Survival Analysis

KM • Cox • diagnostics • explicit censoring and estimand clarity

TipGoal

Deliver time-to-event analyses that are estimand-aligned, with explicit event/censoring rules and reproducible outputs.

Deliverables

Specifications / Notes

• Survival Analysis Spec / Notes

Programming

• Survival Analysis Program

Programs and Outputs (HTML)

• Survival Output Report

What you’ll find here

• Purpose: KM/Cox workflows consistent with clinical reporting expectations
• Traceability: ADTTE-like dataset → event/censoring definitions → model → outputs
• Typical outputs: KM plot + at-risk, medians, HR/CI, PH diagnostics, sensitivity checks

Core deliverables (typical)

• Kaplan–Meier plot (with at-risk table)
• Event counts and median survival (with CI)
• Cox model hazard ratio (HR) with CI (stratified when required)
• Diagnostics: proportional hazards checks (when applicable)

QC checklist

• Event definition: Precisely stated; consistent across outputs
• Censoring rules: Explicit and testable
• Time origin: Clear (randomization, first dose, etc.)
• Sensitivity: Alternative censoring or model variants (as needed)