Which domain is used to handle data gaps during sensitivity analysis?

• A. Missing data
• B. Follow-up
• C. Exposure
• D. Disease outcome

Answer: (A)

Handling data gaps in sensitivity analysis centers on missing data. When information is missing, results can be biased if we don’t account for why and how the gaps occur. The missing data domain provides a structured way to test how conclusions might change under different missingness scenarios and imputation approaches, such as multiple imputation under MAR or pattern-mixture models for MNAR. By exploring best/worst-case patterns or imputing missing values under plausible assumptions, we assess the robustness of the finding to the gaps in the data.

Other domains address different sources of bias: follow-up focuses on biases from participants lost to study follow-up, while exposure and disease outcome domains deal with measurement error or misclassification of exposure or outcome, not the missingness itself.