Gastroenterology is the specialty that learned to take the unlikely answer seriously.

In 1984, The Lancet published a short paper from Perth, in Western Australia. Biopsies had been taken from the stomachs of 100 consecutive patients coming for gastroscopy, and curved bacteria were found in 58 of them. The bacteria were present in almost every patient with active chronic gastritis, a duodenal ulcer or a gastric ulcer. Ulcers were blamed on stress and acid at the time, and the idea that a germ caused them was met with disbelief. Twenty-one years later it won the Nobel Prize.

In 2013, the New England Journal of Medicine published a trial of a treatment most doctors would once have refused to try. Patients whose Clostridium difficile infection kept coming back were given a short course of vancomycin, then an infusion of donor stool through a tube into the duodenum; others were given vancomycin alone, or vancomycin with bowel lavage. The trial was stopped early. The first infusion cured 13 of 16 patients, against 4 of 13 on vancomycin alone and 3 of 13 with lavage. The strangest treatment in the trial was the one that worked.

Gastroenterologists had long removed polyps on the belief that it prevented cancer. In 1993 the National Polyp Study put numbers to the belief: among 1,418 patients who had adenomas removed and were followed for an average of 5.9 years, colorectal cancer occurred 76 to 90% less often than expected. In 2012 a follow-up of 2,602 of the study’s patients who had adenomas removed, traced for up to 23 years, found deaths from colorectal cancer 53% below the number expected. A procedure done on belief had become one done on evidence.

These are three of the thousands of studies that shaped modern gastroenterology. They changed the field’s mind about what causes an ulcer, what cures an infection and what prevents a cancer, and each time the answer was in a paper years before it was in a textbook.

It is still happening. Recently, in June 2025, the New England Journal of Medicine published a trial in fatty liver disease that has turned to inflammation and scarring. The trial gave 1,197 adults a weekly injection of semaglutide or placebo. It reported on the first 800 at 72 weeks. The liver inflammation resolved in 62.9% on semaglutide, against 34.3% on placebo. Scarring improved in 36.8%, against 22.4%. Patients lost 10.5% of their body weight, against 2.0%. The cost was stomach upset, more common on the drug. A liver disease that had no drug now had one. Practice follows the literature as it goes. It is a good culture. It has exactly one cost.