A new conditional performance score for the evaluation of adaptive group sequential designs with sample size...

Area of Science:

• Clinical Trials
• Biostatistics
• Statistical Methodology

Background:

• Standard clinical trials fix sample size at the outset, relying on potentially variable historical data.
• Adaptive group sequential designs permit sample size recalculation mid-trial via interim analyses.
• Current methods lack standardized performance assessments for adaptive designs' sample size adjustments.

Purpose of the Study:

• To review existing sample size recalculation rules and performance measures in adaptive designs.
• To introduce a novel conditional performance score for evaluating adaptive trial designs.
• To assess standard recalculation rules using the proposed score via simulations.

Main Methods:

• Overview of sample size recalculation rules and performance metrics.
• Development of a new conditional performance score.
• Monte-Carlo simulations to apply the score to various recalculation rules.

Main Results:

• Identified a gap in standardized performance evaluation for adaptive trial sample size rules.
• Proposed a new conditional performance score integrating key metrics.
• Demonstrated application of the score to assess common recalculation strategies.

Conclusions:

• A comprehensive performance evaluation is crucial for adaptive clinical trial designs.
• The proposed conditional performance score offers a valuable tool for assessing sample size recalculation rules.
• Further research and standardization are needed for optimal adaptive design implementation.