All Tools Bookmark


Facebook Share Twitter Share

AF: Model-Based Estimation of Confounder-Adjusted Attributable Fractions

Estimates the attributable fraction in different sampling designs adjusted for measured confounders using logistic regression (cross-sectional and case-control designs), conditional logistic regression (matched case-control design) and Cox proportional hazard regression (cohort design with time-to-event outcome). The variance of the estimator is obtained by combining the delta method with the the sandwich formula.