Preservation of Type I Error for Partially-Unblinded Sample Size Re-Estimation

Area of Science:

• Biostatistics
• Clinical Trial Design
• Statistical Methodology

Background:

• Sample size re-estimation (SSR) is crucial for adaptive clinical trials, enabling adjustments based on accrued data.
• Current SSR methods use blinded or unblinded approaches, aiming to preserve the Type I error rate.
• A need exists for SSR methods that balance operational feasibility with statistical rigor.

Purpose of the Study:

• To propose and evaluate a partially-unblinded method for sample size re-estimation (SSR) in clinical trials.
• To assess the impact of this method on Type I error rates for binary and continuous endpoints.
• To explore and clarify mathematical expressions for SSR under various variance scenarios, including dual variance.

Main Methods:

• Developed a partially-unblinded SSR approach, using interim data without the effect size to maintain operational blinding.
• Conducted proof-of-concept and simulation studies to validate the method's performance.
• Investigated SSR mathematical expressions for homogeneity, heterogeneity, and dual variance scenarios for binary and continuous data.

Main Results:

• The proposed partially-unblinded SSR method effectively preserves the Type I error rate.
• Demonstrated the method's applicability to both binary and continuous endpoints.
• Derived and clarified dual variance mathematical expressions for SSR, showing they offer a compromise between homogeneity and heterogeneity, bounded sample size estimates, and suitability for adaptive designs.

Conclusions:

• Partially-unblinded SSR offers a viable strategy for adaptive trial design, maintaining statistical integrity.
• The developed method provides flexibility in sample size adjustments without compromising the Type I error rate.
• The findings extend the utility of SSR methods, particularly the dual variance approach, to a broader range of clinical trial designs.