Power and sample-size calculations for trials that compare slopes over time: Introducing the slopepower command

Area of Science:

• Biostatistics
• Clinical Trial Design
• Longitudinal Data Analysis

Background:

• Clinical trials aiming to slow disease progression often analyze continuous outcomes by comparing changes over time (slopes) between treatment groups.
• Accurate sample-size calculations for these trials require estimates of between- and within-subject variability, which are frequently unknown.
• The algebraic complexity of sample-size calculations for slope outcome trials can be a barrier to effective trial planning.

Purpose of the Study:

• To introduce slopepower, a user-friendly command for sample-size and power calculations in trials comparing slope outcomes.
• To provide a practical tool based on linear mixed-model methodology for planning clinical trials with longitudinal data.
• To demonstrate the utility of slopepower in determining necessary sample sizes for various trial designs.

Main Methods:

• The slopepower command utilizes linear mixed-model methodology, building upon established statistical approaches for longitudinal data.
• It operates in two stages: first, estimating mean slopes, variances, and covariances from user-supplied data using a linear mixed model.
• Second, it combines these estimates with user-defined parameters for treatment effectiveness and future trial design to compute sample size or power.

Main Results:

• The slopepower command offers a streamlined approach to sample-size and power calculations for trials analyzing slope outcomes.
• It simplifies the complex algebra previously required for such designs, making trial planning more accessible.
• The tool facilitates the comparison of sample sizes needed across different potential trial designs.

Conclusions:

• The slopepower command is a valuable tool for researchers designing clinical trials that evaluate disease progression over time.
• It enhances the efficiency and accuracy of sample-size determination, crucial for resource allocation and study validity.
• This command aids in optimizing trial design by enabling informed comparisons of different planning scenarios.