Power and sample size for randomized phase III survival trials under the Weibull model

Area of Science:

• Biostatistics
• Clinical Trial Design
• Survival Analysis

Background:

• Phase III clinical trials are critical for drug evaluation, requiring robust statistical methods for accurate results.
• Traditional nonparametric methods like the log-rank test may not fully leverage distributional assumptions for optimal power.
• The Weibull distribution is frequently used to model survival data in clinical trials.

Purpose of the Study:

• To propose and evaluate two novel parametric tests for designing two-arm phase III survival trials.
• To compare the performance of these parametric tests against the nonparametric log-rank test.
• To derive power and sample size formulas for the proposed parametric tests.

Main Methods:

• Development of two parametric statistical tests tailored for the Weibull survival model.
• Simulation studies to compare the power and Type I error rates of the proposed tests against the log-rank test.
• Derivation of analytical formulas for sample size and power calculations under the Weibull model.

Main Results:

• The proposed parametric tests demonstrate competitive or superior performance compared to the log-rank test in simulation studies.
• Sample size and power formulas for the parametric tests are derived, facilitating trial planning.
• The sensitivity analysis indicates the impact of misspecifying the Weibull shape parameter on sample size requirements.

Conclusions:

• The proposed parametric tests provide a valid and potentially more efficient alternative for designing phase III survival trials under the Weibull model.
• These methods allow for effective study design, including planning duration and handling non-uniform entry and loss to follow-up.
• The derived formulas enhance the precision of sample size estimations in Weibull-modeled survival studies.