Reasoning GP · Start here

The SCA, explained properly

Everything a candidate actually needs to know about the Simulated Consultation Assessment — the format, the marking arithmetic most guides get wrong, why candidates really fail, and a 12-week plan built directly on the training tools in this site. Written the way an examiner would brief a friend.

12 × 12
cases × minutes each
3 min
records window between cases
3 × 0–3
domains graded per case
/ 126
total (CM weighted ×1.5)
~9 : 3
video : audio case mix

01The format — a relay, not a sprint

You sit the SCA remotely, from a booked private room at a GP surgery (usually your own), on your own computer via the RCGP's online exam platform, with a remote invigilator watching throughout. No exam centre, no travel — and no corridor between stations to collect yourself in.

Twelve consultations arrive back-to-back. Before each one you get 3 minutes with the patient's records — demographics, past history, medications, recent entries, and the reason for contact. Then the role-player (a trained actor) connects and the 12 minutes start. When the buzzer goes, it ends — finished or not — and the next records appear.

3 minrecords 12 min — consultationopening → gathering → sharing → management → safety-net → close 3 minrecords 12 min — consultation…× 12, no pause button

What the 3 minutes are actually for

  • Mine, don't skim. The records contain deliberate signal: the medication started four weeks ago, the DNA'd reviews, the nurse's note, the carer flag. Candidates who restate what the notes already say waste consultation minutes and signal template-thinking.
  • Predict the station. Ask: what is this case probably testing? Where's the trap? What must I have done by minute 7? Our case briefs train exactly this — every Hot Seat case opens with an exam-format record extract.
  • Most cases are the patient — but expect some to be a carer, a relative, another professional, or a results/data discussion. Physical examination is never assessed; findings are given if you ask for the right ones.
The skill nobody practises: recovering between stations. A collapsed case must be left in its own 12 minutes — examiners mark each case independently, and the diet is lost more often by one bad station bleeding into three than by the bad station itself. The Mock Exam Circuit exists to train this.

02How it's actually marked

Every case is graded by an examiner in three domains, each Clear Fail (0) · Fail (1) · Pass (2) · Clear Pass (3). Clinical Management is weighted ×1.5 in the arithmetic — so each case carries 10.5 marks and the diet totals 126. There is no fixed pass mark: it's set fresh each sitting by borderline-regression standard-setting, recently landing around the high 70s. And crucially, you don't pass or fail individual cases — only the cumulative total counts.

Data Gathering & Diagnosis

Hypothesis-driven questioning that visibly responds to this patient — not a recited systems template. Records used, red flags screened, and a working diagnosis committed to out loud.

Clinical Management×1.5

A safe, specific, UK-primary-care plan at the correct urgency, shared as options. The weighting means a management zero is the single most expensive event in the exam.

Relating to Others

Cues answered in the moment they're offered, ICE woven into the plan (not bolted on), plain language, genuine shared decisions. Formulaic empathy reads as no empathy.

The myth that fails people: "good communication can carry weak medicine." It cannot — domains don't compensate, and CM is worth half as much again. The reverse myth fails the rest: "the medicine is what matters." A doctor-centred interrogation caps Relating at borderline no matter how correct the plan. The exam wants both, and the weighting says management slightly more.

Train your marking eye before you train anything else: the grade descriptors show what each grade sounds like, and You Be the Examiner tests whether your judgement matches an examiner's — because until it does, your self-assessment of practice runs is noise.

The exact phrases examiners pick from

When a domain doesn't reach the standard, examiners select from a fixed list of feedback statements. Reverse-engineer them — each is a thing to not do.

  • Data Gathering & diagnosis: data gathering insufficient for a safe assessment; existing record information under-used; psychosocial information missed; unsystematic/disorganised gathering; poor prioritisation across multiple problems; implications of abnormal findings not grasped; differentials inadequately generated or tested; decision-making illogical, incorrect or incomplete.
  • Clinical management & medical complexity: referral, investigation or prevention/health-promotion plan not reflective of current practice; medical management of risk inadequate; implications of comorbidity under-considered; uncertainty (including the patient's) managed poorly; inadequate follow-up/continuity/safety-netting; ineffective time management.
  • Relating to others: non-verbal cues/active listening not shown; agenda, health beliefs or preferences under-explored; circumstances/cultural differences not responded to; explanations not adapted to the person; a judgemental approach; insufficient respect or sensitivity; ownership of decisions inappropriate; teamwork/others' roles not recognised; safeguarding concerns not recognised or acted on.

03The 12 clinical experience groups

Cases are sampled from the RCGP's twelve clinical experience groups. The college is explicit that the selection isn't a ranking — prepare across all of them, because the blueprint guarantees breadth. The groups, with what they tend to test:

👶
Children & young peopleConsulting through parents without losing the child; safeguarding antennae; adolescent confidentiality.
⚧️
Gender, reproductive & sexual healthContraception, menopause/HRT, pregnancy red flags (ectopic!), sensitive history-taking.
📈
Long-term conditions & cancerReviews that negotiate rather than lecture; 2WW conversations; multimorbidity trade-offs.
🧓
Older adultsFrailty, delirium vs dementia, polypharmacy, carers as second patients, end-of-life honesty.
🧠
Mental health & addictionRisk assessment that's specific, not scripted; alcohol/substance conversations without judgement.
🚨
Urgent & unscheduled careTelephone triage, disposition decisions, the case where today-vs-tomorrow IS the mark.
🛡️
Health disadvantage & vulnerabilitiesSafeguarding, capacity, veterans, homelessness, communication difficulties.
🌍
Ethnicity, culture & diversityBeliefs that shape management (fasting, family decision-making) — respected and planned with, not around.
📋
New & undifferentiated presentationsThe classic GP case: committing to a sensible working diagnosis under uncertainty.
💊
Prescribing & pharmacologySafe, specific prescribing: interactions, monitoring, deprescribing, the drug-seeking conversation.
🔬
Investigations & resultsExplaining abnormal results honestly; not hiding behind tests; rational investigation.
⚖️
Professional & ethical dilemmasConfidentiality, third-party calls, colleagues as patients, dubious paperwork — reasoning out loud beats the "right answer".

Every case in The Hot Seat is tagged to its groups and filterable — and your heat-map shows which groups you've never practised.

04Why candidates actually fail

Pass rates differ sharply between UK graduates and IMGs — a differential the college itself studies. Examiners' reports and trainer experience point at the same handful of patterns, and none of them is a knowledge deficit:

1
The recited template. "Any fever, weight loss, night sweats…" fired regardless of presentation. It feels safe; it reads as not thinking. Fix: hypothesis first, questions second — every Hot Seat case trains the discriminating question.
2
Management arrives at minute 11. A beautiful history with no plan scores worse than an adequate history with a good plan — CM is weighted ×1.5. Fix: summarise by 6, manage by 7. The time maps in every case brief drill this rhythm.
3
Formulaic empathy. "I'm sorry to hear that, it must be very difficult" deployed on a timer convinces no one. Fix: respond to the specific thing said — the "don't say it like this" contrasts exist for exactly this.
4
Cues left on the table. The under-stated British cue — "I've been a bit off, you know" — sails past. Each station plants them deliberately. Fix: the cue-recognition checklists in the Hot Seat feedback, and You Be the Examiner, where you watch cues being missed by someone else.
5
Secondary-care reflexes. "I'll refer you" / "I'll discuss with my senior" where a UK GP manages in-house. The SCA examines an independent GP. Fix: learn what primary care holds — our algorithms and protocols are the reference layer.
6
Hedging the diagnosis. Refusing to commit ("it could be several things…") reads as not knowing. Examiners reward a sensible working diagnosis said plainly, with uncertainty managed honestly.
7
Jargon and the unexplained plan. "Stage 2 hypertension per NICE" means nothing to a patient. Fix: the 45-Second Explainer — three a day until plain English is the default register.

05A 12-week training plan

Built on one principle: calibrate first, drill second, simulate third. Practising before you can recognise good is how people rehearse their mistakes for twelve weeks.

Wks 1–2
Calibrate. Read the grade descriptors until you can hear them. Mark all three transcripts in You Be the Examiner — repeat until you're within one grade of the examiner everywhere. Read three playbooks.
Wks 3–6
Drill. Two Hot Seat cases a week, full feedback loop: scorecard → cues & time discipline → word-pictures → re-run anything that landed in Fail territory. Three explainer drills a day — they take four minutes. Log every case in Examiner Marking.
Wks 7–8
Find the cold cells. Open your heat-map. Practise ONLY the unpractised groups and the weakest domain for two weeks. If you have role-play partners or real (consented, de-identified) consultations, run them through Real-Consultation Feedback — your real habits differ from your performance habits.
Wks 9–11
Simulate. One Mock Exam Circuit per week under true timing — build from 3 cases to the full bank. Grade every case immediately after, while it's warm. Re-read the playbook for any consultation type that rattled you.
Wk 12
Taper. No new material. One light circuit early in the week, explainer drills daily, sleep guarded. Re-read your own reflection notes — by now they're the most personalised revision resource that exists for you.

06On the day

  • The room is your responsibility. Book it well in advance, test the kit on the same computer, same network, same chair, days before. A "do not disturb" sign and a colleague guarding the door are not paranoia.
  • Tech checklist: wired connection if possible, camera at eye level, face lit from the front, headset tested, phone OUT of the room except as instructed for invigilation.
  • Between cases, breathe out the last one. Three minutes is enough: ten seconds to let it go, the rest to mine the next records. Nobody else knows how the last station went — including, often, the candidate. Examiners watch one case each; case 7's examiner never saw case 6.
  • The buzzer is not failure. Plenty of passing consultations are cut off mid-close. What can't be missing is the management — another reason the minute-7 rule matters.
  • If technology fails, follow the invigilator's instructions and keep evidence (times, screenshots where permitted). Interrupted diets are handled by the college — your job is only to document and stay calm.

07The platform & exam-day logistics

The SCA runs on the RCGP's Osler Online platform, from your own booked private room. Two sessions run per day and you're allocated one: AM 08:45–13:25 or PM 13:20–18:05. Knowing the exact choreography removes a whole layer of exam-day stress.

The exam-day timeline

  • Candidate registration (45 min). Log in to Osler; the invigilator does your ID check (government photo ID — passport or UK driving licence; a selfie emailed to [email protected] is the fallback if your camera won't focus) and a 360° environment check — they may take up to 30 minutes to appear.
  • Rounds 1–6 (1h 30). 15 minutes per station: 3 minutes reading then a 12-minute video or telephone consultation. A 15-second gap between stations to compose yourself (not a break).
  • Comfort break (10 min) after case 6 — a blue coffee-cup icon and countdown appear.
  • Rounds 7–12 (1h 30), same structure.
  • Wrap-up & reruns (45 min). You stay under exam conditions and may not log out until the invigilator gives permission.

Navigating the platform

  • Before you start: an automatic device check (mic, video, internet), then position your face until the on-screen oval turns blue (yellow = not aligned), and an audio test. Use the walkthrough as many times as you like beforehand.
  • During a case: your brief and your private notes carry over from reading time; you can minimise/close the brief and expand notes using the on-screen icons. A red one-minute warning then a 5-second countdown signals the end; completed stations get a green tick.
  • Telephone cases (about 3 of the 12) show a phone image — you hear the role-player but don't see them, and they/​the examiner don't see you, though you stay on camera for the invigilator.
  • To reach the invigilator: the hand-raise button (mid-consultation) or the direct-message function. You must message them when leaving and returning during the break.

Your desk — and reruns

  • Allowed: a water bottle, a wipeable whiteboard + marker, and a paper BNF. Banned: any clinical guidance, charts or posters on the walls — clear them before the environment check. No phone use during the break.
  • Reruns (45 min reserved). Granted for genuine disruption — repeated signal drops or severe external noise. Flag any issue immediately — for rounds 1–6 during the break, for rounds 7–12 in the 5 minutes straight after case 12 — or a lead-examiner review can't be guaranteed. A rerun appears as an Ad-hoc Station at the foot of your schedule.
  • If tech fails: press refresh (≈15 s to reconnect) or log out and back in (you return to your station). IT helpline: 020 3188 7680 — keep your phone to hand, as they'll call you.

Booking, attempts & kit

  • 9 sittings a year (monthly except July, August & December), bookable up to 12 months ahead on Fourteen Fish. A maximum of 4 attempts across any version (CSA/RCA/SCA) — usually no more than 2 in ST3 — so time it well and don't resit before you've fixed the flagged areas.
  • Kit: updated Chrome or Edge; a PC, laptop or Mac only (no phones/tablets); the same device and location as your device check. Wired headphones are fine and often fix sound issues; Bluetooth is not permitted.
  • No physical examination is required — cases are written without it, or examination findings are supplied in the pre-reading by another professional.

What the cases emphasise

  • Expect uncertain diagnosis, polypharmacy and comorbidity — this is not one problem done well in 12 minutes. Role-players may be the patient, a parent/carer or a health/social-care professional.
  • Cases reach into less-everyday territory: domestic abuse, FGM, transgender care, assisted-dying requests, safeguarding.
  • Of the 12 blueprint areas, the college flags a top 5 likely in every diet: under-19s; gender, reproductive & sexual health; long-term conditions; older adults; mental health.

08Quick answers

Do I need to pass a minimum number of cases?
No. Only the cumulative total out of 126 counts. One disastrous case is survivable; the same weakness repeating across twelve is not — which is why the domain averages on your heat-map predict the result better than any single mock.
Is there a "preferred" consultation model?
The college is explicit that there isn't — no script, no model, no magic phrases. The descriptors reward behaviours (cues answered, plans shared, safety-nets specific), not structures. Any model that produces those behaviours is fine; reciting any model instead of consulting is not.
Will I be examined on physical examination?
No — examination skills live in workplace-based assessment. In the SCA you ask for findings and they're provided. The skill being tested is knowing which findings you need and what they change.
How is the audio (telephone) case different?
Roughly a quarter of the diet is audio-only. Your ears replace your eyes: verbalise what you'd normally show ("I'm writing this down", "take your time"), check understanding more often, and treat disposition (today vs tomorrow vs 999) as the centre of gravity — telephone cases are very often urgent-care cases.
When should I sit it?
In ST3, when your trainer agrees your ordinary surgeries look like passing consultations — the SCA samples your daily work, so the best preparation is consulting well daily and then exam-shaping it for 8–12 weeks. Booking is through your RCGP account; places go quickly, so plan the diet date before you plan the revision.
I failed. What now?
Your feedback names the weak domains — believe it over your memory of the day. Take your results letter to The Resit Clinic: it turns the letter and an honest questionnaire into named diagnoses and a personalised 8-week treatment plan. Rebuild from calibration, not volume — resitters who simply do more cases usually rehearse the same habits harder. And use your deanery's support — targeted help after a fail is what it exists for.