A taxonomy of estimands for regulatory clinical trials with discontinuations.

The National Research Council Panel on Handling Missing Data in Clinical Trials recommended that protocols for clinical trials 'explicitly define... causal estimands of primary interest'. In discussions with sponsors of clinical trials since the publication of the National Research Council report, the expression causal estimands has been the subject of confusion. It may not be entirely clear what the National Research Council panel meant, and in any case, it has not been clear how this recommendation might be put in practice. This paper's purpose is to say how the working group understands it and how we think it should be put in practice. We classify possible choices of estimand according to their usefulness for regulatory purposes in various clinical settings. Published 2015. This article is a U.S. Government work and is in the public domain in the USA. Published 2015. This article is a U.S. Government work and is in the public domain in the USA.