This workflow is for a higher specialty trainee (typically ST4 and above) sitting the SCE in Geriatric Medicine who wants to use BMJ OnExamination as a structured first-pass bank while revising around a clinical rota. It addresses the written best-of-five paper — the only component the SCE has. The principal limitation to plan around: BMJ OnExamination is a fixed question bank with difficulty filters and peer analytics, not a proprietary adaptive engine, so the "adaptivity" has to come from you.
What BMJ OnExamination offers for SCE Geriatric Medicine right now
| Item | Detail (vendor-reported unless stated) |
|---|---|
| Product | SCE in Geriatric Medicine question bank |
| Question count | 245+ questions (vendor-reported) |
| Access period | Subscription tiers from 1 to 12 months |
| Price | 1 month £69.99; 3 months £99.99; 6 months £159.99; 12 months £194.99 (vendor-reported) |
| Adaptive/AI features | None described. Difficulty selection, themed mock tests, group leaderboards, offline app access |
| Components supported | The written SCE paper only (there is no clinical component to reproduce) |
| Last checked | 20 July 2026 |
Figures are vendor-reported and move; confirm the live count, access window and price on the BMJ OnExamination product page before you buy. The single most important thing to note is that BMJ OnExamination does not advertise an adaptive difficulty algorithm. It gives you difficulty tags, a themed mock and a leaderboard — useful, but the sequencing decisions are yours. Treat any talk of "adaptive practice" as a study strategy you run, not a feature the platform provides.
Exam anchor: the SCE format and the Geriatric Medicine blueprint
Every SCE, regardless of specialty, has the same shape: two papers of 100 best-of-five questions each, 200 questions in total, three hours per paper, sat in one day, computer-based on the Surpass platform at a test centre, one mark per correct answer and no negative marking. Only the blueprint changes. For Geriatric Medicine, the Federation of the Royal Colleges of Physicians blueprint spreads those 200 marks across the whole specialty curriculum. The heaviest domains are cognitive issues (delirium and dementia, around 20 questions), falls and poor mobility (around 16), stroke care (around 15), rehabilitation and multidisciplinary teamworking (around 14), and continence, orthogeriatrics/osteoporosis and neurology (around 10 each). Sizeable general-medicine-in-older-people domains — cardiovascular, infection, respiratory, endocrine, renal — sit alongside core geriatric topics such as comprehensive assessment, planning transfer of care, basic science, nutrition, tissue viability and palliative care. That is the official requirement. A vendor's topic list is a proxy for it, not a substitute, so map the bank onto the blueprint yourself rather than trusting a headline total.
Baseline week: measure before you personalise the feed
Before you let any bank shape your revision, take a small, blueprint-stratified unseen sample so you have a defensible starting point. Draw roughly 40–60 questions spread across the blueprint in rough proportion to its weighting — a few on delirium and falls, one or two on the smaller domains — and sit them timed, mixed and without help. Record first-attempt accuracy by domain, not just an overall percentage. This baseline is your reference point for everything that follows, and it is why you should keep a reserve of questions you have never seen: once you have worked through a bank, its score tells you about recognition, not recall. Your Q-bank percentage is not your exam score, and a baseline taken on unseen items is the cleanest signal you will get early on.
First pass: set domain floors so nothing hides behind a rising average
The failure mode with any large bank is a comfortable overall percentage that conceals unattempted syllabus areas. A rising average can mean you are getting better, or simply that you keep answering questions in the domains you already like. Prevent this by setting a minimum number of attempted questions per blueprint domain — a floor — before you allow yourself to chase the headline number. Use BMJ OnExamination's topic filters to work through the bank domain by domain, weakest first, and log attempts against each blueprint area. The rule is simple: no domain drops below its floor, and low-volume areas (nutrition, tissue viability, sensory impairment, the palliative-care slice) get attempted early, because they are the ones a self-directed learner quietly skips.
Error taxonomy: label every miss before you re-read anything
Reviewing means diagnosing why a question was wrong, not re-reading the explanation until it feels familiar. Give every miss one code:
- Knowledge gap — you did not know the fact or guideline.
- Misread stem — you knew it but misread the question (age, timing, "most likely" versus "next step").
- Premature closure — you locked onto a diagnosis and stopped weighing alternatives.
- Guideline error — you applied out-of-date or wrong-jurisdiction guidance.
- Calculation error — creatinine clearance, drug dosing, fluid balance.
- Time-pressure error — you rushed or ran out of time.
The distribution tells you what to fix. A pile of knowledge gaps means read more; a pile of misread stems means slow down and annotate; a pile of premature-closure errors means practise generating a differential before you commit.
Review interval: match the response to the error type
Not every miss deserves the same treatment. A one-off knowledge gap on a rare condition may need nothing more than a single spaced review. A recurring guideline error deserves a short source read against current NICE, SIGN or CKS guidance and the SmPC/eMC for the drug detail. A premature-closure pattern deserves a fresh transfer question that tests the same principle in a different clinical wrapper, so you rehearse the reasoning rather than memorising one stem. Reserve immediate repeats for questions where you made a careless slip you can correct on the spot; everything else is better spaced.
Mixed-block switch: objective criteria, not a feeling
Move from topic-filtered practice to timed random blocks when you can demonstrate three things: every blueprint domain is above its floor; your first-attempt accuracy on the last two mixed samples is stable rather than swinging; and you are inside roughly three to four weeks of the exam. Topic-filtered blocks cue you to the subject — you know a continence question is coming because you selected continence. The real paper does not, so the back half of your preparation should be increasingly random, timed and mixed to remove that scaffolding.
Exit criteria: what "ready" actually looks like
Readiness is not bank completion. Stop when you meet a defined set of signals: every domain attempted above its floor with acceptable accuracy; stable first-attempt performance across at least two mixed unseen samples; pacing that lands you comfortably inside three hours per 100 questions; retention holding on spaced re-tests of previously missed items; and a calibration check against official material. Finishing 100% of a bank is not on that list. A completed bank measures exposure; the exit criteria measure transfer.
Worked example: a seven-day plan around clinical work
Here is a realistic week for a trainee doing this alongside a rota, using BMJ OnExamination for one defined job — structured first-pass and review — and iatroX for unseen transfer measurement. No proprietary algorithms are assumed on either side.
- Monday (45 min): BMJ OnExamination topic block on falls and poor mobility, timed. Code every miss.
- Tuesday (45 min): Review Monday's misses; write one corrective line each. Short source read on the multifactorial falls assessment.
- Wednesday (30 min): iatroX — a small timed, mixed, unseen block to measure transfer, not to learn. Log first-attempt accuracy by domain.
- Thursday (45 min): BMJ OnExamination block on delirium and dementia, timed; code misses.
- Friday (rest or 20 min): Spaced review of the two oldest sets of missed items only.
- Saturday (90 min): One longer timed mixed block on BMJ OnExamination across several domains; full error-coding.
- Sunday (45 min): iatroX unseen mixed block as the week's readiness check; compare against last week; decide next week's domain floors.
The division of labour matters. The competitor bank does the heavy lifting of first-pass exposure and review; iatroX supplies the unseen measurement layer so your weekly readiness signal is not contaminated by questions you have already seen. This is the two-Q-bank rule applied to Geriatric Medicine: one bank to learn on, one to measure on.
Decision checklist: continue, supplement, switch or stop
- Continue BMJ OnExamination if your domain floors are filling, your error codes are shifting from knowledge gaps toward slips, and your unseen accuracy is rising.
- Supplement with a second source if one or two blueprint domains stay stubbornly weak after focused work, or if you have exhausted the bank's questions in a heavy domain.
- Switch primary bank only for a measurable reason — a domain the bank barely covers, or dated explanations — never for novelty.
- Stop adding new material and move to pure mixed-mock mode once you meet the exit criteria; more first-pass volume past that point adds fatigue, not marks.
Bottom line
BMJ OnExamination is a reasonable structured first-pass bank for SCE Geriatric Medicine, provided you supply the adaptivity it does not: domain floors, honest error-coding, spaced review and a reserved pool of unseen questions for weekly measurement. Used that way — competitor bank to learn, iatroX to measure — you get a preparation you can actually read, rather than a rising percentage you have to trust. Compare your options on the iatroX comparison hub and treat completion as a milestone, not a finish line.
FAQ
Is BMJ OnExamination enough for SCE Geriatric Medicine on its own? For many candidates a single well-worked bank plus official material is workable, but "enough" depends on your baseline and how honestly you review, not on the bank alone. BMJ OnExamination gives structured first-pass coverage; what it cannot give you is a measurement of retention on questions you have never seen, because once you have worked through it every score reflects recognition. Pair it with a reserved unseen pool for timed measurement before you call it sufficient.
Which SCE Geriatric Medicine component does BMJ OnExamination not reproduce well? The SCE has only the written best-of-five paper, so there is no clinical or oral component to miss — the more useful answer is that BMJ OnExamination does not reproduce a genuinely adaptive or blueprint-audited experience. It offers difficulty tags and a themed mock, but it will not force you into the low-volume domains (nutrition, tissue viability, sensory impairment) unless you deliberately select them, so the coverage discipline is yours to impose.
How many BMJ OnExamination questions should I complete per day for SCE Geriatric Medicine? The right figure is set by your review capacity, not by the bank size — most trainees revising around clinical work sustain roughly 20–30 timed questions on a weekday evening with a longer block at the weekend. With a bank of 245+ questions (vendor-reported, 20 July 2026), the constraint is rarely raw volume but whether you review each miss properly, so favour fewer questions fully coded over large untended blocks.
When should I stop using BMJ OnExamination and move to mixed mocks? Move to mixed timed mocks when every blueprint domain is above its floor, your first-attempt accuracy on unseen samples is stable across at least two checks, and you are within three to four weeks of the exam. That switch matters because topic-filtered blocks cue you to the subject; the real paper is random, so the closing phase of revision should remove the scaffolding rather than lean on it.
How should I combine BMJ OnExamination with iatroX without duplicating practice? Give each tool one job: BMJ OnExamination for first-pass learning and review, iatroX for timed, mixed, unseen measurement. iatroX is not a specialty-specific Geriatric SCE bank — it is the cross-specialty knowledge and unseen-MCQ layer that tells you whether what you revised on BMJ OnExamination actually transfers. Do not re-test the same items on both; use one to learn and the other to check, which is exactly how the two-Q-bank rule avoids destroying your calibration.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 20 July 2026. Question counts, prices and features are vendor-reported and change frequently; confirm current figures on the BMJ OnExamination product page before purchase. Disclosure: iatroX operates a competing question bank; its role here is confined to unseen cross-specialty measurement and spaced retrieval, jobs BMJ OnExamination's SCE Geriatric product does not claim to do. Corrections are welcome via the feedback route on iatrox.com.
References: The Federation of the Royal Colleges of Physicians — SCE in Geriatric Medicine specialty page and blueprint (thefederation.uk); BMJ OnExamination SCE in Geriatric Medicine product and SCE Resources pages (onexamination.com); NICE, SIGN and CKS guidance plus the SmPC/eMC for medicines detail; iatroX internal references — Your Q-Bank Percentage Is Not Your Exam Score and the two-Q-bank rule.