What examiners penalise in viva answers is usually not obscure knowledge gaps. It is unsafe, unstructured thinking spoken out loud. If your response jumps to a diagnosis, ignores immediate priorities, or never lands on a clear plan, you can sound less competent than you are.
What examiners penalise in viva answers most
Across official spoken assessments, the scoring themes are remarkably similar. The Simulated Consultation Assessment (SCA) marks data gathering and diagnosis, clinical management and medical complexity, and relating to others; MRCP(UK) Part 2 Clinical Examination (PACES23) and the MRCOG Part 3 Clinical Assessment also assess communication, information gathering, clinical judgement or applied clinical knowledge, and patient welfare or safety. Taken together, those official descriptors point to the same pattern: examiners reward answers that are safe, clear, and patient-centred, not just factually correct ones.
Official descriptors also show what weaker answers look like. PACES marksheets flag unclear or unsystematic explanations, jargon, failure to involve the patient, incomplete or incorrect management, poor listening and empathy, and behaviour that jeopardises safety. RCGP guidance likewise marks down outdated or dangerous plans, vague safety netting, unrealistic management, and language the patient cannot follow.
The practical takeaway is simple: examiners often penalise the shape of your reasoning as much as the content.
Unsafe answers are punished first
If a stem contains immediate risk, your first job is to sound safe. A 67-year-old with crushing chest pain, an agitated septic child, or a postpartum woman with headache and hypertension should trigger priorities before detail.
A safe opening is short and deliberate:
- identify the immediate threat
- state what you will do now
- give the likely problem and the dangerous alternatives you must not miss
- outline the next investigations and escalation
- finish with review, disposition, or safety netting
Do not open with a long differential dump. Open with a plan. For chest pain, a better first line is: My immediate concern is acute coronary syndrome, so I would assess ABCDE, monitor the patient, get an ECG promptly, treat pain, involve senior help early, and keep other life-threatening causes such as aortic dissection and pulmonary embolism in mind while I refine the history and examination.
That answer reassures the examiner in the first 15 seconds. It sounds like someone they would trust on call.
Unstructured answers leak marks
Official marking language rewards structure. RCGP's global skills material emphasises progressing through the consultation and leaving time for management, while PACES descriptors explicitly reward explanations that are clear and structured. Rambling is not neutral. It makes you harder to mark well.
When you are unsure, use a five-part spine:
- priority and urgency
- top diagnosis or problem list
- focused information you still need
- management now
- review, escalation, and communication
This can sound very natural. In a surgical-style stem about right upper quadrant pain, you might say: This is biliary sepsis until proven otherwise. I would first assess for sepsis and peritonism, check observations, take a focused history for duration, vomiting, jaundice and comorbidity, send bloods including inflammatory markers and liver tests, give analgesia and fluids, start antibiotics if infection seems likely, and discuss early with the surgical team if the patient is systemically unwell. I would explain the working diagnosis to the patient and make the next decision point clear.
Notice what that does. It does not pretend certainty. It shows organised, safe uncertainty.
Ignoring the patient costs marks, even in clinical stations
Many candidates lose marks because they answer the clinical problem but not the human problem. In official guidance, RCGP expects a person-centred approach and checking understanding, GMC Good medical practice 2024 emphasises partnership, respect, clear communication and keeping patients safe, and MRCOG Part 3 even uses lay examiners on some tasks to assess communication, patient safety and information gathering from the patient's perspective.
So if a GP-style consultation about headache turns into a polished migraine monologue, but you never ask why the patient is worried, you have probably missed the station. If the real concern is that her sister died from a brain tumour, your technically correct answer can still feel tone-deaf.
A simple rescue framework is:
- acknowledge the concern
- explore what the patient thinks, fears, or needs
- explain the likely problem in plain English
- negotiate the plan
- check understanding and agreement
The same principle applies when the role-player is a colleague. A referral or handover answer should still be structured, respectful, and explicit about risk. Examiners are listening for whether you can adapt your language to the room, not deliver the same speech to every station.
Vague management exposes weak clinical judgement
Candidates often know the diagnosis but still bleed marks because the plan is foggy. Saying I would do some bloods, maybe start treatment, and discuss with seniors is too thin. It tells the examiner that you have not prioritised, not weighed risk, and not decided what matters now.
This matters because official guidance repeatedly ties passing performance to safe, up-to-date, realistic management. The RCGP clinical management toolkit says poor performance includes plans that are outdated, incorrect, dangerous, unrealistic, or vague on follow-up and safety netting; it also notes that failure commonly stems from not managing conditions according to current guidance and evidence.
A stronger management answer usually covers:
- what I think is happening
- what I need to rule out urgently
- what I will do in the next minutes or today
- what I will tell the patient or team
- what would make me escalate, review, or change course
For example, in a paediatric oral station about a wheezy 3-year-old, do not stop at likely viral-induced wheeze. Say how sick the child is, what observations matter, whether you would try inhaled bronchodilator treatment now, what would push you towards admission, and what safety-net advice the parent must leave with.
Good judgement sounds specific. Even when you are uncertain, your next step should not be.
Common mistakes
- opening with a diagnosis before stating urgency or immediate risk
- giving every differential the same weight
- answering the case you hoped for rather than the question asked
- forgetting drugs, allergies, pregnancy status, comorbidity, or functional impact
- using jargon with a patient or actor
- failing to say what you would actually do next
- offering no disposition, follow-up, or safety net
- becoming flustered when challenged instead of reprioritising calmly
- talking for too long before making a decision
- missing the patient agenda, emotion, or reason for consulting
Practice workflow
Do not just collect more question banks. Rehearse the exact behaviours examiners mark.
A useful weekly routine is:
- two or three short mock vivas with a peer, trainer, or AI tool
- one stem rehearsed first as a 60-second opening, then as a full answer
- immediate debrief in four buckets: safety, structure, communication, judgement
- one immediate re-run of the same stem after feedback
If you are sitting a named exam, map your errors to the official domains. For the Simulated Consultation Assessment (SCA), that means data gathering and diagnosis, clinical management and medical complexity, and relating to others. For MRCP(UK) Part 2 Clinical Examination (PACES23) and MRCOG Part 3, it is useful to tag errors under information gathering, communication, clinical judgement or applied clinical knowledge, and patient safety or welfare.
After each mock, ask three blunt questions:
- Did I sound safe in the first 20 seconds?
- Could the examiner hear my structure?
- Did I answer this patient's or colleague's real concern?
That is the level of debrief that changes marks.
Summary
- Safety beats polish. If the case is high risk, sound safe before you sound clever.
- Structure is visible. Signposting and prioritising make your knowledge easier to reward.
- Patient-centred language is not decorative. It is part of the mark.
- Uncertainty is acceptable if your reasoning, escalation, and safety netting are clear.
- The best practice is short, repeated, and debriefed against examiner domains.
References
- https://www.rcgp.org.uk/mrcgp-exams/simulated-consultation-assessment/marking-and-results
- https://www.rcgp.org.uk/mrcgp-exams/simulated-consultation-assessment/toolkit-clinical-management
- https://www.rcgp.org.uk/mrcgp-exams/simulated-consultation-assessment/toolkit-data-gathering
- https://www.rcgp.org.uk/mrcgp-exams/simulated-consultation-assessment/toolkit-global-skills
- https://www.mrcpuk.org/sites/default/files/documents/PACES23%20Consultation%20scenario%20writing%20guidance.pdf
- https://www.mrcpuk.org/sites/default/files/documents/PACES23%20-%20Communication%201.pdf
- https://www.rcog.org.uk/careers-and-training/exams/mrcog-our-specialty-training-exam/mrcog-part-3/mrcog-part-3-format/
- https://www.gmc-uk.org/professional-standards/good-medical-practice-2024