Two-Stage Adaptive Designs for Three-Treatment Bioequivalence Studies

Area of Science:

• Pharmaceutical Sciences
• Biostatistics
• Clinical Trial Design

Background:

• Bioequivalence (BE) studies commonly use crossover trials, requiring within-person variance specification for sample size determination.
• Specifying within-person variance is challenging, driving interest in adaptive designs for BE trials.
• Existing adaptive designs primarily focus on two-treatment BE studies, limiting their application.

Purpose of the Study:

• To extend adaptive design methodology to multi-treatment bioequivalence trials.
• To address the need for adaptive designs in three-treatment BE studies.
• To provide a framework for designing adaptive multi-treatment BE trials.

Main Methods:

• Extension of previously proposed two-treatment adaptive BE trial designs to three-treatment scenarios.
• Discussion of specific design considerations for multiple comparisons in multi-treatment BE trials.
• Extensive simulation studies to evaluate the performance of the proposed adaptive designs.

Main Results:

• Demonstration that two-stage adaptive designs with favorable statistical operating characteristics can be developed for three-treatment BE trials.
• Comparison of the performance of proposed adaptive procedures through simulation.
• Identification of effective adaptive strategies for three-treatment BE study designs.

Conclusions:

• The proposed methodology successfully extends adaptive designs to three-treatment bioequivalence studies.
• Adaptive multi-treatment BE trials can achieve desirable statistical properties.
• This work provides valuable tools for designing more complex bioequivalence studies.