Abstract:
|
In recent years, various outcome adaptive randomization (AR) methods have been used to conduct comparative clinical trials, in which current patient outcomes are used to randomize future patients. AR remains controversial, however, and some of its properties are not well understood by the clinical trials community. This talk presents results of a computer simulation to evaluate properties of a 200-patient trial conducted using two Bayesian AR methods compared to a fairly randomized design. The simulations show that AR has several undesirable properties. These include a surprisingly high probability of a sample size imbalance in the wrong direction, with many more patients assigned to the inferior treatment arm, the opposite of the intended effect. Compared to a fairly randomized design, AR produces less reliable final inferences, including a greatly overestimated actual treatment effect difference. These problems decrease potential benefit to future patients, and may also decrease benefit to patients enrolled in the trial. When considering whether to use AR, these problems should be weighed against its putative ethical benefit, which in fact may be illusory.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.