This audit is for endocrinology and diabetes higher specialty trainees using BMJ OnExamination's SCE bank and its performance analytics to gauge readiness. Checked on 20 July 2026, this is one of the specialties BMJ OnExamination genuinely covers, so the audit is real. The principal limitation is interpretive: the dashboard's headline average and pass-likelihood indicator are built on self-selected practice and do not, on their own, tell you whether you will pass.
What BMJ OnExamination offers for SCE Endocrinology and Diabetes right now
BMJ OnExamination's SCE in Endocrinology & Diabetes lists 625 practice questions in best-of-five format aligned to the Royal College of Physicians curriculum, with mock tests described as letting you "test myself against recent exam themes and ones curated by BMJ's editorial team." Pricing is tiered by subscription length from £59.99 for one month up to £199.99 for twelve months, across eleven duration options (vendor-reported, 20 July 2026). At platform level the vendor describes "Performance Feedback & Peer Comparison" showing "how likely you are to pass" with graphs that benchmark you against peers, plus "High Impact Questions" the vendor states are algorithmically ranked, and filtering by topic, type and difficulty. No explicit numeric predicted score or confidence-marking feature is documented—treat the pass-likelihood readout as a coarse indicator, not a calibrated prediction.
| Attribute (BMJ OnExamination, Endo & Diabetes SCE) | Vendor-reported, 20 July 2026 |
|---|---|
| Question count | 625, best-of-five |
| Price | £59.99 (1 month) to £199.99 (12 months); eleven tiers |
| Curriculum alignment | RCP curriculum for the SCE |
| Analytics | Peer comparison, pass-likelihood indicator, difficulty filter |
| "Adaptive" element | High Impact Questions ranking (vendor-described) |
| Mock exam | Yes, editorial and recent-theme mocks |
The SCE Endocrinology and Diabetes exam: the official anchor
The SCE is two papers of 100 best-of-five questions each (200 total), each three hours, on one day, computer-based at a test centre, one mark per correct answer, no negative marking. The exam runs once a year—the next sittings are 11 November 2026 and 10 November 2027—and UK trainees typically take it in their penultimate year, with no formal entry requirements. Only the blueprint changes by specialty. The Federation's published Endocrinology & Diabetes blueprint distributes the 200 questions roughly as follows (Federation blueprint; confirm the current weighting on the exam-body page):
| Blueprint domain | Approx. questions |
|---|---|
| Thyroid | 32 |
| Complications of diabetes | 29 |
| Diabetes mellitus (diagnosis, general management, delivery of care) | 25 |
| Other | 25 |
| Adrenal | 22 |
| Hypothalamus and pituitary | 22 |
| Gonads | 19 |
| Parathyroid / calcium | 16 |
| Diabetic emergencies and management during acute illness/surgery | 10 |
Define every metric before you trust it
First-attempt accuracy is your percentage on items you have never seen and is the only figure that behaves like the exam. Repeat accuracy is your percentage on items already attempted, inflated by recognition. Percentile / peer benchmark compares you with other BMJ OnExamination users—self-selected, not the SCE cohort. The pass-likelihood indicator ("how likely you are to pass") is a low-resolution readout on curated practice, not a validated prediction. Coverage is the share of the bank attempted, not the share of the blueprint mastered. Difficulty is typically a cohort statistic—how often others got it right—rather than an intrinsic property, and it drives the High Impact Questions ranking. Time per item is the least gameable and often the most useful number: below roughly 1.8 minutes on untimed practice means your accuracy is not exam-comparable.
Selection bias: why adaptive-style feeds distort the average
An algorithmic feed that surfaces "high impact" items, plus your own habit of revising weak topics, redoing flagged questions and reading explanations mid-attempt, means your practice is a sample of your attention, not of the blueprint. Weak domains get over-sampled—good for learning, misleading for measurement—and the raw percentage becomes incomparable with a mixed unseen block. A trainee who has drilled thyroid and diabetes to 90% but barely touched gonads, parathyroid or the pituitary can show a flattering average that a balanced 200-item mock would puncture. This is the core reason a Q-bank percentage is not an exam score: the denominator is curated.
Blueprint audit: your attempts versus the official weighting
Tally your attempts by domain and compare with the blueprint table above rather than trusting the home-screen average. The endocrinology-specific trap is that the two big diabetes-and-thyroid blocks (accounting for well over 80 of the 200 questions between diabetes, its complications and thyroid) are also the areas trainees find most comfortable, so they get over-practised while calcium/parathyroid, gonads and pituitary dynamic-function testing are under-practised despite reliable representation. Build the matrix explicitly using the completion-is-not-coverage method, and treat any domain below your target attempt count as unaudited regardless of your overall percentage.
The readiness test: five conditions for a credible signal
A number is a readiness signal only when the items are unseen, the block is timed at roughly 1.8 minutes per item, the content is mixed across the whole blueprint, you work with no assistance, and the sample is large—prefer 100-plus items and discount any single short run as noise. BMJ OnExamination's mock tests can satisfy the timed and mixed conditions within the endocrine blueprint; the caution is to run them as first attempts under exam discipline, not as reviewed sets, and to read the resulting first-attempt figure rather than the cumulative average on your home screen.
Override rules: force what the feed under-samples
Override the High-Impact-style feed on a schedule. Force the low-count domains that still appear every diet—parathyroid/calcium, gonads, the "diabetic emergencies/perioperative" slice—into each week. Force calculation and data-interpretation items: dynamic function tests (short Synacthen, oral glucose tolerance, water deprivation, insulin tolerance), thyroid function panels and calcium correction are the exam's quantitative spine and are easy for an engagement-optimised feed to under-sample. Force current-guideline therapeutics—GLP-1 receptor agonists, SGLT2 inhibitors, insulin regimens, levothyroxine and steroid replacement—read against the SmPC/eMC and current NICE and specialty-society (for example ABCD/Diabetes UK, Society for Endocrinology) guidance rather than ward habit. The feed optimises for engagement; the blueprint samples the whole curriculum.
Worked example: turning the dashboard into next week's quotas
Suppose the dashboard shows 84% overall: 300 attempts in diabetes/thyroid at 90%, 20 in parathyroid/calcium at 55%, 10 in gonads at 60%, none in pituitary dynamic testing, and a mean 62 seconds per item. The 84% is not a readiness signal—it is an average over an unbalanced, untimed, self-selected sample. Convert it into quotas, not a pass prediction:
| Signal | Reading | Next week's action |
|---|---|---|
| 90% diabetes/thyroid, very high volume | Over-sampled comfort zone | Cap at 20 timed unseen items |
| 55% parathyroid/calcium, low volume | Real weakness, under-tested | 40 items, spaced |
| No pituitary dynamic-testing items | Blind spot in a quantitative area | 30 items, unseen, timed |
| 62s per item | Pace inflated by untimed practice | All blocks to timed mode |
No pass percentage is produced, because none is defensible from this data.
Seven-day plan: BMJ for one job, iatroX for another
Give BMJ OnExamination the coverage-and-mock job it does well for this specialty, and give iatroX the unseen-measurement job. iatroX is not an endocrinology-specific SCE bank; it is the cross-specialty UK/MRCP-level knowledge and unseen-MCQ measurement layer that sits alongside a specialty bank—useful for the general-medicine reasoning and prescribing safety that underlie endocrine stems, and for spaced retrieval of your misses.
| Day | BMJ OnExamination (coverage + mock) | iatroX (unseen measurement) |
|---|---|---|
| Mon | 40 items: parathyroid/calcium + gonads | — |
| Tue | Error review; retrieval notes | 20 unseen mixed items, timed |
| Wed | 40 items: pituitary + adrenal | — |
| Thu | 40 items: diabetes complications | 20 unseen mixed items, timed |
| Fri | 40 items: dynamic function + calculations | — |
| Sat | 100-item timed mixed mock in BMJ | — |
| Sun | Spaced re-test of the week's misses | 30 unseen mixed items, timed |
No item is practised twice across platforms, keeping the measurement signal clean per the two-Q-bank rule.
Three mistakes this audit is designed to stop
First, reading the "how likely you are to pass" indicator as a pass prediction—it is a coarse signal on curated data. Second, letting the algorithmic feed and your own preferences over-sample diabetes and thyroid so the average rises while calcium, gonads and pituitary stay thin. Third, practising untimed with explanations open and then treating the resulting percentage as exam-comparable; it is not.
Decision checklist: continue, supplement, switch or stop
Continue with BMJ OnExamination if your blueprint-coverage matrix is filling and your first-attempt, timed accuracy is rising. Supplement with iatroX for unseen measurement and, if a domain stays thin, a second blueprinted endocrine bank—added carefully to avoid overlap. Switch primary banks only for a measurable, persistent coverage gap. Stop adding questions when timed, mixed, first-attempt accuracy plateaus and errors are careless rather than knowledge gaps, and move to exam-condition mocks. Every branch rests on a measurable gap, not novelty or sunk cost.
Bottom line
For SCE Endocrinology and Diabetes, BMJ OnExamination is a real and reasonable coverage-and-mock resource, and its analytics are worth using—provided you read first-attempt, timed, mixed figures and ignore the flattering cumulative average and the coarse pass-likelihood indicator. Audit your attempts against the blueprint, force the quantitative and low-count domains, and use iatroX to measure transfer on unseen items. Practice data is not readiness; disciplined reading of the dashboard turns it into something close.
FAQ
Is BMJ OnExamination enough for SCE Endocrinology and Diabetes on its own? It is a legitimate primary coverage bank for this specialty with 625 curriculum-aligned questions and mocks (vendor-reported, 20 July 2026), but "enough on its own" depends on disciplined use: without a blueprint audit and unseen, timed measurement, its dashboard can flatter you, so most candidates pair it with a source of unseen mixed practice.
Which SCE Endocrinology and Diabetes component does BMJ OnExamination not reproduce well? The exam-condition measurement of readiness: its cumulative average and pass-likelihood indicator are built on self-selected practice and do not reproduce the calibration you get from unseen, timed, mixed blocks, and quantitative dynamic-function interpretation can be under-sampled unless you force it.
How many BMJ OnExamination questions should I complete per day for SCE Endocrinology and Diabetes? For a trainee working clinically, around 40 items a day weighted toward the domains your coverage audit flags, plus a short timed unseen block, is sustainable and more useful than large untimed batches; the aim is spaced, reviewed coverage of the whole blueprint rather than a completion percentage.
When should I stop using BMJ OnExamination and move to mixed mocks? Move to full mixed mocks once your blueprint-coverage matrix is complete and your first-attempt, timed accuracy on unseen items has stabilised; run BMJ's own mock under exam discipline as a first attempt, and let that first-attempt figure—not the home-screen average—be the trigger.
How should I combine BMJ OnExamination with iatroX without duplicating practice? Use BMJ OnExamination for blueprinted coverage and its mocks, and reserve iatroX for unseen, timed, mixed measurement of transfer, keeping the banks separate so no item is practised twice and the iatroX figure remains an uncontaminated readiness signal.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 20 July 2026. Vendor figures (question counts, pricing tiers, feature and alignment descriptions) are vendor-reported and change without notice; verify the current position on onexamination.com before purchase. Endocrine prescribing facts should be taken from the SmPC/eMC and current NICE, CKS and specialty-society guidance. Disclosure: iatroX operates a competing UK question bank and clinical-knowledge platform; here its role is confined to jobs BMJ OnExamination does not claim—cross-specialty knowledge and unseen-MCQ measurement—and iatroX is not a specialty-specific SCE Endocrinology bank. Corrections are welcome via the feedback route on iatrox.com. References: Federation of Royal Colleges of Physicians (MRCP(UK)) SCE format and SCE Endocrinology & Diabetes blueprint; BMJ OnExamination product and features pages (onexamination.com); and the iatroX framework pieces on Q-bank percentage and blueprint-coverage matrices. Compare tools on the iatroX comparison hub.
Run a fresh, timed unseen block in iatroX and decide learn, retest, simulate or stop →