Sample size calculation for randomized selection trials with a time-to-event endpoint and a margin of practical...

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Survival Analysis

Background:

• Selection trials compare experimental treatments without a control arm.
• Existing sample size methods are limited for time-to-event endpoints in these trials.
• Dichotomizing time-to-event endpoints like progression-free survival alters the clinical question and may reduce statistical power.

Purpose of the Study:

• To develop sample size calculation theory for selection trials with time-to-event endpoints.
• To provide practical tools for researchers designing such studies.
• To address the gap in sample size methodologies for non-controlled comparative trials.

Main Methods:

• Developed theoretical framework for sample size calculation assuming exponential or Weibull distributions for time-to-event data.
• Created a free web application for sample size calculations.
• Developed an R package for sample size determination in selection trials.

Main Results:

• Established a theoretical basis for sample size calculations in selection trials with time-to-event endpoints.
• Provided accessible computational tools (web app and R package) for practical application.
• Addressed the limitations of dichotomizing time-to-event data for sample size and analysis.

Conclusions:

• The developed methods and tools enable robust sample size determination for selection trials using time-to-event endpoints.
• Researchers can now accurately design studies comparing experimental treatments without a control arm.
• The findings support improved trial design and statistical power for progression-free survival analysis.