Adaptive Designs for Clinical Trials with Highly Successful Treatments
- PDF / 468,882 Bytes
- 7 Pages / 504 x 720 pts Page_size
- 1 Downloads / 204 Views
Drug Information Journal,Vol. 35, 1087-1093, 2001 Printed in the USA. All rights reserved.
ADAPTIVE DESIGNS FOR CLINICAL TRIALS WITH HIGHLY SUCCESSFUL TREATMENTS* ANASTASIA IVANOVA,PHD Assistant Professor of Biostatistics, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina
WILLIAM F. ROSENBERGER, PHD Associate Professor of Mathematics and Statistics, Adjunct Associate Professor of Epidemiology and Preventive Medicine, University of Maryland, Baltimore County and Baltimore, Baltimore, Maryland
We compare the perjormance of two adaptive designs and equal allocation in a clinical trial with two highly successful treatments and binary outcomes. The measure of interest in the trial is the odds ratio. The goal of the adaptive design is to decrease the total number of failures compared to equal allocation while keeping the power at the same level. One design is based on sequential maximum likelihood estimation, the other on an urn model. Wefind that the urn model produces a better procedure than the sequential maximum likelihood approach and equal allocation, in that it yields fewer expected treatment failures, maintains the power of the asymptotic test, and is more powerful when the Fisher S exact test is used. We conclude that adaptive designs have attractive properties when both treatments are highly successful. Key Words: Ethics; Measures o f association; .Optimal allocation; Power; Urn models
INTRODUCTION RANDOMIZING PATIENTS in equal proportions to experimental and control therapies in a clinical trial has become so embedded in the culture of clinical research that it is a useful exercise to examine afresh the basis for equal allocation. For, if there were no substantial reasons for equal allocation to treatments, one could reasonably presume that patients would prefer to have a greater
Reprint address: Professor Anastasia V. Ivanova, Department of Biostatistics, University of North Carolina at Chapel Hill, CB7420, Chapel Hill, NC 27599-7420. *Grant support: For Professor Rosenberger, R29DK51015-05 from the National Institute of Diabetes and Digestive and Kidney Diseases. This was an invited paper presented by the second author at the Pharmaceutical Research and Manufacturers AssociationFd and Drug Administration Workshop, November 1, 2000, Washington, District of Columbia.
than 50% chance of being assigned the better treatment; and one could reasonably presume that physicians would prefer that their patients have a greater than 50% chance of being assigned the better treatment. In some diseases where recruitment of patients is especially difficult, advertising a 2 : 1 or 3 : 1 allocation favoring the experimental therapy can be a recruitment tool. In the seminal text by Friedman et al. (l), two reasons for advocating equal allocation are given: 1. Equal allocation maximizes the power of tests of the treatment effect, and 2. Equal allocation is consistent with the state of equipoise that precedes the trial. Let us examine both of these statements more carefully. Equal allocation may, in f
Data Loading...
