Get access

A simple Bayesian decision-theoretic design for dose-finding trials

Authors

  • Shenghua Kelly Fan,

    Corresponding author
    • Department of Statistics and Biostatistics, California State University at East Bay, Hayward, CA 94542, U.S.A.
    Search for more papers by this author
  • Ying Lu,

    1. Cooperative Studies Program Coordinating Center, VA Palo Alto Health Care System, Mountain View, CA 94030, U.S.A.
    2. Department of Health Research and Policy, Stanford University, Stanford, CA 94305-5405, U.S.A.
    Search for more papers by this author
  • You-Gan Wang

    1. Centre for Applications in Natural Resource Mathematics, The University of Queensland, Australia
    Search for more papers by this author

Shenghua Kelly Fan, Department of Statistics and Biostatistics, California State University at East Bay, Hayward, CA 94542, U.S.A.

E-mail: kelly.fan@csueastbay.edu

Abstract

A flexible and simple Bayesian decision-theoretic design for dose-finding trials is proposed in this paper. In order to reduce the computational burden, we adopt a working model with conjugate priors, which is flexible to fit all monotonic dose-toxicity curves and produces analytic posterior distributions. We also discuss how to use a proper utility function to reflect the interest of the trial. Patients are allocated based on not only the utility function but also the chosen dose selection rule. The most popular dose selection rule is the one-step-look-ahead (OSLA), which selects the best-so-far dose. A more complicated rule, such as the two-step-look-ahead, is theoretically more efficient than the OSLA only when the required distributional assumptions are met, which is, however, often not the case in practice. We carried out extensive simulation studies to evaluate these two dose selection rules and found that OSLA was often more efficient than two-step-look-ahead under the proposed Bayesian structure. Moreover, our simulation results show that the proposed Bayesian method's performance is superior to several popular Bayesian methods and that the negative impact of prior misspecification can be managed in the design stage. Copyright © 2012 John Wiley & Sons, Ltd.

Ancillary