Sample sizes for clinical trials with time-to-event endpoints and competing risks

Area of Science:

• Clinical Trials
• Biostatistics
• Epidemiology

Background:

• Clinical trials often evaluate time-to-event endpoints in the presence of competing risks.
• Accurate sample size calculation is crucial for treatment evaluation and statistical power.
• Traditional survival analysis methods may need adaptation for competing risks scenarios.

Purpose of the Study:

• To present a formula for sample size computation in clinical trials with two competing outcome states.
• To outline statistical methods for determining patient recruitment numbers based on event rates.
• To provide tools like nomograms for visualizing trial duration and size alternatives.

Main Methods:

• Utilizing Schoenfeld's formula for calculating the required number of primary endpoint events.
• Incorporating accrual period, trial duration, and event-specific hazard functions for patient number determination.
• Applying a formula for sample size computation tailored for two competing risks.

Main Results:

• A method is presented for calculating the necessary number of patients for trials with competing risks.
• The study demonstrates how to translate prior study information into model parameter assumptions.
• Nomograms are proposed as aids for interdisciplinary committees in trial planning and monitoring.

Conclusions:

• The proposed statistical methods facilitate sample size determination for clinical trials with competing risks.
• The 4D trial serves as a practical example illustrating the application of these methods.
• Effective planning and monitoring of clinical trials are enhanced through clear communication of trial parameters.