Power and Sample Size for Longitudinal Models in R - The longpower Package and Shiny App

Area of Science:

• Biostatistics
• Clinical Trial Design
• Medical Research Methodology

Background:

• Longitudinal studies are essential in medical research for tracking changes over time.
• Accurate sample size calculation is vital for ensuring sufficient statistical power and valid inferences in these studies.
• Existing methodologies for sample size determination in longitudinal studies are complex and depend on various factors.

Purpose of the Study:

• To review and present sample size formulas for continuous longitudinal data.
• To introduce a practical tool (Shiny app) for researchers to calculate sample sizes.
• To facilitate sample size estimation for Alzheimer's disease clinical trials using ADNI data.

Main Methods:

• Literature review of sample size calculation methodologies for longitudinal studies.
• Development of sample size formulas for continuous longitudinal data.
• Creation of a Shiny application for interactive sample size estimation.
• Utilizing Alzheimer's Disease Neuroimaging Initiative (ADNI) data for specific parameter estimation.

Main Results:

• Formulas and illustrative examples for calculating sample sizes for continuous longitudinal data are provided.
• A Shiny app is introduced, enabling researchers to input pilot estimates for sample size calculations.
• The app can estimate pilot parameters for Alzheimer's studies using ADNI data.
• Sample size and power curves can be generated to assess study design operating characteristics.

Conclusions:

• The developed package and Shiny app simplify sample size determination for longitudinal studies.
• These tools enhance the ability of researchers to design robust clinical trials, particularly for Alzheimer's disease.
• The resources aid in assessing study design efficiency for longitudinal continuous outcomes.