Trimmed means for symptom trials with dropouts.

Dropouts from randomized trials, often for lack of efficacy or toxicity, have usually been handled as 'missing data'. We suggest that they are instead complete observations, just not numeric ones. We propose an exact test of the hypothesis of no drug effect, taking all randomized patients into account, based on a readily interpretable statistic. The method also copes with a drug that is toxic in some patients but beneficial to others, a difficult problem for standard methods. A robust conclusion of efficacy can be drawn with no assumptions other than randomization. Published 2016. This article is a U.S. Government work and is in the public domain in the USA. Published 2016. This article is a U.S. Government work and is in the public domain in the USA.