Sample size re-estimation in a superiority clinical trial using a hybrid classical and Bayesian procedure

Area of Science:

• Biostatistics
• Clinical Trial Design
• Statistical Inference

Background:

• Sample size and power calculations are crucial for continuous endpoints.
• Traditional methods for sample size re-estimation often ignore uncertainty in parameter estimates from pilot studies.

Purpose of the Study:

• To propose a hybrid classical and Bayesian method for sample size re-estimation in two-stage study designs.
• To formally integrate prior beliefs and pilot study data into sample size adjustments.
• To address the uncertainty in parameter estimates often overlooked by traditional methods.

Main Methods:

• Utilizes a hybrid classical and Bayesian approach for sample size re-estimation.
• Introduces Conditional Expected Power (CEP) as a measure of power, averaging the power curve using prior distributions.
• Implements a procedure to determine the second-stage sample size for achieving desired interim power based on observed first-stage results.

Main Results:

• The proposed method formally integrates prior beliefs and pilot data, accounting for parameter uncertainty.
• Conditional Expected Power (CEP) provides a robust measure for interim power calculations.
• Evaluations compare the proposed CEP re-estimation method against three traditional approaches.

Conclusions:

• The hybrid Bayesian and classical method offers a more robust approach to sample size re-estimation in clinical trials.
• The CEP re-estimation method acknowledges and handles the inherent uncertainty in parameter estimates at interim analysis.
• This approach enhances the reliability of sample size adjustments in two-stage study designs.