| Autor: |
Chatton A; INSERM UMR 1246 - SPHERE, Université de Nantes, Université de Tours, Nantes, France.; A2COM-IDBC, Pacé, France., Le Borgne F; INSERM UMR 1246 - SPHERE, Université de Nantes, Université de Tours, Nantes, France.; A2COM-IDBC, Pacé, France., Leyrat C; INSERM UMR 1246 - SPHERE, Université de Nantes, Université de Tours, Nantes, France.; Department of Medical Statistics & Cancer Survival Group, London School of Hygiene and Tropical Medicine, London, UK., Gillaizeau F; INSERM UMR 1246 - SPHERE, Université de Nantes, Université de Tours, Nantes, France.; Centre Hospitalier Universitaire de Nantes, Nantes, France., Rousseau C; INSERM UMR 1246 - SPHERE, Université de Nantes, Université de Tours, Nantes, France.; Centre Hospitalier Universitaire de Nantes, Nantes, France.; INSERM CIC1414, CHU Rennes, Rennes, France., Barbin L; Centre Hospitalier Universitaire de Nantes, Nantes, France., Laplaud D; Centre Hospitalier Universitaire de Nantes, Nantes, France.; Centre de Recherche en Transplantation et Immunologie INSERM UMR1064, Université de Nantes, Nantes, France., Léger M; INSERM UMR 1246 - SPHERE, Université de Nantes, Université de Tours, Nantes, France.; Département d'Anesthésie-Réanimation, Centre Hospitalier Universitaire d'Angers, Angers, France., Giraudeau B; INSERM UMR 1246 - SPHERE, Université de Nantes, Université de Tours, Nantes, France.; INSERM CIC1415, CHRU de Tours, Tours, France., Foucher Y; INSERM UMR 1246 - SPHERE, Université de Nantes, Université de Tours, Nantes, France. Yohann.Foucher@univ-nantes.fr.; Centre Hospitalier Universitaire de Nantes, Nantes, France. Yohann.Foucher@univ-nantes.fr. |
| Abstrakt: |
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. |
