swCRTdesign

swCRTdesign implements design, simulation, and analysis methods for stepped wedge cluster-randomized trials to estimate statistical power and model complex intra-cluster and time-dependent correlation structures.

Key Features:

• Implementation: Implemented as an R package providing computational routines for stepped wedge trial design, simulation, and analysis.
• Random effects modeling: Accounts for correlations arising from clusters, time within clusters, and treatment effects within clusters using a random effects model.
• Power calculation: Computes statistical power for stepped wedge trial designs and generates power curves across design parameters.
• Simulation: Simulates stepped wedge trial data to support estimation of power and assessment of potential trial outcomes under varied scenarios.
• Summaries and plotting: Produces descriptive summaries and visualizations, including design plots and result plots.
• Design flexibility: Accommodates a wide array of stepped wedge configurations and custom correlation structures for tailored analyses.

Scientific Applications:

• Evaluation of interventions: Supports design and analysis of healthcare stepped wedge trials evaluating interventions across clusters over time.
• Trial planning and optimization: Facilitates power estimation and selection of stepped wedge designs for study planning.
• Simulation studies: Enables simulation-based studies to assess power and understand potential outcomes under various scenarios.
• Analysis accounting for correlation: Enables analysis that accounts for intra-cluster and time-dependent correlation structures in stepped wedge trials.

Methodology:

Implemented in R; uses random effects models to represent cluster, within-cluster time, and treatment correlations; provides simulation of stepped wedge trial data, statistical power calculations, and functions to generate descriptive summaries and plots.