An Improved Bayesian Pick-the-Winner (IBPW) Design for Randomized Phase II Clinical Trials

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Drug Development

Background:

• Phase II clinical trials are crucial for identifying effective drug candidates for further development.
• Efficient trial designs with small sample sizes and early stopping are essential.
• Randomized multi-arm designs are increasingly used, but balancing efficiency and accuracy remains a challenge.

Purpose of the Study:

• To introduce an improved randomized two-stage Bayesian pick-the-winner (IBPW) design for phase II clinical trials.
• To formalize hypothesis testing for winner selection and optimize sample sizes and decision cut-offs.
• To strictly control type I and type II errors in the winner selection process.

Main Methods:

• Development of the improved randomized two-stage Bayesian pick-the-winner (IBPW) design.
• Formalization of winner-selection based hypothesis testing.
• Optimization of sample sizes and decision cut-offs under flexible hypotheses.
• Simulation studies to evaluate operating characteristics.

Main Results:

• The IBPW design formalizes winner-selection hypothesis testing.
• Optimized sample sizes and decision cut-offs strictly control type I and type II errors.
• Simulation studies show improved operating characteristics for winner selection compared to existing methods.
• The IBPW design retains desirable features of the original Bayesian pick-the-winner design.

Conclusions:

• The IBPW design offers a statistically rigorous and efficient approach for phase II randomized multi-arm trials.
• This design improves upon the Bayesian pick-the-winner method by offering better control over statistical errors.
• The IBPW design facilitates more reliable selection of drug candidates for phase III testing.