A practical simulation method to calculate sample size of group sequential trials for time-to-event data under...

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Survival Analysis

Background:

• Group sequential designs are common in clinical trials.
• Accurate sample size estimation is critical, especially for survival trials with complex data characteristics.
• Existing methods face challenges due to distributional ambiguity, censored data, and information time definitions.

Purpose of the Study:

• To propose a practical simulation-based method for sample size estimation in multi-stage, two-arm survival group sequential designs.
• To incorporate the Weibull distribution and discontinuation probabilities for more accurate estimates.
• To provide a SAS program for the proposed method.

Main Methods:

• A simulation-based approach is employed for sample size calculation.
• The Weibull distribution is considered alongside the exponential distribution for survival data.
• The method incorporates discontinuation probabilities and calculates assessment indexes for interim analysis planning.
• The study explores the impact of the shape parameter on sample size under the Weibull distribution.

Main Results:

• The simulation method provides accurate sample size estimates for survival group sequential trials.
• The incorporation of discontinuation probability improves estimate accuracy.
• Assessment indexes aid in determining the number and timing of interim analyses.
• A method to estimate the Weibull shape parameter using median survival time and hazard ratio is proposed.

Conclusions:

• The proposed simulation method offers a practical solution for sample size estimation in survival group sequential trials.
• The method is robust, accounts for complex survival data features, and can be adapted for adaptive designs.
• It provides valuable insights into the effects of the shape parameter and aids in trial design optimization.