Master Population Health & Ethics
for AMC Cat 1

What the AMC Cat 1 Tests in Population Health & Ethics

This exam tests your ability to apply epidemiological principles and ethical reasoning to clinical scenarios. You must interpret screening test characteristics (sensitivity, specificity, PPV, NPV) using real thresholds like HbA1c ≥48 mmol/mol for diabetes or fasting glucose ≥7.0 mmol/L. You need to calculate NNT for interventions such as statins in primary prevention (e.g., atorvastatin 20 mg) and understand number needed to harm (NNH) for drugs like warfarin (major bleed ~2% per year). Ethical questions focus on valid consent (Gillick competence for under-16s), refusal of life-saving treatment (competent adult), and mandatory reporting (notifiable diseases: meningococcal meningitis, tuberculosis, COVID-19). You must know the four principles (autonomy, beneficence, non-maleficence, justice) and how to balance them in resource allocation, e.g., NICE threshold of £20,000–30,000 per QALY.

High-Yield Concepts

• Screening Test Performance: Sensitivity = TP/(TP+FN); Specificity = TN/(TN+FP). For a condition with 5% prevalence (e.g., undiagnosed HIV), a test with 99% sensitivity and 95% specificity yields PPV = (0.05*0.99)/(0.05*0.99 + 0.95*0.05) ≈ 51%. Always recalculate PPV with disease prevalence.
• Number Needed to Treat (NNT): NNT = 1/ARR. For primary prevention with simvastatin 40 mg over 5 years, absolute risk reduction for major vascular events is about 2.0% (ARR = 0.02), so NNT = 50. For secondary prevention, NNT ≈ 20. Use absolute risk, not relative risk, for clinical decisions.
• Vaccination and Herd Immunity: Herd immunity threshold = 1 – (1/R0). For measles (R0 = 12–18), threshold is 92–94%. With MMR vaccine efficacy ~97% (two doses), coverage must exceed 95% to prevent outbreaks. Current UK schedule: first dose at 12 months, second at 3 years 4 months.
• Informed Consent and Capacity: Use the Mental Capacity Act 2005 (England/Wales): a person lacks capacity if unable to understand, retain, weigh, or communicate a decision. For children <16, Gillick competence is assessed—if competent, they can consent but not necessarily refuse life-saving treatment (courts may overrule).
• Confidentiality and Disclosure: Breach confidentiality only if required by law (e.g., notifiable diseases: meningococcal septicaemia, TB, food poisoning) or if there is serious risk of harm to an identifiable third party (e.g., patient with HIV refusing to disclose to partner). Document rationale. Use GMC guidance.
• Resource Allocation and QALYs: NICE uses a cost-effectiveness threshold of £20,000–30,000 per QALY gained. For example, sofosbuvir for hepatitis C (SVR >95%) was initially rejected due to high cost (~£35,000/QALY) but later approved after price negotiation. Understand opportunity cost and the 'rule of rescue'.
• Epidemiological Measures of Association: Relative risk (RR) = incidence in exposed / incidence in unexposed. Attributable risk (AR) = incidence in exposed – incidence in unexposed. For smoking and lung cancer, RR ≈ 15–30, AR ≈ 0.1–0.2 per year (depending on pack-years). Odds ratio approximates RR for rare diseases (prevalence <10%).
• Mandatory Reporting and Notifiable Diseases: Under the Health Protection (Notification) Regulations 2010, doctors must notify the local health protection team of suspected cases of specified diseases: acute meningitis, measles, mumps, rubella, TB, food poisoning, and COVID-19. Failure to report is a legal offence.

Common Traps in Population Health & Ethics Questions

• Confusing sensitivity with PPV: a highly sensitive test (e.g., D-dimer for PE) has low PPV in low-prevalence populations.
• Thinking that NNT is fixed: it depends on baseline risk—NNT is lower in high-risk patients (e.g., secondary prevention vs primary).
• Assuming a competent adult can always refuse treatment: if refusal leads to death and pregnancy is involved, courts may intervene to protect the fetus (e.g., forced caesarean in extremis).
• Believing that all breaches of confidentiality are unethical: disclosure is justified to prevent serious harm (e.g., a violent patient with psychosis threatening to kill).
• Misapplying Gillick competence: a child may be Gillick-competent to consent to contraception but not to refuse chemotherapy for leukaemia.
• Forgetting that screening tests with high specificity (e.g., 99%) still produce false positives if prevalence is very low (e.g., 0.1%).

How to Revise Population Health & Ethics for the AMC Cat 1

Focus on interpreting 2x2 tables and calculating sensitivity, specificity, PPV, NPV, and NNT/NNH from raw numbers. Practice ethical scenarios where you must identify the principle (e.g., autonomy vs beneficence) and justify your decision using GMC or legal frameworks. For population health, know the UK vaccination schedule and screening programmes (e.g., NHS bowel cancer screening: FIT every 2 years ages 60–74). Questions often present a clinical vignette (e.g., a 45-year-old with incidental hypertension) and ask for the most appropriate ethical action or the epidemiological term. Revise the Mental Capacity Act and Gillick competence with case examples. Do not memorise obscure statistics; instead, practise rapid calculation of ARR and NNT from trial data (e.g., 5-year event rates 8% vs 6%).

Practise it: MedLumen has 100 Population Health & Ethics questions for the AMC Cat 1, each with a full explanation and references.