G-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: a comparative simulation study

Abstract Controlling for confounding bias is crucial in causal inference. Distinct methods are currently employed to mitigate the effects of confounding bias. Each requires the introduction of a set of covariates, which remains difficult to choose, especially regarding the different methods. We cond...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Arthur Chatton, Florent Le Borgne, Clémence Leyrat, Florence Gillaizeau, Chloé Rousseau, Laetitia Barbin, David Laplaud, Maxime Léger, Bruno Giraudeau, Yohann Foucher
Formato: article
Lenguaje:EN
Publicado: Nature Portfolio 2020
Materias:
R
Q
Acceso en línea:https://doaj.org/article/694ea9cc77664ee690089eec2f9159a9
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:694ea9cc77664ee690089eec2f9159a9
record_format dspace
spelling oai:doaj.org-article:694ea9cc77664ee690089eec2f9159a92021-12-02T17:52:42ZG-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: a comparative simulation study10.1038/s41598-020-65917-x2045-2322https://doaj.org/article/694ea9cc77664ee690089eec2f9159a92020-06-01T00:00:00Zhttps://doi.org/10.1038/s41598-020-65917-xhttps://doaj.org/toc/2045-2322Abstract Controlling for confounding bias is crucial in causal inference. Distinct methods are currently employed to mitigate the effects of confounding bias. Each requires the introduction of a set of covariates, which remains difficult to choose, especially regarding the different methods. We conduct a simulation study to compare the relative performance results obtained by using four different sets of covariates (those causing the outcome, those causing the treatment allocation, those causing both the outcome and the treatment allocation, and all the covariates) and four methods: g-computation, inverse probability of treatment weighting, full matching and targeted maximum likelihood estimator. Our simulations are in the context of a binary treatment, a binary outcome and baseline confounders. The simulations suggest that considering all the covariates causing the outcome led to the lowest bias and variance, particularly for g-computation. The consideration of all the covariates did not decrease the bias but significantly reduced the power. We apply these methods to two real-world examples that have clinical relevance, thereby illustrating the real-world importance of using these methods. We propose an R package RISCA to encourage the use of g-computation in causal inference.Arthur ChattonFlorent Le BorgneClémence LeyratFlorence GillaizeauChloé RousseauLaetitia BarbinDavid LaplaudMaxime LégerBruno GiraudeauYohann FoucherNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 10, Iss 1, Pp 1-13 (2020)
institution DOAJ
collection DOAJ
language EN
topic Medicine
R
Science
Q
spellingShingle Medicine
R
Science
Q
Arthur Chatton
Florent Le Borgne
Clémence Leyrat
Florence Gillaizeau
Chloé Rousseau
Laetitia Barbin
David Laplaud
Maxime Léger
Bruno Giraudeau
Yohann Foucher
G-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: a comparative simulation study
description Abstract Controlling for confounding bias is crucial in causal inference. Distinct methods are currently employed to mitigate the effects of confounding bias. Each requires the introduction of a set of covariates, which remains difficult to choose, especially regarding the different methods. We conduct a simulation study to compare the relative performance results obtained by using four different sets of covariates (those causing the outcome, those causing the treatment allocation, those causing both the outcome and the treatment allocation, and all the covariates) and four methods: g-computation, inverse probability of treatment weighting, full matching and targeted maximum likelihood estimator. Our simulations are in the context of a binary treatment, a binary outcome and baseline confounders. The simulations suggest that considering all the covariates causing the outcome led to the lowest bias and variance, particularly for g-computation. The consideration of all the covariates did not decrease the bias but significantly reduced the power. We apply these methods to two real-world examples that have clinical relevance, thereby illustrating the real-world importance of using these methods. We propose an R package RISCA to encourage the use of g-computation in causal inference.
format article
author Arthur Chatton
Florent Le Borgne
Clémence Leyrat
Florence Gillaizeau
Chloé Rousseau
Laetitia Barbin
David Laplaud
Maxime Léger
Bruno Giraudeau
Yohann Foucher
author_facet Arthur Chatton
Florent Le Borgne
Clémence Leyrat
Florence Gillaizeau
Chloé Rousseau
Laetitia Barbin
David Laplaud
Maxime Léger
Bruno Giraudeau
Yohann Foucher
author_sort Arthur Chatton
title G-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: a comparative simulation study
title_short G-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: a comparative simulation study
title_full G-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: a comparative simulation study
title_fullStr G-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: a comparative simulation study
title_full_unstemmed G-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: a comparative simulation study
title_sort g-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: a comparative simulation study
publisher Nature Portfolio
publishDate 2020
url https://doaj.org/article/694ea9cc77664ee690089eec2f9159a9
work_keys_str_mv AT arthurchatton gcomputationpropensityscorebasedmethodsandtargetedmaximumlikelihoodestimatorforcausalinferencewithdifferentcovariatessetsacomparativesimulationstudy
AT florentleborgne gcomputationpropensityscorebasedmethodsandtargetedmaximumlikelihoodestimatorforcausalinferencewithdifferentcovariatessetsacomparativesimulationstudy
AT clemenceleyrat gcomputationpropensityscorebasedmethodsandtargetedmaximumlikelihoodestimatorforcausalinferencewithdifferentcovariatessetsacomparativesimulationstudy
AT florencegillaizeau gcomputationpropensityscorebasedmethodsandtargetedmaximumlikelihoodestimatorforcausalinferencewithdifferentcovariatessetsacomparativesimulationstudy
AT chloerousseau gcomputationpropensityscorebasedmethodsandtargetedmaximumlikelihoodestimatorforcausalinferencewithdifferentcovariatessetsacomparativesimulationstudy
AT laetitiabarbin gcomputationpropensityscorebasedmethodsandtargetedmaximumlikelihoodestimatorforcausalinferencewithdifferentcovariatessetsacomparativesimulationstudy
AT davidlaplaud gcomputationpropensityscorebasedmethodsandtargetedmaximumlikelihoodestimatorforcausalinferencewithdifferentcovariatessetsacomparativesimulationstudy
AT maximeleger gcomputationpropensityscorebasedmethodsandtargetedmaximumlikelihoodestimatorforcausalinferencewithdifferentcovariatessetsacomparativesimulationstudy
AT brunogiraudeau gcomputationpropensityscorebasedmethodsandtargetedmaximumlikelihoodestimatorforcausalinferencewithdifferentcovariatessetsacomparativesimulationstudy
AT yohannfoucher gcomputationpropensityscorebasedmethodsandtargetedmaximumlikelihoodestimatorforcausalinferencewithdifferentcovariatessetsacomparativesimulationstudy
_version_ 1718379172272799744