Multiple-arm superiority and non-inferiority designs with various endpoints

Area of Science:

• Biostatistics
• Clinical Trial Design
• Statistical Methodology

Background:

• Multi-arm dose-response superiority trials are common for continuous and binary data.
• Non-inferiority designs have been recently explored, primarily in two-arm trials.
• Existing methods lack a unified approach for diverse endpoints and trial designs.

Purpose of the Study:

• To develop a unified asymptotic formulation for sample size calculation in k-arm clinical trials.
• To extend sample size calculations to include non-inferiority designs and survival endpoints.
• To provide a practical tool for power and sample size determination in various trial settings.

Main Methods:

• Derivation of a unified asymptotic formulation for sample size calculation.
• Application to superiority and non-inferiority designs for continuous, binary, and survival endpoints.
• Development of a closed-form solution based on contrast tests.

Main Results:

• A unified sample size calculation method is presented for k-arm trials (k>0).
• The method successfully covers single-arm and k-arm (k>=2) designs with survival endpoints.
• Closed-form solutions for power and sample size calculations are derived.

Conclusions:

• The proposed method offers a comprehensive approach to sample size calculation for diverse clinical trial designs and endpoints.
• This formulation addresses a gap in the statistical literature, particularly for survival data in multi-arm trials.
• The availability of application examples and a SAS program facilitates practical implementation.