This workflow is for final-year UK medical students and F1s using BMJ OnExamination to prepare for the Prescribing Safety Assessment (PSA). It addresses the whole paper — all eight question types — but with one governing constraint: the bank is small, so the plan is built to spend items deliberately and keep a slice unseen for a genuine readiness check. The principal limitation is not quality; it is quantity. A finite pool cannot be both your teaching set and your mock, so you have to decide in advance which questions you burn learning and which you protect for measurement.
What BMJ OnExamination offers for the PSA right now
BMJ OnExamination does sell a dedicated PSA product — this is not a coverage gap. The figures below are vendor-reported and were last checked on 19 July 2026; confirm the live numbers on the product page before purchase.
| Feature | Vendor-reported detail (19 July 2026) |
|---|---|
| Question pool | Around 110 PSA practice questions (a deliberately focused set, not a mega-bank) |
| Access options | 3 months £6.99, 6 months £9.99, 9 months £12.99, 12 months £14.99 |
| Free access | 10 free questions per day before purchase |
| Personalisation | "Revision plans" that target weaknesses, plus feedback and peer comparison |
| Mocks | Timed mocks curated around recent exam themes by BMJ's editorial doctors |
| Extras | Competitive leaderboard, offline app download |
The load-bearing figure is that pool of roughly 110 questions. That is enough to teach the patterns of every question type, but it is not enough to spend carelessly. Note too that the "revision plan" personalisation is a weakness-targeting and peer-comparison feature; this article makes no claim about any underlying difficulty-adapting algorithm, and you should verify exactly what the plan does on the product page rather than assume it behaves like a large adaptive bank.
The exam anchor: the PSA blueprint
The PSA is a two-hour, computer-based assessment of 60 items worth 80 marks in total, standard-set by a modified Angoff method and valid for two years once passed. It is sat by final-year students and some F1s. The 60 items are drawn from eight question types — Prescribing, Prescription Review, Planning Management, Communicating Information, Calculation Skills, Adverse Drug Reactions, Drug Monitoring and Data Interpretation — spread across seven clinical domains: Medicine, Surgery, Elderly Care, Paediatrics, Psychiatry, Obstetrics and Gynaecology, and General Practice.
Three points matter for a workflow. First, the eight types are not equally weighted — the prescribing-writing items carry a large share of the marks — so verify the current mark allocation on the official blueprint and weight your practice towards the high-tariff types. Second, the PSA is an open-reference exam: you have an online medicines reference available during the assessment, so the skill being tested is safe, efficient use of a reference under time pressure, not blind recall. For revision, treat the Summary of Product Characteristics via the electronic medicines compendium (eMC), alongside NICE and CKS, as your authoritative source for doses, interactions and monitoring. Third, the official practice papers are the calibration gold standard — a small, finite resource you should protect for near-exam rehearsal rather than burn early.
Baseline week: sample before you personalise
Before you let any revision plan shape your feed, take a small blueprint-stratified baseline: roughly two questions from each of the eight types, unseen, timed, unassisted. You are not trying to pass; you are locating your worst two or three question types and your calculation speed. Record first-attempt accuracy by type. This baseline decides where your finite BMJ OnExamination items are best spent, and it gives you a reference point the personalised feed will otherwise obscure once it starts steering you towards your errors.
First pass: set domain floors
On your first pass through the bank, work by question type, not at random, and set a floor for each of the eight so a rising overall score cannot hide an untouched type. A simple rule: no type may sit at zero attempted while any other type is fully drilled. Because prescribing and prescription-review items carry the most marks, give them the largest share of your first pass, but do not let calculation skills, drug monitoring or adverse drug reactions fall below their floor. The point of floors is that BMJ OnExamination's weakness-targeting will happily push you deeper into two or three types; floors stop that convenience from becoming a blind spot on exam day.
An error taxonomy for prescribing
When you get a PSA item wrong, the correction depends entirely on why. Label every error as one of:
- Knowledge gap — you did not know the drug, dose, interaction or monitoring rule.
- Misread stem — you missed the allergy, the renal function, the age or the "which is most appropriate" framing.
- Premature closure — you fixed on the first plausible drug and stopped checking contraindications.
- Guideline error — you followed an outdated or wrong recommendation; correct it against NICE, CKS or the SmPC via the eMC.
- Calculation error — an arithmetic or unit slip in a dose, rate or conversion.
- Time-pressure error — you knew it but rushed, or ran out of time before reaching it.
These are not interchangeable, and they do not deserve the same fix.
Review interval: what each error earns
- Knowledge gap → a short, targeted source read (SmPC/eMC, NICE or CKS), then a spaced re-test in a few days on a different item testing the same rule — not an immediate repeat of the same question.
- Misread stem → no new content; a process fix. Build a stem-checklist habit (allergies, renal and hepatic function, age, pregnancy, current drugs) and rehearse it on your next block.
- Premature closure → deliberate practice generating a differential of options before choosing, on transfer questions rather than the failed item.
- Guideline error → correct the source, then re-test on an unseen item after a spacing interval.
- Calculation error → drill a dedicated set of calculations to rebuild speed and accuracy; this is the type most improved by volume.
- Time-pressure error → not a content problem at all; fix it in timed mixed blocks, below.
The discipline is to avoid immediately re-answering a question you just saw. In a 110-item pool, an instant repeat converts a scarce unseen item into a memorised one and inflates your score without improving prescribing.
The mixed-block switch
Move from type-filtered practice to timed random blocks when two conditions are met: every question type is above its floor, and your first-attempt accuracy on your two weakest types is climbing rather than static. The trigger is coverage plus trajectory, not a target percentage. In the final week to ten days, most of your practice should be full-length, timed, mixed and unassisted, because the PSA tests judgement across types under a clock, and that is precisely what type-filtered drilling does not rehearse.
Exit criteria
You are ready to stop grinding questions — not when the bank is 100% complete — when all of the following hold:
- Coverage floor met — every question type and every clinical domain attempted, none starved.
- Stable first-attempt performance — your unseen first-attempt accuracy is steady across types, not propped up by repeats.
- Pacing — you finish timed blocks inside two minutes per item on average with time to spare for calculations.
- Retention — you can re-derive a rule days later without the explanation open.
- Official-material calibration — you have sat at least one official PSA practice paper late, under exam conditions, and scored comfortably above the threshold.
Bank completion is not on that list. Completion of a small pool is easy and tells you little; calibrated performance on unseen and official material is the real exit signal.
A seven-day plan for students and foundation doctors
BMJ OnExamination does one job here — teaching the eight question types with feedback — while an unseen bank measures transfer. No proprietary-algorithm claims are made.
| Day | BMJ OnExamination job | Unseen-measurement job (iatroX) |
|---|---|---|
| Mon | First pass: prescribing and prescription review to floor | — |
| Tue | Planning management + communicating information | 15 unseen mixed items, timed |
| Wed | Calculation skills drill + drug monitoring | — |
| Thu | Adverse drug reactions + data interpretation | — |
| Fri | Review error log; re-test misses as transfer items | 15 unseen mixed items, timed |
| Sat | One full timed mock (protect unseen items) | — |
| Sun | — | Official practice paper OR fresh mixed PSA block; log by type |
Keep the measurement blocks on a bank you do not revise from — iatroX runs free UK-core PSA content for exactly this — so BMJ OnExamination's finite pool stays intact for its own mock and your readiness reading is never a memory test. This is the two-Q-bank rule: learn on one, measure on the other, so calibration stays clean.
Decision checklist: continue, supplement, switch or stop
- Continue the first pass while any question type sits below its floor or your weakest types are still improving.
- Supplement with a second, unseen bank as soon as you realise your 110-item pool is being spent on both teaching and testing — split the jobs before you contaminate your mock.
- Switch your primary resource only if you find recurring factual errors or the explanations are not moving your unseen scores after a fair trial — not on novelty or a fancier dashboard elsewhere. Compare options on the iatroX comparison hub.
- Stop buying questions when your exit criteria are met and an official practice paper confirms it; further items add little, and rest plus one final timed rehearsal serves you better.
Frequently asked questions
Is BMJ OnExamination enough for the PSA on its own? It can be a sufficient teaching resource, because its curated, editorially-reviewed set covers all eight question types at a realistic standard. The honest limitation is the pool size — around 110 questions, vendor-reported on 19 July 2026 — which is too small to serve as both your learning bank and your mock. Pair it with the official practice papers and a separate unseen bank for measurement, and it is enough; used alone for everything, you will run out of clean items.
Which PSA component does BMJ OnExamination not reproduce well? No bank fully reproduces the calculation-under-time-pressure element or the experience of using the in-exam medicines reference efficiently, and a small pool cannot give you the sheer calculation volume some candidates need. It also cannot substitute for the official practice papers as a calibration standard. Use the bank to learn the patterns, then rehearse pacing and reference-use in full timed mocks and on the official material.
How many BMJ OnExamination questions should I complete per day for the PSA? Because the pool is finite, quality beats speed: 15 to 25 questions a day worked thoroughly, with a source read for every knowledge gap, will exhaust the bank sensibly over a fortnight without wasting items. Resist the urge to blitz all 110 in a weekend — you would convert your only mock into memorised content and lose your readiness check.
When should I stop using BMJ OnExamination and move to mixed mocks? Switch the balance towards timed mixed mocks once every question type is above its floor and your weakest types are improving on first attempts. In the last week to ten days, prioritise full-length timed papers and the official practice material over new topic-filtered questions, since pacing and cross-type judgement are what remain to be trained.
How should I combine BMJ OnExamination with iatroX without duplicating practice? Give each a single job. Learn the eight question types on BMJ OnExamination, then measure transfer on fresh, timed, mixed PSA blocks in iatroX that you never revise from. Because the pools differ, you avoid re-testing memorised items, protect BMJ OnExamination's small bank for its own mock, and keep an uncontaminated readiness signal.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026; vendor figures (question count, prices, feature names) are vendor-reported on that date and change between cycles — verify the current numbers on the BMJ OnExamination product page. UK medicines information in this article refers to the Summary of Product Characteristics via the electronic medicines compendium (eMC), alongside NICE and CKS. Disclosure: iatroX operates a competing PSA question bank; this article confines iatroX's role to unseen readiness measurement, the job BMJ OnExamination's finite pool cannot safely perform for you. Corrections are welcome via the feedback route on iatrox.com.
References: Prescribing Safety Assessment — official format and practice papers (prescribingsafetyassessment.ac.uk); BMJ OnExamination PSA product page (onexamination.com, vendor-reported); iatroX PSA bank; Your Q-Bank Percentage Is Not Your Exam Score; the two-Q-bank rule.