A Nonparametric Bayesian Approach for Adjusting Partial Compliance in Sequential Decision-Making

• Activity Number: 208 - Personalized and Precision Medicine
• Type: Contributed
• Date/Time: Tuesday, August 10, 2021 : 1:30 PM to 3:20 PM
• Sponsor: Biometrics Section
• Author(s): Indrabati Bhattacharya* and Ashkan Ertefaie and Andrew Gordon Wilson and Brent Johnson and James Mckay and Kevin Lynch
• Companies: University of Rochester and University of Rochester and New York University and University of Rochester and University of Pennsylvania and University of Pennsylvania
• Keywords: partial compliance; nonparametric Bayesian; Markov chain Monte Carlo

Abstract

Existing methods in estimating the mean outcome under a given Dynamic Treatment Regime (DTR) rely on intention-to-treat (ITT) analyses which estimate the effect of following a certain DTR regardless of compliance behavior of patients. There are two major concerns with ITT analyses: (1) the estimated effects are often biased toward the null effect; (2) the results are not generalizable and reproducible due to the potential differential compliance behavior. These are particularly problematic in settings with high level of non-compliance such as substance use disorder treatments. Due to the relatively low level of compliance in such studies, ITT analyses essentially estimate the effect of being randomized to a certain treatment sequence which is not of interest. We fill this important gap by defining the target parameter as the mean outcome under a dynamic treatment regime given potential compliance strata. We propose a flexible nonparametric Bayesian approach, which consists of a Gaussian copula model for the potential outcomes. Our simulations show the usefulness of this approach in practice and illustrate the robustness of our estimator in nonlinear and non-Gaussian settings.