EUVIMEDOpdag EUVIMED
Open access · CC BY 4.0

How Effective Are Machine Learning and Doubly Robust Estimators in Incorporating High‐Dimensional Proxies to Reduce Residual Confounding?

Mohammad Ehsanul Karim iD, Yang Lei

Pharmacoepidemiology and Drug Safety · 2025

DOI: 10.1002/pds.70155 ↗

At a glance

Journal article in Pharmacoepidemiology and Drug Safety, 2025. Written by Mohammad Ehsanul Karim and 1 co-authors. Institution: School of Population and Public Health, University of British Columbia, Vancouver, British Columbia, Canada. Major MEDLINE topics: Machine Learning, Observational Studies as Topic. Cited 4 times so far (Crossref); 43 references. The full text is freely available (open access, CC BY 4.0). Funded by: Natural Sciences and Engineering Research Council of Canada, MEK's Natural Sciences and Engineering Research Council of Canada (NSERC).

AbstractSubjectsBibliografiske oplysningerCiteRelated studies

Abstract (original language)

Background

Residual confounding presents a persistent challenge in observational studies, particularly in high‐dimensional settings. High‐dimensional proxy adjustment methods, such as the high‐dimensional propensity score (hdPS), are widely used to address confounding bias by incorporating proxies for unmeasured confounders. Extensions of hdPS have integrated machine learning, such as LASSO and super learner (SL), and doubly robust estimators, such as targeted maximum likelihood estimation (TMLE). However, the comparative performance of these methods, especially under different learner configurations and high‐dimensional proxies, remains unclear.

Method

We conducted plasmode simulations to evaluate the performance of standard methods, SL, TMLE, and double cross‐fit TMLE (DC‐TMLE) under varying exposure and outcome prevalence scenarios. Learner libraries included: 1 learner (logistic regression), 3 learners (logistic regression, MARS, and LASSO), and 4 learners (adding XGBoost, a non‐Donsker learner). Metrics included bias, coverage, and variability.

Results

Methods without proxies exhibited the highest bias and poorest coverage, highlighting the critical role of proxies in confounding adjustment. Standard methods incorporating high‐dimensional proxies showed robust performance, achieving low bias and near‐nominal coverage. TMLE and DC‐TMLE reduced bias but exhibited worse coverage compared to standard methods, particularly with larger learner libraries. Notably, DC‐TMLE, expected to address under‐coverage issues, failed to perform adequately in high‐dimensional settings with non‐Donsker learners, further emphasizing the instability introduced by complex libraries.

Conclusion

Our findings underscore the utility of high‐dimensional proxies in standard methods and the importance of tailoring learner configurations in SL and TMLE to ensure reliable confounding adjustment in high‐dimensional contexts.

Subjects

MeSH terms (MEDLINE)

Machine LearningObservational Studies as TopicHumansConfounding Factors, EpidemiologicPropensity ScoreLikelihood FunctionsBiasComputer SimulationLogistic Models

Author keywords

TMLEhdPShigh‐dimensionplasmode simulationproxy adjustment

Authors and institutions

  • School of Population and Public Health, University of British Columbia, Vancouver, British Columbia, Canada.
  • Centre for Advancing Health Outcomes, University of British Columbia, Vancouver, British Columbia, Canada.
  • Department of Statistics, University of British Columbia, Vancouver, British Columbia, Canada.

ORCID profiles: Mohammad Ehsanul Karim

Funding

  • Natural Sciences and Engineering Research Council of Canada
  • MEK's Natural Sciences and Engineering Research Council of Canada (NSERC)

Cite

APA

Karim, M. E., & Lei, Y. (2025). How Effective Are Machine Learning and Doubly Robust Estimators in Incorporating High‐Dimensional Proxies to Reduce Residual Confounding?. Pharmacoepidemiology and Drug Safety. https://doi.org/10.1002/pds.70155

Vancouver

Karim ME, Lei Y. How Effective Are Machine Learning and Doubly Robust Estimators in Incorporating High‐Dimensional Proxies to Reduce Residual Confounding?. Pharmacoepidemiology and Drug Safety. 2025. doi:10.1002/pds.70155

BibTeX
@article{Karim2025,
  author = {Karim, Mohammad Ehsanul and Lei, Yang},
  title = {How Effective Are Machine Learning and Doubly Robust Estimators in Incorporating High‐Dimensional Proxies to Reduce Residual Confounding?},
  journal = {Pharmacoepidemiology and Drug Safety},
  year = {2025},
  doi = {10.1002/pds.70155}
}
Start for free →