In previous chapters, most of our efforts were directed at the assessment of bioequivalence in average bioavailability for the standard 2 2 crossover design for comparing two formulations of a drug product. The standard two-sequence, twoperiod crossover, however, is not useful in the presence of carryover effects. In addition, it does not provide independent estimates of intra-subject variabilities. To account for these disadvantages, in practice, it is of interest to consider a higher-order crossover design. A higher-order crossover design is defined as a crossover design in which either the number of periods or the number of sequences is greater than the number of formulations to be compared. The most commonly used higher-order designs for comparing two formulations include a four-sequence, two-period design (or Balaam’s design), a two-sequence, three-period design, and a four-period design with two or four sequences. Some of these designs were briefly described as designs A, B, and C in Section 2.5. In this chapter, statistical methods for assessing bioequivalence of average bioavailability from these experimental designs are discussed. Consider the following general model for a higher-order crossover design:

Yijk ¼ mþ Gk þ Sik þ Pj þ F(j,k) þ C(j1,k) þ eijk, (9:1:1)

where i ¼ 1, 2, . . . , nk j ¼ 1, . . . , J, k ¼ 1, . . . , K Yijk, m, Pj, F( j,k), C( j1,k), Sik, and eijk are defined as those in model 2.5.1 Gk is the fixed effect of sequence k.