This adaptive workflow is for emergency medicine trainees using Rosh Review (now part of Blueprint Prep) while preparing for the Royal College of Physicians and Surgeons of Canada (RCPSC) certification in Emergency Medicine. It addresses the written multiple-choice component only. The principal limitation, stated plainly: Rosh's EM bank is calibrated to the US ABEM Qualifying Exam and the US EM Model, not the Royal College — and its adaptive feed, left unmanaged, can quietly starve low-volume Canadian blueprint domains while your overall score climbs.
The calibration gap: an ABEM bank against a Royal College exam
Rosh Review's emergency medicine bank is built for the American Board of Emergency Medicine and structured to the US EM Model. The clinical reasoning transfers; the specifics — guideline sources, drug naming, medico-legal and system context, and above all the blueprint weighting — are American. The Royal College examines to its own Objectives of Training in Emergency Medicine, and it pairs the written component with a separate Applied examination Rosh does not touch. So there are two gaps to manage, not one: a jurisdiction gap (US content you must localise) and an adaptivity gap (an algorithm optimising for a US-shaped feed). Name both before you start.
What Rosh Review / Blueprint Prep offers for RCPSC EM right now
Vendor-reported, last checked 19 July 2026. Rosh Review is a Blueprint Prep brand; verify counts and prices on roshreview.com.
| Item | Rosh Review EM Initial Certification (vendor-reported, 19 July 2026) |
|---|---|
| Questions | ~3,000 (Basic) to ~5,000 (Standard/Premium) EM Qualifying-Exam items |
| Price (USD, 1-yr) | Basic ~599; Standard ~699; Premium ~798 |
| Adaptive/analytics | Performance dashboard with vendor-reported projected scores / "pass probability"; illustrated explanations; Premium adds a Mock Qualifying Exam |
| Calibrated to | The US ABEM Qualifying Exam / US EM Model |
| RCPSC components | Written-MCQ knowledge layer only; no Applied/oral rehearsal |
Treat "pass probability" as a vendor-reported figure calibrated to the US exam — not a prediction for the Royal College written, still less the Applied.
The RCPSC Emergency Medicine exam anchor
RCPSC certification in Emergency Medicine has two separate parts: a computer-based written multiple-choice component and a separate Applied examination (oral/OSCE-style), at the same test centre. The Royal College does not publish an accessible page with exact question counts, sessions or timing — verify the current structure and question count on royalcollege.ca. The blueprint is the Royal College's, not the US EM Model, and it includes lower-volume domains that a US-tuned adaptive feed will under-serve if you let it. The Applied component sits outside any MCQ bank entirely.
Baseline week: a blueprint-stratified unseen sample before you personalise
Before you let the platform adapt to you, take a short, blueprint-stratified sample of unseen questions across every Royal College EM domain — including the low-volume ones. The reason is specific to adaptive banks: once the algorithm starts personalising, it optimises against your recent behaviour, and a domain you never see cannot register as a gap. A stratified baseline, taken cold, gives you the domain map the algorithm will not — a map you then defend. Record it as a coverage baseline, not a score.
First pass: set domain floors the algorithm cannot hide
The core discipline of adaptive revision is the domain floor: a minimum number of attempted, reviewed questions in every blueprint domain, set by you, that the algorithm is not allowed to skip. Without floors, an adaptive feed can push your overall percentage up while a low-volume domain — toxicology, environmental emergencies, an uncommon paediatric presentation — goes almost untouched, because the algorithm sees little marginal gain there. A rising overall score then hides an unattempted syllabus area. Floors convert "the algorithm decides" into "the algorithm optimises within a coverage constraint I set." Check floors weekly; a domain below floor gets manual, topic-filtered attention regardless of what the feed suggests.
An error taxonomy for emergency medicine
Adaptive review is only as good as your classification of misses. Separate them:
- knowledge gaps (you did not know it);
- misread stems (you knew it, read it wrong under time);
- premature closure (anchored before the stem finished);
- guideline errors (right approach, wrong or US-specific guideline);
- calculation errors (dose, rate, score);
- time-pressure errors (right with time, wrong at pace).
Each demands a different fix, and an adaptive algorithm cannot tell them apart — it only sees right or wrong. You supply the taxonomy; it makes your review, not the feed, the engine of improvement.
Review interval: transfer question, spacing, source read — or nothing
Not every miss deserves the same response, and immediately re-drilling the same item is usually the weakest option, because it trains recognition. Decide by error type:
- knowledge gap — a short source read (a current Canadian or international EM guideline), then a transfer question days later;
- premature closure or misread — timed mixed practice, not more facts;
- guideline or localisation error — switch the reference to Canadian guidance and re-test the principle;
- calculation error — drill that calculation to fluency;
- a one-off slip — spaced review, or nothing.
Reserve immediate repeats for genuine fluency drills; everything else is better served by a spaced, unseen transfer question.
Mixed-block switch
Reduce topic-filtered practice and increase timed random blocks once every domain has cleared its floor and your first-attempt performance is holding. Objective triggers: all floors met; first-attempt accuracy on unseen mixed blocks stable across two sittings; pacing within target. Until floors are met, keep some topic-filtered work for the domains below floor; after, the centre of gravity should shift decisively to timed, mixed, unseen blocks that mimic exam conditions.
Exit criteria
Stop adding new Rosh work when, together: the coverage floor is met in every Royal College domain; first-attempt performance on unseen mixed blocks is stable; pacing is right; retention holds on spaced re-tests; and your performance calibrates against any official Royal College material you can obtain. Bank completion and vendor "pass probability" are not on this list. The Applied component, separately, must be readied through case-based oral/OSCE practice Rosh does not provide.
A seven-day plan for a trainee working clinically
Rosh for one job (adaptive first-pass learning with floor management and coded review) and iatroX for unseen transfer measurement. No proprietary-algorithm claims; ordinary spaced retrieval.
- Monday (30 min): check the domain floors; any domain below floor gets a 15-item topic-filtered Rosh block. Code the misses.
- Tuesday (30 min): source reads for the top two miss-codes; localise to Canadian or international EM guidance.
- Wednesday (25 min): an iatroX transfer set — unseen items on the same principles, timed.
- Thursday (20 min): spaced review of resolved misses only.
- Friday (40 min): a fresh, timed, mixed iatroX block across all EM domains — the week's clean measurement.
- Saturday (45 min): Applied practice — verbalised resuscitation and case management with a peer or supervisor. No MCQs.
- Sunday: rest, or a short floor review.
Rosh items are never re-shown in iatroX; iatroX only measures on unseen questions. One bank learns, one measures — the two-Q-bank pattern.
Decision checklist: continue, supplement, switch or stop
- Continue Rosh if domain floors are rising and your coded review is changing behaviour.
- Supplement (iatroX, Canadian or international EM guidelines) if errors are transfer or localisation errors, or a domain sits stubbornly below floor.
- Switch the job to oral/OSCE case practice for anything the Applied exam tests — Rosh cannot reach it.
- Stop new questions once coverage floors, unseen first-attempt performance, pacing and retention are stable; protect a mock and rest.
Decide on measurable gaps — a domain floor, an error trend, weeks to exam — not on a rising percentage or a vendor pass-probability number.
Bottom line
Rosh Review / Blueprint Prep is a large, well-illustrated EM bank — a strong written-knowledge option for RCPSC EM provided you remember it is an ABEM product you localise, and that its adaptivity optimises for a US-shaped feed. Set domain floors so no low-volume Royal College domain hides behind a rising score, classify your errors yourself, measure on unseen items, and rehearse the Applied component where a bank cannot. Run it that way and the algorithm works for your coverage instead of against it.
Frequently asked questions
Is Rosh Review and Blueprint Prep enough for RCPSC Emergency Medicine on its own? No. It is a strong written-knowledge bank, but it is calibrated to the US ABEM Qualifying Exam and the US EM Model, so it needs localising to Canadian practice, and it does not address the RCPSC Applied (oral/OSCE-style) examination. It can serve the written component's knowledge layer; the Royal College blueprint's specifics and the oral both sit outside it.
Which RCPSC Emergency Medicine component does Rosh Review and Blueprint Prep not reproduce well? The Applied examination. A single-best-answer bank cannot assess verbalised resuscitation management, structured reasoning under examiner questioning or the OSCE-style stations; those need case-based oral and simulation practice with a supervisor. Confine Rosh to the written-knowledge layer.
How many Rosh Review and Blueprint Prep questions should I complete per day for RCPSC Emergency Medicine? There is no correct fixed number, and with a 3,000–5,000 item bank (vendor-reported) volume matters less than coverage and review depth. A workable pattern is a small daily block driven by your domain floors — feeding under-covered domains first — with every miss coded and actioned. Chasing a high daily count while low-volume domains stay below floor is exactly the failure mode this workflow exists to prevent.
When should I stop using Rosh Review and Blueprint Prep and move to mixed mocks? When every Royal College domain has cleared its floor and first-attempt performance on unseen items is stable across two sittings — not when the bank is "finished" or when the vendor pass-probability looks reassuring. Reserve unseen, timed mixed blocks and any mock for a genuine readiness check near the exam.
How should I combine Rosh Review and Blueprint Prep with iatroX without duplicating practice? Give them separate jobs and never share items. Use Rosh for adaptive first-pass learning and coded review; keep iatroX as the clean, unseen bank for transfer questions and timed mixed measurement. An item seen in Rosh must not reappear as "measurement" in iatroX. That separation — one bank to learn, one to measure — is the two-Q-bank rule, and it is also what prevents duplicated practice.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026. Vendor-reported figures (Rosh/Blueprint question counts, prices, "pass probability", features) are labelled as such and change without notice — confirm them on roshreview.com. Disclosure: iatroX operates a competing question bank; here it is confined to the unseen written-MCQ measurement job Rosh does not claim for the Royal College, and it does not reproduce the RCPSC Applied examination. Corrections via the feedback route on iatrox.com.
References: Royal College of Physicians and Surgeons of Canada — Emergency Medicine examination format and Objectives of Training (royalcollege.ca); Rosh Review / Blueprint Prep EM Initial Certification pages (roshreview.com); the iatroX RCPSC hub (https://www.iatrox.com/canada); "Your Q-Bank Percentage Is Not Your Exam Score" (https://www.iatrox.com/blog/qbank-percentage-not-your-exam-score); the completion-is-not-coverage blueprint-matrix pillar (https://www.iatrox.com/blog/question-bank-completion-is-not-coverage-how-to-build-a-blueprint-coverage-matrix-for-any-medical-exam); the iatroX comparison hub (https://www.iatrox.com/compare).
Run a fresh, timed RCPSC Emergency Medicine block in iatroX →