Structural Nested Models and G-Estimation: The Partially Realized Promise

• Activity Number: 401
• Type: Invited
• Date/Time: Tuesday, August 11, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Stijn Vansteelandt* and Marshall Joffe
• Companies: Ghent University and University of Pennsylvania
• Keywords: causal inference ; G-estimation ; time-varying confounding ; propensity score ; time-varying exposure

Abstract

Structural nested models (SNMs) and the associated method of G-estimation were first proposed by James Robins over two decades ago as approaches to modelling and estimating the joint effects of a sequence of treatments or exposures. The models and estimation methods have since been extended to dealing with a broader series of problems, and have considerable advantages over the other methods developed for estimating such joint effects. Despite these advantages, the application of these methods in applied research has been relatively infrequent; we view this as unfortunate. In this talk, after a brief overview of the models and estimation methods as developed, primarily by Robins, I will provide insight into their advantages over other methods. I will then discuss remedies for two limitations of these methods that have hindered them to be more broadly adopted: the use of standard software for fitting SNMs and accounting for right censoring in the analysis of survival times.