Most candidates who fail the viva do not fail because they know too little. They fail because the examiner cannot hear a safe, structured clinical mind at work. If you do not think out loud, if your priorities appear late, or if you spend the station proving knowledge instead of answering the task, you can underperform badly despite knowing the medicine. Current official marking frameworks in spoken assessments reward demonstrated structure, judgement, safety and communication, not private insight.
Across specialties, examiners are not just asking, 'Do you know the diagnosis?' They are asking, 'Can you recognise the task, gather the right information, prioritise risk, explain yourself clearly, and leave the patient safe?' MRCP(UK) Part 2 Clinical Examination (PACES23), the MRCGP Simulated Consultation Assessment (SCA), and MRCOG Part 3 Clinical Assessment all separate these elements into markable domains.
Why this gap matters
In a written paper, you get credit for the final answer. In a viva or structured oral station, you get credit for the route you take, the order you take it in, and whether your reasoning sounds safe enough for real practice.
That is why a candidate with slightly less factual recall can outperform a more knowledgeable one. If their answer is clearer, more structured and more obviously patient-centred, the examiner has something to reward. That is not unfair. It is the point of the format.
A useful way to think about it is this: the markable unit is not the fact. It is the performance of clinical judgement.
Why you fail the viva even when you know the medicine
You answer the topic, not the task
A classic error is hearing a familiar stem and launching into a textbook. The examiner asks about chest pain, headache, postoperative confusion or reduced fetal movements, and you give a mini-lecture instead of a targeted clinical response.
Official descriptors repeatedly reward a structured approach to information gathering, problem-solving and management. So before you answer, decide what kind of task you have actually been given: diagnosis, immediate management, explanation, prioritisation, ethics, or escalation.
A simple opening frame helps:
- 'This is mainly a prioritisation and immediate management problem.'
- 'My immediate concerns are the time-critical and life-threatening causes.'
- 'The key information I need next is...'
- 'My working diagnosis is... but I also need to exclude...'
- 'I would now investigate and treat in parallel by...'
For example, with a 67-year-old with central chest pain, a weak answer starts with a long list of causes. A better answer starts with priorities: assess stability, consider acute coronary syndrome first, exclude the immediately dangerous alternatives, and explain what you would do in the next few minutes. That sounds like a doctor making decisions, not a candidate reciting revision notes.
You keep your reasoning private
In spoken exams, silent reasoning does not score. PACES23 guidance makes clear that it is your responsibility to demonstrate the required skills within the station, and the consultation guidance defines high-value performance as clear, structured communication plus sensible judgement.
This is where 'think out loud' matters. Not rambling. Not narrating every thought. Just making the important parts audible.
Try this sequence when you are stuck:
- working diagnosis
- dangerous alternatives
- feature that makes you think that
- immediate next step
- escalation or safety-net if you are wrong
So instead of saying, 'I would do some bloods and imaging,' say, 'My leading concern is biliary sepsis. I would assess for sepsis now, send blood cultures and liver tests, start antibiotics if indicated, and ask for senior surgical review early if the patient is unstable or obstructed.' That answer is easier to trust.
You sound unsafe when you rush
Candidates often know the right final diagnosis but omit the risky part of the answer: what must not be missed, what cannot wait, and when to call for help. Official standards repeatedly reward red-flag recognition, effective prioritisation, safe and defensible management, patient welfare and communication that reduces risk.
Good viva language sounds like a registrar. It contains uncertainty, but not vagueness.
Use phrases like:
- 'My first priority is patient safety.'
- 'The red flags I need to exclude are...'
- 'If the patient is unstable, I would escalate immediately and start...'
- 'If initial findings do not fit my working diagnosis, I would broaden to...'
- 'I would explain the uncertainty and give clear follow-up or safety-netting.'
That last line matters. In many exams, a decent differential without a safe plan still fails to reassure the examiner.
Pacing is part of the mark
Time management is not separate from knowledge. In a viva, pacing is part of clinical judgement. The RCGP SCA feedback states that a common reason for running out of time is spending too long on data gathering and then rushing management, explanations and follow-up. Its toolkit describes completing data gathering by 6-7 minutes as good performance in a 12-minute consultation. PACES23 guidance also reminds candidates that they are responsible for demonstrating the required skills within the allotted time.
You are not being rewarded for suspense.
A practical time map for many viva answers looks like this:
- first 15-20 seconds: define the task and immediate priorities
- next 30-45 seconds: give your leading diagnosis or differential with ranking
- next 30-45 seconds: investigations and treatment
- final 15-20 seconds: escalation, complications, follow-up, safety-netting, or patient explanation
In a consultation-style station, the same principle applies. Do not spend most of the case gathering data and then bolt on a rushed management plan in the final minute. Examiners notice when the safe bit arrives too late.
Communication that feels professional, not scripted
Sometimes the medicine is broadly right, but the delivery sounds wooden. PACES23 rewards clear, structured, jargon-free explanation that addresses the patient's concerns. The SCA standards reward person-centred communication, and the relating-to-others toolkit warns specifically against false, formulaic empathy.
So do not paste stock phrases over the case. Examiners can hear it.
Better communication usually has five moves:
- start with what the patient already understands
- name the main concern in plain English
- explain your working diagnosis or plan clearly
- check whether the plan is realistic for this person
- confirm understanding and what happens next
Compare these two openings.
Weak: 'I am sorry to hear that. We will do some tests. It could be many things.'
Stronger: 'You are worried this headache could be something serious. From what you have told me, I need to check for red flags first. If those are absent, the most likely cause is a primary headache disorder, and I would talk you through treatment and when to seek urgent help.'
That sounds calmer, safer and more useful. It also helps the examiner follow your structure.
Common mistakes
These are the failure patterns that make a knowledgeable candidate look weak. They mirror official feedback themes around insufficiently targeted data gathering, rushed management, failure to address concerns, poor structure and unsafe follow-up.
- opening with a long preamble instead of immediate priorities
- giving an unranked differential that never lands on a working diagnosis
- keeping your reasoning in your head and hoping the examiner will infer it
- sounding definitive when the case is actually uncertain
- forgetting to say what you would do if the patient deteriorates
- using jargon when the task is explanation or counselling
- performing formulaic empathy that does not fit the patient in front of you
- spending so long gathering information that management becomes rushed
- ending without a summary, plan or safety-net
Practice workflow
Do fewer cases, better. One properly debriefed mock viva is worth far more than an evening of passive reading.
Official SCA feedback encourages candidates to watch recordings, role play scenarios with supervisors or peers, and review feedback with a supervisor. The principle transfers well to other spoken assessments: perform, review, repeat.
A workable weekly loop looks like this:
- choose three short stems from different specialties
- spend 30 seconds planning each answer on paper
- answer aloud for 90 seconds without stopping
- repeat the same stem with examiner interruption after 30 seconds
- debrief on structure, safety, clarity and pace before you discuss factual gaps
When you debrief, ask three blunt questions:
- What did I make explicit?
- What stayed hidden in my head?
- At what point did the answer stop sounding safe and organised?
If you use AI for rehearsal, use it for interruption drills, follow-up questions and timing. Do not outsource your clinical standards to it. Check uncertain clinical content against current guidelines, local teaching and your trainers.
Summary
- Knowing the medicine is necessary, but in a viva it is not enough.
- Examiners reward visible reasoning: structure, prioritisation, safety and communication.
- Start by identifying the task, not by downloading everything you know about the topic.
- Think out loud in a disciplined way so the examiner can mark your judgement.
- Pace the station so management, explanation and safety-netting are not squeezed into the final seconds.
References
- https://www.mrcpuk.org/sites/default/files/documents/Candidate%20guide%20notes.pdf
- https://www.mrcpuk.org/sites/default/files/documents/PACES23%20Consultation%20scenario%20writing%20guidance.pdf
- https://www.rcgp.org.uk/mrcgp-exams/simulated-consultation-assessment/marking-and-results
- https://www.rcgp.org.uk/mrcgp-exams/simulated-consultation-assessment/feedback-statements
- https://www.rcgp.org.uk/mrcgp-exams/simulated-consultation-assessment/toolkit-global-skills
- https://www.rcgp.org.uk/mrcgp-exams/simulated-consultation-assessment/Toolkit-relating-others
- https://www.rcog.org.uk/careers-and-training/exams/mrcog-our-specialty-training-exam/mrcog-part-3/mrcog-part-3-format/