Sample Size Re-Estimation in Blinded Hybrid-Control Design Using Inverse Probability Weighting.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Real-World Data Analysis

Background:

• Hybrid control designs integrating historical and real-world data are increasingly used in clinical studies.
• Pre-specifying information borrowing from historical controls is crucial but can be challenged by covariate distribution discrepancies.
• Such discrepancies can limit effective data borrowing, potentially reducing statistical power below target levels.

Purpose of the Study:

• To propose novel sample size re-estimation strategies for hybrid control designs.
• To address the challenge of maintaining statistical power when significant baseline covariate differences emerge between current and historical studies.
• To ensure the robustness of clinical trial evaluations despite potential data heterogeneity.

Main Methods:

• Development of two sample size re-estimation strategies applicable during blinded clinical studies.
• Utilization of inverse probability weighting (IPW) based on assignment probabilities to current or historical studies.
• Simulation studies to evaluate the performance of the proposed strategies under varying covariate distributions.

Main Results:

• The proposed IPW-based strategies effectively adjust sample size upwards when large discrepancies in baseline covariates are detected.
• These adjustments help prevent a loss of statistical power that would otherwise occur.
• Simulations demonstrated the strategies' ability to maintain targeted power levels.

Conclusions:

• The presented sample size re-estimation methods offer a practical solution for hybrid control designs facing covariate imbalances.
• These strategies enhance the reliability and power of clinical trials utilizing external data.
• The case study illustrates the feasible application of these methods in real-world clinical research.