Handling time varying confounding in observational research

By Mohammad Ali Mansournia, M. Etminan, G. Danaei, J. Kaufman, Gary S. Collins. 2017

333 citations recorded by the source corpus.

Many exposures of epidemiological interest are time varying, and the values of potential confounders may change over time leading to time varying confounding. The aim of many longitudinal studies is to estimate the causal effect of a time varying exposure on an outcome that requires adjusting for time varying confounding. Time varying confounding affected by previous exposure often occurs in practice, but it is usually adjusted for by using conventional analytical methods such as time dependent Cox regression, random effects models, or generalised estimating equations, which are known to provide biased effect estimates in this setting. This article explains time varying confounding affected by previous exposure and outlines three causal methods proposed to appropriately adjust for this potential bias: inverse-probability-of-treatment weighting, the parametric G formula, and G estimation.

Subfields: Statistics and Probability, Causal Inference & Discovery

How researchers see this paper

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

In Statistics and Probability

In Causal Inference & Discovery

Citations in the corpus

Time-series forecasting with deep learning: a survey (2020, background): “Without any adjusting for timedependent confounders, straightforward estimation techniques can results in biased results, as shown in [78].”
Time Series Deconfounder: Estimating Treatment Effects over Time in the Presence of Hidden Confounders (2019, methodology): “MSMs use inverse probability of treatment weighting (IPTW) to adjust for the time-dependent confounding bias present in observational datasets (Mansournia et al., 2017; Bica et al., 2020b).”
Estimating Counterfactual Treatment Outcomes over Time Through Adversarially Balanced Representations (2020, background): “In Figure 6 (a), the covariate X is a time-dependent confounder because it affects the treatment assignments and at the same time, its value is changed by past treatments (Mansournia et al., 2017), as illustrated by the red arrows.”
A tool to assess risk of bias in non-randomized follow-up studies of exposure effects (ROBINS-E) (2024, positive): “The second variant additionally addresses time-varying confounding, which can occur when the exposure changes over time and follow-up time is split according to exposure level (Hernán and Robins, 2020; Mansournia et al., 2017).”