Power and sample size determination in clinical trials with multiple primary continuous correlated endpoints

Area of Science:

• Biostatistics
• Clinical Trial Design
• Statistical Analysis

Background:

• Increasing use of multiple primary correlated endpoints in clinical trials.
• Necessity of controlling family-wise error rate (FWER) for valid statistical inference.
• Need for practical methods for sample size calculation and data analysis in such scenarios.

Purpose of the Study:

• To provide formulas for sample size computation and data analysis for trials with multiple correlated endpoints.
• To present methods for controlling the family-wise error rate (FWER).
• To offer practical tools for researchers designing and analyzing complex clinical trials.

Main Methods:

• Development of formulas for sample size and data analysis.
• Discussion of two distinct approaches: an individual union-intersection procedure and a global multivariate model.
• Utilization of simulation studies and real-world applications for validation.

Main Results:

• Formulas and methodologies for sample size and FWER control are provided.
• Both individual and global procedures are detailed and illustrated.
• The R package rPowerSampleSize is introduced as a practical implementation.

Conclusions:

• The study offers essential statistical tools for clinical trials with multiple correlated endpoints.
• The provided methods ensure appropriate control of the family-wise error rate (FWER).
• Researchers can utilize these formulas and the associated R package for enhanced trial design and analysis.