Sample size adaptation designs and efficiency comparison with group sequential designs.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Statistical Inference

Background:

• Group sequential designs (GSDs) are established methods for interim analyses in clinical trials.
• Sample size adaptation designs (SSADs) offer flexibility but require rigorous efficiency validation.
• Comparing the efficiency of SSADs and GSDs is crucial for optimizing clinical trial resource utilization.

Purpose of the Study:

• To systematically present sample size adaptation designs (SSADs).
• To provide analytical proof of the efficiency advantage of general SSADs over group sequential designs (GSDs).
• To introduce a class of sample size mapping functions for defining SSADs.

Main Methods:

• Development of theorems describing SSAD properties within a two-stage adaptive clinical trial framework.
• Derivation of sufficient conditions to analytically prove efficiency.
• Utilizing weighted combination tests for SSADs.

Main Results:

• Analytical proof demonstrates that SSADs based on weighted combination tests are uniformly more efficient than GSDs across a range of true treatment differences.
• Fully adaptive SSADs can achieve comparable statistical power to GSDs with reduced average sample sizes.
• Substantial sample size savings are achievable with SSADs.

Conclusions:

• SSADs provide a statistically powerful and more efficient alternative to GSDs in clinical trial design.
• The proposed SSAD framework allows for significant optimization of sample size, leading to cost and time efficiencies.
• Practical guidance and examples are provided for implementing efficient SSADs.