Clinical assessment, diagnosis, and differential diagnosis
Domain 2 - Clinical assessment, diagnosis, and differential diagnosis
This domain examines hypothesis-driven history and examination, deliberate differential construction, and the probabilistic interpretation of tests. The focus is on decisions that change management.
Core rules
- Start with pretest probability derived from history, age, sex, and risk factors.
- Use test characteristics correctly: a sensitive test is useful when negative only if pretest probability is low; a specific test is useful when positive.
- Apply likelihood ratios: a positive LR above 10 or a negative LR below 0.1 substantially changes probability.
- Build a differential that includes common, dangerous, and treatable conditions, ranked by probability and by harm if missed.
- Reassess when the clinical course does not fit the working diagnosis.
Worked example
A 55-year-old smoker has pleuritic chest pain and dyspnoea after a long flight. The differential includes pulmonary embolism, acute coronary syndrome, pneumonia, pneumothorax, and musculoskeletal pain. Wells score is 4.5, so PE is likely. D-dimer is not the right first test in a likely patient because a negative result does not exclude PE at this pretest probability; CT pulmonary angiography is indicated. If the patient were low-risk (Wells below 2), a negative high-sensitivity D-dimer could exclude PE.
Traps
- Anchoring on the first diagnosis and abandoning the differential.
- Ordering tests that will not alter management.
- Ignoring base rates or interpreting a test without pretest probability.
- Confusing "ruled out" with "not yet investigated".