Duffy-Santner confidence intervals for the two-stage three-outcome design

Area of Science:

• Biostatistics
• Clinical Trial Design
• Pharmaceutical Research

Background:

• The three-outcome phase II clinical trial design is increasingly utilized.
• Existing confidence interval (CI) methodology was not explicitly defined for the binary endpoint in this design.
• Standard phase II trials with binomial endpoints commonly employ the sequential binomial confidence interval (SBCI) algorithm.

Purpose of the Study:

• To define and describe the construction of confidence intervals for the three-outcome phase II clinical trial design.
• To adapt the sequential binomial confidence interval (SBCI) algorithm for use in three-outcome settings.
• To validate the application of SBCI methodology in three-outcome clinical trials.

Main Methods:

• Review of existing sequential binomial confidence interval (SBCI) literature.
• Examination of SBCI software implementations.
• Application of the SBCI algorithm to the three-outcome phase II clinical trial design framework.
• Conducting simulations to assess CI coverage.

Main Results:

• A clear methodology for constructing confidence intervals in three-outcome phase II trials was described.
• The sequential binomial confidence interval (SBCI) algorithm was successfully adapted for the three-outcome setting.
• Simulation results demonstrated that the SBCI algorithm provides appropriate coverage for the three-outcome design.

Conclusions:

• The sequential binomial confidence interval (SBCI) algorithm is suitable for calculating confidence intervals in three-outcome phase II clinical trials.
• This work provides essential statistical methodology for the growing use of three-outcome trial designs.
• Proper application of the SBCI algorithm ensures reliable statistical inference in these advanced trial designs.