Numerous estimation procedures employed in causal inference often rely on accurately measured data. However, the prevalence of measurement errors in practical studies may yield biased effect estimates. It is common to employ validation samples to rectify such biases in the measurement error literature. This article focuses on the estimation of the average causal effect with a misclassified binary treatment in a primary population of interest. By leveraging a validation sample with covariates, an error-prone version of treatment and a true treatment recorded, we provide identifiability results under certain conditions. Building on identifiability, we explore three classes of estimators, each demonstrating consistency and asymptotic normality within distinct model sets. Furthermore, we propose a multiply robust estimation approach for the treatment effect based on the semiparametric theory framework. The multiply robust estimator retains consistent under any one of the listed model sets and achieves the semiparametric efficiency bound, provided all models are correct. We demonstrate the satisfactory performance of the proposed estimators through simulation studies and a real data analysis.
RobinsJM. Association, causation, and marginal structural models. Synthese1999; 121: 151–179.
2.
RobinsJMHernanMABrumbackB. Marginal structural models and causal inference in epidemiology. Epidemiology2000; 11: 550–560.
3.
LuncefordJKDavidianM. Stratification and weighting via the propensity score in estimation of causal treatment effects: a comparative study. Stat Med2004; 23: 2937–2960.
4.
LindenAAdamsJL. Using propensity score-based weighting in the evaluation of health management programme effectiveness. J Eval Clin Pract2010; 16: 175–179.
5.
YiGY. Statistical Analysis with Measurement Error or Misclassification: Strategy, Method and Application. New York: Springer, 2017.
6.
VanderWeeleTJLiY. Simple sensitivity analysis for differential measurement error. Am J Epidemiol2019; 188: 1823–1829.
7.
ShepherdBEYuC. Accounting for data errors discovered from an audit in multiple linear regression. Biometrics2011; 67: 1083–1091.
8.
XuYKimJKLiY. Semiparametric estimation for measurement error models with validation data. Canadian Journal of Statistics2017; 45: 185–201.
9.
AmorimGTaoRLotspeichS, et al.Two-phase sampling designs for data validation in settings with covariate measurement error and continuous outcome. J R Stat Soc Ser A: Stat Soc2021; 184: 1368–1389.
10.
TaoRLotspeichSCAmorimG, et al.Efficient semiparametric inference for two-phase studies with outcome and covariate measurement errors. Stat Med2021; 40: 725–738.
11.
AmorimGTaoRLotspeichS, et al.Three-phase generalized raking and multiple imputation estimators to address error-prone data. Stat Med2024; 43: 379–394.
12.
BoatmanJAVockDMKoopmeinersJS, et al.Estimating causal effects from a randomized clinical trial when noncompliance is measured with error. Biostatistics2018; 19: 103–118.
13.
YanagiT. Inference on local average treatment effects for misclassified treatment. Econom Rev2019; 38: 938–960.
14.
JiangZDingP. Measurement errors in the binary instrumental variable model. Biometrika2020; 107: 238–245.
15.
Van der WeeleTJValeriLOgburnEL. The role of measurement error and misclassification in mediation analysis: mediation and measurement error. Epidemiology2012; 23: 561–564.
16.
ChengCSpiegelmanDLiF. Mediation analysis in the presence of continuous exposure measurement error. Stat Med2023; 42: 1669–1686.
ZhuYGultchinLGrettonA, et al.Causal inference with treatment measurement error: a nonparametric instrumental variable approach. In Uncertainty in Artificial Intelligence. PMLR, pp.2414–2424.
21.
GravelCAPlattRW. Weighted estimation for confounded binary outcomes subject to misclassification. Stat Med2018; 37: 425–436.
22.
ShuDYiGY. Causal inference with measurement error in outcomes: bias analysis and estimation methods. Stat Methods Med Res2019; 28: 2049–2068.
23.
ShuDYiGY. Inverse-probability-of-treatment weighted estimation of causal parameters in the presence of error-contaminated and time-dependent confounders. Biometrical J2019; 61: 1507–1525.
24.
ShuDYiGY. Weighted causal inference methods with mismeasured covariates and misclassified outcomes. Stat Med2019; 38: 1835–1854.
25.
EdwardsJKColeSRFoxMP. Flexibly accounting for exposure misclassification with external validation data. Am J Epidemiol2020; 189: 850–860.
26.
RichardsonDBKeilAPEdwardsJK, et al.A bespoke instrumental variable approach to correction for exposure measurement error. Am J Epidemiol2022; 191: 1954–1961.
27.
LurbetMFRojanoBWhittaker BrownSA, et al.Obesity trends among asthma patients in the united states: a population-based study. Ann Glob Health2019; 85: 10.
28.
LiuJLiW. A semiparametric method for evaluating causal effects in the presence of error-prone covariates. Biometrical J2021; 63: 1202–1222.
29.
CarrollRJRuppertDStefanskiLA, et al.Measurement Error in Nonlinear Models: a Modern Perspective. Boca Raton: Chapman & Hall/CRC Press, 2006.
30.
WangLTchetgenE. Bounded, efficient and multiply robust estimation of average treatment effects using instrumental variables. J R Stat Soc Ser B: Stat Methodol2018; 80: 531–550.
31.
SunBMiaoW. On semiparametric instrumental variable estimation of average treatment effects through data fusion. Stat Sin2022; 32: 569–590.
LewbelA. Estimation of average treatment effects with misclassification. Econometrica2007; 75: 537–551.
34.
RidgewayGKovalchikSAGriffinBA, et al.Propensity score analysis with survey weighted data. J Causal Inference2015; 3: 237–249.
35.
NeweyWKMcFaddenD. Large sample estimation and hypothesis testing. Handbook Economet1994; 4: 2111–2245.
36.
Van der VaartAW. Asymptotic Statistics. 3. Cambridge: Cambridge University Press, 2000.
37.
TsiatisAA. Semiparametric Theory and Missing Data. 2. New York: Springer, 2006.
38.
BangHRobinsJM. Doubly robust estimation in missing data and causal inference models. Biometrics2005; 61: 962–973.
39.
TanZ. Regression and weighting methods for causal inference using instrumental variables. J Am Stat Assoc2006; 101: 1607–1618.
40.
SunBLiuLMiaoWet al.Semiparametric estimation with data missing not at random using an instrumental variable. Stat Sin2018; 28: 1965–1983.
41.
ShiXMiaoWNelsonJC, et al.Multiply robust causal inference with double-negative control adjustment for categorical unmeasured confounding. J R Stat Soc Ser B: Stat Methodol2020; 82: 521–540.
42.
RobinsJM. Correcting for non-compliance in randomized trials using structural nested mean models. Commun Stat - Theory Methods1994; 23: 2379–2412.
43.
PetersenMLvan der LaanMJ. Causal models and learning from data: integrating causal modeling and statistical estimation. Epidemiology2014; 25: 418–426.
44.
ChernozhukovVChetverikovDDemirerM, et al.Double/debiased machine learning for treatment and structural parameters. Econom J2018; 21: C1–C68.
45.
DíazI. Machine learning in the estimation of causal effects: targeted minimum loss-based estimation and double/debiased machine learning. Biostatistics2019; 21: 353–358.
46.
HinesODukesODiaz-OrdazK, et al.Demystifying statistical learning based on efficient influence functions. Am Stat2022; 76: 292–304.
47.
JiangZYangSDingP. Multiply robust estimation of causal effects under principal ignorability. J R Stat Soc Ser B: Stat Methodol2022; 84: 1423–1445.
48.
BraunDZiglerCDominiciF, et al.Using validation data to adjust the inverse probability weighting estimator for misclassified treatment. Harvard University Biostatistics Working Paper Series Working Paper 2012016; 1–16.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
