This workflow is for UK final-year medical students and foundation doctors using Geeky Medics for the Prescribing Safety Assessment (PSA) who keep re-attempting the same comfortable cardiology and infection cases and want a balanced, blueprint-led set instead. One honest clarification first: Geeky Medics does not sell a product it calls a "PSA Simulator." It sells a PSA Question Bank with three timed mock papers, and the "simulation" is that timed mock mode. The principal limitation is that a recognition bank under-trains blank-field prescription construction.
What Geeky Medics offers for the PSA right now
The current-state box below is what a candidate needs before building a plan. Vendor-reported figures are labelled and dated; verify them on the product page before you buy, because counts and prices change.
| Feature | What we found (vendor-reported, 20 July 2026) |
|---|---|
| Product name | "PSA Question Bank" plus "Mock PSA exams" — not marketed as a "simulator" |
| Question count | 600+ on the overview page; 800+ on the purchase page — described as "not AI-generated" |
| Mock papers | Three timed mock exams "mapped to the format and curriculum of the PSA", across all eight question types |
| Question types covered | All eight: Prescribing, Prescription Review, Planning Management, Providing Information, Calculation Skills, Adverse Drug Reactions, Drug Monitoring, Data Interpretation |
| Adaptive / AI features | None advertised; explicitly "not AI-generated" |
| Price | £19.99, annual subscription with auto-renewal |
| Extras | Specialty and category filtering, performance analytics, a notebook of 350+ topic summaries, recorded webinars |
The two different question counts on two Geeky Medics pages are a good reason to treat all vendor figures as provisional and check the live number yourself. The absence of adaptive delivery is not a criticism — it simply means the balancing has to be done by you, using the filters, rather than by an algorithm.
Exam anchor: what the official blueprint actually requires
The PSA is 60 items in two hours, 200 marks, standard-set by modified Angoff and delivered by the British Pharmacological Society and MSC Assessment. Prescribing carries 80 marks — 40% of the total — in just eight items; the remaining seven question types share the other 120 marks. Items span seven clinical domains: Medicine, Surgery, Elderly Care, Paediatrics, Psychiatry, Obstetrics & Gynaecology and General Practice. The official practice papers on the PSA website are the calibration standard; Geeky Medics' claim to map its mocks "to the format and curriculum of the PSA" is a vendor claim you should sense-check against those official papers, not a substitute for them.
Build a representative case matrix
Before you answer anything, decide what a balanced set looks like, then use the specialty and category filters to assemble it. Spread your set across five axes so you are not just repeating familiar scenarios:
- Blueprint domain: at least one Prescribing item in each of the seven clinical domains.
- Acuity: mix acute prescribing (sepsis, acute kidney injury, DKA) with chronic and preventive prescribing.
- Patient age: include paediatric weight-based dosing and elderly renal-adjusted dosing, not just fit adults.
- Complexity: include polypharmacy and interaction-heavy cases, not only single-drug decisions.
- Communication challenge: include Providing Information items where the task is explaining a medicine safely.
Write the target counts down before you start. A set that is 60% cardiology and respiratory is the comfortable-scenario trap; the fix is a quota, not good intentions.
Record a clean baseline
Do one full timed mock — one of the three Geeky Medics papers — without pausing, without restarting, and without reading the rubric or explanations first. Preserve that attempt as your baseline. The point is to capture how you perform under exam conditions before the platform's feedback and your own re-reading contaminate the result. Note your completion (did you reach item 60), your per-section timing, and your raw accuracy by question type.
Score twice and record disagreements
Score the mock first the way the platform scores it. Then score it again yourself against the official PSA domains and the criteria the official practice papers use — especially for the Prescribing items, where "safe as written, no omission" is the standard, not "the right drug". Record every disagreement between the two scores. Those disagreements are the most valuable output of the whole exercise: they are the places where an automated or recognition-style mark and the real exam standard diverge, and they tend to cluster in prescription construction and dose exactness. If you are relying on any automated feedback for constructed answers, calibrate it first using this method.
Convert feedback into two behaviours, not a list
After each case, resist the urge to write a long improvement list you will never read. Convert the feedback into exactly two observable behaviours for the next case. For example: "specify a review date on every anticoagulant prescription" and "check renal function before dosing any renally-cleared drug". Two behaviours per case, logged with an error code, accumulate into a targeted profile far faster than transcribing explanations. Do not copy the whole explanation into your notes — that is re-reading, not retrieval.
Use deliberate variation
When you have found a weakness, do not re-attempt the same item until you remember its answer. Take the same clinical principle and apply it to a different patient: change the agenda, add a comorbidity, alter the renal function, or add time pressure. Geeky Medics' category filters let you pull a different item that tests the same principle. The aim is transfer — can you apply the rule to an unseen case — not recall of one specific question.
Exit standard
You are ready to move on from Geeky Medics as your main tool when three things are true at once: your score is stable across the mocks and unseen items you have not memorised; your own official-criteria scoring and the platform's scoring largely agree, so your self-assessment is calibrated; and you have no recurrent safety-critical omission in prescription construction. A high percentage on re-seen questions with none of these three is not readiness.
Worked example: a seven-day plan
This gives Geeky Medics one defined job — high-volume recognition practice and timed mocks — and uses iatroX for unseen transfer measurement. No claim is made about proprietary algorithms; this is ordinary spaced practice.
- Day 1: Baseline. One full timed Geeky Medics mock, no pauses. Score twice, log disagreements.
- Day 2: Prescribing-only sets across three under-practised domains (say Paediatrics, O&G, Psychiatry). Two behaviours per case.
- Day 3: Calculation Skills and Drug Monitoring sets; check every figure against the SmPC/eMC.
- Day 4: Transfer day. Attempt a small set of unseen iatroX PSA items testing the same principles you missed on Days 2–3. Compare the error type, not the wording.
- Day 5: Prescription-construction reps — ten blank-field prescriptions from stems, marked against official criteria or by a prescriber.
- Day 6: Second full timed Geeky Medics mock. Compare completion and per-section timing with Day 1.
- Day 7: Review the week's error log; convert the recurring codes into next week's quotas.
Decision checklist: continue, supplement, switch or stop
- Continue with Geeky Medics as your volume engine if your unseen score is rising and your coverage matrix still has empty cells to fill.
- Supplement with dedicated construction practice and a second unseen bank once recognition accuracy is high but your official-criteria and platform scores disagree — that gap is construction, which the bank under-trains.
- Switch your emphasis to timed full papers when domain floors are met and the marginal new question stops changing your error profile.
- Stop buying more resources when new material stops teaching you anything; the fix is then repetition of weak behaviours and calibration, not novelty.
Frequently asked questions
Is Geeky Medics PSA Simulator enough for Prescribing Safety Assessment on its own? Geeky Medics is a strong, low-cost recognition bank with three timed mocks, and it is a reasonable single primary resource — but no recognition bank alone is enough, because the PSA also tests blank-field prescription construction, formulary navigation and pacing that a bank under-trains. Pair it with the official practice papers for calibration and with deliberate construction practice marked to the official criteria. Treat the £19.99 annual price (vendor-reported, 20 July 2026) as buying volume, not a complete preparation.
Which Prescribing Safety Assessment component does Geeky Medics PSA Simulator not reproduce well? The Prescribing question type is the least well reproduced by any recognition-oriented bank, because the real exam gives you a blank prescription and marks whether your entry is complete and safe as written, whereas a bank typically presents structured options. Geeky Medics' mocks cover all eight question types, but the highest-stakes skill — constructing an unambiguous prescription with correct dose, route and frequency from nothing — needs blank-field practice and a prescriber's eye, not more multiple-choice reps.
How many unseen Geeky Medics cases or stations should I preserve for final Prescribing Safety Assessment calibration? Preserve at least one full mock you have never attempted for the final week, and keep a pool of unseen single items in your weakest two domains, so your last measurement is genuinely unseen and timed rather than a re-run of memorised questions. Because Geeky Medics offers three mocks, a workable rule is to use two during preparation and protect the third, unopened, for a final timed calibration alongside an official practice paper.
When should I stop using Geeky Medics and move to mixed mocks? Move to mixed timed mocks once you have met a floor in every clinical domain and your single-topic accuracy has plateaued — even if you have not finished the bank. Continuing topic-filtered blocks past that point trains recognition in isolation, whereas the exam demands unsignposted, mixed, timed performance. The trigger is coverage and a plateau, not completion of every item.
How should I combine Geeky Medics with iatroX without duplicating practice? Give each tool one job and never answer the same item twice across them. Use Geeky Medics for volume, filtered topic blocks and its timed mocks; use iatroX only for unseen, timed transfer measurement — a fresh block on the principle you just missed, so you are testing whether the learning transferred rather than repeating a competitor's question. Compare the error type across the two, not the wording, and log the divergence. This is the two-Q-bank rule applied to the PSA.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 20 July 2026. Geeky Medics figures — the question count (600+/800+, "not AI-generated"), the three mock papers, and the £19.99 annual auto-renewing price — are vendor-reported from the Geeky Medics PSA product and purchase pages on 20 July 2026 and may change; verify them before purchase. The PSA format is taken from the official blueprint and practice papers. Medicines facts should be checked against the SmPC/eMC, with NICE and CKS for practice guidance.
Disclosure: iatroX operates a competing PSA question bank, so it is a competitor to Geeky Medics. We have confined iatroX's role here to unseen, timed transfer measurement — a job the Geeky Medics bank does not claim — and have been positive about Geeky Medics where warranted. Corrections are welcome via the feedback route on iatrox.com.
References: Geeky Medics PSA Question Bank and Mock PSA exams pages; PSA blueprint and practice papers (prescribingsafetyassessment.ac.uk and the BPS Assessment learner portal); SmPC/eMC and NICE/CKS for medicines content. Internal: the iatroX PSA bank, the comparison hub, and Your Q-Bank Percentage Is Not Your Exam Score.
Try an unseen Prescribing Safety Assessment case in iatroX →