Evidence-Based Practice and Quality Improvement
Appraising Evidence
Candidates should understand the hierarchy of evidence, from systematic reviews and randomized trials to observational studies and expert opinion, and that the question type determines the best design (therapy: randomized trial; diagnosis: cross-sectional accuracy study; prognosis: cohort; harm: cohort or case-control). Know PICO framing, internal and external validity, bias types (selection, measurement, confounding, publication), and the meaning of confidence intervals and statistical versus clinical significance.
Applying and Improving
Understand shared decision-making that integrates evidence, clinician judgment, and patient values. Know quality improvement frameworks such as PDSA cycles, root cause analysis, and the distinction between quality assurance and quality improvement. Recognize common measures (structure, process, outcome) and the danger of gaming metrics.
Traps
Frequent traps: assuming a statistically significant result is clinically important, confusing association with causation, and generalizing trial results to populations excluded from the trial. Another trap is choosing a case series to answer a therapy question, or treating a surrogate endpoint as a patient-important outcome.
Worked Example
Rule of thumb: match the design to the question, then ask about effect size and confidence interval width. A trial showing a small statistically significant mortality difference with a wide confidence interval crossing clinically meaningful harm should not be presented to a patient as clear benefit.