Data-Driven Prior Distributions for A Bayesian Phase-2 COPD Dose-Finding Clinical Trial. (3rd July 2018)

• Record Type: Journal Article
• Authors: Novick, Steven; Ho, Shuyen; Best, Nicky
• Is Part Of: Statistics in biopharmaceutical research. Volume 10:Number 3(2018)
• Page Start: 166
• Page End: 175
• Publication Date: 2018-07-03
• Subjects: Analysis prior -- Bayesian -- Clinical trials -- Design prior -- Dose finding -- Historical data -- Phase 2; Pharmacy -- Statistical methods -- Periodicals; Pharmaceutical biotechnology -- Statistical methods -- Periodicals; Biopharmaceutics -- Periodicals; Biometry -- Periodicals; Pharmacy -- Statistical methods; Periodicals; 615.190727
• DOI: 10.1080/19466315.2018.1462728
• Languages: English
• ISSNs: 1946-6315

Abstract

ABSTRACT: The prior distribution reflects knowledge and uncertainty of the modeled parameters. Determining the prior distribution for a dose-finding clinical trial can be influential in its design and analysis. Using the planning of a phase 2 trial for COPD with a dose-response curve as a case study, we illustrate the use of relevant historical data for the nonlinear curve mean-model parameters as well as consideration for terms to characterize between-trial and within-trial variability. Through a predictive inference exercise, a data-driven informative prior distribution is constructed for the future study. We share our strategies on how to obtain informative Bayesian priors for both design and analysis of dose-finding clinical trials using relevant historical data and deal with the associated issues.