Synopsis
Elicit a safe, person-centred clinical history, transform it into a discriminating problem representation, and communicate uncertainty and priorities across written and clinical examinations.
- Begin with the patient's agenda and an open account, then clarify chronology, severity, functional effect and urgent symptoms without losing the person's own meaning.
- Build hypotheses while listening, but test them with neutral questions and keep at least one credible alternative until discordant data have been explained.
- A problem representation is a one-sentence synthesis of who the patient is, the tempo and syndrome, relevant context and the discriminating positives and negatives.
Reasoning priorities
Capture the patient's account before testing diagnostic hypotheses.
Preserve the person's words for key symptoms, then clarify timing, quality, severity and associations with neutral questions; leading questions can manufacture confirmation.
Worked reasoning
A 72-year-old reports three days of worsening breathlessness; inputs include orthopnoea, ankle swelling, missed diuretic tablets, no pleuritic pain, and reduced walking distance.
- Organise the inputs by tempo and mechanism: acute-on-chronic congestion is supported by orthopnoea, oedema, functional decline and interrupted treatment, while absent pleuritic pain weakens but does not eliminate embolic disease.
- Add relevant context and severity rather than every detail: older age, established treatment suggesting prior fluid disease, and inability to walk the usual distance materially alter risk and urgency.
- State the representation: a 72-year-old with three days of progressive congestive breathlessness and functional decline after missing diuretic therapy, with orthopnoea and oedema but no pleuritic pain.
- Use the representation to prioritise physiological assessment and focused cardiac-respiratory examination while retaining infection, arrhythmia and pulmonary embolism as alternatives requiring targeted evidence.
- Verify by reading the sentence back against the raw history: every included feature changes probability or urgency, no unsupported diagnosis is asserted, and the patient confirms the timeline and medicine interruption.