Figure 5: a) Final form of the proposed Williams design of order 5. It is ensured that no subject has two placebo treatments during the last two periods, if the two assignments of placebo to the codes A–E is done in a way that no neighboring codes (AB, BC, CD, DE and AE) are mapped to placebo.
b) The resulting three designs that can be obtained from the assessments of the codes (M – open-label moxifloxacin, P – placebo, X, Y: The two doses of the new investigational drug).