How Our Questions Are Written and Reviewed

Our standard for a question

We write to the structure of the sample questions in the ICBSE Guide to the Intercollegiate MRCS examination. In practice that means:

• A short clinical scenario. Every clause in the stem either sets the scene or changes the answer. Anatomy, physiology and pathology are asked through a patient, not as a bare fact.
• Same-category options. Five nerves, five vessels, five causes of a raised potassium, five next steps, so the answer cannot be found by eliminating options of the wrong type.
• One defensible best answer. Several options may be plausible; exactly one is best for this patient under current UK practice, and the explanation says why the others are not.
• A clear lead-in. Which structure is most likely injured, which mechanism explains the finding, which is the most likely diagnosis, which is the most appropriate next step.
• Blueprint mapping. Every question is mapped to a topic of the ICBSE Part A blueprint and the ISCP Core Surgical Training curriculum, so you can see coverage and performance by domain and topic on the syllabus, and the bank is weighted in the blueprint’s proportions.

Difficulty is calibrated to the level the ICBSE describes for Part A: the knowledge a surgical trainee should have achieved two to three years after qualification. A question is hard because its best options are close — separated by a relation, a threshold or a detail in the stem — not because the stem is obscure or the fact preclinical. Difficulty labels are set on that basis and kept honest, because the adaptive engine depends on them.

Explanations

Each explanation justifies the best answer, explains why each distractor is wrong or less good for this patient, and ends with a reference. Where the question turns on a number — a dose, a vertebral level, a threshold in a guideline — the explanation gives it and says where it comes from.

Where the facts come from

Explanations are referenced to current UK guidance and standard texts, including:

• NICE guidelines, for the management questions of Paper 2
• Advanced Trauma Life Support (ATLS), 10th edition, for trauma
• Royal College of Surgeons of England guidance on surgical practice
• The British National Formulary (BNF) for drugs and doses
• Last’s Anatomy: Regional and Applied and Gray’s Anatomy for Students for anatomy
• Ganong’s Review of Medical Physiology and Guyton and Hall Textbook of Medical Physiology for physiology
• Robbins and Cotran Pathologic Basis of Disease for pathology
• Bailey and Love’s Short Practice of Surgery for the principles of surgery in general

We write to the current version of UK guidance, because the examination is set to present-day UK practice. Where a textbook and current UK guidance differ, the guidance is the answer; where two standard texts differ on a point of anatomy or physiology, the explanation says so and names the one we follow.

Review and sign-off

Questions are drafted and typeset to our house style, then checked by the editorial team against the reference each explanation cites: the best answer, the discriminator and any figures quoted must match current UK guidance or the named text. Specialist review is ongoing rather than complete. Where a domain has been through it we say so; where it has not, we do not imply otherwise, and we do not name reviewers who have not reviewed. Reviewers can change any part of an item — stem, options, best answer, explanation or difficulty — and their edits replace the original.

The free sample on free questions is drawn from the same bank and held to the same standard, so it is representative of what you get when you subscribe.

How we use AI

We use AI tools openly, for two jobs: drafting questions and explanations to a written house style, and keeping a bank of this size consistent in format, terminology and referencing across 13 domains. AI is never the final word. It does not decide what is correct; the cited reference and the human checker do.

AI drafts can contain errors, as can human ones. That is why every item is checked against its reference, why review status is published rather than assumed, and why the correction process below exists.

Reporting an error

If you think a question is wrong, out of date or ambiguous, flag it in the app so you can find it again and tell us through the contact form, with enough detail to identify the question and your reasoning, ideally with a reference. Every report is triaged. Where an error is confirmed, the question is corrected or retired, and the change is logged against the question so that the fix is traceable.

Spelling and typography

Questions and explanations use UK English spelling and UK drug names as they appear in the BNF. Ranges take an en-dash (2–3 years), dimensions take a multiplication sign (3 × 2 cm), and units are written the way UK results are reported. None of this changes an answer, but consistency means you are never slowed by the presentation of a question.

What we will not claim

• No guaranteed pass. No question bank can guarantee a result, and we will not suggest otherwise.
• No pass-mark percentages. The Part A pass mark is set by standard setting for each sitting, so we do not quote one.
• No invented exam frequencies. The blueprint says how many questions each area carries; it does not say which topics will appear, and neither do we.
• No recalled or official questions. Our questions are original. We are independent of the ICBSE and the Royal Colleges.
• No review we have not had. Specialist review is ongoing and its status is published; we do not claim consultant or named-reviewer sign-off that has not taken place.
• No fake reviews. We do not publish invented testimonials, ratings or pass rates.