FREEMEDSITE MRCS ACADEMY · ARTICLE 22 · ETHICS, STATISTICS & AUDIT

Ethics, Medical Statistics and Clinical Audit Made Easy for MRCS Part A

A clear exam-first guide to surgical consent, capacity, confidentiality, safety culture, diagnostic statistics, clinical trials and the audit cycle. Learn 12 structured modules, track 52 revision objectives and practise 40 original five-option clinical SBAs.

12Teaching modules

52Saved objectives

40Original SBAs

MRCS Part A · Papers 1 & 2Practical formulasUK professional principlesBrowser-local progress

How to use: Read each lesson, work the formula examples, tick topics you can explain, then practise a single question at a time. Questions are original educational items, not copied exam recalls.

Important scope: Ethics scenarios use UK professional guidance. Capacity, consent law and safeguarding duties have jurisdiction-specific differences across England, Wales, Scotland and Northern Ireland. This page supports MRCS revision but does not replace current clinical or legal advice. The 2021 official MRCS indicative breakdown includes data interpretation/audit and medico-legal aspects.

On this page

1. Four principles and why patient context matters2. Consent: material risks, alternatives and shared decisions3. Capacity and refusal of treatment4. Emergency decisions and overall benefit5. Confidentiality, clinical images and data protection6. Professionalism, complaints, candour and safety culture7. Clinical research, service evaluation and audit8. The 2-by-2 diagnostic table9. Predictive values, likelihood ratios and prevalence10. Absolute risk, relative risk, odds ratio, NNT and NNH11. Bias, p values, uncertainty and trial design12. Clinical audit cycles and quality improvement52-topic saved revision checklist40 original clinical SBAsFAQsReferences and Academy navigation

1. Four principles and why patient context matters

Autonomy is respect for a patient’s informed choices; beneficence promotes welfare; non-maleficence considers potential harms; justice concerns fair access and resource allocation. These principles can conflict: a patient with capacity can refuse a beneficial intervention. Clinical decisions also depend on current professional standards and jurisdiction-specific law.

MRCS clinical pearl: Exam priority: identify which principle is in tension, then choose the safe, lawful, patient-centred next step.

2. Consent: material risks, alternatives and shared decisions

Consent is a continuing, understandable dialogue. Explain the diagnosis, expected benefits, relevant harms, reasonable alternatives and the option of no treatment where appropriate. Ask what matters to this particular person: a rare voice change may be material to a singer. Document what was discussed; signatures do not replace communication.

MRCS clinical pearl: Exam trap: “commonest risk” and “material risk” are not interchangeable.

Original schematic: informed, supported and voluntary consent is a process.

3. Capacity and refusal of treatment

For an adult, start with a presumption of capacity. Assess the person’s ability to understand, retain, use or weigh relevant information and communicate a choice, with appropriate decision/time specificity under the applicable law. Optimise communication and reversible illness first. A choice the doctor considers unwise is not in itself evidence of incapacity.

MRCS clinical pearl: Exam priority: separate capacity from agreement with medical advice.

4. Emergency decisions and overall benefit

An unconscious patient may require immediate lifesaving intervention without express consent if no valid applicable refusal is known. Act proportionately under the relevant legal framework, involve appropriate colleagues, seek known preferences when feasible, and document why delay was unsafe. For nonurgent decisions in those lacking capacity, carefully consult proxies or those close to the patient as applicable.

MRCS clinical pearl: Key caution: the laws and formal proxies vary among UK nations; do not memorise one universal family-consent rule.

5. Confidentiality, clinical images and data protection

Identifiability can arise from dates, tattoos, unusual conditions or combinations of details. Share relevant information securely with care teams on a proper basis; disclosure for social media or teaching usually requires additional safeguards. Use minimal necessary information and approved systems. Pseudonymised information may still be personal data.

MRCS clinical pearl: Exam trap: removing a patient name does not necessarily create anonymous data.

6. Professionalism, complaints, candour and safety culture

Patients harmed by an error deserve timely truthful explanation, an appropriate apology, clear follow-up and safety reporting. The professional duty of candour is separate from the question of legal liability. Near misses can expose systems weaknesses. Raise concerns about colleagues whose performance creates patient risk through appropriate channels.

MRCS clinical pearl: Exam priority: protect patients, communicate honestly, record and learn.

7. Clinical research, service evaluation and audit

Research aims to answer a defined question about effectiveness or generate new knowledge and may involve allocation to treatments. Audit compares practice against an explicit standard. Service evaluation describes current activity without necessarily measuring compliance against a benchmark. Quality improvement iteratively tests changes. Approval and information governance depend on design and organisational requirements.

MRCS clinical pearl: The project label does not determine its governance classification.

8. The 2-by-2 diagnostic table

Among patients with disease, test positives are true positives and test negatives are false negatives. Among patients without disease, test negatives are true negatives and test positives are false positives. Sensitivity = TP/(TP+FN). Specificity = TN/(TN+FP). The tested reference standard and target population matter.

MRCS clinical pearl: Sensitivity conditions on disease; specificity conditions on disease absence.

Original schematic. Sensitivity conditions on disease; specificity conditions on disease absence.

Sensitivity = TP / (TP + FN)   ·   Specificity = TN / (TN + FP)

Worked example: If 85 of 100 infected patients test positive, sensitivity is 85%. If 160 of 200 uninfected patients test negative, specificity is 80%.

9. Predictive values, likelihood ratios and prevalence

PPV = TP/(TP+FP); NPV = TN/(TN+FN). They change with pretest prevalence even when sensitivity and specificity are constant. LR+ = sensitivity/(1-specificity). LR- = (1-sensitivity)/specificity. Post-test odds = pretest odds × likelihood ratio.

MRCS clinical pearl: A positive test result needs interpretation within the clinical pretest probability.

LR+ = Se / (1 − Sp)   ·   LR− = (1 − Se) / Sp

Worked example: Se 0.90, Sp 0.80 ⇒ LR+ = 4.5. A positive result multiplies pretest odds by 4.5 rather than creating an automatic diagnosis.

10. Absolute risk, relative risk, odds ratio, NNT and NNH

ARR = risk in control - risk in intervention. RR = exposed risk / unexposed risk. RRR = 1 - RR (when protective). NNT = 1/ARR using proportion units, rounded up conventionally. NNH = 1/absolute risk increase. For a case-control 2×2 table, odds ratio = ad/bc.

MRCS clinical pearl: Always write rates as decimals for NNT and distinguish absolute from relative benefits.

ARR = CER − EER   ·   NNT = 1 / ARR   ·   RR = EER / CER

Worked example: 20% control risk versus 10% intervention risk means ARR 0.10, NNT 10, RR 0.5 and relative risk reduction 50%.

11. Bias, p values, uncertainty and trial design

A p value is the tail probability of data at least this extreme under a null model; it is not the probability a hypothesis is true. A 95% CI reports an interval method with 95% long-run coverage, not a 95% posterior probability for a fixed parameter. Random sequence generation and concealment of future allocations are distinct; intention-to-treat respects assigned groups. Watch spectrum, lead-time and selection biases.

MRCS clinical pearl: For RR and OR, a CI crossing 1 usually fails the corresponding two-sided 5% threshold.

12. Clinical audit cycles and quality improvement

Choose an evidence-based standard, define eligible cases and measurable criteria, collect baseline data, compare with the standard, change the process and re-audit. Use QI methods such as PDSA for smaller iterative tests. Measure outcomes, processes and balancing harms, and examine performance over time. Protect patient data throughout.

MRCS clinical pearl: An audit cycle is not complete simply because results were presented at a meeting.

Original schematic: reporting baseline results without re-measurement does not demonstrate improvement.

52-topic interactive revision checklist

Tick a topic only when you can explain it without notes and solve its clinical scenario. Progress remains on this device and is not linked to your account.

0 / 52 topics completed

Browser-local only; clearing site data or using private browsing may erase progress.

Ethical framework

Autonomy versus beneficence in refusalNon-maleficence and treatment harmsJustice and limited resourcesProfessional standards versus personal beliefs

Consent

Discuss individually material surgical risksOffer reasonable alternatives including no treatmentDocument adequate shared decision-makingDistinguish clinical from teaching consentRecognise coercion and absent voluntariness

Capacity

Presume adult capacity and optimise communicationAssess understanding and retentionAssess use or weighing and communicationRecognise time- and decision-specific incapacityRespect a capacitous adult refusal

Emergency decisions

Act in overall benefit when capacity is absentRecognise advance decisions and legal proxiesConsult those close to patient when feasibleDocument urgency, proportionality and reasoning

Confidentiality

Recognise indirect patient identifiersUse minimum necessary information for direct careHandle clinical images and teaching disclosuresRecognise pseudonymised versus anonymous data

Candour & professionalism

Explain adverse events honestly and apologiseReport near misses and learn from systemsEscalate concern about impaired colleaguesManage conflicts of interest transparently

Study governance

Distinguish audit versus researchDistinguish service evaluation and QISeek approvals before experimental allocationKnow information governance and lawful data handling

Diagnostic test metrics

Fill in TP, FN, TN and FPCalculate sensitivityCalculate specificityCalculate predictive valuesRecognise prevalence effects

Bayesian thinking

Calculate LR+Calculate LR-Relate pretest and post-test oddsRecognise spectrum effects on accuracy

Risk measures

Calculate RR and ARRCalculate NNT and NNHCalculate OR from 2 by 2 tableSeparate absolute and relative reduction

Uncertainty & trials

Interpret confidence intervals and null valuesInterpret p value without probability fallacySeparate type I and type II errorsAllocation concealment and intention-to-treatIdentify lead-time and spectrum bias

Audit & QI

Define explicit audit standards and denominatorsPlan change and repeat the auditUse PDSA cycles appropriatelyMonitor balancing measures and run charts

40 original MRCS clinical SBAs

Choose one of A–E, press Check answer, study every explanation, then press Next Question. Only one question is shown at a time when JavaScript runs. Without JavaScript, all questions and answers remain readable.

0 / 40Marked

0Correct

N/AAccuracy

Question 1 of 40

Question 1 of 40 · Material risk and consent

Q01. A 47-year-old professional singer is offered thyroid surgery. The consultant explains bleeding and infection but omits permanent voice-pitch change because it is uncommon. What best completes consent?

A. State that risks under 1% never need discussionB. Obtain a signature without revisiting the conversationC. Discuss risks material to this patient, including loss of singing range, and reasonable alternativesD. Ask the anaesthetist to decide which surgical risks matterE. Avoid mentioning risks to minimise preoperative anxiety

View detailed explanation, examiner trap and viva

Correct answer: C. Discuss risks material to this patient, including loss of singing range, and reasonable alternatives

Why correct: Materiality depends on the individual’s circumstances and what a reasonable person in that position would regard as significant; the discussion should address their professional voice use.

Why other options are incorrect:

A. State that risks under 1% never need discussion: A numerical threshold cannot replace individual assessment of material risk.

B. Obtain a signature without revisiting the conversation: A signed form does not repair inadequate shared decision-making.

D. Ask the anaesthetist to decide which surgical risks matter: The responsible clinical team must discuss the proposed intervention and material options.

E. Avoid mentioning risks to minimise preoperative anxiety: Withholding material information undermines informed choice.

Exam pearl: The patient’s priorities affect what is material, even for low-probability outcomes.

Examiner trap: Montgomery-style material risk is not identical to the most common risk.

Viva extension: Would a reasonable alternative be discussed? Yes—including surveillance or another suitable strategy.

Question 2 of 40 · Consent and alternatives

Q02. A 68-year-old with an asymptomatic small abdominal aortic aneurysm asks whether surveillance is possible. The clinician describes only elective repair. What should occur?

A. Explain surveillance and other reasonable options with individual benefits and harmsB. Explain only the operation the surgeon prefersC. State that consent is valid once the patient signs the formD. Withhold rupture-risk information to avoid distressE. Assume all aneurysms have an identical treatment threshold

View detailed explanation, examiner trap and viva

Correct answer: A. Explain surveillance and other reasonable options with individual benefits and harms

Why correct: Shared decision-making includes reasonable alternatives and the option of no intervention where appropriate.

Why other options are incorrect:

B. Explain only the operation the surgeon prefers: Clinician preference is insufficient to exclude reasonable alternatives.

C. State that consent is valid once the patient signs the form: Consent requires an adequate discussion and voluntary informed choice.

D. Withhold rupture-risk information to avoid distress: Information about relevant benefit and harm supports choice.

E. Assume all aneurysms have an identical treatment threshold: Management varies with anatomy, risk and patient factors.

Exam pearl: Consent is a process rather than a single form.

Examiner trap: Reasonable alternative does not mean every conceivable treatment.

Viva extension: How does shared decision-making accommodate patient values? Elicit priorities and align options.

Question 3 of 40 · Capacity assessment

Q03. An adult with delirium after surgery refuses a CT scan. On questioning, they cannot retain the explanation for long enough to decide. What is the correct interpretation?

A. Declare lifelong incapacity because of a delirium diagnosisB. Assume refusal proves incapacityC. Treat a relative as automatically authorised to consentD. Proceed as if all postoperative patients lack capacityE. Assess decision-specific capacity at that time and address reversible causes

View detailed explanation, examiner trap and viva

Correct answer: E. Assess decision-specific capacity at that time and address reversible causes

Why correct: Capacity is functional, decision-specific and time-specific; delirium may temporarily impair retention and decision-making.

Why other options are incorrect:

A. Declare lifelong incapacity because of a delirium diagnosis: A diagnosis alone cannot establish incapacity for all future decisions.

B. Assume refusal proves incapacity: An unwise or unwanted choice alone does not establish incapacity.

C. Treat a relative as automatically authorised to consent: Relatives do not automatically have legal decision-making authority.

D. Proceed as if all postoperative patients lack capacity: Presume capacity until appropriately assessed.

Exam pearl: Under England/Wales law: understand, retain, use/weigh and communicate relevant information.

Examiner trap: Capacity can recover: reassess when the condition changes.

Viva extension: How would you support a decision? Treat pain/delirium and use clear, accessible information.

Question 4 of 40 · Competent refusal

Q04. An alert adult with major upper GI bleeding understands that refusing transfusion may cause death. They consistently decline blood for deeply held beliefs. What should the team do?

A. Transfuse because the treatment is life-savingB. Respect the capacitous refusal while offering acceptable alternatives and documenting discussionC. Ask a relative to replace the patient’s decisionD. Assume a religious decision indicates incapacityE. Discharge immediately without treating haemorrhage

View detailed explanation, examiner trap and viva

Correct answer: B. Respect the capacitous refusal while offering acceptable alternatives and documenting discussion

Why correct: A competent adult may refuse a treatment despite grave consequences; check that the refusal is informed and voluntary.

Why other options are incorrect:

A. Transfuse because the treatment is life-saving: Urgency alone does not override a valid capacitous refusal.

C. Ask a relative to replace the patient’s decision: Relatives cannot overrule a capacitous adult.

D. Assume a religious decision indicates incapacity: Personal beliefs do not mean incapacity.

E. Discharge immediately without treating haemorrhage: Offer all acceptable medical and surgical alternatives and supportive care.

Exam pearl: Competent adult refusal should be respected even if death is a foreseeable consequence.

Examiner trap: Refusal of one therapy is not refusal of every form of care.

Viva extension: What must be documented? Capacity, information, specific treatment refused and alternatives.

Question 5 of 40 · Emergency best interests

Q05. An unconscious patient with suspected ruptured abdominal aortic aneurysm needs immediate surgery. No advance refusal or authorised representative is known. Which approach is appropriate?

A. Wait for written consent from any available relativeB. Do nothing because no consent form is signedC. Assume the unconscious patient has permanently lost autonomyD. Deliver immediately necessary treatment for overall benefit and document the reasoningE. Perform unrelated elective procedures during the emergency

View detailed explanation, examiner trap and viva

Correct answer: D. Deliver immediately necessary treatment for overall benefit and document the reasoning

Why correct: Where capacity is absent and delay risks death, treatment may proceed under the relevant emergency legal framework while considering known preferences.

Why other options are incorrect:

A. Wait for written consent from any available relative: This can create dangerous delay and relatives are not automatically proxies.

B. Do nothing because no consent form is signed: Emergency treatment can be lawful despite absent express consent.

C. Assume the unconscious patient has permanently lost autonomy: Only the current decision and situation are being addressed.

E. Perform unrelated elective procedures during the emergency: Authority extends only to justified, necessary treatment.

Exam pearl: Emergency incapacity is not a waiver of proportionality or patient wishes.

Examiner trap: Investigate advance decisions where feasible without jeopardising rescue.

Viva extension: What if a valid applicable advance refusal of the specific treatment is known? Seek urgent legal/senior guidance and ordinarily respect it.

Question 6 of 40 · Teaching and consent

Q06. Before a non-urgent examination under anaesthesia, a trainee asks to perform an additional pelvic examination solely for teaching. It is not required for care and has not been discussed. Best action?

A. Proceed because the patient consented to general anaesthesiaB. Do not perform it for teaching without appropriate specific consentC. Proceed because the examination is low riskD. Ask another clinician for informal permissionE. Document it afterward and assume retrospective consent

View detailed explanation, examiner trap and viva

Correct answer: B. Do not perform it for teaching without appropriate specific consent

Why correct: Teaching procedures that are not necessary for clinical care require the patient to be told and given an appropriate choice.

Why other options are incorrect:

A. Proceed because the patient consented to general anaesthesia: Anaesthesia consent does not authorise unrelated teaching examinations.

C. Proceed because the examination is low risk: Low risk does not remove consent requirements.

D. Ask another clinician for informal permission: A colleague cannot substitute for the patient’s consent.

E. Document it afterward and assume retrospective consent: Consent must ordinarily precede the procedure.

Exam pearl: A patient can decline teaching participation without losing access to care.

Examiner trap: Do not conflate treatment consent with research or teaching consent.

Viva extension: How would you document permission? State purpose, personnel and agreed limits.

Question 7 of 40 · Confidentiality on social media

Q07. A surgeon wants to post an unusual postoperative photograph on a teaching account. The face is absent but a rare tattoo and admission date could identify the patient. What is best?

A. Post because the patient’s face is not visibleB. Post because social media is used for educationC. Post when the surgeon has removed the patient name onlyD. Ask a colleague rather than the patient if sharing is acceptableE. Avoid posting identifiable information without a valid basis and appropriate explicit permission

View detailed explanation, examiner trap and viva

Correct answer: E. Avoid posting identifiable information without a valid basis and appropriate explicit permission

Why correct: Indirect identifiers may identify a patient; privacy and professional obligations apply to clinical images.

Why other options are incorrect:

A. Post because the patient’s face is not visible: Identifiability extends beyond facial features.

B. Post because social media is used for education: Educational value does not itself authorise disclosure.

C. Post when the surgeon has removed the patient name only: Removing a name may still leave recognisable identifiers.

D. Ask a colleague rather than the patient if sharing is acceptable: Colleague agreement is not a substitute for the correct disclosure basis.

Exam pearl: Anonymisation means re-identification must not reasonably be possible.

Examiner trap: A case can remain identifiable from context.

Viva extension: What safer alternatives exist? Use fully de-identified authorised educational material.

Question 8 of 40 · Direct-care information sharing

Q08. A ward doctor shares relevant allergy and medication information with the on-call anaesthetist treating the same patient. Which statement is most accurate?

A. Sharing any personal information is always forbiddenB. All clinical details should be sent to the entire hospitalC. Relevant confidential information can ordinarily be shared within the direct-care team on a proper basisD. A patient’s public social-media profile removes confidentiality dutiesE. Only an outside lawyer can authorise information exchange between treating professionals

View detailed explanation, examiner trap and viva

Correct answer: C. Relevant confidential information can ordinarily be shared within the direct-care team on a proper basis

Why correct: Sharing relevant data for direct care commonly relies on appropriate implied consent and confidentiality safeguards, depending on the circumstances.

Why other options are incorrect:

A. Sharing any personal information is always forbidden: Confidentiality permits necessary, proportionate direct-care sharing.

B. All clinical details should be sent to the entire hospital: Only necessary information should go to appropriate recipients.

D. A patient’s public social-media profile removes confidentiality duties: Public information does not negate clinical professional obligations.

E. Only an outside lawyer can authorise information exchange between treating professionals: Routine appropriate care sharing does not need such authorisation.

Exam pearl: Use the minimum necessary information with relevant professionals.

Examiner trap: Implied consent for direct care does not authorise unrelated publicity.

Viva extension: What if the patient objects to a specific disclosure? Assess and respect the objection or identify a justified exception.

Question 9 of 40 · Duty of candour

Q09. After a medication error prolongs admission, the patient is stable and asks why recovery was delayed. What should the responsible team do?

A. Explain promptly and honestly what is known, apologise and follow reporting and review requirementsB. Wait until litigation begins before discussing the harmC. Conceal the event to protect the junior prescriberD. Attribute definite blame before facts are establishedE. Assume an apology is an admission of legal liability

View detailed explanation, examiner trap and viva

Correct answer: A. Explain promptly and honestly what is known, apologise and follow reporting and review requirements

Why correct: Professional candour requires openness with patients when things go wrong, as well as appropriate safety reporting and follow-up.

Why other options are incorrect:

B. Wait until litigation begins before discussing the harm: Candour is a professional duty, not dependent on legal proceedings.

C. Conceal the event to protect the junior prescriber: Concealment obstructs learning and patient trust.

D. Attribute definite blame before facts are established: Candour should accurately separate known facts from ongoing investigation.

E. Assume an apology is an admission of legal liability: A genuine apology is part of appropriate compassionate communication.

Exam pearl: Disclose and apologise; distinguish established facts from unresolved causes.

Examiner trap: Candour applies to harm or distress, not only catastrophic incidents.

Viva extension: What else follows? Document communication, report the event and reduce recurrence.

Question 10 of 40 · Near-miss learning

Q10. A wrong-side operation is prevented in time out before incision. No patient harm occurs. What is the best safety response?

A. Delete records because no injury occurredB. Punish one staff member without analysing processesC. Ignore the event if the WHO checklist eventually succeededD. Report the near miss and investigate systems contributing to the errorE. Stop using the safety checklist because it found a problem

View detailed explanation, examiner trap and viva

Correct answer: D. Report the near miss and investigate systems contributing to the error

Why correct: Near misses offer valuable opportunities for system learning and hazard reduction even when harm is avoided.

Why other options are incorrect:

A. Delete records because no injury occurred: Absence of harm does not remove learning value.

B. Punish one staff member without analysing processes: A systems approach asks how multiple defences failed.

C. Ignore the event if the WHO checklist eventually succeeded: The intercepted error still signals risk.

E. Stop using the safety checklist because it found a problem: This is evidence that a barrier prevented harm.

Exam pearl: Near misses should inform improvement rather than disappear.

Examiner trap: A successful safety barrier does not mean there was no underlying hazard.

Viva extension: How is learning shared? Through local incident reporting and improvement feedback.

Question 11 of 40 · Raising concerns

Q11. A junior repeatedly observes a colleague appearing intoxicated during operations and is worried about patient safety. Best response?

A. Wait for definite patient injuryB. Post accusations on social mediaC. Arrange informal cover-up of the colleague’s mistakesD. Act promptly to protect patients and raise concerns through appropriate senior or organisational channelsE. Assume senior staff are exempt from concerns processes

View detailed explanation, examiner trap and viva

Correct answer: D. Act promptly to protect patients and raise concerns through appropriate senior or organisational channels

Why correct: Patient safety takes priority; raise substantiated concerns and avoid unsafe confrontation or cover-up.

Why other options are incorrect:

A. Wait for definite patient injury: Serious foreseeable risk merits early escalation.

B. Post accusations on social media: Public accusations breach professionalism and are not safe escalation.

C. Arrange informal cover-up of the colleague’s mistakes: Concealment increases risk.

E. Assume senior staff are exempt from concerns processes: Duty to act applies regardless of hierarchy.

Exam pearl: Patient safety concerns should be raised promptly and proportionately.

Examiner trap: Do not equate escalation with unproven public accusation.

Viva extension: What should be documented? Objective observations, actions and escalation.

Question 12 of 40 · Research ethics

Q12. A team proposes randomising patients to two surgical approaches to determine which has superior function, beyond ordinary service evaluation. Before enrolment, what is essential?

A. Call it an audit to bypass reviewB. Obtain appropriate research approvals and informed participant consent or valid lawful authorisationC. Obtain consent only after collecting study outcomesD. Use identifiable data publicly because surgery is routineE. Assume an existing departmental quality project approves the trial

View detailed explanation, examiner trap and viva

Correct answer: B. Obtain appropriate research approvals and informed participant consent or valid lawful authorisation

Why correct: Interventional research requires appropriate protocol review, governance and consent arrangements, not simply departmental audit approval.

Why other options are incorrect:

A. Call it an audit to bypass review: Labelling does not change the activity’s purpose or governance requirements.

C. Obtain consent only after collecting study outcomes: Prospective participation requires appropriate prior processes.

D. Use identifiable data publicly because surgery is routine: Data protection and confidentiality still apply.

E. Assume an existing departmental quality project approves the trial: Research requires its own appropriate review.

Exam pearl: Audit tests care against a standard; research may test hypotheses and generate new generalisable knowledge.

Examiner trap: Intent and design matter more than the project title.

Viva extension: How do you check classification? Seek research governance or ethics office advice.

Question 13 of 40 · Conflicts of interest

Q13. A clinician receives funding from a device company and is recommending that device in a hospital procurement meeting. What should happen?

A. Declare the relevant financial interest and follow impartial procurement governanceB. Hide the relationship because the device is effectiveC. Allow the sponsor to write the committee recommendationD. Assume a small payment cannot affect judgementE. Disclose only if a competitor complains

View detailed explanation, examiner trap and viva

Correct answer: A. Declare the relevant financial interest and follow impartial procurement governance

Why correct: Transparency allows conflicts to be managed and decisions to focus on evidence and patient benefit.

Why other options are incorrect:

B. Hide the relationship because the device is effective: Efficacy does not remove conflict disclosure obligations.

C. Allow the sponsor to write the committee recommendation: Independent appraisal is necessary.

D. Assume a small payment cannot affect judgement: Perceived and actual conflicts require appropriate management.

E. Disclose only if a competitor complains: Disclosure should be proactive.

Exam pearl: A conflict of interest can exist without proven misconduct.

Examiner trap: Transparency is necessary but may not be sufficient; recusal may be appropriate.

Viva extension: What matters in device appraisal? Comparative evidence, safety, costs and governance.

Question 14 of 40 · Screening: sensitivity

Q14. A new sepsis screen is evaluated against a reference standard: 100 patients truly have sepsis and 85 screen positive; 200 do not have sepsis. What is the sensitivity?

A. 15%B. 42.5%C. 92.5%D. Cannot be calculated without knowing specificityE. 85%

View detailed explanation, examiner trap and viva

Correct answer: E. 85%

Why correct: Sensitivity equals true positives divided by all patients with the disease: 85/100 = 85%.

Why other options are incorrect:

A. 15%: This is the false-negative proportion among those with disease.

B. 42.5%: This incorrectly divides positives by the nondiseased group.

C. 92.5%: This would require additional information and is not 85/100.

D. Cannot be calculated without knowing specificity: The diseased group counts alone are sufficient.

Exam pearl: Sensitivity = TP/(TP+FN).

Examiner trap: Sensitivity is not the same as positive predictive value.

Viva extension: What is the false-negative rate here? 15%.

Question 15 of 40 · Screening: specificity

Q15. In a validation cohort, 160 of 200 patients without postoperative infection have a negative screening test. What is specificity?

A. 20%B. 40%C. 80%D. 160%E. It requires disease prevalence

View detailed explanation, examiner trap and viva

Correct answer: C. 80%

Why correct: Specificity is TN/(TN+FP) = 160/200 = 80%.

Why other options are incorrect:

A. 20%: This is the false-positive rate.

B. 40%: This does not match the nondiseased group denominator.

D. 160%: Percentages of this kind cannot exceed 100%.

E. It requires disease prevalence: Specificity can be calculated from the nondiseased group alone.

Exam pearl: Specificity = TN/(TN+FP).

Examiner trap: Specificity describes how well disease absence is recognised.

Viva extension: How many false positives? 40.

Question 16 of 40 · Predictive value and prevalence

Q16. A diagnostic test has unchanged sensitivity and specificity in two hospitals. Disease prevalence is much lower in Hospital B. What usually happens to positive predictive value in B?

A. Positive predictive value always increasesB. Sensitivity must fallC. Specificity must increaseD. Negative predictive value always fallsE. Positive predictive value decreases

View detailed explanation, examiner trap and viva

Correct answer: E. Positive predictive value decreases

Why correct: When disease becomes rarer, a larger proportion of positive results may be false positives even if test sensitivity and specificity stay constant.

Why other options are incorrect:

A. Positive predictive value always increases: Lower prevalence generally reduces PPV.

B. Sensitivity must fall: Sensitivity is a conditional test characteristic and need not change.

C. Specificity must increase: Specificity can remain the same.

D. Negative predictive value always falls: NPV usually improves as prevalence decreases, all else equal.

Exam pearl: Prevalence changes predictive values.

Examiner trap: Do not assume a positive screen means the same post-test risk everywhere.

Viva extension: How might NPV change with lower prevalence? Typically rise.

Question 17 of 40 · Positive likelihood ratio

Q17. A screening instrument has sensitivity 90% and specificity 80%. What is its positive likelihood ratio?

A. 0.18B. 0.5C. 4.5D. 1.125E. 9.0

View detailed explanation, examiner trap and viva

Correct answer: C. 4.5

Why correct: LR+ = sensitivity/(1 - specificity) = 0.90/0.20 = 4.5.

Why other options are incorrect:

A. 0.18: Multiplication of two probabilities is not the LR+ formula.

B. 0.5: This is not calculated using sensitivity and false-positive rate.

D. 1.125: This incorrectly divides 0.90 by 0.80.

E. 9.0: This would require a 10% false-positive rate, not 20%.

Exam pearl: LR+ uses the false-positive rate in the denominator.

Examiner trap: LR+ is not the same as relative risk.

Viva extension: If pretest odds are 1:4, what are post-test odds? About 4.5:4 = 1.125:1.

Question 18 of 40 · Absolute risk reduction

Q18. A trial reports wound infection in 20% of control patients and 12% of intervention patients. What is absolute risk reduction?

A. 40%B. 8 percentage pointsC. 0.6D. 12%E. 32%

View detailed explanation, examiner trap and viva

Correct answer: B. 8 percentage points

Why correct: ARR = control event rate - intervention event rate = 20% - 12% = 8%.

Why other options are incorrect:

A. 40%: This is relative risk reduction, not absolute risk reduction.

C. 0.6: This is the relative risk (12/20), not ARR.

D. 12%: This is the intervention event rate.

E. 32%: Adding the rates does not measure reduction.

Exam pearl: ARR and RRR can convey very different impressions.

Examiner trap: 8% is not the same as 8% relative reduction.

Viva extension: What is RRR here? 40%.

Question 19 of 40 · Number needed to treat

Q19. Control surgical-site infection risk is 20%, compared with 10% after an intervention. What is the number needed to treat to prevent one infection?

A. 2B. 5C. 20D. 10E. 100

View detailed explanation, examiner trap and viva

Correct answer: D. 10

Why correct: ARR = 0.20 - 0.10 = 0.10, so NNT = 1/0.10 = 10.

Why other options are incorrect:

A. 2: This is unrelated to the absolute risk difference.

B. 5: That would correspond to ARR of 20%.

C. 20: That would correspond to ARR of 5%.

E. 100: That would correspond to ARR of 1%.

Exam pearl: Always use ARR as a proportion in the NNT formula.

Examiner trap: Do not use relative risk reduction as the denominator.

Viva extension: How do you round an NNT? Conventionally round up to the next whole person.

Question 20 of 40 · Relative risk

Q20. In a cohort study, postoperative DVT develops in 30/150 immobilised patients and 10/100 mobile patients. What is relative risk for the immobilised group?

A. 2.0B. 0.5C. 3.0D. 0.10E. 1.0

View detailed explanation, examiner trap and viva

Correct answer: A. 2.0

Why correct: Risk in exposed = 30/150 = 0.20 and risk in unexposed = 10/100 = 0.10; RR = 2.

Why other options are incorrect:

B. 0.5: This reverses the risk comparison.

C. 3.0: This compares event counts without correcting for denominator.

D. 0.10: This is risk in the unexposed group, not RR.

E. 1.0: The observed risks are not equal.

Exam pearl: RR compares two probabilities, not raw event counts.

Examiner trap: Unequal group sizes make raw counts misleading.

Viva extension: What is the absolute risk difference? 10 percentage points.

Question 21 of 40 · Odds ratio

Q21. A case-control study of post-op thrombosis finds 40 exposed and 20 unexposed among cases, and 20 exposed and 40 unexposed among controls. What is the odds ratio?

A. 4.0B. 2.0C. 1.0D. 0.25E. 8.0

View detailed explanation, examiner trap and viva

Correct answer: A. 4.0

Why correct: OR = (40 x 40)/(20 x 20) = 4.0.

Why other options are incorrect:

B. 2.0: This may arise from comparing exposed proportions only.

C. 1.0: The exposure odds clearly differ.

D. 0.25: This is the reciprocal of the stated comparison.

E. 8.0: The cross-product ratio is 4, not 8.

Exam pearl: For a 2 by 2 case-control table, OR = ad/bc.

Examiner trap: Case-control sampling generally does not directly yield incidence risk.

Viva extension: When can OR approximate RR? When the outcome is rare under suitable assumptions.

Question 22 of 40 · Confidence intervals

Q22. An RCT finds RR 0.82 for recurrence with a 95% confidence interval of 0.60 to 1.12. What is the best conventional interpretation?

A. The intervention is proven ineffectiveB. There is an 82% chance the treatment worksC. There is a 95% probability the fixed true RR lies in this specific intervalD. The interval includes 1, so a two-sided 5% statistical significance threshold is not metE. The result proves the treatment is harmful

View detailed explanation, examiner trap and viva

Correct answer: D. The interval includes 1, so a two-sided 5% statistical significance threshold is not met

Why correct: For relative measures, the null is 1; the reported 95% confidence interval crosses it.

Why other options are incorrect:

A. The intervention is proven ineffective: Failure to demonstrate significance does not prove equivalence.

B. There is an 82% chance the treatment works: RR=0.82 is not a probability of benefit.

C. There is a 95% probability the fixed true RR lies in this specific interval: The frequentist CI interpretation concerns interval procedures over repeated samples.

E. The result proves the treatment is harmful: The point estimate is below 1 and uncertainty spans benefit and harm.

Exam pearl: RR/OR null value is 1; mean difference null value is 0.

Examiner trap: Non-significant does not mean no effect.

Viva extension: What might improve precision? Larger well-designed studies.

Question 23 of 40 · P values

Q23. A study reports p=0.03 for a predefined difference in complications. Which statement is valid?

A. There is a 3% probability the null hypothesis is trueB. There is a 97% probability treatment is clinically worthwhileC. Under the null model, data at least this extreme would be observed with probability about 3%D. The study must be free of biasE. Every patient will benefit

View detailed explanation, examiner trap and viva

Correct answer: C. Under the null model, data at least this extreme would be observed with probability about 3%

Why correct: The p value is defined conditional on the null hypothesis and model assumptions.

Why other options are incorrect:

A. There is a 3% probability the null hypothesis is true: A p value is not a posterior probability.

B. There is a 97% probability treatment is clinically worthwhile: Clinical importance is not determined by a p value alone.

D. The study must be free of bias: Statistical significance does not prevent systematic error.

E. Every patient will benefit: An average effect does not imply uniform individual benefit.

Exam pearl: Statistical significance is not synonymous with clinical importance.

Examiner trap: Pre-specified endpoints and multiplicity matter.

Viva extension: What else must be reported? Effect size, uncertainty, harms and validity.

Question 24 of 40 · Type I error

Q24. A study concludes that an intervention reduces infection when in truth no benefit exists. Which statistical error occurred?

A. Type II (beta) errorB. ConfoundingC. Selection biasD. Measurement biasE. Type I (alpha) error

View detailed explanation, examiner trap and viva

Correct answer: E. Type I (alpha) error

Why correct: A false positive occurs when a true null hypothesis is rejected.

Why other options are incorrect:

A. Type II (beta) error: Type II is failure to detect a true effect.

B. Confounding: Confounding is a systematic mixing of exposure and outcome associations.

C. Selection bias: Selection bias concerns who enters/continues in study groups.

D. Measurement bias: Measurement bias stems from systematic inaccuracies of observation.

Exam pearl: Alpha controls the conventional false-positive threshold for a test.

Examiner trap: Do not interchange alpha and beta.

Viva extension: What is power? 1-beta.

Question 25 of 40 · Type II error and power

Q25. An underpowered trial misses a real 20% relative reduction in infection. Which error is most likely?

A. Type I errorB. Type II (beta) errorC. Observer biasD. Confounding by indicationE. Ecological fallacy

View detailed explanation, examiner trap and viva

Correct answer: B. Type II (beta) error

Why correct: A false negative is failure to reject a false null hypothesis.

Why other options are incorrect:

A. Type I error: Type I corresponds to a spurious positive finding.

C. Observer bias: Observer bias is related to assessments, not the generic meaning of inadequate power.

D. Confounding by indication: Confounding by indication can affect nonrandomised treatment comparisons.

E. Ecological fallacy: Ecological fallacy infers individual associations from group data.

Exam pearl: Power = 1 - beta.

Examiner trap: A small negative trial may be inconclusive rather than definitive.

Viva extension: How can power be increased? Adequate sample size and event rates, among other factors.

Question 26 of 40 · Randomisation and allocation concealment

Q26. An RCT uses computer randomisation but surgeons can view the next group assignment before enrolling each patient. What key design protection has failed?

A. Outcome assessor blindingB. Intention-to-treat analysisC. Allocation concealmentD. Sample-size calculationE. External validation

View detailed explanation, examiner trap and viva

Correct answer: C. Allocation concealment

Why correct: Seeing assignments before enrolment may influence recruitment and create selection bias, despite a random sequence.

Why other options are incorrect:

A. Outcome assessor blinding: Blinding outcome assessors occurs after allocation and addresses measurement bias.

B. Intention-to-treat analysis: Intention-to-treat concerns how allocated participants are analysed.

D. Sample-size calculation: Sample size is separate from hiding future assignment.

E. External validation: External validity is generalisability across settings.

Exam pearl: Random sequence generation and allocation concealment are distinct safeguards.

Examiner trap: Sealed opaque sequential envelopes may conceal allocation only if properly used.

Viva extension: Why is allocation concealment important even with blinded outcomes? It protects enrolment integrity.

Question 27 of 40 · Intention to treat

Q27. In a pragmatic RCT comparing fixation methods, 15 patients cross from their assigned operation to another. What analysis best preserves randomisation for the primary estimate?

A. Exclude all crossovers from analysisB. Assign all crossover patients to the procedure receivedC. Analyse only perfect protocol adherersD. Select whichever analysis gives the smallest p valueE. Analyse participants in the groups to which they were randomised

View detailed explanation, examiner trap and viva

Correct answer: E. Analyse participants in the groups to which they were randomised

Why correct: Intention-to-treat generally retains initial groups and reduces post-randomisation selection bias.

Why other options are incorrect:

A. Exclude all crossovers from analysis: Exclusion after allocation undermines randomisation.

B. Assign all crossover patients to the procedure received: As-treated analysis changes group balance.

C. Analyse only perfect protocol adherers: Per-protocol may answer a different question but is not primary ITT.

D. Select whichever analysis gives the smallest p value: This promotes selective reporting.

Exam pearl: Intention to treat estimates the effect of assignment under trial conditions.

Examiner trap: Missing outcome data still requires principled handling.

Viva extension: When is per-protocol useful? As a prespecified supplementary sensitivity analysis.

Question 28 of 40 · Screening bias

Q28. A new MRI screening test is evaluated only among patients with obvious advanced disease and completely healthy volunteers. What is a key concern?

A. Lead-time bias onlyB. No bias because there is a reference standardC. Publication bias necessarilyD. Spectrum bias may exaggerate diagnostic performanceE. Random error has been eliminated

View detailed explanation, examiner trap and viva

Correct answer: D. Spectrum bias may exaggerate diagnostic performance

Why correct: Extreme cases and controls do not reflect the clinical grey zone in which tests are used.

Why other options are incorrect:

A. Lead-time bias only: Lead-time bias concerns earlier diagnosis lengthening measured survival without changing death time.

B. No bias because there is a reference standard: Choice of participants can still distort sensitivity and specificity.

C. Publication bias necessarily: Publication bias concerns which studies are published.

E. Random error has been eliminated: Any finite sample has sampling variability.

Exam pearl: Diagnostic accuracy depends on an appropriate clinical spectrum.

Examiner trap: Case-control diagnostic studies may overestimate clinical performance.

Viva extension: What sample would be better? Consecutive appropriate patients with suspected disease.

Question 29 of 40 · Lead-time bias

Q29. A cancer-screening programme detects tumours two years earlier, and measured survival from diagnosis rises, but time of death is unchanged. What explains the observation?

A. Length-time bias aloneB. Lead-time biasC. Recall biasD. Attrition biasE. Observer-expectancy bias

View detailed explanation, examiner trap and viva

Correct answer: B. Lead-time bias

Why correct: Earlier diagnosis can lengthen observed survival time without delaying death.

Why other options are incorrect:

A. Length-time bias alone: Length-time bias concerns over-representation of indolent conditions.

C. Recall bias: Recall bias involves differential reporting of past exposure.

D. Attrition bias: Attrition arises from differential loss to follow-up.

E. Observer-expectancy bias: This relates to assessor expectations influencing measurements.

Exam pearl: Screening outcomes should favour mortality or relevant patient-centred endpoints.

Examiner trap: Longer survival after diagnosis may be an artefact.

Viva extension: What is overdiagnosis? Identification of disease that would not become symptomatic.

Question 30 of 40 · Number needed to harm

Q30. A new analgesic causes clinically important nausea in 15% versus 5% with standard analgesia. What is number needed to harm?

A. 10B. 2C. 5D. 20E. 100

View detailed explanation, examiner trap and viva

Correct answer: A. 10

Why correct: Absolute risk increase = 0.15 - 0.05 = 0.10, so NNH = 1/0.10 = 10.

Why other options are incorrect:

B. 2: That implies a 50% absolute harm increase.

C. 5: That implies a 20% risk increase.

D. 20: That implies a 5% risk increase.

E. 100: That implies a 1% risk increase.

Exam pearl: NNH = 1 / absolute risk increase.

Examiner trap: A treatment can have both a favourable NNT and an unfavourable NNH.

Viva extension: How does clinical decision-making use both? Compare benefits, harms and patient values.

Question 31 of 40 · Diagnostic decision

Q31. A negative test has sensitivity 98% in a low-prevalence clinic, but a patient has high clinical suspicion of malignancy. What is the safest interpretation?

A. A negative result rules out every diseaseB. Interpret the negative test in the context of pretest probability and pursue appropriate further assessmentC. Negative predictive value is fixed across settingsD. A positive result would always confirm diseaseE. Sensitivity alone determines treatment benefit

View detailed explanation, examiner trap and viva

Correct answer: B. Interpret the negative test in the context of pretest probability and pursue appropriate further assessment

Why correct: Even highly sensitive tests can miss disease; clinical risk and likelihood ratios determine residual post-test probability.

Why other options are incorrect:

A. A negative result rules out every disease: No imperfect test excludes disease with certainty.

C. Negative predictive value is fixed across settings: NPV varies with prevalence and clinical pretest probability.

D. A positive result would always confirm disease: Positive results can be false positives.

E. Sensitivity alone determines treatment benefit: Treatment decisions depend on more than diagnostic test sensitivity.

Exam pearl: Pretest probability matters even for a negative test.

Examiner trap: Do not confuse test accuracy with clinical safety of ruling out.

Viva extension: What helps quantify the change? Negative likelihood ratio and post-test odds.

Question 32 of 40 · Clinical audit standards

Q32. A surgical unit measures the proportion of emergency laparotomies receiving antibiotics within a locally approved target derived from guidance. What type of activity is this?

A. A randomised clinical trialB. A case reportC. Basic laboratory researchD. Clinical audit against an explicit standardE. A qualitative ethnographic study

View detailed explanation, examiner trap and viva

Correct answer: D. Clinical audit against an explicit standard

Why correct: Audit compares existing practice with a defined criterion and identifies gaps.

Why other options are incorrect:

A. A randomised clinical trial: No participants are assigned to experimental interventions.

B. A case report: The unit measures service-level performance, not one exceptional patient.

C. Basic laboratory research: It assesses clinical service delivery.

E. A qualitative ethnographic study: It measures compliance rather than culture through observation/interview methods.

Exam pearl: Audit needs an explicit standard or criterion.

Examiner trap: Descriptive data collection alone may be a service evaluation, not a completed audit cycle.

Viva extension: What is the next step after discovering a gap? Implement change and re-measure.

Question 33 of 40 · Audit cycle

Q33. After an audit finds only 60% compliance with surgical VTE assessments, staff education is introduced. What should happen next to close the audit loop?

A. Repeat measurement against the same standard after the interventionB. Publish the baseline figure onlyC. Change the standard to the observed performanceD. Stop measuring because staff were educatedE. Randomise patients to withholding VTE assessment

View detailed explanation, examiner trap and viva

Correct answer: A. Repeat measurement against the same standard after the intervention

Why correct: Re-audit demonstrates whether the implemented change improved compliance.

Why other options are incorrect:

B. Publish the baseline figure only: Publication alone does not establish improvement.

C. Change the standard to the observed performance: An evidence-based standard should not be weakened merely to improve apparent compliance.

D. Stop measuring because staff were educated: Education is an intervention, not an outcome measure.

E. Randomise patients to withholding VTE assessment: Withholding standard care is not an appropriate audit response.

Exam pearl: Measure → compare → change → remeasure.

Examiner trap: An audit without re-audit may leave improvement unverified.

Viva extension: What if the result remains poor? Reassess barriers and plan another cycle.

Question 34 of 40 · PDSA and QI

Q34. A theatre team tests a revised sign-in prompt on one operating list, reviews failures the next day and modifies it before scaling. Which method is most closely illustrated?

A. Meta-analysisB. Case-control samplingC. Plan–Do–Study–Act cycleD. Cross-sectional prevalence surveyE. Kaplan–Meier analysis

View detailed explanation, examiner trap and viva

Correct answer: C. Plan–Do–Study–Act cycle

Why correct: A small iterative test followed by study and adaptation is typical quality-improvement methodology.

Why other options are incorrect:

A. Meta-analysis: Meta-analysis pools estimates from multiple studies.

B. Case-control sampling: Case-control design samples based on outcome status.

D. Cross-sectional prevalence survey: This assesses a snapshot rather than iterating change.

E. Kaplan–Meier analysis: This estimates time-to-event survival curves.

Exam pearl: PDSA is iterative; it is not a substitute for measuring outcomes.

Examiner trap: Clinical audit and QI overlap but have different immediate emphases.

Viva extension: What should be tracked? Process, outcome and balancing measures.

Question 35 of 40 · Audit denominator

Q35. An audit reports that 44 of 50 eligible patients received an appropriate VTE risk assessment. What is compliance?

A. 44%B. 6%C. 12%D. 110%E. 88%

View detailed explanation, examiner trap and viva

Correct answer: E. 88%

Why correct: Compliance = 44/50 x 100 = 88%.

Why other options are incorrect:

A. 44%: The numerator is a count, not the percentage.

B. 6%: Six represents patients missed, corresponding to 12% non-compliance.

C. 12%: This is non-compliance, not compliance.

D. 110%: The count cannot yield more than 100% compliance.

Exam pearl: Define the eligible denominator before starting data collection.

Examiner trap: Changing exclusions after seeing results introduces bias.

Viva extension: What is the missed proportion? 12%.

Question 36 of 40 · Audit sampling

Q36. A ward audit selects only weekday elective operations to assess antibiotic prophylaxis performance across emergency and elective surgery. The major problem?

A. It increases the study sample to all surgeryB. It automatically creates an RCTC. It eliminates confoundingD. It directly measures patient satisfactionE. Selection bias and poor representativeness of the target service

View detailed explanation, examiner trap and viva

Correct answer: E. Selection bias and poor representativeness of the target service

Why correct: Weekday elective lists may differ from night/weekend emergencies, limiting the assessment of overall compliance.

Why other options are incorrect:

A. It increases the study sample to all surgery: The design excludes a large service segment.

B. It automatically creates an RCT: No random treatment allocation has occurred.

C. It eliminates confounding: Selective inclusion does not remove bias.

D. It directly measures patient satisfaction: The measurement concerns antibiotic processes.

Exam pearl: Your audit sampling frame must match the population you want to improve.

Examiner trap: Convenience samples may systematically miss weak points.

Viva extension: What is one better approach? Include predefined representative case types and shifts.

Question 37 of 40 · Balancing measures

Q37. A hospital reduces length of stay after hernia surgery. Which balancing measure helps detect unintended harm?

A. Unplanned readmission within 30 daysB. Number of leaflets printedC. Staff attendance at the launch presentationD. Monthly count of clinic postersE. Total beds in the hospital

View detailed explanation, examiner trap and viva

Correct answer: A. Unplanned readmission within 30 days

Why correct: Earlier discharge may be beneficial, but readmission identifies a possible trade-off and should be tracked.

Why other options are incorrect:

B. Number of leaflets printed: Printing is not a patient safety balance.

C. Staff attendance at the launch presentation: Attendance measures implementation, not unintended harm.

D. Monthly count of clinic posters: Posters do not capture patient outcomes.

E. Total beds in the hospital: Bed count is not a sensitive balancing measure for discharge safety.

Exam pearl: Use process, outcome and balancing measures together.

Examiner trap: One improved metric can hide deterioration elsewhere.

Viva extension: What other balancing measures apply? ED returns, complications or patient-reported burden.

Question 38 of 40 · Audit governance

Q38. A resident exports identifiable clinical audit data to a personal unencrypted laptop for convenience. Which action is most appropriate?

A. Proceed because audit never involves confidential informationB. Stop the insecure transfer and use approved access, minimisation and information-governance arrangementsC. Upload the file to a public sharing serviceD. Share raw identifiers with any interested traineeE. Assume a pseudonym is equivalent to complete anonymity

View detailed explanation, examiner trap and viva

Correct answer: B. Stop the insecure transfer and use approved access, minimisation and information-governance arrangements

Why correct: Audit is a legitimate improvement activity but still requires appropriate data protection, permissions and secure handling.

Why other options are incorrect:

A. Proceed because audit never involves confidential information: Audit datasets can contain identifiable health data.

C. Upload the file to a public sharing service: This increases privacy risk.

D. Share raw identifiers with any interested trainee: Access should be authorised and limited.

E. Assume a pseudonym is equivalent to complete anonymity: A linkable pseudonym may still be personal data.

Exam pearl: Data protection applies to audit and QI, not just research.

Examiner trap: Pseudonymised does not necessarily mean anonymous.

Viva extension: How could risk be reduced? Use approved systems and minimum necessary variables.

Question 39 of 40 · Research versus audit

Q39. Two teams both record outcomes after appendicectomy. Team A checks local adherence to a recognised standard; Team B allocates patients to untested intervention to estimate efficacy. Which statement is most accurate?

A. Both automatically require no review because they involve routine patientsB. Both are necessarily the same because they collect outcomesC. Team A cannot publish its findingsD. Team A is primarily audit; Team B is interventional research requiring appropriate governanceE. Team B becomes audit by using a local protocol

View detailed explanation, examiner trap and viva

Correct answer: D. Team A is primarily audit; Team B is interventional research requiring appropriate governance

Why correct: Activity classification depends on whether care is assessed against a standard or experimental intervention is assigned to answer a research question.

Why other options are incorrect:

A. Both automatically require no review because they involve routine patients: The trial requires research processes.

B. Both are necessarily the same because they collect outcomes: Purpose and design differ.

C. Team A cannot publish its findings: Audit reports may be disseminated with appropriate governance.

E. Team B becomes audit by using a local protocol: A local label does not change the experimental nature.

Exam pearl: Clarify project classification before collecting or allocating patients.

Examiner trap: Publication intention alone does not define research.

Viva extension: Who can clarify borderline projects? Research governance and clinical audit leads.

Question 40 of 40 · Run chart interpretation

Q40. After a change in theatre practice, a run chart shows eight consecutive monthly points all above the pre-intervention median. What is the most appropriate interpretation?

A. This proves the intervention caused the changeB. It proves the data were fabricatedC. This suggests a non-random shift that warrants further investigationD. The median must immediately be deletedE. No measurement error is possible

View detailed explanation, examiner trap and viva

Correct answer: C. This suggests a non-random shift that warrants further investigation

Why correct: A sustained run on one side of a stable baseline median suggests special-cause change, though the exact run rule depends on the framework used.

Why other options are incorrect:

A. This proves the intervention caused the change: Temporal patterns alone do not eliminate competing causes.

B. It proves the data were fabricated: An extended run can occur from genuine process change.

D. The median must immediately be deleted: Retain a baseline to interpret shifts before updating it appropriately.

E. No measurement error is possible: Run charts do not establish perfect data quality.

Exam pearl: Run charts examine patterns over time, not just before/after means.

Examiner trap: Do not infer causation from an uncontrolled shift alone.

Viva extension: What further checks help? Case-mix, concurrent changes, data quality and sustainability.

Answers and score save in this browser. Your checklist and quiz are stored separately.

Frequently asked questions

Are the ethics MCQs based on the law of one UK country?

They teach broadly relevant GMC principles. Capacity and some consent processes differ between England, Wales, Scotland and Northern Ireland, so use current local law for real cases.

Which statistics formulas should I memorise?

Sensitivity, specificity, PPV, NPV, LR+, LR−, RR, ARR, RRR, OR, NNT and NNH are core high-yield concepts. Prioritise choosing the right denominator.

Is a clinical audit the same as research?

No. Audit generally measures care against explicit standards; interventional research evaluates questions under a research protocol. Governance classification can be nuanced.

Do progress ticks sync to another device?

No. Progress uses browser localStorage on the current browser and website origin only.

How do I answer ethics SBAs under time pressure?

Identify capacity, voluntariness, risk of harm, immediate danger and any confidentiality concern. Reject options that conceal harm or bypass patient involvement.

References and MRCS Academy

Published clinical and professional references supporting the concepts in this original educational article. No actual examination recall text has been reproduced.

1. Intercollegiate MRCS — current candidate guidance
2. GMC — Seven principles of decision making and consent
3. GMC — Confidential information for direct care
4. GMC — Good Medical Practice: patients, partnership and communication
5. NHS England — Clinical audits and registries best practice (2026)
6. NICE — Principles for best practice in clinical audit
7. BMJ — Diagnostic tests: sensitivity and specificity
8. BMJ — Diagnostic tests: predictive values
9. BMJ — Diagnostic tests: likelihood ratios

Published MRCS Academy lessons

Article 01
Start MRCS Here Article 02
Complete Syllabus Article 03
Books and Resources Article 04
12-Week Study Plan Article 05
SBA Question Strategy Article 06
Common Preparation Mistakes Article 07
Surgical Anatomy Overview Article 08
Upper Limb Anatomy Article 09
Inguinal & Abdominal Wall Article 10
Head & Neck Surgical Anatomy Article 11
Cardiovascular & Respiratory Physiology Article 12
Renal & Acid–Base Physiology Article 13
GI, Endocrine & Neurophysiology Article 14
Lower Limb Surgical Anatomy Made Easy for MRCS Part A Article 15
Thoracic, Pelvic and Visceral Surgical Anatomy Article 16
Surgical Pathology, Inflammation and Wound Healing Made Easy for MRCS Part A Article 17
Surgical Pharmacology and Microbiology Made Easy for MRCS Part A Article 18
Perioperative Care, Anaesthesia and Surgical Complications Made Easy for MRCS Part A Article 19
Trauma, Shock and Burns Made Easy for MRCS Part A Article 20
Acute Abdomen and Emergency Surgical Conditions Made Easy for MRCS Part A Article 21
Common Surgical Specialty Presentations Made Easy for MRCS Part A

Previous lesson: Article 21 — Common Surgical Specialty Presentations

Next lesson: Article 23 — Full MRCS Part A Mock Examinations (coming next; its URL will be added once published).

FreeMedSite MRCS Academy. This page is for education and revision, not individual medical or legal advice. Review date: 17 October 2026. © FreeMedSite — original educational text, quiz and schematic diagrams.