Regulatory authorities require demonstration of bioequivalence through
comparisons of different pharmacokinetic parameters, the area under t
he plasma concentration-time curve (AUC), the maximum plasma concentra
tion (C-max), and the time to reach peak concentration (T-max). The ap
plicability and validity of regulatory requirements have been widely c
riticized on statistical and clinical relevance grounds. For most nonc
omplicated absorption models, the AUC correlates well with the extent
of absorption. However, in nonlinear models of absorption, in mechanis
ms involving recycling of drugs, and for drugs with long half-life, th
e use of total AUC (from zero to infinity) can give erroneous and clin
ically irrelevant results since the area is mostly determined by elimi
nation phase or by recycling. The calculation of total AUC also involv
es prolonged sampling, adding to the cost and risks associated with bi
oequivalence studies. The use of C-max or T-max as a measure of rate o
f absorption, to correlate with clinical relevance, is widely criticiz
ed on logical, technical, and statistical grounds. For drugs used on a
multiple-dose basis, C-max and T-max evaluations become redundant sin
ce the average plateau concentration is not affected by these paramete
rs. To resolve the drawbacks in the traditional methodology of bioequi
valence evaluation, the use of partial areas in lieu of total AUC, T-m
ax and C-max is suggested. This study investigates the logic and robus
tness of the partial-area method in establishing bioequivalence. We co
nclude that the 5 h AUC is a more relevant parameter to establish napr
oxen bioequivalence than AUC(inf). We recommend against using symmetri
cal confidence intervals and report excellent agreement among several
methods of calculating confidence intervals, probability values, and n
onparametric tests. We suggest that a single-point short-term AUC is a
better indicator of the bioequivalence of generic products than the t
otal AUC, C-max and T-max as required currently by the regulatory auth
orities. (C) 1997 by John Wiley & Sons, Ltd.