Structural nested models and G-estimation: the partially realized promise

By Stijn Vansteelandt, M. Joffe. 2014

207 citations recorded by the source corpus.

Structural nested models (SNMs) and the associated method of G-estimation were first proposed by James Robins over two decades ago as approaches to modeling 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. To remedy this, we provide an overview of the models and estimation methods as developed, primarily by Robins, over the years. We provide insight into their advantages over other methods, and consider some possible reasons for failure of the methods to be more broadly adopted, as well as possible remedies. Finally, we consider several extensions of the standard models and estimation methods.

Subfields: Management Science and Operations Research, Statistics and Probability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Management Science and Operations Research

In Statistics and Probability

Citations in the corpus

G-Net: A Deep Learning Approach to G-computation for Counterfactual Outcome Prediction Under Dynamic Treatment Regimes (2020, methodology): “G-methods include gcomputation [4–6], structural nested models [7, 8], and marginal structural models [9, 10].”
Learning When-to-Treat Policies (2019, methodology): “…overview of the literature on dynamic decision rules is beyond the scope of this article; instead, we refer the reader to recent review papers (Vansteelandt and Joffe 2014; Kosorok and Laber 2019; Clifton and Laber 2020; Levine et al. 2020), as well as textbooks by Chakraborty and Moodie…”
Q-Learning: Theory and Applications (2020, methodology): “Robins (2004) introduced a related semiparametric estimator of the optimal policy,but this method has not been used as widely as Q-learning (Vansteelandt & Joffe 2014), perhaps because of its relative complexity.”
Learning Cost-Effective and Interpretable Treatment Regimes (2017, unknown): “Regression-based methods [27, 31, 28, 32, 43, 27, 25] model the conditional distribution of the outcomes given the treatment and characteristics of patients and choose the treatment resulting in the best possible outcome for each individual.”
An introduction to g methods. (2016, background): “If, at each time point, the exposure is conditionally independent of the potential outcomes (sequential exchangeability) then the conditional covariance between the exposure and potential outcomes is zero.(14) Formally, these conditional independence relations can be written as:”
On the Differences between General Cross-Lagged Panel Model and Random-Intercept Cross-Lagged Panel Model: Interpretation of Cross-Lagged Parameters and Model Choice (2020, background): “…estimates of causal parameters can be obtained when either a treatment assignment model or a model for outcome that would be observed if the treatment were stopped from a specific time can be correctly specified (the doubly robust property; see Vansteelandt & Joffe, 2014; Usami, 2020 for details).”
Towards optimal doubly robust estimation of heterogeneous causal effects (2020, background): “…models, for example partially linear models assuming τ ( x ) to be constant [Robins et al., 1992, Robinson, 1988], or structural nested models in which τ ( x ) followed some known parametric form [Robins, 1994, van der Laan, 2005, van der Laan and Robins, 2003, Vansteelandt and Joffe, 2014].”
Interactive Q-learning for Quantiles (2017, methodology): “Structural nested distribution models (SNDMs) estimated using g-estimation facilitate estimation of point treatment effects on the cumulative distribution function of the outcome (Robins 2000; Vansteelandt and Joffe 2014).”