Testing and estimation of proportion (or risk) ratio under the matched-pair design with multiple binary endpoints

Area of Science:

• Biostatistics
• Clinical Trials
• Statistical Modeling

Background:

• The proportion ratio (PR) is a key metric for comparing experimental and control treatments in randomized clinical trials.
• Analyzing multivariate binary matched-pair data presents unique statistical challenges.
• Existing methods may lack precision or power, especially with limited sample sizes.

Purpose of the Study:

• To develop and evaluate statistical procedures for testing and estimating proportion ratios in multivariate binary matched-pair data.
• To compare the performance of asymptotic and permutation-based methods.
• To provide tools for analyzing treatment effects in complex clinical trial designs.

Main Methods:

• Development of asymptotic and permutation-based procedures for testing equality of treatment effects.
• Derivation of confidence intervals for proportion ratios under a mixed-effects exponential risk model.
• Monte Carlo simulations to assess Type I error rates and power.

Main Results:

• Asymptotic and permutation-based tests demonstrate good Type I error control for large sample sizes.
• Permutation-based tests are particularly useful for small numbers of matched pairs.
• Weighted linear average estimators enhance statistical power and precision when treatment effects are unidirectional.

Conclusions:

• The proposed statistical methods offer robust tools for analyzing proportion ratios in multivariate binary matched-pair data.
• Permutation-based approaches provide a valuable alternative for studies with limited sample sizes.
• The findings are applicable to real-world clinical trial data, such as analyzing adverse events in drug trials.