Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Sensitivity analyses for unmeasured confounding assuming a marginal structural model for repeated measures.

Babette A Brumback1, Miguel A Hernán, Sebastien J P A Haneuse

  • 1Department of Biostatistics, UCLA School of Public Health, Los Angeles, CA 90095, USA. brumback@ucla.du

Statistics in Medicine
|February 26, 2004
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Global Sensitivity Analysis for Studies Extending Inferences From a Randomized Trial to a Target Population.

Statistics in medicine·2026
Same author

Nonparametric identification is not enough, but randomized controlled trials are.

Observational studies·2025
Same author

Rejoinder: Nonparametric identification is not enough, but randomized controlled trials are.

Observational studies·2025
Same author

Minimax rates for heterogeneous causal effect estimation.

Annals of statistics·2025
Same author

Erratum: Comparison of the Test-negative Design and Cohort Design With Explicit Target Trial Emulation for Evaluating COVID-19 Vaccine Effectiveness.

Epidemiology (Cambridge, Mass.)·2025
Same author

The Target Trial Framework for Causal Inference From Observational Data: Why and When Is It Helpful?

Annals of internal medicine·2025

This study introduces a new framework to assess how unmeasured confounding affects causal effect estimates from marginal structural models (MSMs). Findings show that even moderate unmeasured confounding can alter conclusions, highlighting the importance of sensitivity analyses in time-varying treatment studies.

Area of Science:

  • Causal inference
  • Biostatistics
  • Epidemiology

Background:

  • Marginal structural models (MSMs) and inverse probability of treatment weighted (IPTW) estimators are used for time-varying treatments.
  • Sensitivity of IPTW estimators to unmeasured confounding is a critical concern.

Purpose of the Study:

  • To investigate the sensitivity of IPTW estimators to unmeasured confounding.
  • To introduce a novel framework for sensitivity analyses using a nonidentifiable model.
  • To present augmented IPTW estimators for MSM parameters.

Main Methods:

  • Developed a new framework for sensitivity analyses quantifying unmeasured confounding.
  • Introduced augmented IPTW estimators for MSM parameters.
  • Applied methods to analyze zidovudine therapy's effect on CD4 counts in HIV-infected men.

Related Experiment Videos

Main Results:

  • Under no unmeasured confounding, the 95% confidence interval for treatment effect included zero.
  • With moderate unmeasured confounding, the 95% confidence interval no longer included zero.
  • The analysis of zidovudine therapy's effect on CD4 counts demonstrated sensitivity to unmeasured confounding.

Conclusions:

  • The study's findings indicate that analyses using MSMs can be sensitive to unmeasured confounding.
  • The proposed framework facilitates sensitivity analyses for time-varying treatment effects.
  • Encourages broader application of sensitivity analyses in similar research contexts.