MMRM

Longitudinal endpoints • covariance structure • estimand-aligned inference

TipGoal

Analyze change-from-baseline longitudinal endpoints using MMRM, aligned with clinical trial conventions and estimand-focused reporting.

Deliverables

Specifications / Notes

• MMRM Analysis Spec / Model Notes

Programming

• MMRM Analysis Program

Programs and Outputs (HTML)

• MMRM Output Report

What you’ll find here

• Purpose: Repeated-measures inference for continuous endpoints across visits
• Traceability: Analysis dataset → visit mapping → model specification → results
• Typical outputs: LS-means by visit, treatment differences, CI/p-values, fit diagnostics

Model components (typical)

• Fixed effects: TRT, VISIT, TRT×VISIT
• Baseline covariate (and optionally baseline×visit)
• Subject as repeated unit
• Covariance: UN / AR(1) / CS (per SAP or justified via diagnostics)

QC checklist

• Baseline definition: Consistent baseline window and parameter alignment
• Visit mapping: Windowing rules clearly documented
• Convergence: Diagnostics captured; sensitivity to covariance choice
• Sensitivity: Alternative covariance structures / model variants (when needed)