Equivalence testing and equivalence limits of metered-dose inhalers and dry powder inhalers measured by in vitro impaction

In this study, criteria for the acceptability of comparative in vitro equivalence testing are proposed. Furthermore, the following equivalence limits for in vitro impaction methods are postulated: the 90% confidence interval (CI) of the in vitro deposition ratio of the test product and the reference product should lie within 0.80-1.20. The aim of this study was to challenge these limits by applying them to in vitro impaction results of several groups of pressurized metered-dose inhalers and dry powder inhalers containing salbutamol and beclomethasone dipropionate. The deposition results were obtained with the Twin Impinger. All products had a marketing authorization in the Netherlands and were considered therapeutically equivalent within each group. The postulated equivalence limits/group were challenged by fictitiously assigning a preparation as a test product or reference product and calculating the 90% CI of the deposition ratio of the test and reference products. All possible combinations of products within a group were tested. The products were considered equivalent if the 90% CI of the quotient lay within 0.80-1.20. In most cases, the quotient of the test product and reference product remains within 0.80-1.20, but due to a high variability in the deposition results of several products, the 90% CI of the quotient sometimes falls outside the proposed equivalence limits. It is concluded that the equivalence limits postulated are rather conservative, with respect to accepting equivalence. The limits can therefore serve as a prudent predictor of equivalence within the acceptability criteria proposed, but have to be further validated. (C) 2000 Elsevier Science B.V. All rights reserved.