Sample Size Determination for Response-Adaptive Randomization With Recurrent Event Responses and Unequal Follow-Up...

Area of Science:

• Clinical Trials
• Biostatistics
• Medical Research Methodology

Background:

• Recurrent event data are common in clinical trials, and the doubly biased coin design (DBCD) has shown benefits.
• Existing literature lacks comprehensive methods for sample size determination specifically for DBCD.
• Patient dropout and unequal follow-up times can significantly impact trial outcomes and sample size calculations.

Purpose of the Study:

• To develop and present methods for calculating sample size for clinical trials employing DBCD.
• To propose a sample size determination procedure that accommodates unequal follow-up times due to patient dropout.
• To quantify the impact of dropouts on sample size and treatment allocation.

Main Methods:

• Development of theoretical methods for sample size computation under DBCD.
• Incorporation of unequal follow-up times and patient dropout into the sample size determination procedure.
• Derivation of theoretical results to assess the influence of dropouts on sample size and allocation ratios.

Main Results:

• The proposed methods provide accurate sample size calculations for DBCD, considering unequal follow-up.
• Failure to account for patient dropout leads to an underestimation of the required sample size.
• Theoretical results quantify the impact of dropouts on both sample size and treatment allocation proportions.

Conclusions:

• The developed sample size determination procedure is crucial for the validity of clinical trials using DBCD.
• The procedure accurately accounts for patient dropout and unequal follow-up, ensuring adequate statistical power.
• Simulation studies and a clinical example demonstrate the practical utility and advantages of the proposed method.