Efficiency and robustness of causal effect estimators when noncompliance is measured with error

Research output: Contribution to journalArticle

Abstract

Estimating causal effects from randomized controlled trials is often complicated due to participant noncompliance to randomized treatment. Although there are a variety of methods to estimate causal effects in the presence of noncompliance, they generally make the assumption that noncompliance is measured without error. This is frequently an untenable assumption, particularly when noncompliance is based on participant self-report. To overcome this issue, we treat compliance as an unobserved variable and show how to estimate the probability of compliance given a biomarker of treatment and the other observed data. We present inverse probability weighted estimators, regression-based estimators, and a doubly-robust augmented estimator that rely on the estimated probability of compliance rather than an indicator of compliance. We investigate the finite-sample properties of the estimators and their efficiency and robustness under correctly specified or misspecified models, and we apply the estimators to a recently completed trial of very low nicotine content cigarettes.

Original languageEnglish (US)
Pages (from-to)4126-4141
Number of pages16
JournalStatistics in Medicine
Volume37
Issue number28
DOIs
StatePublished - Dec 10 2018

Fingerprint

Noncompliance
Causal Effect
Compliance
Robustness
Estimator
Misspecified Model
Randomized Controlled Trial
Nicotine
Tobacco Products
Self Report
Regression Estimator
Biomarkers
Randomized Controlled Trials
Estimate

Keywords

  • causal inference
  • inverse probability weighting
  • measurement error
  • noncompliance
  • randomized controlled trials

PubMed: MeSH publication types

  • Journal Article
  • Research Support, N.I.H., Extramural
  • Research Support, U.S. Gov't, P.H.S.

Cite this

@article{fa7a0dd79f174317be922a5b7d82306c,
title = "Efficiency and robustness of causal effect estimators when noncompliance is measured with error",
abstract = "Estimating causal effects from randomized controlled trials is often complicated due to participant noncompliance to randomized treatment. Although there are a variety of methods to estimate causal effects in the presence of noncompliance, they generally make the assumption that noncompliance is measured without error. This is frequently an untenable assumption, particularly when noncompliance is based on participant self-report. To overcome this issue, we treat compliance as an unobserved variable and show how to estimate the probability of compliance given a biomarker of treatment and the other observed data. We present inverse probability weighted estimators, regression-based estimators, and a doubly-robust augmented estimator that rely on the estimated probability of compliance rather than an indicator of compliance. We investigate the finite-sample properties of the estimators and their efficiency and robustness under correctly specified or misspecified models, and we apply the estimators to a recently completed trial of very low nicotine content cigarettes.",
keywords = "causal inference, inverse probability weighting, measurement error, noncompliance, randomized controlled trials",
author = "Boatman, {Jeffrey A.} and Vock, {David M} and Joe Koopmeiners",
year = "2018",
month = "12",
day = "10",
doi = "10.1002/sim.7922",
language = "English (US)",
volume = "37",
pages = "4126--4141",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "28",

}

TY - JOUR

T1 - Efficiency and robustness of causal effect estimators when noncompliance is measured with error

AU - Boatman, Jeffrey A.

AU - Vock, David M

AU - Koopmeiners, Joe

PY - 2018/12/10

Y1 - 2018/12/10

N2 - Estimating causal effects from randomized controlled trials is often complicated due to participant noncompliance to randomized treatment. Although there are a variety of methods to estimate causal effects in the presence of noncompliance, they generally make the assumption that noncompliance is measured without error. This is frequently an untenable assumption, particularly when noncompliance is based on participant self-report. To overcome this issue, we treat compliance as an unobserved variable and show how to estimate the probability of compliance given a biomarker of treatment and the other observed data. We present inverse probability weighted estimators, regression-based estimators, and a doubly-robust augmented estimator that rely on the estimated probability of compliance rather than an indicator of compliance. We investigate the finite-sample properties of the estimators and their efficiency and robustness under correctly specified or misspecified models, and we apply the estimators to a recently completed trial of very low nicotine content cigarettes.

AB - Estimating causal effects from randomized controlled trials is often complicated due to participant noncompliance to randomized treatment. Although there are a variety of methods to estimate causal effects in the presence of noncompliance, they generally make the assumption that noncompliance is measured without error. This is frequently an untenable assumption, particularly when noncompliance is based on participant self-report. To overcome this issue, we treat compliance as an unobserved variable and show how to estimate the probability of compliance given a biomarker of treatment and the other observed data. We present inverse probability weighted estimators, regression-based estimators, and a doubly-robust augmented estimator that rely on the estimated probability of compliance rather than an indicator of compliance. We investigate the finite-sample properties of the estimators and their efficiency and robustness under correctly specified or misspecified models, and we apply the estimators to a recently completed trial of very low nicotine content cigarettes.

KW - causal inference

KW - inverse probability weighting

KW - measurement error

KW - noncompliance

KW - randomized controlled trials

UR - http://www.scopus.com/inward/record.url?scp=85052685934&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85052685934&partnerID=8YFLogxK

U2 - 10.1002/sim.7922

DO - 10.1002/sim.7922

M3 - Article

VL - 37

SP - 4126

EP - 4141

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 28

ER -