Predicting study duration in clinical trials with a time-to-event endpoint.

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Survival Analysis

Background:

• Event-driven clinical trials compare survival functions using formulas like Freedman or Schoenfeld to determine the number of events.
• Sample size and study duration are derived from required events but are not uniquely determined, leading to practical challenges.
• Current methods lack effective visualization for sample size and duration relationships and uncertainty evaluation.

Purpose of the Study:

• To develop a graphical approach for visualizing the relationship between sample size and study duration in clinical trials.
• To derive methods for evaluating and updating the uncertainty in study duration based on observed events.

Main Methods:

• Developed a graphical method to illustrate the interplay between sample size and study duration.
• Derived the probability density function for study duration.
• Created a method to update this probability density function using observed event data (information time).

Main Results:

• A novel graphical approach effectively visualizes the relationship between sample size and study duration.
• The probability density function of study duration and its update mechanism provide a quantitative measure of uncertainty.
• These tools aid in exploring various combinations of sample size and duration based on practical constraints.

Conclusions:

• The proposed graphical and statistical methods enhance the operational management of clinical trials with time-to-event endpoints.
• Improved visualization and uncertainty evaluation contribute to more informed decision-making regarding sample size and study duration.
• These advancements are expected to optimize resource allocation and trial conduct.