Repeated measures in clinical trials: analysis using mean summary statistics and its implications for design

Area of Science:

• Biostatistics
• Clinical Trial Design
• Statistical Analysis

Background:

• Repeated measurements are common in clinical trials.
• Traditional analyses may not fully leverage pre- and post-treatment data.
• Efficient statistical methods are needed for analyzing such data.

Purpose of the Study:

• To explore the utility of simple summary statistics for analyzing repeated measurements in two-treatment randomized clinical trials.
• To quantify the superiority of analysis of covariance (ANCOVA) over alternative methods.
• To provide guidance on the design of repeated measures studies, including sample size determination.

Main Methods:

• Utilizing pre-treatment and post-treatment means as summary statistics.
• Applying analysis of covariance (ANCOVA) for data analysis.
• Developing a simple model for covariance structure between time points.
• Presenting methods for sample size calculations in repeated measures designs.

Main Results:

• ANCOVA demonstrates reduced variance and avoidance of bias compared to analyzing post-treatment means or mean changes.
• The benefits of multiple pre-treatment measurements in study design are quantified.
• Examples from clinical trials illustrate the practical application and benefits of the proposed methods.
• The compound symmetry assumption is shown to be a useful simplification for planning, though not required for the analysis.

Conclusions:

• Simple summary statistics, particularly ANCOVA, provide a robust and efficient method for analyzing repeated measurements in clinical trials.
• The study offers practical recommendations for the design and analysis of repeated measures studies.
• Flexibility in design to accommodate alternative correlation structures is advised when necessary.