|
Abstract:
|
(*) Student Paper Award Winner. One can greatly increase the efficiency of a randomized controlled trial by integrating information obtained from other trials with similar treatment arms. A number of methods have been proposed to facilitate borrowing in clinical trials in the Bayesian paradigms, including hierarchical models and, more recently, multisource exchangeability models (MEMs). When the total sample size of the supplementary trials is considerably larger than that of the primary trial, care must be taken when integrating supplementary data to avoid overwhelming the data of the primary trial. In this paper, we address this issue by deriving a set of priors, which we refer to as "capping priors," that can limit the amount of information, characterized by the effective supplementary sample size (ESSS), that is dynamically borrowed via MEMs at a pre-specified ESSS threshold. We demonstrate the behavior of this technique via a simulation study. Then, we apply our method to data from a randomized trial of very low nicotine content (VLNC) cigarettes in smokers with serious mental illnesses, and augment the trial with data from three supplemental trials of VLNC cigarettes.
|
