Constructing g-computation estimators: two case studies in selection bias

TL;DR

This paper explores how to adapt g-computation estimators to complex selection bias scenarios in epidemiology, illustrating the process through two case studies and emphasizing practical implementation and theoretical properties.

Contribution

It demonstrates how to construct g-computation estimators for complex biases, providing a framework for translating causal diagrams into estimators with theoretical and simulation validation.

Findings

Adapted g-computation estimators for treatment-induced selection bias

Adapted g-computation estimators for biases without a joint adjustment set

Simulation studies illustrating estimator performance

Abstract

G-computation is a useful estimation method that can be adapted to address various biases in epidemiology. However, these adaptations may not be obvious for some complex causal structures. This challenge is an example of the much wider issue of translating a causal diagram into a novel estimation strategy. To highlight these challenges, we consider two recent cases from the selection bias literature: treatment-induced selection and co-occurrence of biases that lack a joint adjustment set. For each case study, we show how g-computation can be adapted, describe how to implement that adaptation, show some general statistical properties, and illustrate the estimator using simulation. To simplify both the theoretical study and practical application of our estimators, we express the proposed g-computation estimators as stacked estimating equations. These examples illustrate how epidemiologists can translate identification results into a g-computation estimator and study the theoretical and finite-sample properties of a novel estimator.