If a question-bank explanation tells you the "correct" blood-pressure threshold, first-line antibiotic or screening interval, that answer is often only correct in one country. This pillar gives you the VERIFY loop: a six-step, exam-agnostic routine for checking guideline-sensitive Q-bank explanations against the guideline your exam actually uses in the UK, US, Canada or Australia. Use it to build a reusable Guideline Ledger, catch mismatches before they cost marks, and know which single-best-answer key your blueprint expects. It is designed for candidates using a bank written for one market to sit an exam set in another.
Why the "correct answer" moves when you cross a border
Clinical guidelines are national. Exams test national practice. Question banks are written for one market but sold and used globally, and every explanation carries a publication date that quietly ages against a moving guideline. Put those four facts together and you get a predictable problem: the same clinical stem can have a different "single best answer" in London, Boston, Toronto and Melbourne, and your bank may be keying the wrong one for your exam.
This is not an abstract worry, and it is not about pass rates. It is about specific, high-frequency facts that marking schemes treat as deterministic. Consider a handful that recur across almost every knowledge exam:
- Hypertension thresholds. The 2017 ACC/AHA guideline redefined hypertension in the US at 130/80 mmHg. NICE in the UK still diagnoses stage 1 hypertension at a clinic reading of 140/90 mmHg (confirmed on ambulatory or home monitoring at 135/85 mmHg). Hypertension Canada anchors diagnosis on automated office readings with its own thresholds. A stem that asks "is this patient hypertensive?" therefore has three defensible answers depending on the flag on the paper.
- First-line antihypertensive. NICE runs an age- and ethnicity-directed algorithm: an ACE inhibitor or ARB for a patient under 55 without type 2 diabetes and not of Black African or African-Caribbean family origin, and a calcium-channel blocker for those aged 55 and over or of Black African or African-Caribbean origin. A US-authored explanation may key a thiazide-type diuretic, or treat any of four classes as acceptable, reflecting the ACC/AHA menu rather than NICE's directed choice.
- Cardiovascular risk and statins. The UK offers a statin at a QRISK3 ten-year risk of 10% or more. The US uses the Pooled Cohort Equations with a 7.5% threshold for discussion. Canadian practice has leaned on the Framingham Risk Score and Canadian Cardiovascular Society thresholds, and Australia uses its own absolute cardiovascular disease risk calculator. Feed the same patient into four different risk engines and you can get "treat" in one country and "borderline" in another with identical bloods.
- Screening intervals. Australia's National Cervical Screening Program uses a primary HPV test every five years from ages 25 to 74. The UK screens from 25 as well, but with historically shorter intervals that are now being extended towards five years for HPV-negative results. A bank written before that change may still key a three-year interval for a 30-year-old.
- Antibiotic first choice. For an uncomplicated lower urinary tract infection in a non-pregnant woman, NICE (NG109) keys nitrofurantoin first-line where renal function allows. US guidance from the IDSA lists nitrofurantoin, trimethoprim-sulfamethoxazole, fosfomycin or pivmecillinam as first-line options. Same patient, different keyed drug.
- Nomenclature. Adrenaline and epinephrine, paracetamol and acetaminophen, salbutamol and albuterol, lidocaine and lignocaine, furosemide and frusemide are the same molecules under different names. A candidate who does not recognise the pair can mis-read an otherwise easy item.
Underneath all of this sits a different medicines reference in each market. The UK's is the Summary of Product Characteristics (SmPC) via the electronic Medicines Compendium (eMC). Australia has the Australian Medicines Handbook and Therapeutic Guidelines; Canada has the Compendium of Pharmaceuticals and Specialties; the US relies on approved product labelling and clinical references such as UpToDate and Lexicomp. When an explanation cites a dose or an interaction, it is quietly citing one of these, and the number can differ.
The mechanism is always the same: a bank explanation is a compressed reflection of one guideline set, at one moment in time, for one country. VERIFY is the routine that decompresses it.
The VERIFY loop
Run any guideline-sensitive explanation through six steps.
- V — Vet the claim. Ask first whether the claim is even jurisdiction-sensitive. Thresholds, first-line drugs, screening ages and intervals, treatment targets, staging systems and terminology usually are. Mechanism, anatomy, physiology, pharmacodynamics and ECG interpretation usually are not. Only sensitive claims earn the rest of the loop.
- E — Establish the source. Identify which body the explanation is leaning on, even when it does not say. A 130/80 threshold signals ACC/AHA; a QRISK figure signals NICE; an "AOBP" reading signals Hypertension Canada. Name the implicit source.
- R — Relocate to your jurisdiction. Find the equivalent current guideline for the same question in the country whose exam you are sitting. This is the single step most candidates skip.
- I — Identify the delta. State the exact difference in one line: the number, the drug, the age, the interval. "US 130/80; UK clinic 140/90, ambulatory 135/85."
- F — Fix the currency. Confirm both guidelines are current versions and record the date you checked. Guidelines are revised; a "correct" answer from 2021 can be superseded within the same country.
- Y — Yield to the blueprint. Decide which answer your exam's blueprint actually rewards, and record it. If you are sitting PLAB 1 or the MRCGP AKT you yield to UK guidance even when your bank is American; if you are sitting the USMLE you yield to US guidance even when your instinct is British.
The output of VERIFY is not a vague sense of doubt. It is a row in a ledger.
The Guideline Ledger (reusable, embeddable)
This is the template. Copy it once and reuse it on every exam-specific study block; every child article in this framework can link back to this table rather than reproducing it.
| Claim under review | Source it reflects (as written) | My-jurisdiction equivalent | The delta | Currency checked | Answer my exam wants |
|---|---|---|---|---|---|
| Diagnose hypertension at 130/80 | ACC/AHA 2017 (US) | NICE NG136 (UK) | US 130/80; UK clinic 140/90, ABPM 135/85 | 19 Jul 2026 | UK: 140/90 |
| First-line: thiazide diuretic | ACC/AHA menu (US) | NICE age/ethnicity rule (UK) | UK <55, no T2DM: ACEi/ARB | 19 Jul 2026 | UK: ACEi/ARB |
| Screen cervix every 3 years | Older UK/US cytology era | Australia NCSP | AU: primary HPV, 5-yearly, 25–74 | 19 Jul 2026 | AU: 5 years |
| First-line UTI: TMP-SMX | IDSA (US) | NICE NG109 (UK) | UK: nitrofurantoin first | 19 Jul 2026 | UK: nitrofurantoin |
Four columns are the discipline; the last two are the decision. Keep a running ledger across a study block and a pattern emerges — you learn which topics in your bank are systematically "foreign", and you stop being surprised by them in the exam.
Worked example 1 — a written MCQ (MRCP(UK) Part 1 or PLAB 1)
A best-of-five item describes a 48-year-old man of white European family origin, no diabetes, with confirmed hypertension, and asks for the first-line agent. Your bank's explanation keys a thiazide-type diuretic and cites a US lineage. You are sitting PLAB 1, aligned to the GMC's MLA content map and UK practice, or MRCP(UK) Part 1 with its clinical pharmacology and therapeutics component.
Run VERIFY. Vet: first-line drug choice is guideline-sensitive — yes. Establish: the explanation reflects the ACC/AHA menu. Relocate: NICE NG136 governs your exam. Identify the delta: for a patient under 55 without type 2 diabetes and not of Black African or African-Caribbean origin, NICE directs an ACE inhibitor or ARB, not a thiazide. Fix currency: confirm you are reading the current NICE algorithm. Yield: your UK exam wants ramipril or an ARB. One ledger row, and a whole class of "wrong" bank answers is now predictable rather than confusing. The same discipline handles a PLAB 1 or UKMLA stem on statin thresholds, where the QRISK 10% cut-off is the keyed answer even if the explanation reaches for a US risk figure.
Worked example 2 — a structured-response item (MCCQE Part I or RACGP KFP)
Structured and short-menu formats expose jurisdiction even more sharply than best-of-five, because they reward the precise local answer rather than the closest plausible option. Take the MCCQE Part I, now a multiple-choice examination assessing readiness for supervised practice in Canada, or an RACGP Key Feature Problem, where each case is scored on a small number of decisive management steps.
A case gives a 55-year-old with a fixed lipid profile and asks whether to start pharmacotherapy. Vet: this is a risk-threshold decision — highly jurisdiction-sensitive. Establish: your bank explanation applies the US Pooled Cohort Equations at 7.5%. Relocate: for a Canadian exam you relocate to the Framingham-based Canadian Cardiovascular Society framework; for an Australian exam you relocate to the national absolute cardiovascular risk calculator and its thresholds. Identify the delta: the same patient can cross the treatment line under one engine and fall short under another. Fix currency: confirm which calculator version the current national guideline uses. Yield: the key feature the marker is looking for is "apply the correct national risk tool and act on its threshold", not "quote a US percentage". The structured format punishes the imported answer directly.
Worked example 3 — a clinical simulation (USMLE Step 3 CCS or MRCGP SCA)
Simulations test sequencing, and sequencing is where jurisdiction hides in plain sight. In the USMLE Step 3 computer-based case simulations, you enter orders in free text against US practice; the "next best step" for an outpatient UTI, a lipid decision or an immunisation is US-normed. In the MRCGP SCA, twelve simulated consultations are judged against UK general practice, where NICE and CKS define the acceptable management and the acceptable prescription.
Suppose you drill both on one platform. A bank explanation for a simulated respiratory case recommends a management path calibrated to one country. Vet, Establish, Relocate: confirm whether the ordering behaviour, the antibiotic and the follow-up interval belong to the jurisdiction you are being marked in. Identify and Fix: note where the imported path diverges — a different first-line antibiotic, a different safety-net interval, a different threshold to investigate. Yield: in the MRCGP SCA you are rewarded for the UK-appropriate, NICE-and-CKS-consistent plan, and the simulation format means an imported plan loses marks silently, without a red "incorrect" flag to warn you. Because iatroX is a question-bank and knowledge platform rather than a consultation simulator, treat the simulator as the arena and VERIFY as the way you sanity-check the underlying knowledge the arena assumes.
Failure modes and where not to use VERIFY
The loop is powerful, which makes it easy to over-apply. Watch for these.
- Do not verify jurisdiction-invariant content. The Krebs cycle, the brachial plexus, the mechanism of a loop diuretic and the ECG in hyperkalaemia do not change at a border. Running VERIFY on them wastes time and manufactures false doubt.
- Do not "correct" the bank towards your home country when you are sitting a foreign exam. An IMG using a US bank to prepare for PLAB should relocate to UK guidance; a UK graduate using a UK bank for the USMLE should relocate to US guidance. The point is to reconcile to your exam, not to assume British answers are universally right.
- Do not over-index on nomenclature at the expense of substance. Recognise adrenaline and epinephrine as the same drug, then move on.
- Beware in-country lag, not just cross-border difference. A bank explanation can be "the wrong answer" in its own country simply because a national guideline has been updated since it was written. The "F" step exists precisely for this.
- Beware house-style answers no guideline supports. Occasionally a bank asserts a threshold or a first choice that no current national body actually recommends. If you cannot relocate the claim to a real guideline, treat the explanation, not your knowledge, as suspect.
VERIFY is a scalpel for guideline-sensitive claims, not a blanket suspicion of every explanation.
The evidence hierarchy you are verifying against
When two sources disagree, rank them. From most to least authoritative for exam purposes:
- Your exam's official blueprint or content map. The GMC MLA content map was updated in January 2026 and applies from September 2026; the ABIM, MCC, AMC and RACGP publish their own. This defines what "correct" means for you.
- The current national clinical guideline for the jurisdiction of your exam — NICE, CKS and SIGN in the UK; ACC/AHA, IDSA, USPSTF and ACIP in the US; Diabetes Canada, Hypertension Canada and the Canadian Cardiovascular Society in Canada; the NHMRC, Therapeutic Guidelines and the RACGP in Australia.
- The national medicines reference for doses and interactions — the SmPC via the eMC in the UK, and each country's equivalent handbook or labelling.
- Peer-reviewed guideline summaries and review articles.
- The bank explanation.
- AI-generated commentary, which sits at the bottom until grounded and dated.
A bank explanation never outranks a current national guideline. Verifying is simply making the ladder explicit.
Do this in the next seven days
Pick one bank you actually use and one exam you are actually sitting. Answer twenty questions in the relevant systems. As you review, flag every guideline-sensitive claim — expect five to ten. Build the Guideline Ledger above and fill a row for each. Verify at least five against a primary source rather than the bank's word. Then, and this is the step that turns audit into revision, test the identified gaps on unseen items: run a fresh, timed block on the mismatched topics in a measurement bank and see whether you now select the answer your exam wants under time pressure. Compare banks deliberately with a comparison hub rather than by feel, and read your scores through the lens of why your Q-bank percentage is not your exam score.
An iatroX worked example (vendor-neutral)
Here is how the loop looks with iatroX in the stack, described plainly and without any proprietary-algorithm claim. When a claim looks jurisdiction-sensitive, you can put the question to Ask iatroX, which is built to be citation-first: it surfaces the source behind an answer so you can see which guideline it rests on and confirm the jurisdiction, rather than trusting an unsourced assertion. You then relocate to your national guideline yourself. To test the gap, you run unseen, timed blocks on iatroX Boards; because the UK-core banks are free, iatroX works naturally as the "measurement bank" in the two-Q-bank rule — one bank to drill, one to measure on genuinely unseen items. iatroX is one worked example of the framework, not the framework itself; the VERIFY loop is deliberately platform-neutral so it holds whichever banks you own.
Frequently asked questions
Does this framework work for every medical exam? It works for any exam that tests national clinical practice, which is almost all of them — the UK's MRCP, PLAB, UKMLA, MRCGP and PSA; the US USMLE, ABIM and ABEM; Canada's MCCQE and CCFP; Australia's AMC and RACGP. The proportion of guideline-sensitive items rises with how "clinical" the paper is: a pharmacology-and-therapeutics paper or a general-practice paper is dense with them, while a pure basic-science paper has fewer. Where an item is jurisdiction-invariant, VERIFY simply returns "not sensitive" at the first step and you move on, so the loop never does harm — it only ever adds signal.
How often should the framework should be updated? Update a specific ledger row whenever the national guideline behind it is revised, and re-verify any claim if the guideline body has issued a new version since your bank explanation was written. In practice that means a light re-check each revision cycle and a prompt one whenever a major guideline changes — a new hypertension, lipid or screening recommendation should trigger an immediate pass over the affected rows. Anchor to structural dates too: the GMC MLA content map updated in January 2026 and applying from September 2026 is exactly the kind of event that should send you back through your ledger.
Which metrics are valid across different Q-banks? The comparable metric is not raw cumulative percentage, which is bank-specific and inflated by repeats; it is your first-pass accuracy on unseen, timed, blueprint-weighted items, and the direction of that number over time. For this framework specifically, add a "jurisdiction-mismatch rate" — the share of guideline-sensitive items where your bank keyed a foreign answer — because it tells you how much your bank is pulling you away from your exam's expected key. Percentiles and predicted scores do not travel across banks and should never be compared between products.
How should AI-generated feedback be verified? Treat AI feedback as the least authoritative source until proven otherwise, because general models tend to default to US guidance and can present it with unwarranted confidence. Force the model to state the jurisdiction and cite the specific guideline, then check that the guideline is real, current and correct for your exam. This is the same discipline set out in the guide to calibrating AI-graded feedback before you trust the score and in how to audit an AI medical exam tutor for grounding, answer leakage, hallucinations and retention. Grounding is everything: an ungrounded answer cannot be relocated, and an answer you cannot relocate you cannot trust.
How does iatroX implement the framework? iatroX contributes two of the pieces the loop needs and nothing it cannot support. Ask iatroX is citation-first, so it is built to expose the source behind a claim, which is what makes the "Establish" and "Relocate" steps quick. iatroX Boards provides unseen, timed UK-core blocks so you can test a mismatched topic once you have verified it, acting as the measurement bank in a two-bank setup. iatroX does not claim to tell you which national guideline governs your exam — that judgement stays with you and your blueprint — and it makes no proprietary-algorithm claim about doing so.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026. Guideline thresholds and exam formats change; verify any current figure against the primary source before you rely on it, and treat vendor-reported claims as vendor-reported and dated. Disclosure: iatroX operates a question bank that competes with the products a reader might audit with this framework; the framework is deliberately platform-neutral, and iatroX's role here is confined to the citation-first verification and unseen-measurement jobs described, not to deciding which national guideline applies to your exam. Corrections are welcome via the feedback route on iatrox.com.
Version and update history: v1.0, published and clinician-reviewed 19 July 2026 (Dr Kolawole Tytler). Review cycle: at least annually and on any major guideline change.
References: NICE and CKS (nice.org.uk); the American College of Cardiology and American Heart Association 2017 high blood pressure guideline; IDSA (idsociety.org); Diabetes Canada, Hypertension Canada and the Canadian Cardiovascular Society; the National Health and Medical Research Council and Therapeutic Guidelines (Australia); Cancer Council Australia and the National Cervical Screening Program; the GMC MLA content map (gmc-uk.org); and the exam bodies for MRCP(UK), PLAB, MCCQE, RACGP and the USMLE. Internal: why your Q-bank percentage is not your exam score and the two-Q-bank rule.