Epidemiology
Core concepts
Epidemiology studies the distribution and determinants of health states in populations and applies that knowledge to control health problems. A candidate must command measures of frequency (incidence, prevalence), association (risk ratio, rate ratio, odds ratio), and impact (attributable risk, population attributable fraction). Incidence needs a population at risk and a defined time period; prevalence is a snapshot reflecting both incidence and duration. Descriptive studies generate hypotheses, while cohort, case-control, and trial designs test them. Causal judgement requires attention to chance, bias, confounding, and temporality; Bradford Hill viewpoints are aids, not a checklist.
Traps
Typical errors include using the total population instead of the population at risk, confusing case fatality with mortality rate, reading an odds ratio as a risk ratio when the outcome is common, the ecological fallacy, and confusing confounding with effect modification. Screening questions often hide lead-time and length-time bias. Outbreak questions usually turn on the correct attack-rate denominator.
Worked example
In a cohort of 500 exposed and 500 unexposed workers, 50 exposed and 10 unexposed develop disease. Exposed risk = 50/500 = 10%; unexposed risk = 10/500 = 2%; risk ratio = 5.0; risk difference = 8 percentage points. If exposure prevalence in the community is 20%, the population attributable fraction is Pe(RR−1)/[1+Pe(RR−1)] = 0.2(4)/[1+0.8] = 0.44, or 44%.
Rule of thumb
Always ask: who is the denominator, over what time, and is the comparison fair?