MedExamHelperHow to pass the viva
What an examiner is actually scoring
A viva is scored by an examiner who has been listening to candidates for several hours, often across several days. The examiner is tired, the marking sheet is dense, and the candidate who earns full marks does so by making the marking easier. That is the frame for everything below. You are not optimising for the perfect answer. You are optimising for the answer the examiner can mark quickly and confidently.
There are two marking shapes you will meet. Domain-by-domain rubrics score each visible component separately. Global-rating rubrics score the whole impression on a single scale. Domain rubrics reward structure; global rubrics reward fluency. Both reward the same underlying thing: the candidate knows what they are doing, can say what they are doing, and stops when they have done it.
The first 15 seconds: framing the answer
The examiner has not yet decided what mark to give you. They are listening for the shape of the answer before the content lands. The first sentence is the only sentence that buys you the next two minutes of attention.
A first sentence that works does three things. It restates the question in your own words, names the framework you will use, and commits to a working answer rather than a hedge. "I think this is acute coronary syndrome, and I will work through the history, the examination, and the immediate management" is a higher-scoring opening than "there are several possible causes, but it depends". The first sentence tells the examiner where to mark. The second leaves the examiner guessing.
A useful opening template:
- Restate: "You have asked about X."
- Commit: "The most likely cause is Y."
- Signpost: "I will cover the supporting features, the features that would change my mind, and the immediate management."
Three lines, ten seconds. That is enough.
Signposting: telling the examiner where you are
Signposting is the explicit naming of where you are in the answer. It is not filler. It is the only mechanism the examiner has for following your logic when they are also marking.
A simple signposting pattern: state the section, give the content, close the section. "Looking at supporting features: the history of central chest pain, radiation to the left arm, and associated nausea. That supports the working diagnosis. Moving to features that would change my mind: the pain is reproducible on palpation, which would push me away from ACS." Each section is named as it opens and named as it closes. The examiner can mark each section without reconstructing your thinking.
Signposting also recovers you when you lose the thread. A mid-answer pause followed by "let me come back to the immediate management, because that is the priority" is a higher-scoring recovery than silence or a topic shift. The examiner marks the recovery as composure, and composure is on the sheet.
Structure: the 30-second, 2-minute, and 5-minute answer
Not every viva answer is the same length. Match the structure to the time you have.
A 30-second answer. Used for the opening question or a follow-up clarification. Three sentences: working answer, one supporting feature, one feature that would change your mind. Do not list differentials. Do not hedge. The candidate who fills 30 seconds with caveats scores the same as the candidate who gives no answer.
A 2-minute answer. The default. Use a four-part shape: opening commitment, supporting features, features that change your mind, immediate management. Each part is signposted. The examiner can mark each part independently.
A 5-minute answer. Used for "talk me through how you would manage this patient from arrival to discharge" prompts. The four-part shape extends with a fifth section on escalation and a closing line on safety-netting. This is the only answer where you should volunteer a full differential list. Elsewhere, commit first and qualify second.
Prioritisation and safety-netting
Every viva has a right answer, a safer answer, and a too-clever answer. The safer answer is the one the examiner marks as safe. It names the working diagnosis, the immediate management, the trigger for escalation, and what the patient should do if things change at home. The examiner is not looking for the most sophisticated answer; they are looking for the answer that keeps the patient safe.
A safety-netting line that scores: "If the pain returns at rest, lasts more than fifteen minutes, or is associated with breathlessness, the patient should call an ambulance and not drive themselves." That is a safety-netting line. "They should seek medical attention if concerned" is a hedge. The first lands on the marking scheme; the second does not.
Common traps
The first trap is volunteering differentials the examiner did not ask for. The examiner asked "what is the most likely diagnosis"; the candidate answers with five differentials and no commitment. The marking scheme scores commitment first, breadth second.
The second trap is hedging. "Possibly, maybe, I think it could be" reads as uncertainty. The examiner cannot mark uncertainty on a domain rubric. A committed wrong answer scores higher than a hedged right answer, because the committed wrong answer shows the candidate can be reasoned with by the next question, and the hedged right answer does not.
The third trap is the runaway answer. The candidate fills the time, does not stop, does not pause for the examiner, and overruns the next station. The examiner marks the overrun as poor time management. Close when you have covered the shape; do not pad.
The fourth trap is the silence. The examiner asked; the candidate has not answered; the silence grows. Silence is read as uncertainty. A recovery line — "let me commit to the working diagnosis first, then qualify it" — beats silence.
How marking criteria read what you say
A marking scheme does not score what you know. It scores what you said. Two candidates with identical knowledge can score differently because one structured the answer and the other did not.
The defence is to speak in rubric language. If the rubric lists "supporting features", say "supporting features". If the rubric lists "escalation triggers", say "escalation triggers". The examiner is matching your words to their sheet, not to their memory of you. Speak the rubric's language and the marks land.
Generic feedback — "good answer, but consider X" — is useless for viva practice because it does not tell you which rubric line you missed. Practise with a tool that cites your transcript against a marking scheme, and you will see which rubric lines you are not speaking to.
Practising outside the exam room
The viva is a speaking skill. You cannot learn it by reading. You need to hear yourself answer, and you need someone or something to mark what you said.
Record yourself answering a timed prompt. Listen back. Count the signposts, the hedges, the safety-netting lines. The first recording will be uncomfortable; the fifth will be better; the tenth will be a different candidate.
Peer practice is more useful than solo practice. A colleague playing the examiner will follow up on a weak answer in a way your recording will not. The follow-up question is where the viva is won or lost.
App-based mock vivas are useful when the marking scheme is visible and the feedback cites your transcript. They are less useful when the feedback is a generic score. Insist on transcript-cited feedback; anything less is practice without feedback, which feels like work and moves you nowhere. See the med-school study system that actually sticks for the wider frame.
Where this pillar sits
This page is the hub for the viva cluster. The other pillars complete the picture: complete OSCE station library for the OSCE-side practice, the clinical skills procedure handbook for the procedure stations, and exam format guides by country for the formats that include a viva.