Variable importance measures for heterogeneous treatment effects
Biometrics · 08.10.2025 · 81 (4)
DOI: 10.1093/biomtc/ujaf140 ↗Kurz gefasst
Fachartikel in der Zeitschrift Biometrics, online erschienen am 24.12.2025. Verfasst von Oliver J Hines und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Epidemiology, Columbia University, New York, NY 10032, United States. Hauptthemen laut MEDLINE: Machine Learning, Precision Medicine. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 45 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: Wellcome Trust, Medical Research Council, Royal Society-Welcome Trust.
Abstract (Originalsprache)
Motivated by applications in precision medicine and treatment effect heterogeneity, recent research has focused on estimating conditional average treatment effects (CATEs) using machine learning (ML). CATE estimates may represent complicated functions that provide little insight into the key drivers of heterogeneity. Therefore, we introduce nonparametric treatment effect variable importance measures (TE-VIMs), based on the mean-squared error (MSE) in predicting the individual treatment effect. More precisely, TE-VIMs represent the increase in MSE when variables are removed from the CATE conditioning set. We derive efficient TE-VIM estimators which can be used with any CATE estimation strategy and are amenable to ML estimation. We propose several strategies to calculate these VIMs (eg, leave-one out, or keep-one in), using popular meta-learners for the CATE. We study the finite sample performance through a simulation study and illustrate their application using clinical trial data.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Epidemiology, Columbia University, New York, NY 10032, United States.
- Department of Statistical Science, University College London, London, WC1E 6BT, United Kingdom.
- Department of Mathematics, Computer Science and Statistics, Ghent University, 9000 Ghent, Belgium.
ORCID-Profile: Oliver J Hines, Karla Diaz-Ordaz
Förderung
- Wellcome Trust
- Medical Research Council
- Royal Society-Welcome Trust
Zitieren
APA
Hines, O. J., Diaz-Ordaz, K., & Vansteelandt, S. (2025). Variable importance measures for heterogeneous treatment effects. Biometrics, 81(4). https://doi.org/10.1093/biomtc/ujaf140
Vancouver
Hines OJ, Diaz-Ordaz K, Vansteelandt S. Variable importance measures for heterogeneous treatment effects. Biometrics. 2025;81(4). doi:10.1093/biomtc/ujaf140
BibTeX
@article{Hines2025,
author = {Hines, Oliver J and Diaz-Ordaz, Karla and Vansteelandt, Stijn},
title = {Variable importance measures for heterogeneous treatment effects},
journal = {Biometrics},
year = {2025},
volume = {81},
number = {4},
doi = {10.1093/biomtc/ujaf140}
}
EUVIMED