“Full of vagueness, groping, hedging, and ‘shot-gun’ prescriptions” is how Richard Cabot, a physician at Massachusetts General Hospital, characterized many diagnoses at the turn of the twentieth century. One hundred years ago, anxious to address this shortcoming, Cabot published Differential Diagnosis, an anthology of several hundred case studies that demonstrated how to discern the cause of an ailment.

The book stemmed from Cabot’s founding, a decade earlier, of the “clinicopathological conference” (CPC), during which a physician would take to a stage and be presented with the case history of a deceased patient. The physician would then work out the diagnosis aloud, with audience members (mostly physicians and physicians-in-training) questioning him about his methods. In the end he would offer his diagnosis, which was then compared with the results of the patient’s autopsy.

“It was a dramatic event,” says Christopher Crenner, chair of history and philosophy of medicine at the University of Kansas and a Cabot biographer, because participants risked being wrong in public view—and sometimes they were.

Cabot spread the case study approach to teaching by transcribing and publishing cases privately, then sending them to colleagues. When funding for the CPCs ran dry, The New England Journal of Medicine made them a staple—they have appeared in nearly every issue for the past 87 years as Case Records of the Massachusetts General Hospital. (Today an editorial team in the MGH pathology department vets cases for a doctor to present at his or her department’s grand rounds; a written copy of the CPC then appears in NEJM.)

As Cabot’s CPCs gained a wide audience, the method spread across the country and “became a powerful model for physicians to test their knowledge and skill as diagnosticians,” says Crenner.

Moreover, says Nancy Lee Harris, editor of Case Records and a hematopathologist at MGH, “the CPCs are a chance for young physicians to read about how experienced clinicians get to a diagnosis.” Such a skill is essential because, as Cabot wrote, “cases do not often come to us systematically arranged like the account of typhoid in a textbook of practice of medicine.” Doctors must learn to intuit a patient’s condition from a sometimes-limited set of clues.

“Someone once told me that medical school teaches you to think like a scientist,” Harris says, “but the Case Records teach you to think like a doctor.”