Missing data adds uncertainty to results from clinical trials originally intended to definitively assess efficacy & safety. Methods have been developed to impute missing data. However, the very data needed to validate the assumptions of any imputation is, by definition, missing. So the appropriateness of any given imputation often becomes questionable.

Starting from this point, we will discuss alternatives to imputation. For example: 1. What alternatives are there to imputation? 2. How do the estimands from a valid imputation differ from those of a study in which imputation is not performed? 3. How do the estimands from a valid imputation differ from those of a data retrieval study? 4. Is no imputation actually a form of imputation?