Double Robust Efficient Estimators of Longitudinal Treatment Effects: Comparative Performance in Simulations and a Case Study

Linh Tran, Constantin Yiannoutsos, Kara Wools-Kaloustian, Abraham Siika, Mark Van Der Laan, Maya Petersen

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

A number of sophisticated estimators of longitudinal effects have been proposed for estimating the intervention-Specific mean outcome. However, there is a relative paucity of research comparing these methods directly to one another. In this study, we compare various approaches to estimating a causal effect in a longitudinal treatment setting using both simulated data and data measured from a human immunodeficiency virus cohort. Six distinct estimators are considered: (i) an iterated conditional expectation representation, (ii) an inverse propensity weighted method, (iii) an augmented inverse propensity weighted method, (iv) a double robust iterated conditional expectation estimator, (v) a modified version of the double robust iterated conditional expectation estimator, and (vi) a targeted minimum loss-Based estimator. The details of each estimator and its implementation are presented along with nuisance parameter estimation details, which include potentially pooling the observed data across all subjects regardless of treatment history and using data adaptive machine learning algorithms. Simulations are constructed over six time points, with each time point steadily increasing in positivity violations. Estimation is carried out for both the simulations and applied example using each of the six estimators under both stratified and pooled approaches of nuisance parameter estimation. Simulation results show that double robust estimators remained without meaningful bias as long as at least one of the two nuisance parameters were estimated with a correctly specified model. Under full misspecification, the bias of the double robust estimators remained better than that of the inverse propensity estimator under misspecification, but worse than the iterated conditional expectation estimator. Weighted estimators tended to show better performance than the covariate estimators. As positivity violations increased, the mean squared error and bias of all estimators considered became worse, with covariate-Based double robust estimators especially susceptible. Applied analyses showed similar estimates at most time points, with the important exception of the inverse propensity estimator which deviated markedly as positivity violations increased. Given its efficiency, ability to respect the parameter space, and observed performance, we recommend the pooled and weighted targeted minimum loss-Based estimator.

Original languageEnglish (US)
JournalInternational Journal of Biostatistics
DOIs
StatePublished - Jan 1 2019

Fingerprint

Efficient Estimator
Robust Estimators
Treatment Effects
Estimator
Simulation
Conditional Expectation
Nuisance Parameter
Positivity
Misspecification
Treatment effects
Parameter Estimation
Covariates
Causal Effect
Adaptive Learning
Pooling
Mean Squared Error
Exception
Virus
Parameter Space
Two Parameters

Keywords

  • Aiptw
  • causal inference
  • double robust
  • efficient influence function
  • iptw
  • longitudinal treatment
  • multiple testing
  • semiparametric models
  • tmle

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

@article{534d37932de642c68b3c12d3f074143f,
title = "Double Robust Efficient Estimators of Longitudinal Treatment Effects: Comparative Performance in Simulations and a Case Study",
abstract = "A number of sophisticated estimators of longitudinal effects have been proposed for estimating the intervention-Specific mean outcome. However, there is a relative paucity of research comparing these methods directly to one another. In this study, we compare various approaches to estimating a causal effect in a longitudinal treatment setting using both simulated data and data measured from a human immunodeficiency virus cohort. Six distinct estimators are considered: (i) an iterated conditional expectation representation, (ii) an inverse propensity weighted method, (iii) an augmented inverse propensity weighted method, (iv) a double robust iterated conditional expectation estimator, (v) a modified version of the double robust iterated conditional expectation estimator, and (vi) a targeted minimum loss-Based estimator. The details of each estimator and its implementation are presented along with nuisance parameter estimation details, which include potentially pooling the observed data across all subjects regardless of treatment history and using data adaptive machine learning algorithms. Simulations are constructed over six time points, with each time point steadily increasing in positivity violations. Estimation is carried out for both the simulations and applied example using each of the six estimators under both stratified and pooled approaches of nuisance parameter estimation. Simulation results show that double robust estimators remained without meaningful bias as long as at least one of the two nuisance parameters were estimated with a correctly specified model. Under full misspecification, the bias of the double robust estimators remained better than that of the inverse propensity estimator under misspecification, but worse than the iterated conditional expectation estimator. Weighted estimators tended to show better performance than the covariate estimators. As positivity violations increased, the mean squared error and bias of all estimators considered became worse, with covariate-Based double robust estimators especially susceptible. Applied analyses showed similar estimates at most time points, with the important exception of the inverse propensity estimator which deviated markedly as positivity violations increased. Given its efficiency, ability to respect the parameter space, and observed performance, we recommend the pooled and weighted targeted minimum loss-Based estimator.",
keywords = "Aiptw, causal inference, double robust, efficient influence function, iptw, longitudinal treatment, multiple testing, semiparametric models, tmle",
author = "Linh Tran and Constantin Yiannoutsos and Kara Wools-Kaloustian and Abraham Siika and {Van Der Laan}, Mark and Maya Petersen",
year = "2019",
month = "1",
day = "1",
doi = "10.1515/ijb-2017-0054",
language = "English (US)",
journal = "International Journal of Biostatistics",
issn = "1557-4679",
publisher = "Berkeley Electronic Press",

}

TY - JOUR

T1 - Double Robust Efficient Estimators of Longitudinal Treatment Effects

T2 - Comparative Performance in Simulations and a Case Study

AU - Tran, Linh

AU - Yiannoutsos, Constantin

AU - Wools-Kaloustian, Kara

AU - Siika, Abraham

AU - Van Der Laan, Mark

AU - Petersen, Maya

PY - 2019/1/1

Y1 - 2019/1/1

N2 - A number of sophisticated estimators of longitudinal effects have been proposed for estimating the intervention-Specific mean outcome. However, there is a relative paucity of research comparing these methods directly to one another. In this study, we compare various approaches to estimating a causal effect in a longitudinal treatment setting using both simulated data and data measured from a human immunodeficiency virus cohort. Six distinct estimators are considered: (i) an iterated conditional expectation representation, (ii) an inverse propensity weighted method, (iii) an augmented inverse propensity weighted method, (iv) a double robust iterated conditional expectation estimator, (v) a modified version of the double robust iterated conditional expectation estimator, and (vi) a targeted minimum loss-Based estimator. The details of each estimator and its implementation are presented along with nuisance parameter estimation details, which include potentially pooling the observed data across all subjects regardless of treatment history and using data adaptive machine learning algorithms. Simulations are constructed over six time points, with each time point steadily increasing in positivity violations. Estimation is carried out for both the simulations and applied example using each of the six estimators under both stratified and pooled approaches of nuisance parameter estimation. Simulation results show that double robust estimators remained without meaningful bias as long as at least one of the two nuisance parameters were estimated with a correctly specified model. Under full misspecification, the bias of the double robust estimators remained better than that of the inverse propensity estimator under misspecification, but worse than the iterated conditional expectation estimator. Weighted estimators tended to show better performance than the covariate estimators. As positivity violations increased, the mean squared error and bias of all estimators considered became worse, with covariate-Based double robust estimators especially susceptible. Applied analyses showed similar estimates at most time points, with the important exception of the inverse propensity estimator which deviated markedly as positivity violations increased. Given its efficiency, ability to respect the parameter space, and observed performance, we recommend the pooled and weighted targeted minimum loss-Based estimator.

AB - A number of sophisticated estimators of longitudinal effects have been proposed for estimating the intervention-Specific mean outcome. However, there is a relative paucity of research comparing these methods directly to one another. In this study, we compare various approaches to estimating a causal effect in a longitudinal treatment setting using both simulated data and data measured from a human immunodeficiency virus cohort. Six distinct estimators are considered: (i) an iterated conditional expectation representation, (ii) an inverse propensity weighted method, (iii) an augmented inverse propensity weighted method, (iv) a double robust iterated conditional expectation estimator, (v) a modified version of the double robust iterated conditional expectation estimator, and (vi) a targeted minimum loss-Based estimator. The details of each estimator and its implementation are presented along with nuisance parameter estimation details, which include potentially pooling the observed data across all subjects regardless of treatment history and using data adaptive machine learning algorithms. Simulations are constructed over six time points, with each time point steadily increasing in positivity violations. Estimation is carried out for both the simulations and applied example using each of the six estimators under both stratified and pooled approaches of nuisance parameter estimation. Simulation results show that double robust estimators remained without meaningful bias as long as at least one of the two nuisance parameters were estimated with a correctly specified model. Under full misspecification, the bias of the double robust estimators remained better than that of the inverse propensity estimator under misspecification, but worse than the iterated conditional expectation estimator. Weighted estimators tended to show better performance than the covariate estimators. As positivity violations increased, the mean squared error and bias of all estimators considered became worse, with covariate-Based double robust estimators especially susceptible. Applied analyses showed similar estimates at most time points, with the important exception of the inverse propensity estimator which deviated markedly as positivity violations increased. Given its efficiency, ability to respect the parameter space, and observed performance, we recommend the pooled and weighted targeted minimum loss-Based estimator.

KW - Aiptw

KW - causal inference

KW - double robust

KW - efficient influence function

KW - iptw

KW - longitudinal treatment

KW - multiple testing

KW - semiparametric models

KW - tmle

UR - http://www.scopus.com/inward/record.url?scp=85062444708&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85062444708&partnerID=8YFLogxK

U2 - 10.1515/ijb-2017-0054

DO - 10.1515/ijb-2017-0054

M3 - Article

AN - SCOPUS:85062444708

JO - International Journal of Biostatistics

JF - International Journal of Biostatistics

SN - 1557-4679

ER -