Some practical considerations in three-arm non-inferiority trial design

Area of Science:

• Biostatistics
• Clinical Trial Design
• Pharmaceutical Research

Background:

• Non-inferiority trials compare experimental therapies to established active treatments.
• Three-arm trials (experimental, active reference, placebo) enhance validity and assay sensitivity.
• Practical challenges exist in designing and analyzing these complex trials.

Purpose of the Study:

• To offer practical insights from a phase III three-arm non-inferiority trial.
• To guide sample size determination and optimal allocation for non-inferiority testing.
• To compare frequentist and Bayesian statistical approaches for trial analysis.

Main Methods:

• Discusses sample size calculation and allocation strategies for non-inferiority trials.
• Introduces a two-step testing procedure for the 'capturing all possibilities' non-inferiority goal.
• Compares frequentist p-value methods with Bayesian posterior probability analysis.

Main Results:

• Provides R code for sample size determination and allocation.
• Demonstrates the equivalence of the 'capturing all possibilities' goal with a two-step procedure.
• Highlights differences and similarities between frequentist and Bayesian analysis in this context.

Conclusions:

• Offers practical solutions for designing and analyzing three-arm non-inferiority trials.
• The two-step testing procedure is a practical implementation of a key non-inferiority goal.
• Both frequentist and Bayesian methods offer valuable insights, with distinct interpretations.