Created on Tuesday, 10 May 2016. Posted in Methods
The Adaptive Designs Working Group of the MRC Network of Hubs for Trials Methodology Research has produced an online guide on why not to use A+B designs.
The document summarises the main arguments why 3+3 and similar rule based A+B designs are inappropriate for phase I dose escalation studies. These designs are still widely used although superior model-based designs such as the continual reassessment method (CRM) are available.