Epidemiology
Core concepts
Epidemiology questions reward precise definitions. Know the difference between incidence (new cases ÷ population at risk over time) and prevalence (all existing cases ÷ total population at a point or period). Prevalence ≈ incidence × average duration for stable conditions-so longer survival raises prevalence without changing incidence. Distinguish attack rate (incidence in a defined outbreak cohort) from secondary attack rate (new cases among susceptible contacts).
Master the study designs: cohort studies yield incidence and relative risk (RR); case-control studies yield odds ratios (OR) and are efficient for rare outcomes; cross-sectional studies measure exposure and outcome simultaneously and cannot establish temporality; ecologic studies compare group-level data and risk the ecologic fallacy. Understand confounding, effect modification, selection bias, information/recall bias, and Berkson's bias.
Traps
- Asking which measure a design can produce (case-control cannot give incidence).
- Screening questions: sensitivity = TP/(TP+FN); specificity = TN/(TN+FP); positive predictive value depends on prevalence and falls as prevalence drops-a classic trap.
- Confusing crude vs age-adjusted rates; standardization removes age confounding.
- Herd immunity thresholds depend on R0: threshold = 1 − 1/R0.
Worked rule of thumb
RR = [a/(a+b)] ÷ [c/(c+d)] for a 2×2 cohort table. If RR = 1, no association; RR > 1 suggests increased risk; RR < 1 protective. An OR approximates RR only when the outcome is rare (<10%).
Key exam habit
Identify the design first, then the measure it licenses, then check for bias or confounding-most stems hinge on that sequence.