ABSTRACT
This chapter focuses on the design of Phase II dose-finding studies and addresses the question on how to choose the total sample size, doses, and dose allocations. It derives doses and dose allocations based on optimal design considerations. A common problem in dose-finding designs is that the dose range selected is too narrow to identify the complete dose–response curve. The chapter discusses sample size calculation approaches based on estimation of doses of interest or dose-response curves tend to produce considerably larger sample sizes than methods focusing on detection of a dose-response signal. The design minimizing the determinant of the asymptotic covariance matrix will minimize the maximum prediction variance on the considered dose range. The chapter provides an example motivated by a real clinical dose-finding study in chronic obstructive pulmonary disease (COPD) to illustrate the methods.
