Dose Spacing in Early Dose Response Clinical Trial Designs
- PDF / 612,743 Bytes
- 10 Pages / 504 x 719.759 pts Page_size
- 10 Downloads / 199 Views
WZ-X6l5/2002 Copyright 0 ZOO2 Drug Infomiation Association Inc.
DOSE SPACING IN EARLY DOSE RESPONSE CLINICAL TRIAL DESIGNS ANTHONY HAMLETT,PHD Research Fellow. Harvard School of Public Health. Department of Biostatistics. Boston, Massachusetts
NAITEETING,PHD Associate Director. Biostatistics. ffizer Global Research & Development, New London, Connecticut
R. CHOUDARY HANUMARA, PHD Professor of Statistics. Department of Computer Science and Statistics, University of Rhode Island, Kingston. Rhode Island
JEFFREYS. FINMAN,PHD Director, Biostatistics. Pfizer Global Research & Development, New London. Connecticut
Finding the right dose or the right range of doses is one of the most important objectives in the clinical development progrlitn of a new drug. I n designing early Phase 2 dose response studies. one critical question is: “What spacing should be used between test doses.?” I n this article. bve propose an inruitive. model-free approach to address this question. Simulation s are performed to compare the proposed design with optimal designs under logistic and normal models. The proposed method for dose allocation can be applied to various types of dose response studies. Kev Words: Clinical trials: Early phase; Dose response; Dose spacing; Optimal design
INTRODUCTION SELECTING DOSES T O be studied in a dose response clinical trial is an important yet very difficult task. In this article we propose an intuitive approach to help scientists allocate doses in their clinical trials. This new approach is model free and is developed with a few simple assumptions. In order to study the performance of this new approach of dose allocation, simulations are used to compare the new method against dose allocation obtained from optimal designs. Since there are many ways to model a dose
Reprint address: Dr. Naitee Ting. Biostatistics. MS 6025-C2258, PIizer Global Research & Development, 50 Pequot Ave.. New I.ondon, CT 06320.
response curve, simulations are performed on a limited set of models and parameter combinations. Results from these simulations indicate that with certain model/parameter combinations, optimal designs are preferred, while with others, the proposed method is preferred. However, in actual clinical trial designs, the underlying model or parameters are not known, and the studies are designed only with “assumed” models. Hence, the new approach becomes advantageous because no underlying model need be assumed. The second section of this paper provides the background of this problem. The proposed method is introduced in the third section. The fourth section compares the proposed method with optimal designs. The fifth section concludes this paper.
855
Downloaded from dij.sagepub.com at UNIVERSITY OF SASKATCHEWAN LIBRARY on March 18, 2015
A. lliirrrlrtr. N . 7i'Ilg. H. C, lftrrrrirnnrn. cirrd J . S. Firrnrnn
856
BACKGROUND Finding the right dose or the right range of doses is one of the most important objectives in the clinical development program of a new drug. This task is particularly difficult when we try
Data Loading...
