The MRCGP Simulated Consultation Assessment is a performance exam, not a multiple-choice paper, so the honest starting point is a boundary: a question bank can measure the clinical-knowledge substrate your consultations draw on, but it cannot measure the consultation itself. This checklist tells you the minimum evidence to verify — across knowledge coverage and consultation skill — before you decide you have "covered" the SCA and can stop doing new questions. Where a Q-bank is the wrong tool, it says so plainly.
The direct answer, as a checklist
You have covered the SCA when you can tick all of the following, not when you have completed any bank:
- Your underlying clinical-knowledge coverage is complete against the RCGP curriculum, with no under-sampled or stale areas.
- You have deliberately practised the consultation mechanics — voice-only communication, eliciting the patient's agenda, shared decision-making, safety-netting and time management — under realistic conditions.
- You have rehearsed against all three examiner-judged domains, not just diagnosis.
- You have interpretation fluency for images, ECGs, radiographs and laboratory trends where a case might turn on them.
- Your recency is current on guidance-sensitive topics, with the date and jurisdiction of each source recorded.
- You have performance evidence from realistic, timed, unseen practice — including feedback from a person, because no Q-bank can score a consultation.
If any line is unticked, you are not done, regardless of your percentage on knowledge questions.
Current exam snapshot
The SCA consists of 12 simulated remote consultations of 12 minutes each (144 minutes of contact), sat by ST3s over roughly nine diets a year, with trained role-players and cases spanning the breadth of general practice. Each consultation is marked across three domains: Data Gathering and Diagnosis; Clinical Management and Medical Complexity; and Relating to Others. The authoritative reference is the RCGP SCA candidate guidance and its published case examples; treat any third-party case list as a claim to check against that guidance. The fee is substantial (around £1,207; verify the current figure with the RCGP), which is one more reason not to sit before the checklist is green.
Where a Q-bank fits — and where it does not
Be explicit about the tool boundary. A question bank, including iatroX, can help you verify and measure the knowledge a good consultation rests on: the differential you would gather towards, the current first-line management, the red flags you must safety-net, the medicines facts (from the SmPC/eMC) behind a prescribing decision. It cannot assess whether you actually elicited the agenda, built rapport by voice, negotiated a shared plan, or finished in twelve minutes. Those are consultation behaviours that require simulated consultations with role-players, a study group or a trainer, and human feedback. iatroX's own SCA resources are, as of 19 July 2026, still in development and not a consultation simulator; use it for the knowledge substrate and unseen measurement, and use dedicated consultation practice for the performance.
Use the following boundary as your working rule before you rely on any percentage:
| Consultation ingredient | Can a question bank measure it? | Where the real evidence comes from |
|---|---|---|
| The differential you should gather towards | Yes | Blueprint-mapped knowledge items |
| Current first-line management and monitoring | Yes | Knowledge items plus primary sources (NICE, CKS, SIGN, SmPC/eMC) |
| Red flags and safety-netting content | Yes | Knowledge items |
| Eliciting the patient's agenda (ideas, concerns, expectations) | No | Simulated consultations with role-players |
| Building rapport by voice alone | No | Human-graded consultation practice |
| Negotiating a shared, acceptable plan | No | Human-graded consultation practice |
| Finishing within twelve minutes | No | Timed simulated consultations |
| Responding to emotion and uncertainty | No | Human or peer feedback |
The left column of that table is measurable, and it is where iatroX and any knowledge bank earn their place. The right column is the exam's centre of gravity, and it needs people, not questions. Any revision plan that only feeds the measurable half will produce a confident, well-informed candidate who still under-performs on the day.
Blueprint coverage table
Build this table from your own data and keep it live. It audits the knowledge substrate, which is the part a bank can measure:
| Curriculum area | Official emphasis | Questions attempted | First-attempt accuracy | Last reviewed | Confidence (1–5) |
|---|---|---|---|---|---|
| Cardiovascular / respiratory | High | ||||
| Mental health | High | ||||
| Musculoskeletal / pain | Medium | ||||
| Women's / sexual health | Medium | ||||
| Children and young people | Medium | ||||
| Older adults / frailty / polypharmacy | Medium | ||||
| Dermatology | Medium | ||||
| Ear, nose, throat / eyes | Lower | ||||
| Endocrine / metabolic | Medium | ||||
| Neurology | Medium |
A cell that is under-attempted, stale or low-confidence is a knowledge gap to close before it shows up as a fumbled consultation.
Ten domain-level blind spots self-selected practice tends to hide
These are the areas candidates skip when they choose their own cases; each should be checked, and exam-specific clinician review is advisable before you decide you are ready:
- Multimorbidity and polypharmacy in the frail older adult, where the "management" is negotiation, not a single guideline.
- Mental-health presentations with risk assessment and safety-netting.
- Undifferentiated symptoms (fatigue, dizziness, weight loss) with no neat diagnosis.
- Paediatric presentations, including the worried parent and safeguarding cues.
- Sexual and reproductive health, including sensitive-topic communication.
- End-of-life care, ceilings of treatment and difficult conversations.
- Health inequalities, interpreters and remote-consultation limitations.
- Medically unexplained symptoms and shared uncertainty.
- Professional and ethical dilemmas — capacity, confidentiality, third-party concerns.
- Chronic-disease reviews where the task is behaviour change, not diagnosis.
Format checklist: the consultation mechanics
Verify deliberate practice — not just awareness — of each:
- Voice-and-remote consultation: you can gather and safety-net without visual cues, managing silence and pace.
- Patient agenda: you routinely elicit ideas, concerns and expectations, not just symptoms.
- Shared decision-making: you offer options and negotiate rather than instruct.
- Safety-netting: you close every consultation with specific, time-bound advice.
- Domain scoring: you have practised against all three domains and know what each rewards.
- Timing: you reliably complete a focused consultation within twelve minutes.
None of these can be ticked from MCQ practice alone; they need simulated consultations with feedback.
Interpretation checklist
Where a case could hinge on data, confirm fluency: ECGs, chest and other radiographs, skin images, blood-gas and laboratory trends, common calculations, and the ethical/statistical literacy occasionally embedded in a management discussion. A knowledge bank is a reasonable place to rehearse the interpretation itself; the communication of that interpretation to a simulated patient is not.
Recency checklist
List your guidance-sensitive topics — anticoagulation, diabetes and cardiovascular targets, contraception, antimicrobial choices, mental-health pathways — and for each record the current recommendation, the source (NICE, CKS, SIGN, SmPC/eMC) and its date and jurisdiction. Superseded advice delivered confidently in a consultation is a management-domain error. Verify recency against the primary source, using a citation-first tool if helpful, and re-check before the exam.
Performance checklist
Confirm you have, from realistic practice: unseen, timed, mixed knowledge blocks with stable first-attempt accuracy; adequate speed; a falling rate of high-confidence errors; retention of previously corrected material; and — critically — human-graded consultation feedback calibrated against RCGP official material. The knowledge half you can measure with a bank; the consultation half you cannot. Read your Q-bank percentage is not your exam score before you let a number reassure you.
Calibration against official material
The RCGP's own SCA candidate guidance and published case examples are the calibration gold-standard, and nothing a bank or a study group produces should override them. Use them in two ways. First, as a rubric check: after every practice consultation, mark yourself against the three published domains — Data Gathering and Diagnosis, Clinical Management and Medical Complexity, and Relating to Others — rather than against a vague sense of a "good consultation", so your self-assessment speaks the examiner's language. Second, as a currency and register check: the official case examples show the breadth, difficulty and communication register the College expects, and comparing your own practice cases against them exposes whether your self-selected scenarios have drifted too narrow or too comfortable. A candidate who has never calibrated against official material is guessing at the standard, however many questions they have answered. Schedule at least one deliberate calibration pass against the official examples before you decide you are ready, and treat any gap between your self-scores and the official descriptors as a coverage finding to close, not a confidence problem to talk yourself out of.
Three ways candidates misjudge SCA readiness
Mistaking knowledge for consultation skill. A high knowledge percentage feels like readiness, but the SCA scores how you consult, not what you know. Candidates with strong knowledge and untested consultation mechanics are among the most likely to be surprised on the day, because the number they trusted measured the wrong thing.
Practising only the comfortable cases. Left to choose, candidates rehearse the presentations they enjoy and quietly avoid the frail multimorbid patient, the mental-health risk assessment and the difficult conversation — exactly the domains examiners probe. Self-selected practice manufactures false coverage that no completion figure will reveal.
Treating remote consultation as an afterthought. The SCA is sat remotely, so the loss of visual cues, the management of silence, and the discipline of explicit verbal safety-netting are skills in their own right. Candidates who only ever practised face-to-face consulting sometimes find that the format itself, not the medicine, is what trips them.
Each of these is a reason the checklist, not a percentage, decides when you stop: the number can be high while the unticked lines are precisely the ones that carry the marks.
Stop/continue decision tree
- Continue new questions only where the blueprint table shows an under-sampled or weak knowledge area.
- Consolidate (stop new questions, rework errors) where volume is adequate but retention is shaky.
- Simulate — shift effort to full, timed, role-player consultations with feedback — once the knowledge substrate is solid but consultation mechanics are untested.
- Seek teaching where human feedback shows a consultation-behaviour gap a bank cannot fix.
- Rest where blocks are stable, mechanics are rehearsed and errors are careless; more volume then adds fatigue, not readiness.
The trigger is always the measured gap, never a target percentage or the urge to keep clicking.
One-page checklist and a worked example
Copy this and keep it visible:
- Blueprint knowledge coverage complete (no stale/under-sampled cells)
- All three domains deliberately rehearsed
- Voice/remote mechanics, agenda, shared decisions, safety-netting practised
- Interpretation fluency confirmed
- Recency logged with dates and jurisdiction
- Unseen timed knowledge blocks stable
- Human-graded consultation feedback obtained and calibrated
Worked example (invented data). A candidate has completed 900 knowledge questions at a 68% overall average and feels "nearly ready". The blueprint table shows older-adults/frailty attempted only lightly at 52% first-attempt accuracy, and mental-health cases avoided. The performance checklist shows zero human-graded consultations. The decision tree is unambiguous: continue new questions in frailty and mental health to close the knowledge gaps, then switch decisively to simulated consultations with feedback. The 68% average was never the readiness signal; the two empty checklist lines were.
FAQ
How do I know whether I have covered the full MRCGP SCA blueprint? Build the blueprint coverage table above, filling in attempted volume, first-attempt accuracy, last-reviewed date and confidence for every curriculum area, and separately confirm you have rehearsed all three examiner domains and the consultation mechanics. You have covered the blueprint when there are no under-sampled or stale knowledge cells and no untested consultation behaviours — not when a bank reads complete, because bank completion measures the substrate only.
Can one question bank be enough for MRCGP SCA? No, and not because any single bank is weak: the SCA is a consultation performance exam, and no question bank — including iatroX — can assess whether you elicited the agenda, built rapport by voice, negotiated a shared plan or safety-netted in twelve minutes. A bank can make the underlying knowledge solid and measurable; the consultation itself needs simulated practice with role-players and human feedback.
What should I measure instead of my overall Q-bank percentage for MRCGP SCA? For knowledge, measure first-attempt accuracy on unseen, timed, mixed blocks by curriculum area, plus retention and high-confidence error rate. For the exam itself, measure human-graded performance against each of the three domains on realistic, timed consultations. An overall percentage on seen knowledge items tells you little about a domain-scored consultation.
When should I stop doing new MRCGP SCA questions? Stop adding new knowledge questions when your blueprint table has no weak or under-sampled cells and your unseen knowledge blocks are stable; then redirect that time to consultation simulation and feedback. Continuing to do new MCQs after the knowledge substrate is solid is displacement activity — the marginal gain has moved to the consultation, which questions cannot rehearse.
Which MRCGP SCA resource should I use for my weakest component? Match the tool to the component. For a knowledge gap, use a blueprint-mapped bank (such as iatroX) and the primary sources. For a consultation-skill gap — agenda, shared decisions, safety-netting, timing — use simulated consultations with role-players and a trainer or study group, because that is the only setting that can observe and score the behaviour. Do not try to close a communication gap with a question bank, or a knowledge gap with unstructured role-play.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026. Vendor-reported figures (including iatroX's own product status) were taken from public pages on 19 July 2026 and change without notice — verify current figures on the relevant product page. Disclosure: iatroX operates a competing MRCGP knowledge bank (its AKT-level bank) and would compete for revision spend; this checklist confines iatroX to the knowledge substrate and unseen measurement, states plainly that iatroX is not an SCA consultation simulator and that its dedicated SCA resources are in development, and directs consultation practice to human-graded simulation. Corrections: use the feedback route on iatrox.com and we will amend any error or out-of-date figure. (The UK medicines reference throughout is the SmPC/eMC.)
References