Use of the repeated cross-over design in assessing bioequivalence: (within and between subjects variability - Schuirmann Confidence Intervals estimation).

In 1992, the Division of Bioequivalence in the Office of Generic Drugs published a guide to Statistical procedures for bioequivalence studies using a standard two-treatments cross-over design (1). This paper describes the application of the guidelines to a practical protocol and the recent Proc MIXED (SAS) will be shown to be much more convenient than the traditional Proc GLM for theoretical and practical reasons (correct estimation of residuals, analysis of the within-subjects variation, direct calculation of the Schuirmann 90% Confidence Intervals). This new procedure was applied to a study protocol on riluzole (Rilutek) including a replicate design with the within-subject and between-subject variances being estimated on Cmax and AUC biopharmaceutic parameters.