached back to wartime rabbits: in the 1942 search for malaria drugs she had learned that pregnant rabbits broke down quinine poorly and their embryos not at all, and she wondered whether a drug that could harm nerves might also harm a developing child. The nerve reports gave her a reason to ask. In November 1961 the German pediatrician Widukind Lenz linked the epidemic of malformed newborns to thalidomide, and the American application was withdrawn four months later (Bren, 2001).
Meanwhile, the drug moved anyway. During the investigational period Richardson-Merrell distributed more than 2.5 million tablets to over 1,000 American doctors, and nearly 20,000 patients received them, several hundred of them pregnant women (Bren, 2001; Geraghty, 2001).4 Not every doctor had kept records, and many women were never told what they had been given. The FDA formally acknowledged seventeen American children with thalidomide-associated deformities, and in September 2026 NPR reported that a retired FDA historian considers the true number certainly higher, though no exact count exists (Weingartner, 2026).5
Worldwide, estimates pass ten thousand children in forty-six countries (Bren, 2001). Set beside the countries that approved the drug, the American count looks like a victory. For the families inside the smaller count, there is no beside.
The figures make the story sound cleaner than it was. Her refusal held the front door while a side door stood open, because the law of 1960 let companies hand an unapproved drug to doctors as long as it was labeled experimental (National Women’s History Museum). The design of 1960 had a default of yes, a side door for experimental shipments, and one stubborn person standing between them and the public. That is a system that hoped for stubbornness.
Micaiah, in the text, is thrown into prison on bread and water, and the king marches to war regardless (1 Kings 22:27–35). The Bible allows the lone voice to be right and the room to proceed anyway. What Kelsey’s case adds is the rarer outcome, in which a refusal bought just enough time for the evidence to arrive.
None of this makes agreement the enemy. Agreement built on evidence is simply inquiry doing its work; the danger lies where agreement takes the place of the checking. Wide use abroad was not proof for an American application, testimonials were not studies, and her judgment mattered not because she knew the answer but because she would not treat the unproven as proven. Unanimity is a mood, and evidence is a document, and the dissenter’s task is to keep asking which of the two is on the table.
Closing
What One Reviewer Could Not Carry Alone
In the previous installment, we followed Barbara Henry, a teacher who stayed at The Second Desk beside a six-year-old when others left the building. It is a quirk of the calendar that both stories begin in the autumn of 1960. Henry’s faithfulness was presence and Kelsey’s was withholding, and between them they sketch what a decent institution quietly depends on: someone who will not leave, and someone who will not sign.
Seen from that autumn, the lesson looks less like a tribute than a diagnosis. After the Washington Post put Kelsey on its front page that July, Congress took up a stalled bill, and on October 10, 1962, President Kennedy signed the Kefauver-Harris Amendments (Bren, 2001). Another signature, in other words, settled what hers had held in place: the new law required manufacturers to prove that a drug worked as well as that it was safe, brought experimental use under rules for informed consent and the reporting of adverse reactions, and ended approval by default (Meadows, 2006). From then on, the agency’s silence could no longer count as consent.
The correction that lasted was not a call for more Kelseys. It was an attempt to build a system in which a careful reviewer would not be the only thing standing between a default and a patient.
There is a small detail in her biography that I draw no conclusions from. In 1935 she accepted an offer written to a man, signing, at her professor’s suggestion, with her sex noted in brackets, and in 2001 she still could not say whether the offer would have come to a woman (Bren, 2001). I mention it only because it shows how early she understood what a signature can carry.
The silence from the beginning, the one after the stamp, is still being measured. In September 2026, NPR reported that American-born survivors are asking Congress to acknowledge that the country did not escape as lightly as the official count suggests, and that many of their mothers were never told what they had been given (Weingartner, 2026). A refusal can shrink a silence without ever closing it.
Thalidomide itself never disappeared. The agency approved it in 1998 for a severe complication of leprosy, under controls it had never used before: a patient registry, extensive education about the risks, and monthly pregnancy tests for women who could conceive (Bren, 2001). Kelsey, at eighty-six, kept urging caution because, she said, “people forget very soon.” A not yet is the most expensive sentence in a system built to say yes.
The registry is how an institution remembers what one person once refused to forget. Every approval is followed by a silence, and someone is always listening to it. In the autumn of 1960, fewer Americans had to listen than might have, because a new reviewer, called unreasonable by those who wanted a yes, kept saying not yet.
- Linda Bren, “Frances Oldham Kelsey: FDA Medical Reviewer Leaves Her Mark on History,” FDA Consumer 35, no. 2 (March–April 2001): 24–29.
- Michelle Meadows, “Promoting Safe and Effective Drugs for 100 Years,” FDA Consumer, January–February 2006 (Centennial Edition).
- National Women’s History Museum, “Frances Kathleen Oldham Kelsey,” womenshistory.org.
- Karen Geraghty, “Commemorative Issue: Protecting the Public: Dr. Frances Oldham Kelsey,” Virtual Mentor (AMA Journal of Ethics) 3, no. 11 (2001): 418–420.
- Tana Weingartner, “The U.S. has more thalidomide survivors than the government admits,” NPR/WVXU, September 26, 2026.
- Scripture quotations are from the King James Version (public domain).
New essays every week. Delivered quietly.