Decision-theoretic designs for small trials and pilot studies: A review

Area of Science:

• Biostatistics
• Clinical Trial Design
• Decision Theory

Background:

• Small clinical trials are crucial but lack standardized design methodologies.
• Bayesian decision theory presents a potential framework for optimizing these designs.

Purpose of the Study:

• To review and summarize current methodological developments in applying Bayesian decision theory to small clinical trial design.
• To provide an overview of existing decision-theoretic approaches for various trial structures.

Main Methods:

• Literature review of published methods for Bayesian decision-theoretic trial design.
• Categorization of methods based on trial structure: single-stage, multi-stage (with interim analyses), and sequential/programmatic designs.

Main Results:

• Decision-theoretic methods have been successfully applied across diverse areas of small clinical trial design.
• Various methods exist for single-stage, multi-stage, and series of trials, tailored to specific decision-maker perspectives and utility functions.

Conclusions:

• Bayesian decision theory provides a robust framework for designing small clinical trials.
• The choice of method depends on the trial's stage, interim decision points, and the specific utility functions employed.