Evidence-based medicine: a new approach to teaching the practice of medicine

In the early 1990s, most clinical teaching still organized itself around the authority of the senior clinician. Attending physicians taught by example and distilled experience; pathophysiologic reasoning determined what treatment should work; and the notion that a busy practitioner should read and appraise primary literature was largely aspirational. At McMaster University in Hamilton, Ontario, Gordon Guyatt had been building an alternative into his residency program, one that made published research a main tool of clinical reasoning, not an occasional supplement.

Guyatt coined the phrase evidence-based medicine in a March 1991 ACP Journal Club editorial, but the idea reached a wider audience through the 1992 JAMA paper written by the Evidence-Based Medicine Working Group, a collaborative based at McMaster. The paper's argument was direct: unsystematic clinical experience and pathophysiologic reasoning were insufficient guides to patient care; physicians needed to locate, appraise, and apply published research in a systematic and explicit way. It described a framework, not just a slogan.

A year later, in November 1993, JAMA began the Users' Guides to the Medical Literature, a companion series of 25 articles that ran until 2000 with the help of JAMA editor Drummond Rennie. Those guides were practical and specific: they taught physicians how to read a randomized trial, how to interpret confidence intervals, when to trust subgroup analyses, and how to use likelihood ratios at the bedside. The series addressed clinical questions by specialty and study design, giving the framework enough operational detail to actually change how residents read journals.

Medical schools began teaching EBM within a few years, and structured summaries such as ACP Journal Club, produced by McMaster's Health Information Research Unit with the American College of Physicians, gave clinicians appraised research each month. The concept also changed how guidelines were built: instead of relying on expert consensus alone, panels began grading recommendations by the quality of the supporting evidence.

Criticism of EBM has been persistent and occasionally sharp. Detractors argued that the framework devalued clinical judgment, that it worked poorly for patients who did not resemble trial populations, and that industry could manipulate the evidence base to favor commercially favorable conclusions. Later proponents, including Guyatt himself, refined the original framework to explicitly incorporate patient values and clinical context alongside research evidence, acknowledging that applying population-level data to individual patients required interpretation, not just retrieval.