Sample Size Adjustment in Sequential Multiple Assignment Randomized Trials.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Health Services Research

Background:

• Clinical trials often face underpowering due to initial parameter uncertainty.
• Sequential Multiple Assignment Trials (SMARTs) present unique challenges for sample size determination.
• Existing sample size adjustment methods are not readily applicable to SMARTs.

Purpose of the Study:

• To propose a novel sample size adjustment procedure specifically for SMARTs.
• To ensure adequate statistical power in SMARTs despite initial design parameter limitations.
• To optimize resource allocation in clinical trials by investing only in those with promising conditional power.

Main Methods:

• A sample size adjustment procedure based on conditional power is developed for SMARTs.
• Conditional power is derived from a bivariate non-central chi-square distribution.
• Interim analyses are used to re-estimate sample sizes and adapt the trial design.

Main Results:

• The proposed method effectively maintains desirable statistical power, even with initial underpowered sample sizes.
• Simulation studies confirm the procedure's ability to preserve power at the final analysis.
• The approach allows for efficient resource allocation, focusing additional investment on trials with demonstrated potential.

Conclusions:

• The developed sample size adjustment method enhances the robustness of SMARTs.
• This procedure addresses a critical gap in adaptive trial design for complex treatment strategies.
• The method offers a practical solution for improving the efficiency and success rates of clinical trials.