Comparing treatment variances in repeated measures bioavailability trials

Area of Science:

• Pharmacokinetics and Biopharmaceutics
• Statistical Methods in Clinical Trials
• Drug Formulation Analysis

Background:

• Repeated measures bioavailability/bioequivalence studies are crucial for comparing drug formulations.
• Plasma concentration data, including Area Under the Curve (AUC) and Maximum Concentration (Cmax), often exhibit a block compound-symmetric covariance structure.
• Assessing the equality of variances within this covariance matrix is essential for robust study conclusions.

Purpose of the Study:

• To develop and evaluate statistical tests for homogeneity of variances in repeated measures bioavailability/bioequivalence studies.
• To compare the performance of an exact test against four asymptotic tests in terms of Type I error rate control and statistical power.
• To provide practical guidance on selecting appropriate tests for analyzing formulation data.

Main Methods:

• Derivation of an exact statistical test for variance equality.
• Development of four asymptotic tests for the same purpose.
• Monte Carlo simulations to compare the Type I error rate and power of the proposed tests.
• Application of the methods to a real-world bioavailability/bioequivalence study example.

Main Results:

• The study presents an exact test and four asymptotic tests for assessing variance equality in bioavailability/bioequivalence data.
• Simulation results indicate which tests offer optimal control of Type I error rates and superior power.
• The proposed methods are demonstrated with a practical example, facilitating their application.

Conclusions:

• The developed statistical tests provide valuable tools for analyzing variance in bioavailability/bioequivalence studies.
• The comparative analysis aids researchers in selecting the most reliable tests for formulation assessment.
• The findings contribute to more accurate and powerful statistical inference in pharmacokinetic studies.