Online Program Home
My Program

Abstract Details

Activity Number: 14 - Novel Statistical Methods for Network-Based Studies Among People Who Use Drugs
Type: Invited
Date/Time: Sunday, July 28, 2019 : 2:00 PM to 3:50 PM
Sponsor: National Institute on Drug Abuse-NIH
Abstract #300558
Title: Bayesian Auto-G-Computation of Network Causal Effects: Incarceration and Infection in a High Risk Network
Author(s): Isabel Fulcher* and Eric Tchetgen Tchetgen and Ilya Shpitser and Caleb Lareau
Companies: Harvard University and University of Pennsylvania and Johns Hopkins University and Harvard Medical School
Keywords: networks; causal inference; interference; spillover effects

The Networks, Norms, and HIV/STI Risk Among Youth (NNAHRAY) study consists of observational data on a network composed of interconnected sexual and injection drug use partnerships. Like many interpersonal networks, two persons who are not engaged in a partnership may still be associated via transitive connections. Thus, when estimating causal effects of an intervention, statistical methods should account for both (1) long-range outcome dependence between two persons on the network and (2) arbitrary forms of interference whereby one person's outcome is affected by other persons' exposures. In recent work, we developed the auto-g-computation algorithm for causal inference on a single realization of a network of connected units. This algorithm relied on certain coding estimators, which are generally inefficient by virtue of censoring observations and may be unstable even in moderately dense networks. In this work, we develop a Bayesian auto-g-computation algorithm which incorporates data on the entire network and is substantially more efficient than the coding estimator. We then evaluate the effect of prior incarceration on HIV, STI, and Hepatitis C prevalence on the NNAHRAY network.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program