Abstract
A log-normal model is developed for testing π1, the probability that a subject's response will fall within given bioequivalence limits. The model is a parametric analog of Anderson and Hauck's TIER rule. Confidence intervals and hypothesis tests are derived. Statistical power is compared with the that of the TIER rule. The probability of demonstrating mean bioequivalence is shown to greatly exceed that of showing individual bioequivalence.

This publication has 9 references indexed in Scilit: