Power analyses of moment analysis parameter in bioequivalence tests

Simulated and experimental data were used to evaluate the mean residence time (MRT) as a parameter for estimating the rate of bioavailability in bioequivalence tests and to compare MRT with tmax (the time of peak drug concentration), both pharmaceutically and statistically. Although the values of MRT were more dependent on the elimination rate constant than tmax, MRTs were sufficiently sensitive to variations in the absorption rate. The statistical power of MRT infinity obtained by an extrapolation method was lower (especially when the flip-flop phenomenon occurred) than that of MRTt (calculated using data from zero time through the last sampling time). However, the power of MRTt was shown to be comparable to or higher than that of either AUCt or Cmax (the peak drug concentration) from simulated data. These results were also confirmed experimentally using previously obtained data. The effect of variances of plasma concentrations on the power of these parameters was also studied using a simulation technique. Even large variances near the last sampling time did not substantially affect the power of MRTt. Thus MRTt can be used as a parameter for estimating the rate of bioavailability from dosage forms in place of tmax in bioequivalence tests.