I85PREP
⌘ K
ECOM-CR Medical Licensure study materials
ECOM-CR Medical Licensure study resources

ECOM-CR Medical Licensure: Public Health, Epidemiology and Medical Ethics: Measures, Bias and Consent

ECOM-CR Medical Licensure · Work through your exam, one topic at a time.

All guide sections

Public Health, Epidemiology and Medical Ethics: Measures, Bias and Consent

The Public Health, Epidemiology and Medical Ethics domain tests study design, measures of disease frequency and association, screening statistics, bias, and the legal and ethical framework of medical practice.

Core rules

  • Incidence = new cases/population at risk per unit time; prevalence = all cases/population at a point, influenced by duration.
  • Cohort studies give relative risk directly; case–control studies give odds ratios, which approximate RR only when the outcome is rare.
  • Sensitivity = TP/(TP+FN); specificity = TN/(TN+FP); PPV = TP/(TP+FP) and rises with prevalence; NPV falls as prevalence rises.
  • NNT = 1/absolute risk reduction.
  • Bias: selection, information/recall, confounding. Randomisation and blinding reduce them. Confounding is controlled by randomisation, restriction, matching or adjustment - not by increasing sample size.
  • Prevention: primordial, primary, secondary (screening), tertiary (rehabilitation), quaternary (avoiding over-treatment).
  • Screening criteria: important condition, accepted treatment, early detection improves outcome, suitable test with adequate sensitivity and specificity, acceptable to the population, cost-effective.
  • Ethics: autonomy, beneficence, non-maleficence, justice. Valid consent requires capacity, disclosure, understanding and voluntariness. Exceptions include emergencies, statutory reporting and serious risk to others.

Worked example

A screening test is applied to 10,000 people with a prevalence of 1%. Sensitivity is 90% and specificity 95%. Disease present: 100; test positive 90 (TP), false negative 10. Disease absent: 9,900; false positives 5% × 9,900 = 495; true negatives 9,405. PPV = 90/(90 + 495) = 15.4%. Even a good test therefore yields mostly false positives in a low-prevalence population. This is why confirmatory testing is required and why programmes target higher-prevalence groups.

Traps

  • Reporting PPV as a fixed property of the test rather than one dependent on prevalence.
  • Interpreting an odds ratio as a risk ratio when the outcome is common.
  • Confusing confounding with effect modification.
  • Assuming written consent is always required, or that refusal of treatment proves incapacity.