Before you decide you have "covered" SCE Geriatric Medicine, you need evidence, not a feeling and not a bank percentage. This is a checklist, not a study timetable: it defines the minimum you should be able to show — per blueprint domain, per format, per recency-sensitive topic — before you stop doing new questions. If you cannot tick these, you have a content gap, whatever your overall accuracy says.
The direct answer, as a checklist
You have covered SCE Geriatric Medicine when you can demonstrate all of the following:
- Every blueprint domain attempted under timed, unseen conditions, with first-attempt accuracy recorded — not just questions "seen."
- No domain sitting below your personal floor (a margin above the likely pass standard) on mixed, unseen blocks.
- Deliberate, evidenced practice in the four format traps: frailty judgement, mental capacity, polypharmacy/deprescribing and multidisciplinary care.
- Interpretation practice logged for the data types the exam uses.
- Guidance-sensitive topics checked against current UK sources within a known date window.
- Retention confirmed at two weeks, not just on the day.
- Pacing under three hours per 100-question paper.
Everything below expands these into things you can verify.
Current exam snapshot
The SCE in Geriatric Medicine (Federation of the Royal Colleges of Physicians, MRCP(UK)) is two papers of 100 best-of-five questions each — 200 in total — three hours per paper, one day, computer-based, one mark per correct answer, no negative marking. Sat by higher specialty trainees, usually ST4 and above. The structure is fixed; the content is governed by the published blueprint, which is the authority you audit against — not any vendor's "high-yield" list. (Confirm current delivery arrangements on the Federation site, as 2026 introduced changes to how sittings are delivered.)
Build a blueprint coverage table
This is the core instrument. Rebuild it from the official blueprint and populate every row from your own bank data. Do not stop doing new questions while any row is empty or red.
| Domain | Official weight (~/200) | Questions attempted | First-attempt accuracy | Last reviewed | Confidence (R/A/G) |
|---|---|---|---|---|---|
| Cognitive issues (delirium/dementia) | 20 | ||||
| Falls and poor mobility | 16 | ||||
| Stroke care | 15 | ||||
| Rehabilitation and MDT working | 14 | ||||
| Continence / urogenital | 10 | ||||
| Orthogeriatrics and osteoporosis | 10 | ||||
| Planning transfer of care | 9 | ||||
| Palliative care | 9 | ||||
| Basic science | 8 | ||||
| Old-age psychiatry (non-cognitive) | 7 | ||||
| General medicine in older people | ~61 combined | ||||
| Nutrition / tissue viability / sensory | ~13 combined |
The "last reviewed" column matters as much as accuracy: a domain answered well eight weeks ago and untouched since is not covered, it is decaying. Score the confidence column honestly too: red where you would not want that domain to appear first in the paper, amber where you are functional but slow, green where you are quick and accurate on unseen items. A table that is mostly amber is not a table that is ready — it is one telling you precisely which rows to turn green, and in what order, before the exam.
The ten domain-level blind spots self-selected practice tends to hide
Free question choice quietly steers you toward what you already enjoy. These ten areas are the ones trainees most often leave under-practised, and each should be signed off by someone who knows the exam before you call the domain done:
- Mental capacity assessment and best-interests decisions, including deprivation-of-liberty safeguards in practice.
- Deprescribing and anticholinergic burden — applying STOPP/START reasoning, not just naming interactions.
- Delirium versus dementia versus depression discrimination in mixed presentations.
- Frailty scoring and its use in decisions about escalation, surgery and treatment ceilings.
- Continence subtyping and the non-drug management that examiners expect first.
- Orthogeriatric co-management: bone protection, peri-operative medicine and post-fracture pathways.
- Palliative and end-of-life prescribing in older patients, referenced to the SmPC/eMC rather than habit.
- Stroke work beyond hyperacute thrombolysis — secondary prevention, rehabilitation and swallow.
- Community and intermediate-care pathways, discharge planning and continuing-care funding.
- Statistics and ethics items — small in number but easy marks that self-selected practice skips.
Format checklist: the four traps
Verify you have practised these deliberately, because they are reasoning skills, not facts, and a subject-only revision plan misses them:
- Frailty: can you apply a frailty judgement to a management decision, not merely define it?
- Capacity: can you separate capacity from the "unwise decision" it is not, and act accordingly?
- Polypharmacy: can you sequence a deprescribing plan and justify it against current guidance?
- Multidisciplinary care: can you identify the right team action and the correct escalation, which SCE items frequently reward over a drug choice?
Interpretation checklist
Tick each data type you have actually practised interpreting under time:
- Laboratory trends (renal function, electrolytes, bone profile, B12/folate).
- ECGs relevant to falls and syncope (conduction disease, long QT, arrhythmia).
- Radiographs and simple imaging descriptions (fracture patterns, aspiration, cord signs).
- Calculations (creatinine clearance, drug dosing in renal impairment) done correctly under time.
- Ethics, consent and statistics items, which appear across papers.
Recency checklist
Geriatric practice is guidance-sensitive. For each of the following, record the source and its date and jurisdiction, so you are not revising a superseded position:
- Anticoagulation in atrial fibrillation and after falls.
- Osteoporosis treatment thresholds and drug holidays.
- Dementia diagnosis and pharmacological management.
- Delirium prevention and management.
- Secondary stroke prevention.
Cite NICE, CKS, SIGN, the SmPC/eMC and NHS guidance — and log the date you checked, because "I read it once" is not a recency check.
The quiet gap: general medicine in older people
Roughly 61 of the 200 questions are general internal medicine framed in older patients — cardiology, infection, respiratory, endocrine, gastroenterology, renal, musculoskeletal, neurology, dermatology and haematology. Candidates who revise "geriatrics" narrowly, around cognition, falls and continence, routinely under-prepare this slice and lose marks on material they met at MRCP level but have not refreshed in an older-patient context. Audit it explicitly: does each general-medicine system have unseen, timed practice, and are the presentations age-appropriate — atypical infection, polypharmacy interactions, renal dosing, delirium as the presenting feature of physical illness? This is the domain where a cross-specialty MRCP-level bank earns its place, because it holds far more general-medicine volume than a purely geriatric bank, and it is where a narrow revision plan most often springs a leak.
Performance checklist
Coverage is necessary but not sufficient; performance closes the loop:
- Unseen timed mixed blocks above your floor in every domain.
- Speed comfortably inside three hours per paper.
- High-confidence errors driven to near zero — these are the dangerous ones, because you will not review what you were sure of.
- Retention re-tested at two weeks.
- Official-material calibration against the Federation sample questions or a specialty-society mock, sat cold.
Your overall Q-bank percentage is explicitly not the metric here — it blends seen and unseen, easy and hard, and inflates as you re-do items. Measure the columns above instead.
Stop / continue decision tree
- If a domain row is empty or red → continue new questions in that domain.
- If coverage is complete but unseen accuracy is below floor → consolidate: review coded errors, do targeted teaching, re-test.
- If coverage and unseen accuracy are green but pacing fails → simulate: full-length timed papers, no new content.
- If everything is green and retention holds → stop adding questions; protect the gains with light spaced review.
- If high-confidence errors persist → seek teaching, not more questions — you have a calibration problem a bank cannot fix alone.
One-page checklist and a worked example
Copy the block below; it is the whole audit on one page.
SCE GERIATRIC MEDICINE — COVERAGE AUDIT
[ ] Every blueprint domain attempted, timed, unseen
[ ] No domain below personal floor on mixed blocks
[ ] Frailty / capacity / polypharmacy / MDT practised deliberately
[ ] Interpretation: labs, ECG, imaging, calculations, ethics/stats
[ ] Recency logged (source + date + jurisdiction) for AF, bones, dementia, delirium, stroke
[ ] High-confidence errors ~ zero
[ ] Retention re-tested at 2 weeks
[ ] Pacing < 3h per 100-Q paper
[ ] Cold calibration on official/society mock
Worked example (illustrative data). A trainee reports 74% overall and feels ready. The audit shows: cognition green (81% unseen), falls green (78%), but continence red (58% unseen, last reviewed nine weeks ago), transfer-of-care empty, and four high-confidence errors in capacity items. Verdict: not covered. The percentage was carried by two strong domains; the exam samples all of them. The next action is targeted new questions in continence and transfer-of-care, plus a capacity teaching session — not another full mock, and certainly not stopping.
A second profile shows the opposite trap. A trainee reports only 61% overall and assumes they are far off, but the audit reveals even accuracy across every domain, none below floor, near-zero high-confidence errors and retention holding at two weeks — the low headline is simply the product of practising almost entirely on unseen, mixed, timed blocks rather than re-doing easy items. Verdict: closer than the number suggests; the next action is full-length simulation and pacing, not a panic of new content. The two trainees score similarly on paper and need opposite plans, which is the whole point of auditing per domain rather than in aggregate.
Three mistakes this checklist is designed to stop
Mistake one: treating completion as coverage. Finishing a bank tells you that you have seen its questions, not that you have covered the blueprint — a bank can be fully complete and still leave a whole domain thinly sampled. The coverage table exists precisely to separate "seen" from "covered," which is why completion is not coverage.
Mistake two: trusting the headline percentage. A single overall figure blends easy with hard, seen with unseen, and strong domains with weak ones, so it can sit comfortably at 74% while a heavily weighted domain quietly fails. Per-domain, unseen, first-attempt accuracy is the number that generalises; the headline is the number that reassures.
Mistake three: revising by preference. Self-selected practice drifts toward the topics you already enjoy, leaving the blind-spot list — capacity, deprescribing, transfer of care, statistics — under-rehearsed. The blind-spot list and the coverage table exist to force balance, so the exam does not find the domain you avoided.
Bottom line
Coverage is a claim you should be able to evidence, domain by domain, before you stop doing new questions. Rebuild the table from the official blueprint, fill every cell with unseen timed data, check the recency of guidance-sensitive topics, drive high-confidence errors toward zero and confirm retention at two weeks. When every row is green and the signal holds cold on official material, stop and protect the gains. Until then, the empty or red row is your instruction for what to do next — not your overall percentage, and not a feeling of readiness.
Frequently asked questions
How do I know whether I have covered the full SCE Geriatric Medicine blueprint? You know when the coverage table has no empty or red rows: every domain attempted under timed, unseen conditions, first-attempt accuracy recorded, each reviewed recently, and none below your personal floor. "Covered" is a property of evidence per domain, not of a single headline number. Rebuild the table from the official blueprint rather than a vendor's topic list, because only the blueprint defines what the exam can sample.
Can one question bank be enough for SCE Geriatric Medicine? One strong specialty bank can carry most of the specialist core, but a single bank rarely delivers both full blueprint coverage and a renewable supply of genuinely unseen items for measurement — once you have completed a bank, its questions test recognition, not readiness. The pragmatic answer is a two-bank structure: a dedicated geriatric bank to learn and cover the specialist domains, and a second, cross-specialty source for unseen volume and the large general-medicine slice of this particular blueprint.
What should I measure instead of my overall Q-bank percentage for SCE Geriatric Medicine? Measure per-domain first-attempt accuracy on unseen, timed, mixed blocks; the size and location of high-confidence errors; retention at two weeks; and pacing. The overall percentage blends difficulty, mixes seen with unseen, and rises artificially as you repeat items, so it is a comfort metric rather than a readiness one. The per-domain unseen numbers are what generalise to the exam, which is why your Q-bank percentage is not your exam score.
When should I stop doing new SCE Geriatric Medicine questions? Stop when the coverage table is fully green, unseen mixed blocks sit above your floor in every domain, high-confidence errors are near zero, retention holds at two weeks and pacing is comfortable. Past that point, new questions add reassurance rather than marks, and your time is better spent on light spaced review and full-length timed simulation. If any of those conditions fails, the failure tells you precisely which activity to do next.
Which SCE Geriatric Medicine resource should I use for my weakest component? Match the resource to the deficit. If the weakness is specialist knowledge (cognition, falls, orthogeriatrics), a dedicated geriatric bank such as PassGeriMed or StudyPRN's geriatric SCE bank is the right tool. If it is the general-medicine slice or a need for unseen measurement, a cross-specialty MRCP-level bank like iatroX supplies breadth and fresh items. If it is a reasoning or calibration problem — persistent high-confidence errors — the fix is teaching and error review, not another bank.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 21 July 2026. Vendor figures cited are vendor-reported and change; verify counts, prices and features on the relevant product page before relying on them. Disclosure: iatroX operates a cross-specialty UK question bank and therefore competes for revision time with the specialty banks named here; this hub keeps iatroX's role to unseen cross-specialty measurement and does not present it as a geriatric-specific SCE bank. Corrections via the feedback route on iatrox.com. References: Federation of the Royal Colleges of Physicians (MRCP(UK)) SCE Geriatric Medicine blueprint and sample questions; British Geriatrics Society SCE resources; completion is not coverage; the two-Q-bank rule; the iatroX comparison hub.
Complete a fresh SCE Geriatric Medicine baseline in iatroX →