The treatment versus experimentation dilemma in dose finding studies

D. Azriel*, M. Mandel, Y. Rinott

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

45 Scopus citations


Phase I clinical trials are conducted in order to find the maximum tolerated dose (MTD) of a given drug from a finite set of doses. For ethical reasons, these studies are usually sequential, treating patients or groups of patients with the optimal dose according to the current knowledge, with the hope that this will lead to using the true MTD from some time on. However, the first result proved here is that this goal is infeasible, and that such designs, and, more generally, designs that concentrate on one dose from some time on, cannot provide consistent estimators for the MTD unless very strong parametric assumptions hold. Allowing some non-MTD treatment, we construct a randomized design that assigns the MTD with probability that approaches one as the size of the experiment goes to infinity and estimates the MTD consistently. We compare the suggested design with several methods by simulations, studying their performances in terms of correct estimation of the MTD and the proportion of individuals treated with the MTD.

Original languageAmerican English
Pages (from-to)2759-2768
Number of pages10
JournalJournal of Statistical Planning and Inference
Issue number8
StatePublished - Aug 2011


  • Isotonic regression
  • Maximum tolerated dose
  • Phase I trial
  • Stochastic approximation
  • Up-and-down design


Dive into the research topics of 'The treatment versus experimentation dilemma in dose finding studies'. Together they form a unique fingerprint.

Cite this