Decision theoretic designs for phase II clinical trials with multiple outcomes

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Decision Science

Background:

• Phase II clinical trials require evaluating both treatment efficacy and safety.
• Existing phase II designs often lack a decision-theoretic foundation for handling multiple outcomes.
• There is a need for a principled approach to construct phase II designs considering various factors.

Purpose of the Study:

• To develop a Bayesian decision theoretic strategy for constructing phase II clinical trial designs.
• To incorporate both efficacy and safety (adverse events) into the design framework.
• To provide a flexible strategy applicable to various clinical settings.

Main Methods:

• A Bayesian decision theory framework was employed.
• A gain function was defined, including utilities for patient outcomes, reward for promising treatment declaration, and costs for phase II and phase III trials.
• Methods for eliciting gain function parameters and evaluating frequentist operating characteristics were described.

Main Results:

• A novel strategy for phase II design construction based on decision theory was presented.
• The gain function effectively balances efficacy, safety, and economic considerations.
• The strategy was successfully illustrated using a peripheral blood stem cell transplantation trial for multiple myeloma.

Conclusions:

• The proposed Bayesian decision theoretic strategy offers a robust method for phase II clinical trial design.
• This approach allows for a more informed and optimized decision-making process in early-phase drug development.
• The framework provides a valuable tool for researchers designing clinical trials with multiple objectives.