Robust Bayesian sample size determination in clinical trials.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Bayesian Inference

Background:

• Determining adequate sample size is crucial for clinical trial success.
• Bayesian approaches offer a flexible framework for sample size determination.
• Sensitivity to prior assumptions is a key challenge in Bayesian trial design.

Purpose of the Study:

• To develop a robust Bayesian sample size criterion for clinical trials.
• To ensure trial success by achieving a high posterior probability of a treatment effect exceeding a threshold.
• To address the sensitivity of Bayesian methods to prior assumptions.

Main Methods:

• Utilizing a Bayesian approach to define trial success based on posterior probability.
• Implementing a robust sample size criterion considering a class of plausible prior distributions.
• Employing epsilon-contamination priors for their mathematical tractability and flexibility.
• Focusing on the predictive distribution of the posterior probability's lower bound.

Main Results:

• A robust sample size methodology is proposed that accounts for prior uncertainty.
• The method provides a criterion for selecting sample sizes that guarantee a high predictive probability of trial success.
• Demonstrated application in sample size determination for Phase III clinical trials.

Conclusions:

• The proposed robust Bayesian method enhances the reliability of sample size determination in clinical trials.
• This approach mitigates concerns regarding prior assumption sensitivity.
• It offers a practical tool for ensuring sufficient statistical power and successful trial outcomes.