Evaluation of equivalence of two formulations of a drug typically enta
ils the comparison of average bioavailabilities. Recently, however, au
thors have become aware that this may be insufficient to assess indivi
dual bioequivalence, that is, interchangeability of formulations on an
individual basis. This paper outlines a tolerance interval procedure
to assess individual bioequivalence based on a model that includes a s
ubject by formulation interaction. We give methods for several higher-
order cross-over designs along with examples. (C) 1997 by John Wiley &
Sons, Ltd.