Epidemiology EPI202-01 Study Guide - Delta Method, Null Hypothesis, Logistic Regression

• School: Harvard University
• Department: Epidemiology
• Course: Epidemiology EPI202-01
• Professor: Murray Mittleman

Document Summary

Measures: population risk difference (prd) = rt runexp = rd * prev(exposure) = excess incidence in total study population due to exposure. *see roadmap and derivations sheet for most formulas* Measures: prevalence (binomial): # existing cases of a disease / total study population @ a given point. Period: point prevalence + ci over period of time t-year cumulative incidence (binomial): # events that occur between t0 and t1 / Relationships between measures: t-year ci = 1-exp{- ij tj} When ir is constant over t, then t-year ci = 1-exp{-i*t} When i t is small (i. e. < 0. 1) , c ~ i*t: p = (i*d) / (1+i*d) where d = average disease duration, p = prevalence. Or = [c1/(1-c1)] / [c0/(1-c0)] = [c1(1-c0)] / [c0(1-c1)], where c1 / c0 = cir. So when the ci 1, then the or > cir. If binomial distribution: mean = np; variance = npq, if a proportion: mean = p; variance = [p(1-p)]/n.