This workflow is for paediatric trainees preparing for MRCPCH Applied Knowledge in Practice — the second written stage — who want to use BMJ OnExamination without letting a rising overall percentage hide unattempted syllabus areas. It addresses the two same-day single-best-answer papers of sixty questions each. The principal limitation to state up front: BMJ OnExamination is SBA-only and runs no proprietary adaptive engine, so it does not reproduce the MRCPCH Clinical examination, and its "adaptive" feel is really you steering the feed.
What BMJ OnExamination offers for MRCPCH AKP right now
BMJ OnExamination lists a dedicated MRCPCH: AKP product. As of 19 July 2026 it advertises a vendor-reported 330+ single-best-answer, best-of-many and clinical questions (the product's own metadata references a larger combined pool, so treat the headline as a floor and verify the live count on onexamination.com). Access is subscription-based, running from roughly £29.99 for one month to about £139.99 for twelve months in monthly increments (vendor-reported, 19 July 2026), with a free trial of ten questions a day. The feature set includes answer explanations, mock tests curated by the BMJ editorial team, difficulty-level selection, a revision plan that targets weaker areas, peer comparison, group leaderboards and a mobile app with offline access.
Two honest qualifications shape everything below. First, the product is written-SBA only: it maps to the AKP paper and does not contain any clinical-OSCE practice. Second, BMJ OnExamination does not advertise a proprietary adaptive-difficulty algorithm. The personalisation here is candidate-directed — you choose the domain and difficulty and recap your misses. That is a genuine strength for the plan below, because it means no black box decides what you see; but it also means the discipline of setting your own coverage floors falls to you. No proprietary-algorithm claims are made or needed.
MRCPCH Applied Knowledge in Practice: the official shape
AKP is the second written stage of the MRCPCH, sat after Foundation of Practice and Theory and Science and before the MRCPCH Clinical examination. It comprises 120 single-best-answer questions delivered as two separate exams of sixty questions each, both sat on the same day, with two hours and thirty minutes allowed per exam. Every item is single best answer worth one mark, there is no negative marking, and delivery is computer-based at an exam centre or online. That pacing works out at roughly two and a half minutes per question — more generous than the first-stage papers, because AKP rewards applied synthesis: data and image interpretation, growth-chart and investigation reasoning, prescribing decisions and age-specific management rather than isolated recall.
Keep the official requirement separate from third-party claims. The RCPCH publishes the syllabus and a set of official sample questions; those define the blueprint and the house style. A commercial bank's tags, difficulty labels and "high-yield" framing are useful, but they are the vendor's interpretation, not the exam standard. When the two disagree, the RCPCH material wins. For any prescribing item, the UK medicines reference is the Summary of Product Characteristics via the electronic medicines compendium (eMC), read alongside NICE and CKS guidance — not a memorised figure.
Baseline week: measure before you personalise
Before you let difficulty selection or your own topic filters shape the feed, sit a small, blueprint-stratified, unseen sample under timed conditions. Around forty to fifty questions, spread deliberately across the major paediatric domains — neonatology, cardiology, respiratory, gastroenterology, neurology, infectious disease, endocrinology, nephrology, safeguarding and community, and the science-into-practice areas — gives you a first-attempt profile by domain. Record first-attempt accuracy and pace, never cumulative percentage. This baseline is the reference against which everything else is judged, and it stops you optimising a number that mostly reflects your strong topics.
First pass: set domain floors the score cannot hide behind
The failure mode this article is named after is real. If you let a comfortable domain drive practice, the overall percentage climbs while whole syllabus areas go under-attempted. Set explicit domain floors — a minimum number of unseen questions you will attempt in each weak or low-volume area every week — and hold them regardless of how the headline number moves. A rising overall score with an empty safeguarding or neurodisability column is not readiness; it is an averaging artefact. Our guide on why completion is not coverage sets out how to build the matrix that these floors sit inside.
Error taxonomy: name the failure before you fix it
Sort every miss into one category, because the fix differs by type:
- Knowledge gap — you did not know the fact or mechanism.
- Misread stem — you knew it but missed a qualifier, an age, a unit or a negative.
- Premature closure — you locked onto a diagnosis or plan before reading all five options.
- Guideline error — you applied an outdated or non-UK standard.
- Calculation error — you slipped on a weight-based dose, a fluid rate or a centile.
- Time-pressure error — you would have got it with another twenty seconds.
A month of tagged misses tells you whether you have a content problem or a test-craft problem, and the two demand completely different interventions. In AKP specifically, calculation and image-interpretation errors are worth tracking separately, because they cluster and respond to drilling rather than reading.
Review interval: not everything deserves an immediate repeat
Match the response to the error type. A knowledge gap earns a short primary-source read plus a new transfer question, not a re-sit of the same item. A misread stem or a time-pressure error earns timed mixed practice, because the deficit is process, not content. A guideline error earns a check against the current NICE, CKS or SmPC/eMC source and a dated note. A calculation error earns three fresh variants of the same calculation. Reserve immediate repeats for almost nothing — repeating an item you have just seen mostly measures recognition, which is precisely what inflates a bank percentage without moving your real score.
Mixed-block switch: when to stop filtering
Topic-filtered practice is right early, when you are building coverage in weak domains. Switch toward timed, random, full-length blocks once every domain sits above its floor and your first-attempt accuracy in your weakest domain has stabilised across at least two separate sittings. The trigger is objective — floors met, weakest-domain first-attempt stable, pace within target — not a feeling that you have "done enough questions". Random sixty-item blocks then rehearse the interleaving and endurance the two same-day papers actually demand.
Exit criteria: what "ready" looks like, and what it does not
You are ready to stop grinding the bank when five things are simultaneously true: every blueprint domain is above its coverage floor; first-attempt accuracy is stable across mixed unseen blocks rather than concentrated in favourite topics; your pace holds near two and a half minutes per item under timed conditions; your misses on retest are new rather than repeats of the same misconceptions; and your performance holds up on the official RCPCH sample material, which is the calibration gold standard the finite bank cannot replace. Notice that "finished the bank" is absent from that list — bank completion is an activity metric, not a readiness signal.
A seven-day plan: BMJ OnExamination for one job, iatroX for transfer
Give BMJ OnExamination one clear job — structured domain practice with its filters and mock tests — and give iatroX a different job: unseen, timed transfer measurement on items it has never seen you attempt. No proprietary-algorithm claims are needed for this to work.
- Day 1 — baseline. Sit a mixed, timed, unseen block of forty to fifty items in the iatroX UK bank with no assistance. Record first-attempt performance by domain.
- Days 2–4 — targeted first pass. In BMJ OnExamination, work your weakest three domains to their floors, tagging every miss by error type and writing a one-line corrected rule for each.
- Day 5 — calibration. Attempt the official RCPCH AKP sample questions under time and compare your reasoning with the model answers.
- Day 6 — fresh transfer. Sit a new timed iatroX block on different topics and compare your first-attempt trend, not your cumulative percentage — your Q-bank percentage is not your exam score.
- Day 7 — review and plan. Space this week's misses, retest last week's, and set next week's domain floors.
Continue, supplement, switch or stop
Decide on measurable gaps, not novelty or sunk cost. Continue with BMJ OnExamination if your domains are moving toward their floors and your tagged misses are shrinking. Supplement with unseen iatroX measurement if your bank percentage feels high but you have no timed, unseen evidence that the knowledge transfers, applying the two-Q-bank rule so you never run the same item twice. Switch or add a second bank if a whole domain is thin in BMJ's coverage and you cannot get above its floor. Stop topic-drilling a domain once you can answer fresh mixed items in it under time without assistance.
The bottom line is straightforward. BMJ OnExamination is a strong, well-explained SBA bank that is worth including in an AKP revision stack, but it does not adapt for you and it does not measure you against unseen items. You supply the discipline — domain floors, error tags, official-material calibration — and you keep a separate, unseen block for honest measurement. Used that way, it earns its place; used as a percentage to maximise, it flatters you.
FAQ
Is BMJ OnExamination enough for MRCPCH Applied Knowledge in Practice on its own? It is a strong single bank but not a complete preparation on its own. AKP rewards applied synthesis across images, data and age-specific management, so you also need the official RCPCH sample questions to calibrate house style and a separate pool of unseen, timed items to prove the knowledge transfers rather than being recognised. Treat BMJ OnExamination as the domain-practice engine, not the whole plan, and verify its current question count on onexamination.com before you rely on it.
Which MRCPCH Applied Knowledge in Practice component does BMJ OnExamination not reproduce well? It does not reproduce the MRCPCH Clinical examination at all, and you should not expect it to, because it is a written-SBA product. Even within the written domain, a bank you have partly worked through cannot reproduce the fully unseen, timed pressure of two sixty-question papers on the same day, where recognition of previously seen stems quietly inflates your accuracy. Keep an unseen block in reserve for that exact rehearsal.
How many BMJ OnExamination questions should I complete per day for MRCPCH Applied Knowledge in Practice? There is no magic number, and daily volume matters less than the mix, but a sustainable target for a trainee working around clinical shifts is roughly twenty to forty questions a day, done in timed blocks with every miss tagged and reviewed the same day. Two carefully reviewed blocks beat a hundred rushed items, because unreviewed volume trains speed without accuracy and burns through the finite bank you will later want for mock conditions.
When should I stop using BMJ OnExamination and move to mixed mocks? Move to timed mixed blocks and full mocks once every blueprint domain is above its coverage floor and your first-attempt accuracy in your weakest domain has stabilised across at least two separate sittings. At that point, single-topic practice gives diminishing returns, and the pacing, endurance and interleaving data from mixed sixty-item blocks — sat under the real two-and-a-half-minute pace — become the information you actually need to judge readiness.
How should I combine BMJ OnExamination with iatroX without duplicating practice? Apply the two-Q-bank rule: give BMJ OnExamination the job of structured domain practice and give iatroX the separate job of unseen, timed measurement, and never run the same item through both. Duplicating items turns measurement into recognition and inflates your apparent readiness, so keep one bank for learning and one, unseen, for honest transfer scoring.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026; product figures described as vendor-reported were taken from public product listings on that date and change without notice — verify the current count, price and feature set on the vendor's own page. Disclosure: iatroX operates a competing UK question bank, so this article confines the iatroX role to unseen transfer measurement — a job the audited SBA bank does not claim to replace. Corrections are welcome via the feedback route on iatrox.com. References: RCPCH examinations (rcpch.ac.uk/education-careers/examinations) and the official AKP sample questions; the BMJ OnExamination MRCPCH: AKP product page (onexamination.com); the iatroX comparison hub and the iatroX UK bank.
Run a fresh timed MRCPCH Applied Knowledge in Practice block in iatroX →