A Rapidly Converging Algorithm for Exact Binomial Confidence Intervals About the Relative Risk in Follow-up Studies with Stratified Incidence-Density Data

Abstract

A rapidly converging algorithm is given to calculate exact confidence intervals about the adjusted relative risk in follow-up studies with stratified incidence-density data. The network approach that Mehta developed for tables with person-count numerators and denominators is adapted to tables with person-count numerators and person-time denominators. This algorithm updates an earlier program by Guess et al, yielding the same quantities but with running times that are between ten and a hundred times faster. Applications to Poisson regression are discussed.