Pharmaceutical companies conduct bioequivalence studies to demonstrate that a new formulation of a drug product ('Test') is sufficiently similar in rate and extent of absorption to the current formulation ('Reference'). Such similarity permits the efficacy and safety results for the Reference formulation to carry forward to the Test formulation. Anderson and Hauck identified two types of bioequivalence criteria, population and individual. Individual bioequivalence is a criterion for individuals' similarity of bioavailability from the two formulations. We need this criterion to ensure 'switchability', that is, to ensure that a patient who has been taking one formulation can switch to the other with similar safety and efficacy.
Population bioequivalence, the topic of this paper, is a generalization of the current bioequivalence criterion of average bioequivalence. The average bioequivalence criterion is for similar average bioavailabilities of two formulations. Nothing in the average bioequivalence criterion formally considers similarity of variances. Population bioequivalence generalizes average bioequivalence by specifying that the full distributions, not just their means, of bioavailability of the two formulations are sufficiently similar. Population bioequivalence is the criterion applicable whenever patients could not switch between the two formulations, such as for comparing two formulations early in product development, or for comparing subpopulations, such as those defined by gender.