Evidence Based Medicine

Evidence-based medicine (EBM) is the discipline of using the best available research evidence—together with clinical expertise and the patient's values—to make decisions about the care of individual patients.

11 frameworks, from 1948 to today · Clinical Medicine · Medicine & Health

still a working view ended · each leads to its entry below

Overview

Evidence-based medicine (EBM) is the discipline of using the best available research evidence—together with clinical expertise and the patient's values—to make decisions about the care of individual patients1. What is at stake is the reliability of medical action under uncertainty: whether a treatment that looks plausible actually helps, whether a test that reassures actually rules out disease, and whether a guideline built for a population fits the person in the consulting room. EBM grew out of clinical epidemiology, the application of population-based measurement to the patients a clinician actually sees, and it turned that science into a set of habits for reading, judging, and applying research12.

Clinical Epidemiology and the Numerical Turn

Medicine's older reasoning was anatomical and mechanistic: a treatment was justified if it corrected an inferred bodily disturbance. The clinical epidemiology that emerged in the mid-twentieth century asked different questions—how often does disease occur, what raises or lowers risk, how well does a test detect it—and answered them with counts. The British physician Austin Bradford Hill's 1948 randomized trial of streptomycin in pulmonary tuberculosis, run with the Medical Research Council, is the landmark: patients were allocated by a random process so that the two groups were comparable, and the comparison could be attributed to the drug rather than to the patients chosen. The same decades produced the case-control study, which compares people with a disease to similar people without it, and the cohort study, which follows exposed and unexposed groups forward. Bradford Hill's 1965 list of considerations for judging whether an observed association is causal—strength, consistency, temporality, plausibility, and others—gave clinicians a vocabulary for separating real effects from confounding.

The numerical method supplied the arithmetic. Measures such as relative risk, absolute risk reduction, and the number needed to treat (how many patients must be treated for one to benefit) made trial results interpretable at the bedside. The Canadian physician David Sackett, working in the 1970s and 1980s, was central to formalizing clinical epidemiology as a discipline for clinicians rather than statisticians, and his textbook Clinical Epidemiology: A Basic Science for Clinical Medicine (with Brian Haynes, Peter Tugwell, and others, first edition 1985) taught a generation to read the medical literature as a source of evidence about their own patients.

The Randomized Controlled Trial Paradigm

The randomized controlled trial (RCT) became the paradigm case of trustworthy evidence because randomization addresses the central problem of clinical comparison: sicker or healthier patients may be selected into treatment by unmeasured factors. Random allocation, concealed from those enrolling patients, makes the groups exchangeable in expectation. The paradigm matured through the second half of the twentieth century—placebo controls, blinding, intention-to-treat analysis (analyzing patients in the group to which they were assigned, whatever they actually received), and pre-specified outcomes. Large trials such as the ISIS-2 trial of streptomycin and aspirin in heart attack (1988) showed that modest but real effects on death could be detected only with thousands of patients, and that a simple, cheap intervention could be shown to work. The RCT's authority also carried costs: it answers narrow questions about average effects, and its results may not transfer to the complex, multimorbid patients seen in ordinary practice.

Systematic Reviews, Meta-analysis, and the Evidence Hierarchy

Single trials are noisy and sometimes contradictory. The systematic review—a pre-planned, exhaustive search for all relevant studies on a question, with explicit criteria for inclusion and a structured assessment of their quality—was developed to reduce this noise. Meta-analysis, the statistical pooling of results across studies, was applied to medicine from the 1970s and 1980s; the British statisticians Archie Cochrane and Iain Chalmers were decisive advocates, and Chalmers's work on the care of pregnant women led to the Cochrane Collaboration, founded in 1993, which produces and maintains systematic reviews. The Oxford Centre for Evidence-Based Medicine, established in 1995, popularized the evidence hierarchy: a ranking of study designs in which systematic reviews of randomized trials sit at the top, then individual trials, then cohort and case-control studies, then case series and expert opinion. The hierarchy is a heuristic, not a law—a large, well-conducted cohort study can outrank a small, flawed trial—but it gave clinicians a quick way to weigh what they were reading.

Critical Appraisal and the Birth of EBM

The term "evidence-based medicine" was introduced in 1991 by Gordon Guyatt and colleagues at McMaster University, in a series of articles in the Journal of the American Medical Association on reading the medical literature. The movement's manifesto was Sackett and colleagues' 1996 editorial "Evidence based medicine: what it is and what it isn't," which defined it as the conscientious, explicit, and judicious use of current best evidence in decisions about individual patients, integrating evidence with clinical expertise and patient values. Critical appraisal—the structured assessment of a study's validity, results, and applicability—was the core skill3. The PICO formula (Patient, Intervention, Comparison, Outcome) turned a clinical uncertainty into an answerable question, and the "evidence search" became a routine step: formulate the question, find the best evidence, appraise it, apply it, and evaluate the outcome.

Pre-appraised Evidence Services

Not every clinician could or would appraise primary studies. Pre-appraised services condensed the literature into ready-made judgments: the ACP Journal Club (from 1991), Evidence-Based Medicine and Evidence-Based Nursing (from the mid-1990s), and later resources such as Clinical Evidence and UpToDate. The Cochrane Library's systematic reviews and the BMJ's Clinical Evidence offered summaries graded for quality. These services shifted the clinician's task from reading trials to finding and applying trustworthy syntheses—a division of labour that raised its own questions about who appraises the appraisers and how quickly summaries go out of date.

Guidelines, GRADE, and the Evidence-to-Recommendation Problem

Clinical practice guidelines—systematically developed statements to assist decisions about specific circumstances—proliferated from the 1980s, driven by agencies such as the US Agency for Health Care Policy and Research (established 1989) and professional bodies. Early guidelines were often expert consensus in disguise, and the 1990s brought a backlash over conflicting recommendations and undisclosed conflicts of interest. The response was a method for moving from evidence to recommendation. The GRADE approach (Grading of Recommendations Assessment, Development and Evaluation), developed from 2000 by an international group, rates the certainty of evidence for each outcome—high, moderate, low, or very low—and separates that judgment from the strength of the recommendation, which also weighs benefits, harms, patient values, and resource use14. GRADE is now used by many guideline bodies, including the World Health Organization, and it made explicit what earlier guidelines had hidden: that recommendations involve value judgments, not just data.

Shared Decision-Making and Patient Values

EBM's definition placed patient values alongside evidence and expertise, but the early movement was often criticized for neglecting them. Shared decision-making—a process in which clinician and patient exchange information about options, benefits, and harms and reach a decision together—was articulated in the 1980s and 1990s and gained institutional footing with the 1997 founding of the Informed Medical Decisions Foundation in the United States. Decision aids, structured tools presenting the probabilities of outcomes in absolute terms, were developed and tested; trials showed they improve patients' knowledge and reduce decisional conflict, though effects on costly outcomes are less certain. The patient-values strand of EBM thus became a field of its own, and it exposed a tension at EBM's heart: the same evidence can support different decisions for different people.

Knowledge Translation and the Implementation Gap

Knowing the evidence and doing it are different things. Studies repeatedly found that effective treatments were underused and ineffective ones persisted—the "know-do gap." Knowledge translation (also called implementation science) emerged in the late 1990s and 2000s to study how to move evidence into practice: audit and feedback, reminders, educational outreach, and computerized decision support. The Canadian Institutes of Health Research adopted knowledge translation as a strategic priority in 2000, and the field developed its own frameworks and outcome measures. Its central finding is sobering: passive dissemination of guidelines changes little, while active, targeted, locally adapted strategies change more—but no single strategy works everywhere.

The Present

EBM today is the default grammar of clinical evidence. Systematic reviews and meta-analyses underpin guidelines; GRADE structures how recommendations are made; shared decision-making and decision aids are embedded in policy; and knowledge translation supplies the methods for closing the gap between evidence and practice14. The field's open questions are the ones its history created: how to judge evidence from observational and real-world data when trials are impractical; how to handle the rapid growth of preprints and low-quality publications; how to make guidelines trustworthy when industry and advocacy fund them; and how to reconcile population-average evidence with the individual patient in front of the clinician. The numerical method, the randomized trial, the systematic review, and the explicit grading of certainty remain the tools, and the argument over how to use them well is the discipline's continuing work.

References

  1. Evidence-based medicine (introduction, Medical education, Clinical decision-making, Assessing the quality of evidence), Wikipedia
  2. Clinical epidemiology, Wikipedia
  3. Critical appraisal, Wikipedia
  4. GRADE approach (Background and history, introduction), Wikipedia

The frameworks of Evidence Based Medicine

Each framework is a way of working the field has used: when it held, and the question it answers.

  1. 1948-Present

    Clinical Epidemiology

    How often does disease occur and what raises its risk?

    Fly to its tower in Nousfarer

  2. 1948-Present

    Randomized controlled trial

    Does a treatment actually cause better outcomes?

    Fly to its tower in Nousfarer

  3. 1965-Present

    Causal Inference from Observational Data

    When does an observed association reflect a real cause?

    Fly to its tower in Nousfarer

  4. 1970-Present
  5. 1990-Present

    Clinical Practice Guidelines

    What should clinicians do in a given situation?

    Fly to its tower in Nousfarer

  6. 1991-Present
  7. 1991-Present

    Pre-appraised Evidence Services

    Where can busy clinicians find trustworthy answers?

    Fly to its tower in Nousfarer

  8. 1995-Present

    Hierarchy of evidence

    Which kinds of study should we trust most?

    Fly to its tower in Nousfarer

  9. 1997-Present

    Shared decision-making in medicine

    How should clinician and patient decide together?

    Fly to its tower in Nousfarer

  10. 2000-Present

    GRADE approach

    How certain is the evidence and how strong a recommendation?

    Fly to its tower in Nousfarer

  11. 2000-Present

How the ideas connect

Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.

Clinical EpidemiologyRandomized controlled trialCausal Inference from Observational D…Systematic Review and Meta-analysisClinical Practice GuidelinesCritical Appraisal and the Five-Step …Pre-appraised Evidence ServicesHierarchy of evidenceShared decision-making in medicineGRADE approachKnowledge Translation and Implementat…

Beyond the field

Where Evidence Based Medicine's ideas came from in other fields, and what they led to there: the strongest 12 of 73 links.

A city in Nousfarer: a red-carpeted stair climbing terraces to a crowned tower on an acropolis, more towers round it, forest and snowy mountains beyond.
A city in Nousfarer, the game played in the browser.

Evidence Based Medicine is a planet in Nousfarer

Its city has a tower for each of these 11 frameworks. Land, walk in, and climb one concept at a time.

Land on Evidence Based Medicine

Related fields

Fields its ideas are linked to

More in Clinical Medicine

See where Evidence Based Medicine sits on the Map of Human Knowledge