Bayesian sample sizes for exploratory clinical trials comparing multiple experimental treatments with a control.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Bayesian Inference

Background:

• Simultaneous comparison of multiple experimental treatments against a common control is crucial in exploratory clinical trials.
• Traditional frequentist methods often require large sample sizes, potentially limiting early-stage research.
• Incorporating prior knowledge can enhance the efficiency of clinical trial designs.

Purpose of the Study:

• To develop a Bayesian approach for sample size calculation in exploratory clinical trials comparing multiple treatments to a control.
• To ensure study designs provide strong evidence for treatment superiority or a lack of substantial effect.
• To explore methods that allow for uncertainty in standard deviation estimation.

Main Methods:

• A Bayesian framework is employed for simultaneous comparison of multiple treatments against a control.
• Sample size is determined by the probability of concluding treatment promise or lack of efficacy.
• The approach accommodates normally distributed responses with a common standard deviation, initially assumed known, then explored with Bayesian priors.

Main Results:

• The proposed Bayesian method allows for a reduction in sample size compared to conventional frequentist approaches.
• Illustrations demonstrate computed sample sizes, highlighting the practical implications of the Bayesian design.
• The method provides a structured way to integrate prior beliefs into the sample size determination process.

Conclusions:

• The developed Bayesian approach offers an efficient alternative for sample size calculation in exploratory clinical trials.
• This method facilitates stronger conclusions regarding treatment efficacy, supporting informed decisions for Phase III development.
• The flexibility in handling standard deviation uncertainty enhances the robustness of the Bayesian design.