Application of GEE procedures for sample size calculations in repeated measures experiments

Area of Science:

• Biostatistics
• Longitudinal Data Analysis
• Clinical Research Methodology

Background:

• Determining minimum sample size is vital for applied research.
• Existing sample size methods are well-established for single time-point outcomes.
• Longitudinal study sample size calculations are less developed.

Purpose of the Study:

• To adapt the generalized estimating equation (GEE) approach for sample size calculations in longitudinal designs.
• To provide methods for both discrete and continuous outcome variables.
• To address the complexities of longitudinal data analysis.

Main Methods:

• Adaptation of the Liang and Zeger generalized estimating equation (GEE) approach.
• Utilizing the non-central Wald Chi 2 test.
• Employing the damped exponential family for the working correlation matrix.

Main Results:

• Development of sample size calculation methods for longitudinal studies.
• Presentation of a minimum sample size table for binary outcomes.
• Discussion of extensions for unequal allocation, staggered entry, and loss to follow-up.

Conclusions:

• The adapted GEE approach provides a robust method for sample size determination in longitudinal research.
• The methods presented are applicable to various outcome types and study designs.
• This work contributes to the development of more precise sample size calculations for complex longitudinal studies.