This is the implementation companion to our PassMedicine UKMLA analytics audit. That article explains what the dashboard measures; this one is the week-by-week workflow that keeps a self-directed feed from quietly skipping the parts of the GMC content map you find least comfortable — which, with the updated map applying from September 2026, is exactly where marks are won and lost.
What you are working with
As of 19 July 2026, PassMedicine offers low-cost fixed-term UKMLA/MLA access with a large single-best-answer bank, timed and revision modes, and peer-comparison analytics; no AI tutor is advertised. Verify counts and price on the product page. The exam is the MLA's applied knowledge test — the MSC AKT for UK students, PLAB 1 for IMGs — built on the GMC content map (updated January 2026, applying from September 2026; 212 presentations and 430 conditions), plus a separate clinical and professional skills assessment that no Q-bank touches.
The one-page plan
| Phase | Timing | Focus | Exit trigger |
|---|---|---|---|
| Baseline | Week 1 | Unseen map-stratified sample | ~100 items logged across map areas |
| First pass | Weeks 2–6 | Map-floored revision + professional content | Every map area above floor |
| Consolidation | Weeks 7–8 | Timed mixed blocks + error transfer | First-attempt accuracy stable |
| Simulation | Weeks 9–10 | Full timed mixed mocks | Pacing on target; go/no-go |
Baseline before personalising
Spend week one on a map-stratified unseen sample — a spread across the content map's areas of clinical practice, answered honestly before filtered revision biases the picture. This gives you a true starting distribution, which matters more for the MLA than for most exams because the content universe is explicitly published: you can audit against it from day one.
Map-aligned floors, including professional content
Set a minimum question count per map area proportional to the blueprint, and enforce it against the temptation to ride a rising average over unattempted ground. Then set an explicit floor for the content that specialty-filtered practice reliably starves: ethics, law, consent, capacity, safeguarding and prescribing safety. These carry real weight in the MLA design and are the first thing to vanish from a clinically-driven feed. A non-zero weekly floor on professional content is the single highest-value override in this workflow.
Error taxonomy and transfer review
Sort every miss into the six types — knowledge gap, misread stem, premature closure, guideline error, calculation error, time-pressure error — and match the fix to the type rather than defaulting to "revise the topic". For the MLA specifically, watch for a seventh pattern: jurisdiction errors, where an internationally-trained candidate applies a home-country default. Those need UK-guideline calibration, not more general revision. Default to transfer questions over immediate repeats, so you prove the concept rather than memorise the item.
The switch to mixed blocks
Move from filtered practice to timed random blocks when every map-area floor (including professional content) is met, first-attempt accuracy on unseen material has held for two weeks, and pace is within the paper's budget. Then invert the ratio to mostly-mixed, using targeted practice only to patch leaks the blocks expose. Mixed blocks are your measurement instrument; using them before coverage is complete just produces a demoralising number with no diagnostic use.
A seven-day pattern — including for IMGs
Monday: 50 PassMedicine questions in two floor-deficient map areas, explanations read. Tuesday: 30 questions plus focused UK-convention review — NICE-first sequences, UK prescribing safety, safeguarding thresholds (the highest-yield territory for PLAB-route candidates). Wednesday: a timed, unseen 50-question mixed block in iatroX's free UKMLA bank for an uncontaminated signal. Thursday: light error review. Friday: 40 questions on professional-content and low-volume areas. Saturday: a full timed simulation, reviewed same day. Sunday: rest. PassMedicine supplies volume and explanations; iatroX supplies unseen adaptive measurement.
A worked planning example
Eight weeks from your AKT, suppose your position is: overall 66%; cardiology and respiratory strong and heavily practised; psychiatry and paediatrics under 60% on thin samples; ethics and law barely touched; dermatology untouched entirely; timed pace 82 seconds per item. The dashboard says "66%, keep going". The workflow produces a specific fortnight instead.
Psychiatry and paediatrics take the largest clinical quotas, floors set, explanations read. Dermatology and the ethics/law strand — both effectively unexamined — get their first proper sessions, because low-volume and professional content vanish from a self-directed feed exactly when they matter. Pace gets three timed blocks aimed at bringing 82 seconds under the paper's one-minute budget. And one unseen timed mixed block from outside PassMedicine gives you a readiness reading your own filtering did not shape. No overall pass prediction appears, because the 66% was concealing the two liabilities — professional content and pace — most likely to decide a borderline result.
The international-graduate layer
For PLAB-route candidates the workflow carries an extra, high-value job that UK graduates can largely skip: UK-convention calibration. Your clinical knowledge may be strong, but the AKT rewards UK-specific defaults — NICE-first management sequences, UK prescribing safety conventions, UK safeguarding thresholds, and the organisational frameworks of NHS general practice — and these are precisely where a confident, well-trained candidate loses marks to a home-country reflex. Build a standing habit: whenever a management answer surprises you, ask not "did I know the medicine?" but "did I apply the UK default?", and log the divergence. This is usually a larger source of marginal marks for IMGs than any amount of additional clinical volume, and no clinical Q-bank's analytics will flag it for you — it only surfaces when you make jurisdiction an explicit review category.
A concrete instance makes the point. A candidate trained in a system where broad-spectrum antibiotics are first-line for a common infection answers the corresponding AKT question with their home default, gets it wrong, and — crucially — feels no uncertainty doing so, because the answer is correct where they trained. The clinical knowledge is intact; the jurisdiction is not. Only a review habit that asks "is this the UK first-line choice?" catches it, and only a source-grounded check against NICE or CKS confirms the correction. Build that habit into the workflow from week one and the IMG-specific error rate falls faster than any clinical-volume increase could move it.
Continue, supplement, switch or stop
Continue while floors fill and unseen performance climbs. Supplement when bank numbers outpace unseen performance. Switch only for a measurable gap (our ranked UKMLA comparison helps). Stop accumulating volume in the final fortnight; simulate.
Three mistakes this workflow is designed to stop
First, letting the overall percentage stand in for coverage. The MLA publishes its content universe, so "how much of the map have I met?" is answerable — and a rising average over a narrow slice of it is the commonest false comfort in finals-year revision. Set map-area floors and audit against them. Second, starving the professional content. Ethics, law, consent, capacity, safeguarding and prescribing safety carry real weight and are the first thing a clinically-driven feed drops; a non-zero weekly floor on them is the highest-value override here. Third — for international graduates especially — never making jurisdiction an explicit review category. The answer that feels right may be a home-country default, and the only way to catch it is to ask, on every management miss, "did I apply the UK convention?" rather than "did I know the medicine?" A candidate who treats those three as standing rules rarely meets an unpleasant surprise on the map's neglected corners.
Frequently asked questions
Is PassMedicine enough for UKMLA on its own? It can carry AKT volume, but you still need a content-map audit, unseen timed blocks and CPSA preparation the bank does not provide.
Which UKMLA component does PassMedicine not reproduce well? The map's enforced breadth — professional, ethical and prescribing-safety content and low-volume specialties — and the exam's unseen timed delivery.
How many PassMedicine questions should I complete per day for UKMLA? 40–60 map-directed questions on study days, with two protected timed mixed blocks weekly; composition beats raw volume.
When should I stop using PassMedicine and move to mixed mocks? When every map area is above floor, accuracy is stable and pace is on target — the final two to three weeks, given to simulation.
How should I combine PassMedicine with iatroX without duplicating practice? PassMedicine for chosen-area drilling; iatroX (free for UKMLA) for unseen adaptive measurement across the map and Socratic repair on misses.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026; MLA content-map dates are per the GMC; PassMedicine details vendor-published — verify before purchase and confirm your AKT sitting and map version. Disclosure: iatroX offers a free competing UKMLA bank. Corrections via the feedback route on iatrox.com. References: GMC MLA content map (gmc-uk.org); MSC AKT handbook (medschools.ac.uk); PassMedicine product pages; related reading: the PassMedicine UKMLA analytics audit and the UKMLA content map in full.