Optimal treatment allocation in comparative biomedical studies.

In clinical trials involving several treatment protocols, there are usually several objectives of different importance. A general method for finding an optimal design which reflects this practical need is discussed. In particular, we show that the graphical method of Cook and Wong can be applied when there are two objectives. Several illustrative examples are discussed, including multiple-objective designs for comparing several treatment groups with a placebo group. Copyright 2000 John Wiley & Sons, Ltd.