Power and sample size for the S:T repeated measures design combined with a linear mixed-effects model allowing for...

Area of Science:

• Biostatistics
• Clinical Trial Design
• Longitudinal Data Analysis

Background:

• Repeated measures designs are crucial for analyzing longitudinal data in clinical research.
• Existing methods like the simple pre-post design have limitations in handling missing data and sample size efficiency.
• The S:T repeated measures design was previously proposed, combining specific statistical models.

Purpose of the Study:

• To present formulas for calculating statistical power and sample sizes for the S:T repeated measures design.
• To accommodate continuous response variables and linear mixed-effects models.
• To provide a framework for analyzing average treatment effects with missing data.

Main Methods:

• Utilized the S:T repeated measures design framework.
• Employed linear mixed-effects models for data analysis.
• Derived formulas for power and sample size calculations, accounting for missing data under the missing at random assumption.

Main Results:

• Provided explicit formulas for power and sample size determination in the S:T repeated measures design.
• Demonstrated the design's ability to handle missing data effectively.
• Showcased potential sample size reductions compared to simpler designs.

Conclusions:

• The S:T repeated measures design, with linear mixed-effects models, offers a statistically sound approach for continuous outcomes.
• This design facilitates efficient sample size planning and robust analysis in the presence of missing data.
• The presented formulas aid researchers in designing studies and analyzing treatment effects more effectively.