PassMAP is a PARA revision platform built around analytics: performance summaries by system, tracking of repeated errors and weak areas, spaced repetition and progress-over-time views. For a candidate who wants their practice steered toward weak systems and their misses automatically re-surfaced, that feedback loop is genuinely useful. The principal limitation is interpretive: a feed that deliberately over-samples your weak areas produces dashboard numbers that are not comparable with a clean, mixed, unseen block — and PassMAP does not publish a question count or price, so its coverage cannot be verified from the outside. This audit shows how to read its analytics without being misled by them.
This is a narrow child article; for the full coverage framework see the PARA content-gap checklist, and for comparisons the iatroX comparison hub.
Current-state box (vendor-reported, last checked 20 July 2026)
All detail is vendor-reported from passmap.co.uk on the date shown; several key figures are not published.
| Attribute | Vendor-reported detail |
|---|---|
| Question bank size | Not published — the site shows engagement metrics ("25,000+ question interactions in the last 28 days"; "1,000+ submissions in the last 28 days"), which are usage, not inventory. Verify the true bank size. |
| Analytics | Performance summaries by system and question set; early identification of weak areas and repeated errors; progress over time |
| Adaptive/retrieval | "Spaced repetition built into daily revision, automatically"; targeted revision by weak system |
| Mocks | "Timed tests and PARA-style mock exams" (count unspecified — verify) |
| PARA components supported | Written/SBA knowledge only — no OSCE |
| Price / access period | Not published on the page — verify current price and subscription length |
Two honesty flags up front. First, PassMAP does not publish its question count or price, so you cannot verify coverage or value before subscribing — treat both as unknowns to confirm. Second, the "adaptive" element is vendor-described as spaced repetition and weak-area targeting; this audit makes no proprietary-algorithm claims and takes those features at face value pending verification.
Exam anchor
PARA (RCP for the GMC; PA regulation began 13 December 2024) is a KBA of 200 single-best-answer questions (four one-hour papers of 50) plus a 16-station OSCE (14 scored + 2 rest; pass the standard and ≥9 stations), both mapped to the GMC PARA content map (four domains; 18 clinical areas). PassMAP addresses the knowledge component. Its analytics describe your practice; the blueprint defines the exam. When those two disagree, the blueprint wins.
Every metric, defined
Analytics only help if you know what each number does and does not mean:
- First-attempt accuracy — your score on items the first time you see them. The most honest knowledge signal PassMAP can give; insist on seeing it separately from re-review.
- Repeat accuracy — your score on items you have seen before. Rises naturally with exposure and measures memory of the item, not mastery of the topic. Do not mistake it for progress.
- Percentile — your rank against other users. Depends entirely on who else is in the pool and how much they have practised; informative only loosely. (Verify whether PassMAP shows one.)
- Predicted score — any single "you will pass" number. Treat with the most caution; it rests on assumptions you cannot see. This audit presents no pass predictions. (Verify whether PassMAP displays a prediction; if it does, do not act on it as fact.)
- Coverage — the share of the blueprint you have sampled. Only trustworthy if measured against the official 18-area content map, not the platform's own category list.
- Difficulty — the platform's rating of item hardness. Useful for pacing, but vendor-defined; verify how it is derived.
- Time per item — your pace. Directly relevant, because the real KBA runs at roughly one minute per question.
Selection bias: why the feed distorts the average
Here is the core interpretive trap. A platform that "targets weak systems" and re-surfaces your errors is, by design, feeding you a harder-than-average, weak-area-weighted diet. That is good for learning and bad for self-assessment: your rolling accuracy on that feed will read lower and noisier than your true blueprint-wide standing, because you are being fed your worst material on purpose. The opposite error also occurs — if you let the spaced-repetition feed drift toward items you now know, accuracy looks flattering. Either way, the home-screen average is not comparable with a clean, mixed, unseen block that samples the whole blueprint at exam proportions. Never read a targeted feed's percentage as your exam readiness.
Blueprint audit
Do this rather than trust the dashboard: export or tally your attempted-question distribution and lay it against the official content map's 18 areas. Because PassMAP steers toward weak systems, your distribution will likely be skewed toward a few areas you have been drilling and thin in areas the feed has not surfaced — including the low-volume ones (ophthalmology, ENT, haematology, palliative care) that stay hidden. The dashboard's "coverage" figure counts what you have done; the blueprint audit reveals what you have missed. The gap between them is your real content-gap, and it is invisible from the home screen.
Readiness test conditions
A credible readiness signal requires five conditions simultaneously, and a targeted analytics feed satisfies none of them by default:
- Unseen — items you have not practised before, so the score is knowledge, not recall of the item.
- Timed — at the real pace (~1 minute/item across four one-hour papers).
- Mixed — sampling the whole blueprint, not one steered topic.
- No assistance — no hints, no spaced-repetition scaffolding, no immediate feedback mid-block.
- Sufficient sample — enough items (a full paper's worth or more) that the number is not noise.
If any condition is missing, the number is a practice metric, not a readiness signal. This is where a separate, clean, unseen bank earns its place.
Algorithm override rules
Do not let the feed decide everything. Manually force the material an automated weak-area loop under-surfaces:
- Low-volume domains — schedule deliberate blocks in ophthalmology, ENT, haematology and palliative care; the feed will rarely prioritise them.
- Images, ECGs and radiographs — force interpretation practice; algorithms optimise for text-item accuracy and can starve you of image reps.
- Ethics, consent, capacity and statistics — override to include these; they are easy for a feed to under-weight and heavily examined in professional-values items.
- Calculations — schedule them explicitly; a spaced feed may not surface enough.
The principle: use the algorithm for efficiency, but audit it against the blueprint and override wherever the two diverge.
Worked dashboard example
Invented analytics, turned into next week's quotas — no pass prediction:
PassMAP shows: overall 68% (mixed exposure); cardiovascular 76% first-attempt (620 attempts); respiratory 72% (410); acute care 61% (300); mental health 55% (90); haematology 51% (40); ophthalmology 44% (15); ethics untested under time; time/item creeping up in the last block.
Read it: the high-volume rows are fine; the risk sits in the low-volume, low-attempt rows (mental health, haematology, ophthalmology) and in the untested ethics and rising pace. Next week's quotas, driven by the gaps, not the average: three unseen blocks in the weak low-volume areas; one timed image/ECG set; one ethics-and-statistics block under time; and one full mixed, unseen, timed paper to get a clean readiness number the feed cannot give. Notice there is no "you will score X" — only measured gaps converted into targeted work.
Worked example: a seven-day plan
One job for PassMAP (targeted learning and spaced re-surfacing), one for iatroX (clean unseen measurement). No proprietary-algorithm claims.
- Mon: PassMAP — follow the targeted feed into your weakest system; let spaced repetition re-surface prior misses.
- Tue: iatroX — a clean, unseen, timed mixed block on the PARA bank to get a readiness number the feed cannot give.
- Wed: PassMAP — override the feed to force a low-volume area (e.g. ophthalmology) and an image set.
- Thu: iatroX — unseen block in that same low-volume area to check transfer; log high-confidence errors.
- Fri: PassMAP — ethics/statistics and calculations blocks, timed, to override feed under-weighting.
- Sat: iatroX — full unseen, timed, mixed paper for a clean, blueprint-wide signal.
- Sun: Review — run the blueprint audit; set next week's overrides and quotas from the measured gaps.
Decision checklist
- Continue PassMAP while its targeted feed and spaced repetition are demonstrably raising your first-attempt accuracy in weak systems.
- Supplement with a clean, unseen bank for readiness measurement — essential, because a targeted feed cannot produce a credible readiness number — and add OSCE practice separately.
- Switch only for a measurable, persistent problem (unverifiable coverage that leaves a documented blind spot, or analytics you cannot reconcile with the blueprint) — not novelty.
- Stop adding new items when unseen, mixed, timed accuracy is stable across the blueprint; move to consolidation and simulation.
FAQ
Is PassMAP enough for Physician Associate Registration Assessment on its own? No — it trains only the knowledge component and offers no OSCE, so it cannot cover half of PARA, and its bank size and price are not published, so you cannot even verify the knowledge-layer coverage from outside. Its analytics and spaced repetition are a useful learning engine, but a targeted feed cannot double as a readiness measurement, which needs clean unseen blocks. Treat it as a weak-area training tool, with unseen measurement and the OSCE handled separately.
Which Physician Associate Registration Assessment component does PassMAP not reproduce well? The OSCE, completely — PassMAP is a written-SBA analytics platform with no station, examination, communication or procedural practice. Beyond that, the thing its analytics cannot reproduce is a clean readiness signal: because the feed deliberately targets weak areas and re-surfaces misses, its dashboard percentages are not comparable with an unseen, mixed, timed block. So it under-serves both the OSCE and unbiased self-assessment.
How many PassMAP questions should I complete per day for Physician Associate Registration Assessment? Because the bank size and difficulty mix are not published, anchor to function rather than a count: enough targeted items each day to work through your weak systems with full review of misses — commonly 40–60 — plus regular clean unseen blocks elsewhere for measurement. Let the blueprint audit, not the app's daily streak, set the number, weighting under-sampled domains. Guard first-attempt accuracy as the real metric, and remember the visible "interactions" figures are engagement, not your progress. Price and access period are not published — verify before subscribing.
When should I stop using PassMAP and move to mixed mocks? Move to clean mixed mocks as soon as you need a readiness number, which a targeted feed cannot give you honestly — in practice, run unseen mixed blocks in parallel from early on rather than waiting. Stop leaning on PassMAP's average as your headline metric once your weak-system first-attempt accuracy has stabilised; keep the platform for continued weak-area drilling while your readiness signal comes from unseen mocks. The trigger is stable weak-area accuracy plus the need for a blueprint-wide, unbiased score.
How should I combine PassMAP with iatroX without duplicating practice? Split the roles cleanly: PassMAP for targeted, analytics-driven learning and spaced re-surfacing of misses; iatroX for clean, unseen measurement via the PARA bank, so your readiness number is not distorted by a weak-area feed. Do not repeat items across the two — iatroX's value is that its questions are unseen relative to your PassMAP practice, giving an uncontaminated signal. Alternate "train on PassMAP, measure on iatroX," following the two-Q-bank rule.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 20 July 2026. PassMAP does not publish a question count or price on its site; the figures shown there are engagement metrics, and all platform detail is vendor-reported on the date shown — verify the live bank size, price and access period before relying on them. This audit makes no proprietary-algorithm claims and presents no pass prediction. Disclosure: iatroX operates a UK PARA question bank that competes with PassMAP on the knowledge layer; its role here is confined to clean unseen measurement — the job a targeted analytics feed structurally cannot do for itself — and it does not reproduce the OSCE. Corrections are welcome via the feedback route on iatrox.com.
References: passmap.co.uk (product page, vendor-reported); RCP PARA information and OSCE blueprint (rcp.ac.uk); GMC PARA content map (gmc-uk.org); PARA content-gap checklist; how to build a blueprint-coverage matrix; Your Q-Bank Percentage Is Not Your Exam Score; iatroX comparison hub.