Dose-finding in clinical trials


Ying-Kuen Cheung
2010-06-25  10:00 - 12:00
Room 504 , Freshman Classroom Building

In this lecture, we introduce the clinical problem of dose-finding. Briefly, from a statistical viewpoint, dose-finding in clinical trials can be formulated as a percentile estimation problem. However, the practical issues involved in human experimentations necessitate novel methodology apart from the classical bioassay methods. A survey of existing dose-finding methods will be given, with some details on a particular method, called the continual reassessment method. Some connections between these methods and the celebrated stochastic approximation will be made.
Keywords: Clinical trials; ethics; sequential experiments; stochastic approximation.
Some suggested reading:
1. Cheung, YK (2005). Coherence principles in dose-finding studies. Biometrika 92:863-873.
2. O'Quigley et al. (1990). Continual reassessment method: a practical design for phase I clinical trials in cancer. Biometrics 46:33-48.
3. Ratain et al. (1993). Statistical and ethical issues in the design and conduct of phase I and phase II clinical studies of new anticancer agents. Journal of the National Cancer Institute 85:1637-1643.
4. Robbins and Monro (1951). A stochastic approximation method. Annals of Mathematical Statistics 22:400-407.