Bioequivalence and Bioavailability Forum

Ideally, all treated subjects should be included in the statistical analysis. However, subjects in a
crossover trial who do not provide evaluable data for both of the test and reference products […] should not be included.

1. Yes.
   
2. Generally yes. But there is a case where you would drop the subject even if only one period is missing: Two test available and the reference missing.
   
3. By fully replicate you talk about the four-period RTRT|TRTR. No doubt about three missing periods. It’s tricky if both administrations of the test are missing. Although IMHO it is a little bit strange why a subject missing a period should come back for others. But EMA’s Q&A-document contains such a wacky example data set and I have seen a case in the wild. Would you keep the subject in the analysis only to get a better estimate of CV_WR? I would not; this time agreeing with EMA’s GL above.
   Homework: Fully replicate three-period design (RTR|RTR).