10.4 Simulation of Optimal Response-Adaptive Randomization Procedures
10.4.1 Simulating a Single RAR Sequence
If the target allocation ratio is pre-specified (e.g. 1:1 or 2:1), a treatment randomization
sequence can be pre-generated before any subject is enrolled into a trial. In contrast,
a RAR sequence cannot be pre-generated because treatment randomization probabilities
are adaptively modified throughout the trial based on cumulative patient outcome data.
Standard computer packages such as SAS an R have built-in random number generators
that can be used to simulate RAR designs under certain assumptions on outcomes from
patients in the trial. We start with a simulation of a single randomization sequence.
Program
ch10.4.1.sas in the supplemental ...Get Modern Approaches to Clinical Trials Using SAS: Classical, Adaptive, and Bayesian Methods now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
