The 1937 Elixir Sulfanilamide deaths

On October 11, 1937, physicians in Tulsa, Oklahoma, reported to the American Medical Association that an unfamiliar liquid form of sulfanilamide, a new antibacterial drug, was linked to a number of deaths. The association's laboratory asked the manufacturer for the formula, isolated the toxic ingredient and broadcast warnings through newspapers and radio. The ingredient was diethylene glycol, which FDA's account of the episode¹ describes as "a chemical normally used as an antifreeze."

By the time the search was over, the product, sold as Elixir Sulfanilamide, had killed 107 people, many of them children, according to FDA's timeline of food and drug law³. The deaths were spread across 15 states, from Virginia to California.

The episode is usually told as the prologue to the 1938 Food, Drug, and Cosmetic Act, which for the first time required drugs to be shown safe before sale. This case history stays with the poisonings: how a sweet raspberry syrup reached hundreds of pharmacies, how a field force of 239 people tracked it down, and why the only charge available was that the word "elixir" on the label was false. The same solvent has turned up in medicines many times since, which is why it still appears in FDA guidance today.

A new formula, tested for taste

Sulfanilamide had been used safely in tablets and powders for some time. In June 1937, a salesman for a Tennessee drug firm reported demand in the southern states for a liquid version. The company's chief chemist found the drug would dissolve in diethylene glycol, and its control laboratory "tested the mixture for flavor, appearance, and fragrance and found it satisfactory," the 1981 FDA Consumer article² recounts. The firm made a batch and sent out 633 shipments. The first went out in early September.

No one had tested the formula for toxicity, and no law required it. "Selling toxic drugs was, undoubtedly, bad for business and could damage a firm's reputation, but it was not illegal," the FDA history says. Published studies had already reported that diethylene glycol could damage the kidneys, it adds, and "a few simple tests on experimental animals would have demonstrated the lethal properties of the elixir."

Patients, many of them children being treated for sore throats, were ill for about seven to 21 days. FDA's account¹ describes the signs of kidney failure, including stoppage of urine, severe abdominal pain, vomiting and convulsions. There was no known antidote at the time.

Six weeks in the fall of 1937

When FDA inspectors² reached the firm's headquarters and branch offices, they found it had already sent telegrams to more than 1,000 salesmen, druggists and doctors. The messages asked for the product back but did not say it was lethal. At FDA's insistence a second wave went out: "Imperative you take up immediately all elixir sulfanilamide dispensed. Product may be dangerous to life. Return all stocks, our expense," according to the FDA history¹.

How 239 inspectors found the bottles

Practically the agency's entire field force was assigned to the search, joined by state and local health officials. Inspectors started with the firm's shipping records and moved down the chain to distributors, wholesalers and retail drugstores. In one establishment alone, they checked 20,000 sales slips. They tracked down the firm's 200 salesmen to learn where physician samples had gone.

Records were often thin. In East St. Louis, Illinois, 49 prescriptions had been filled, and some carried only notes such as "Betty Jane, 9 months old," according to the FDA Consumer account². Some druggists had sold the elixir without a prescription to buyers whose names they did not know. In Georgia, a druggist said only six ounces of a returned gallon had been dispensed, but 12 ounces were missing when it reached the maker, and local inquiries turned up two more buyers. Both had died. In South Carolina, a doctor said none of his five patients had died, and an inspector later found that four had.

For all that, the search largely worked. Of 240 gallons made and distributed, 234 gallons and 1 pint were recovered, FDA says, and "the remainder was consumed and caused the deaths of the victims."

A charge about a word

FDA's authority to pull the product rested on a technicality. The Pure Food and Drugs Act of 1906⁴ policed adulteration and misbranding, and its penalty for shipping a misbranded drug across state lines was a fine of up to $200 for a first offense. The law said nothing about safety testing. As FDA's history of the 1906 act⁵ notes, the Supreme Court had also ruled that the act did not reach false therapeutic claims.

So the 25 seizures charged misbranding. An "elixir," FDA argued, implied an alcoholic solution, and this one contained no alcohol. "If the product had been called a 'solution' instead of an 'elixir,' no charge of violating the law could have been made," the agency's history says. The firm's owner said that "there was no error in the manufacture of the product."

FDA Commissioner Walter Campbell used the case to press Congress. "It is unfortunate that under the terms of our present inadequate Federal law, the Food and Drug Administration is obliged to proceed against this product on a technical and trivial charge of misbranding," he said, in remarks quoted in the agency's history¹. The only remedy, he argued, was a law that "will require that all medicines placed upon the market shall be safe to use under the directions for use."

What the 1938 law required

A bill to replace the 1906 act had been stalled in Congress for five years, FDA's history of the 1938 act⁶ says. The elixir deaths "reshaped the drug provisions of the new law" and "propelled the bill itself through Congress." Section 505 of the act as enacted⁷ barred interstate shipment of any new drug unless an application was in effect. The application had to include six things.

An application took effect on the 60th day after filing unless FDA postponed it, for no more than 180 days in total, to study it. That was a notification system with a waiting period. Proof of effectiveness came later, with the Kefauver-Harris Amendments of 1962. Even so, the change was large. The new law brought more than 6,000 new drug applications in its first nine years and 13,000 by 1962, FDA's history of the 1938 act's aftermath⁸ says. Within two months of passage, FDA began naming drugs such as the sulfas that could not be labeled for safe use by patients on their own. That debate ended in the Durham-Humphrey Amendment of 1951, which, in the words of FDA's timeline³, restricted to prescription the drugs that "cannot be safely used without medical supervision," and created the modern line between prescription-only and over-the-counter drugs.

The solvent that kept coming back

Diethylene glycol did not stay in 1937. From November 1995 through June 1996, acute kidney failure was diagnosed in 86 children in Haiti after they took acetaminophen syrup made with contaminated glycerin, CDC reported in 1996⁹. In 2022 and 2023, contaminated cough syrups turned up in at least seven countries and were linked to more than 300 deaths in three of them, most of them children under 5, the World Health Organization said in January 2023¹⁰. FDA cites that count in its 2023 guidance on testing high-risk drug components¹¹. That guidance opens its background section with the 1937 elixir and its 107 deaths. The later outbreaks, the excipient supply chain they ran through, and the testing rules they produced, are the subject of our history of diethylene glycol contamination in medicines.