Overview
Introduction: Probability, Danger, and Action
A patient rarely arrives labelled with a single diagnosis.
A patient rarely arrives labelled with a single diagnosis. Chest discomfort may reflect myocardial ischaemia, pulmonary embolism, aortic catastrophe, reflux, or chest-wall pain; headache may be migraine or the first sign of subarachnoid haemorrhage. Differential diagnosis is the disciplined process of ranking those possibilities while protecting the patient from the diagnosis that would cause the greatest harm if missed. Start with a compact problem representation: who the patient is, what changed, how quickly it changed, the most discriminating positive and negative findings, and the relevant risks. Then ask three questions: 1. Which diagnosis is most likely? 2. Which dangerous diagnosis remains plausible? 3. What finding, test, or change in trajectory would move those diagnoses up or down the list? This is more reliable than ordering a broad panel and hoping that one result supplies the answer. A test is useful when its result changes probability or action. A positive or negative result cannot be interpreted independently of pretest probability, timing, assay characteristics, and the patient's stability. Canadian practice adds two safeguards. CTAS assigns acuity from the presenting concern and its...
