Missing Data

TipExecutive summary

This page presents a practical approach to missing data focused on assumption transparency, estimand alignment, and robust interpretation.

Why it matters

Missing data can meaningfully affect estimates and conclusions. The key is not just choosing a method — it is making assumptions explicit and testing whether conclusions remain credible.

What I focus on

  • Aligning missing data strategy to:
    • estimand
    • endpoint type
    • study design
  • Summarizing missingness patterns and reasons
  • Making primary analysis assumptions explicit
  • Planning sensitivity analyses that test key assumptions

Practical contribution areas

  • SAP wording support for missing data assumptions and strategy
  • Review of missingness summaries and risk areas
  • Alignment of missing data handling with ADaM/TLF implementation
  • Clear communication of assumptions, limitations, and robustness

Common pitfalls I help prevent

  • Default method selection without estimand context
  • Weak or unclear assumption documentation
  • Limited characterization of missingness patterns
  • Sensitivity analyses that do not meaningfully test the primary assumption