ABSTRACT
The Balaam design, first proposed by Leslie Balaam in 1968 [1], is a type of higher-order two-period crossover design for comparison of regimen A with regimen B in which subjects are randomly allocated to one of four sequences: AA, AB, BA, or BB. Sequence AA means that the subjects in this group receive regimen A in period 1 followed by regimen A again in period 2. Sequence AB means regimen A followed by regimen B, etc. Thus, all subjects receive a regimen two times. It has features of both parallel design and crossover design. The advantage of the Balaam design as opposed to the simple (AB/BA) crossover design is that it works when the effect of a treatment in the first period could be different from its effect in period 2. Most commonly, this happens when a carryover effect is present. Although still desirable, the washout period is less crucial in this design. However, the requirement of subjects in a Balaam design is twice that in the conventional AB/BA crossover design because now we have AA and BB groups as well.
