|
Abstract:
|
In many controlled clinical trials, when a subject discontinued treatment prematurely, the investigator would withdraw the subject immediately; therefore no scheduled measurements were taken after that time point except perhaps for the return of the end of study follow-up visit. What if the subject was kept on the study schedule and had all planned data collected except terminating the treatment? These data after discontinued treatment could be included in the inference instead of becoming missing. This note quantifies the information gain in a 2-sequence, 2-period, 2-treatment, multiple dose, placebo-controlled crossover design with a very simplistic approach of including all data after treatment discontinuation.
|
