Browse by author
Lookup NU author(s): Professor James WasonORCiD
Full text for this publication is not currently held within this repository. Alternative links are provided below where available.
Two-stage designs are commonly used for Phase II trials. Optimal two-stage designs have the lowest expected sample size for a specific treatment effect, for example, the null value, but can perform poorly if the true treatment effect differs. Here we introduce a design for continuous treatment responses that minimizes the maximum expected sample size across all possible treatment effects. The proposed design performs well for a wider range of treatment effects and so is useful for Phase II trials. We compare the design to a previously used optimal design and show it has superior expected sample size properties. Copyright © Taylor & Francis Group, LLC.
Author(s): Wason JMS, Mander AP
Publication type: Article
Publication status: Published
Journal: Journal of Biopharmaceutical Statistics
Online publication date: 31/05/2012
ISSN (print): 1054-3406
ISSN (electronic): 1520-5711
Publisher: Taylor and Francis
PubMed id: 22651118
Altmetrics provided by Altmetric