An improved multiply robust estimator for the average treatment effect

BMC Medical Research Methodology

Research

Kecheng Wei¹  Ce Wang¹  Chen Huang¹  Yongfu Yu²  Guoyou Qin² 

[1] Department of Biostatistics, Key Laboratory for Health Technology Assessment, National Commission of Health, Key Laboratory of Public Health Safety of Ministry of Education, School of Public Health, Fudan University, Shanghai, China;Department of Biostatistics, Key Laboratory for Health Technology Assessment, National Commission of Health, Key Laboratory of Public Health Safety of Ministry of Education, School of Public Health, Fudan University, Shanghai, China;Shanghai Institute of Infectious Disease and Biosecurity, Shanghai, China;

关键词: Average treatment effect;    Empirical likelihood;    Multiply robust;    Nonparametric model;    Parametric model;   

DOI  :  10.1186/s12874-023-02056-7

 received in 2023-07-08, accepted in 2023-10-01,  发布年份 2023

来源: Springer

【 摘 要 】

Background In observational studies, double robust or multiply robust (MR) approaches provide more protection from model misspecification than the inverse probability weighting and g-computation for estimating the average treatment effect (ATE). However, the approaches are based on parametric models, leading to biased estimates when all models are incorrectly specified. Nonparametric methods, such as machine learning or nonparametric double robust approaches, are robust to model misspecification, but the efficiency of nonparametric methods is low.MethodIn the study, we proposed an improved MR method combining parametric and nonparametric models based on the previous MR method (Han, JASA 109(507):1159-73, 2014) to improve the robustness to model misspecification and the efficiency. We performed comprehensive simulations to evaluate the performance of the proposed method.ResultsOur simulation study showed that the MR estimators with only outcome regression (OR) models, where one of the models was a nonparametric model, were the most recommended because of the robustness to model misspecification and the lowest root mean square error (RMSE) when including a correct parametric OR model. And the performance of the recommended estimators was comparative, even if all parametric models were misspecified. As an application, the proposed method was used to estimate the effect of social activity on depression levels in the China Health and Retirement Longitudinal Study dataset.ConclusionsThe proposed estimator with nonparametric and parametric models is more robust to model misspecification.

【 授权许可】

CC BY   
© BioMed Central Ltd., part of Springer Nature 2023

【 图 表 】

Fig. 1

Fig. 6

Fig. 7

12951_2015_155_Article_IEq73.gif

Fig. 6

12951_2015_155_Article_IEq72.gif

Fig. 3

Fig. 3

Fig. 1

Fig. 2

Fig. 2

Fig. 1

Fig. 2

Fig. 2

Fig. 4

Fig. 1

Fig. 3

Fig. 3

Fig. 1

Fig. 15

Fig. 3

Fig. 14

Fig. 5

Fig. 2

Fig. 13

Fig. 1

Fig. 1

Fig. 1

Fig. 12

Fig. 6

Fig. 3

Fig. 11

Fig. 2

Fig. 9

Fig. 1

12936_2017_1963_Article_IEq41.gif

Fig. 8

Fig. 8

Fig. 6

12951_2016_177_Article_IEq1.gif

12951_2015_155_Article_IEq65.gif

Fig. 4

Fig. 5

Fig. 3

Fig. 1

Fig. 2

Fig. 1

12936_2015_836_Article_IEq14.gif

12951_2017_255_Article_IEq41.gif

12936_2015_836_Article_IEq13.gif

Fig. 2

12936_2017_2014_Article_IEq29.gif

Fig. 2

12936_2017_2045_Article_IEq18.gif

Fig. 2

Fig. 2

Fig. 1

Fig. 1

Fig. 3

Fig. 2

Fig. 2

12936_2016_1182_Article_IEq39.gif

12951_2017_255_Article_IEq40.gif

Fig. 1

Fig. 3

Fig. 4

Fig. 2

Fig. 6

Fig. 2

Fig. 1

Fig. 1

Fig. 4

Fig. 5

Fig. 6

Fig. 5

Fig. 1

12951_2017_255_Article_IEq39.gif

Fig. 1

Fig. 2

12951_2015_155_Article_IEq62.gif

12951_2017_292_Article_IEq1.gif

Fig. 3

Fig. 1

Fig. 10

Fig. 7

Fig. 5

Fig. 1

Fig. 6

Fig. 2

12960_2017_220_Article_IEq2.gif

Fig. 5

Fig. 1

Fig. 4

Fig. 7

Fig. 4

12951_2015_155_Article_IEq58.gif

Fig. 6