Fitting multivariate polynomial growth curves in two-period crossover designs

Area of Science:

• Biostatistics
• Clinical Trial Design
• Statistical Modeling

Background:

• Crossover designs are efficient for clinical trials, allowing within-subject comparisons.
• Repeated measurements over time are common in clinical trial data.
• Analyzing complex response variables in crossover trials can be challenging.

Purpose of the Study:

• To describe statistical methods for analyzing data from two clinical trials using crossover designs.
• To present a two-stage approach for data reduction and analysis.
• To enable hypothesis testing for crossover designs with repeated measurements.

Main Methods:

• Fitting polynomial growth curves for data reduction.
• A two-stage statistical analysis approach.
• Multivariate parametric analysis and testing of the general linear hypothesis.

Main Results:

• The proposed methods allow for data reduction using polynomial growth curves.
• A multivariate parametric analysis is feasible even with small sample sizes typical in crossover designs.
• Standard hypotheses of interest in crossover designs can be effectively tested.

Conclusions:

• The described statistical methods provide a robust framework for analyzing crossover trial data with repeated measurements.
• The approach facilitates efficient data analysis and hypothesis testing, particularly in scenarios with limited sample sizes.
• This methodology supports comprehensive statistical inference for complex clinical trial designs.