Evidence-Based

Evidence-based assessment starts with a clinical question, not a test. First decide whether the patient is stable and identify the diagnosis that must not be missed. Then estimate pre-test probability, select a test or clinical prediction rule that fits the question and the patient's characteristics, interpret the result in that context, and decide what happens next. Reassessment closes the loop because a working diagnosis remains provisional until the patient's trajectory supports it. Research evidence supplies test characteristics, prediction rules, and treatment recommendations, but it does not replace examination, clinical expertise, or the patient's preferences and circumstances. A familiar pattern can speed recognition; atypical findings, conflicting data, or a high-consequence alternative should trigger slower, deliberate reasoning. A negative result is reassuring only when the test and the pre-test probability make it reassuring, and a score is useful only when the patient belongs to the population in which it was validated.