A Sequential procedure for studies comparing multiple doses

An attractive method of dealing with the multiplicity inherent in studies comparing multiple doses is to use a testing strategy in which hypotheses (or sets of hypotheses) are tested in a prespecified order, with rejection at each step a necessary condition for continuation to the next step. Such a strategy naturally lends itself to interim monitoring. We demonstrate how this may be achieved. Copyright © 2002 John Wiley & Sons, Ltd.