Send to

Choose Destination
See comment in PubMed Commons below
Hematol Oncol Clin North Am. 2000 Aug;14(4):849-76, ix.

Therapeutic studies.

Author information

Biometric Research Branch, National Cancer Institute, Bethesda, Maryland, USA.


This article discusses the problems in basic design, conduct, and interpretation associated with phases I, II, and III of the cancer clinical trials and explains the various statistical solutions to these problems. The fundamental problem common to all three trials is achieving a correct and precise answer to the question posed to inform future testing and treatment while protecting trial patients from receiving treatment that has demonstrated excessive toxicity or lack of clinical efficacy. This shared problem gives rise to statistical designs with basic similarities across the three trial types.

[Indexed for MEDLINE]
PubMed Commons home

PubMed Commons


    Supplemental Content

    Full text links

    Icon for Elsevier Science
    Loading ...
    Support Center