The Asymptotic Distribution of the Partial Attributable Risk in Cross-Sectional Studies

The partial attributable risk has been introduced as a tool for partitioning the responsibility for causing an adverse event between various risk factors. It has arisen from epidemiology but is also a valid general risk allocation concept which can for example be applied to data from customer satisfaction surveys. So far, a variance formula for the partial attributable risk has been missing so that confidence intervals were not directly available. This paper provides the asymptotic normal distribution for the partial attributable risk when determined from a cross-sectional study