This audit is for MRCEM Intermediate SBA candidates trying to decide whether BMJ OnExamination's analytics can tell them they are ready. The honest answer reshapes the usual review before we start: as of 19 July 2026, BMJ OnExamination's published emergency-medicine catalogue is a Primary (basic-sciences) SBA product, and we found no dedicated MRCEM Intermediate SBA bank. The principal limitation is therefore not a soft one about dashboards — it is coverage. Read that finding first, then use this method to interrogate whichever bank you actually revise from.
The coverage finding, stated plainly
Before any discussion of percentiles or predicted scores, a coverage audit has to answer one question: does the product contain items for the exam you are sitting? On the publication date, BMJ OnExamination's emergency-medicine collection listed a single SBA product oriented to MRCEM/FRCEM Primary — the basic-sciences examination covering anatomy, physiology, pharmacology, microbiology, pathology and evidence-based medicine, described on the vendor page as "590+ FRCEM SBAQ style questions" at £39.99 for one month's access (vendor-reported, 19 July 2026). We did not find a bank mapped to the MRCEM Intermediate SBA, which is the clinical, management-weighted 180-question examination this article is about.
That distinction matters because the two exams are not interchangeable. MRCEM Primary tests whether you know the science; MRCEM Intermediate SBA tests whether you can apply it under time pressure to undifferentiated emergency presentations. A dashboard built on Primary basic-science items cannot generate a readiness signal for Intermediate clinical decision-making, however polished its analytics look. If you are revising for the Intermediate SBA, treat BMJ OnExamination as a possible tool for shoring up basic-science underpinning that Intermediate stems assume — not as your Intermediate bank. Always verify the current catalogue on the product page, because vendors add and retire products without notice.
What BMJ OnExamination offers for MRCEM right now
Current-state box — last checked 19 July 2026; figures vendor-reported and subject to change:
| Item | What we found (vendor-reported, 19 Jul 2026) |
|---|---|
| Emergency-medicine product | Single SBA bank oriented to MRCEM/FRCEM Primary (basic sciences) |
| Stated question volume | "590+ FRCEM SBAQ style questions" |
| MRCEM Intermediate SBA bank | None found in the published catalogue |
| Access / price | £39.99 for 1 month's access (Primary product) |
| Adaptive / AI features | Free demo billed as "personalised"; no proprietary adaptive engine detailed on the product page |
| Analytics | Progress tracking and performance feedback typical of the BMJ OnExamination platform |
Verify each figure on the BMJ OnExamination emergency-medicine page before relying on it; question counts, prices and access windows move.
Exam anchor: what the Intermediate SBA actually is
The MRCEM Intermediate SBA is set by the Royal College of Emergency Medicine against the 2021 RCEM curriculum. It comprises 180 single-best-answer questions delivered in two 120-minute papers of 90 questions each, usually sat on the same day with a break between, for roughly five hours' total attendance and about 80 seconds per item. There is no negative marking, and the pass mark is set per sitting by modified Angoff, typically landing between 60% and 70%. The SBA draws on Speciality Learning Outcomes 1 and 3 to 7, with a published domain weighting:
| RCEM domain (SLO) | Approx. items | Share |
|---|---|---|
| Care of the complex stable patient (SLO 1) | 55 | 31% |
| Resuscitation / acutely unwell (SLO 3) | 40 | 22% |
| Care of the injured patient (SLO 4) | 30 | 17% |
| Paediatric emergency medicine (SLO 5) | 25 | 14% |
| Procedural skills (SLO 6) | 20 | 11% |
| Complex and challenging situations (SLO 7) | 10 | 6% |
Anything a third-party bank tells you about "coverage" should be read against this official weighting, not against the bank's own internal categories. The official RCEM curriculum and any published sample questions are the gold standard; a vendor's mapping is a claim to be checked.
Define every metric before you trust it
If you do use BMJ OnExamination for Primary consolidation, or when you audit any Intermediate-capable bank, insist on precise definitions for the numbers on the home screen:
- First-attempt accuracy — percentage correct the first time you ever saw an item. This is the only accuracy figure that reflects genuine recall; everything else is contaminated by exposure.
- Repeat accuracy — percentage correct on items you have seen before. High repeat accuracy usually measures memory of the answer, not understanding of the topic.
- Percentile — your rank against other users of that bank, not against the passing standard. The comparison cohort is self-selected and unverified.
- Predicted score — a modelled estimate, only as good as its calibration sample. For an exam the bank does not even cover, a predicted score is meaningless; treat all such figures cautiously.
- Coverage — proportion of the bank's own catalogue attempted. This is completion, not blueprint coverage; a bank can be 100% complete and still under-represent a whole SLO.
- Difficulty — usually the historical percentage-correct across users, a proxy that shifts as the user base changes.
- Time per item — median seconds spent. Useful for pacing against the ~80-second Intermediate budget, but only meaningful in timed, unassisted mode.
Selection bias: why the percentage is not the score
Any bank that lets you filter to weak topics — or that steers you there — will oversample your worst material. That is good for learning and terrible for interpreting a raw percentage. If two-thirds of your recent questions came from your three weakest areas, a home-screen average of 62% is not comparable to a 62% on a mixed, blueprint-proportioned mock. The number moves with the feed, not with your readiness. This is the single most common misreading, and it is why we link every candidate to why your Q-bank percentage is not your exam score. The only percentage worth acting on is one earned on an unseen, timed, mixed block sampled in the official proportions.
Blueprint audit beats the home-screen average
Instead of trusting the average, rebuild your attempted-question distribution against the RCEM weighting. Export or tally what you have actually done, then compare shares. A worked audit for a candidate three weeks out might look like this:
| Domain | Official share | Your attempts | First-attempt acc. | Verdict |
|---|---|---|---|---|
| Complex stable patient | 31% | 44% | 71% | Over-weighted; strong |
| Resuscitation | 22% | 19% | 63% | Roughly on-blueprint |
| Injured patient | 17% | 9% | 58% | Under-sampled; weak |
| Paediatric EM | 14% | 6% | 49% | Badly under-sampled; weak |
| Procedural skills | 11% | 15% | 66% | Over-weighted |
| Complex/challenging | 6% | 7% | 44% | On-blueprint volume; weak accuracy |
The home screen for this candidate might read a comfortable 64%. The audit tells the real story: paediatric EM and the injured patient are both under-sampled and under-performing, and complex/challenging situations (safeguarding, mental capacity, end-of-life, major incident) is weak despite adequate volume. None of that is visible in a single average, and none of it can come from a Primary basic-science bank.
The readiness test: five conditions
A percentage is a credible readiness signal only when all five conditions hold at once:
- Unseen — items you have never attempted, so you are measuring recall, not recognition.
- Timed — at or below the ~80-second Intermediate budget.
- Mixed — random across domains, not filtered to a topic.
- No assistance — no tutor, notes, search or answer options open.
- Adequate sample — enough items (a full 90-question paper, or two, rather than a handful) that the figure is not noise.
If any condition fails, the number is diagnostic feedback, not a readiness signal. A Primary bank fails condition one for the Intermediate exam by definition, because none of its items are Intermediate stems.
Override rules: force the low-volume domains
Whichever bank you use, adaptive or filtered feeds tend to starve the small, high-consequence areas. Set manual overrides so the following are deliberately practised, not left to the feed:
- Paediatric EM — dosing, the unwell child, safeguarding presentations, and paediatric resuscitation.
- Procedural skills — indications, contraindications and complications rather than the manual steps.
- Complex and challenging situations — mental capacity, consent, end-of-life, major incident, and communication under pressure.
- Image and data interpretation — ECGs, radiographs, blood-gas and laboratory trends, which are easy to skip in a text-only feed.
- Calculations — infusion rates, corrected values and paediatric weights, where a single arithmetic slip is a wrong answer.
Worked dashboard example: from analytics to quotas
Take the audit above and convert it into next week's quotas without inventing a pass prediction. The candidate has roughly 450 minutes of protected study across the week. Allocation by measured gap, not by comfort:
- Paediatric EM — 60 unseen items, timed, mixed within the domain (largest gap).
- Injured patient — 45 unseen items, timed.
- Complex/challenging situations — 30 items plus one focused source read on mental-capacity assessment.
- Resuscitation — 30 items to hold the line.
- One mixed, timed 90-item block at week's end, sampled in blueprint proportions, scored as the week's only readiness figure.
The output is a set of quotas and one honest mixed-block score — not a green "ready" light.
A seven-day pattern for busy trainees
Around clinical shifts, give each tool one defined job and do not blur them. Because BMJ OnExamination does not publish an Intermediate bank, its only legitimate job here is optional basic-science repair.
- Mon–Wed: Intermediate-mapped topic practice on an Intermediate-capable bank (for example FRCEMTutor or the iatroX MRCEM SBA bank), 40–60 items a day against your two weakest SLOs.
- Interleaved (optional): if a stem exposes a basic-science gap — a receptor mechanism, an acid-base derivation — a short, targeted session on BMJ OnExamination's Primary material can repair the underpinning. Keep it bounded; it is not your main line.
- Thu: review day — no new topics, only reworking the week's errors from first principles.
- Fri: a fresh, timed, unseen mixed block in iatroX as your transfer-and-measurement layer, scored once and logged. We make no claim to access any competitor's proprietary algorithm; iatroX simply supplies unseen, blueprint-mapped items for a clean measurement.
- Weekend: rest, plus a short read on the single domain that scored worst on Friday.
This is the two-Q-bank rule applied honestly: one bank to learn from, a second, unseen bank to measure on, and no double-counting.
Three mistakes this audit is designed to stop
Mistake one: buying the wrong exam. The names are close enough that candidates purchase a Primary bank for an Intermediate sitting. Check the exam name against the RCEM curriculum, not the marketing.
Mistake two: trusting the average. A rising home-screen percentage on a filtered feed feels like progress and hides your blueprint holes. Only the mixed unseen block counts.
Mistake three: mistaking repeat accuracy for readiness. Getting 90% on items you have already seen tells you your memory works, not that you can handle a novel stem in 80 seconds.
Decision checklist: continue, supplement, switch or stop
- Continue with BMJ OnExamination only if your live need is MRCEM Primary basic sciences and its analytics help you close measured Primary gaps.
- Supplement if you are sitting the Intermediate SBA: you need an Intermediate-mapped bank as your main line and an unseen bank for measurement; BMJ OnExamination can, at most, patch basic-science underpinning.
- Switch if you bought the Primary product expecting Intermediate coverage — that is a coverage mismatch, not a study problem, and no amount of grinding fixes it.
- Stop any resource once your mixed, unseen, timed blocks are stable in the official proportions and your low-volume domains clear their floors. Bank completion is not a stopping rule.
Bottom line
For MRCEM Intermediate SBA, BMJ OnExamination's published emergency-medicine product is the wrong instrument for the job as of 19 July 2026: it targets Primary basic sciences, so its dashboard cannot signal Intermediate readiness. It may still earn a bounded place repairing the science that Intermediate stems assume. Build your readiness picture from an Intermediate-mapped bank plus regular unseen, timed, blueprint-proportioned blocks — and audit the distribution, never the average. Compare tools on the iatroX comparison hub rather than on percentages alone.
FAQ
Is BMJ OnExamination enough for MRCEM SBA on its own? No. More fundamentally, as of 19 July 2026 we could not find a BMJ OnExamination bank for the MRCEM Intermediate SBA at all; its emergency-medicine product is oriented to MRCEM/FRCEM Primary basic sciences (vendor-reported "590+ FRCEM SBAQ style questions", £39.99/month). Even as a general resource it cannot, on its own, prepare you for a clinical, management-weighted 180-question paper it does not cover.
Which MRCEM SBA component does BMJ OnExamination not reproduce well? It does not reproduce the Intermediate SBA at all. The components it therefore cannot rehearse include the applied clinical decision-making of the complex stable patient, resuscitation prioritisation, paediatric emergency dosing and safeguarding, procedural indications, and the complex/challenging situations domain — all of which are Intermediate content rather than Primary basic science.
How many BMJ OnExamination questions should I complete per day for MRCEM SBA? For the Intermediate SBA, the honest answer is zero from this product, because it is a Primary bank. If you separately need Primary basic-science repair, a bounded session of 20 to 30 targeted items is enough to fix an identified gap; your daily Intermediate volume should come from an Intermediate-mapped bank instead.
When should I stop using BMJ OnExamination and move to mixed mocks? If you are on the Intermediate pathway you should move to Intermediate-mapped practice and mixed unseen mocks immediately, using BMJ OnExamination only for occasional basic-science patching. Your stopping signal for all topic-filtered practice is the same: stable performance on timed, unseen, blueprint-proportioned blocks, not completion of any catalogue.
How should I combine BMJ OnExamination with iatroX without duplicating practice? Give each a single, non-overlapping job. Use BMJ OnExamination, if at all, for Primary basic-science underpinning; use an Intermediate-mapped bank as your main learning line; and reserve iatroX for unseen, timed, blueprint-mapped measurement blocks. Because the products address different exams and different jobs, you are measuring on fresh items rather than re-testing memorised ones.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026. Vendor-reported figures (question counts, prices, access periods and feature descriptions) were taken from the providers' public pages on 19 July 2026 and change without notice — verify the current figures on the product page before purchasing. Disclosure: iatroX operates a competing MRCEM SBA question bank; this audit confines iatroX to jobs BMJ OnExamination does not claim to do — unseen transfer measurement and Intermediate-SBA-mapped practice — and does not present it as a like-for-like replacement. Corrections: use the feedback route on iatrox.com and we will amend any error or out-of-date figure.
References