See the published version of this article at BMC Medical Research Methodology.
This is a preprint, a preliminary version of a manuscript that has not completed peer review at a journal. Research Square does not conduct peer review prior to posting preprints. The posting of a preprint on this server should not be interpreted as an endorsement of its validity or suitability for dissemination as established information or for guiding clinical practice.
Learn more about In Review
Research article
Comparison of Methods for Handling Covariate Missingness in Propensity Score Estimation with a Binary Exposure
Donna Coffman
Temple University
Corresponding Author
ORCiD: https://orcid.org/0000-0001-6305-6579
License:
This work is licensed under a CC BY 4.0 License. Read Full License
View the published version at BMC Medical Research Methodology.
Background: Causal effect estimation with observational data is subject to bias due to confounding, which is often controlled for using propensity scores. One unresolved issue in propensity score estimation is how to handle missing values in covariates.
Method: Several approaches have been proposed for handling covariate missingness, including multiple imputation (MI), multiple imputation with missingness pattern (MIMP), and treatment mean imputation. However, there are other potentially useful approaches that have not been evaluated, including single imputation (SI) + prediction error (PE), SI+PE + parameter uncertainty (PU), and Generalized Boosted Modeling (GBM), which is a nonparametric approach for estimating propensity scores in which missing values are automatically handled in the estimation using a surrogate split method. To evaluate the performance of these approaches, a simulation study was conducted.
Results: Results suggested that SI+PE, SI+PE+PU, MI, and MIMP perform almost equally well and better than treatment mean imputation and GBM in terms of bias; however, MI and MIMP account for the additional uncertainty of imputing the missingness.
Conclusions: Applying GBM to the incomplete data and relying on the surrogate split approach resulted in substantial bias. Imputation prior to implementing GBM is recommended.
Keywords
propensity scores, missing data, causal inference, generalized boosted models
Figure 1
Due to technical limitations, full-text HTML conversion of this manuscript could not be completed. However, the manuscript can be downloaded and accessed as a PDF.
This is a list of supplementary files associated with this preprint. Click to download.
Badges
Prescreen
This manuscript passed our Prescreen assessment
Complete author information
Appropriate declaration statements
Potential risks to human health
Prescreen
Peer Review Timeline
CURRENT STATUS: Published
Version 2
Posted 12 Jun, 2020
No comments provided
Editorial decision: Major revision
On 09 Feb, 2020
Review #3 received
Received 07 Feb, 2020
Review #1 received
Received 10 Jan, 2020
Review #2 received
Received 01 Jan, 2020
Reviewer #3 agreed
On 30 Dec, 2019
Reviewer #2 agreed
On 29 Dec, 2019
Reviewers invited
Invitations sent on 13 Dec, 2019
Reviewer #1 agreed
On 13 Dec, 2019
Editor assigned
On 18 Nov, 2019
First submitted
On 17 Nov, 2019
Submission checks complete
On 17 Nov, 2019
Editor invited
On 17 Nov, 2019
Metrics
Comments: 0
PDF Downloads: ...
HTML Views: ...
Subject Areas
Health Economics & Outcomes Research
Learn more about our company.
This preprint is under consideration at BMC Medical Research Methodology. A preprint is a preliminary version of a manuscript that has not completed peer review at a journal. Research Square does not conduct peer review prior to posting preprints. The posting of a preprint on this server should not be interpreted as an endorsement of its validity or suitability for dissemination as established information or for guiding clinical practice.
Learn more about In Review
Research article
Comparison of Methods for Handling Covariate Missingness in Propensity Score Estimation with a Binary Exposure
Donna Coffman
Temple University
Corresponding Author
ORCiD: https://orcid.org/0000-0001-6305-6579
Badges
Prescreen
This manuscript passed our Prescreen assessment
Complete author information
Appropriate declaration statements
Potential risks to human health
Prescreen
Peer Review Timeline
CURRENT STATUS: Published
Version 2
Posted 12 Jun, 2020
No comments provided
Editorial decision: Major revision
On 09 Feb, 2020
Review #3 received
Received 07 Feb, 2020
Review #1 received
Received 10 Jan, 2020
Review #2 received
Received 01 Jan, 2020
Reviewer #3 agreed
On 30 Dec, 2019
Reviewer #2 agreed
On 29 Dec, 2019
Reviewers invited
Invitations sent on 13 Dec, 2019
Reviewer #1 agreed
On 13 Dec, 2019
Editor assigned
On 18 Nov, 2019
First submitted
On 17 Nov, 2019
Submission checks complete
On 17 Nov, 2019
Editor invited
On 17 Nov, 2019
Metrics
Comments: 0
PDF Downloads: ...
HTML Views: ...
Subject Areas
Health Economics & Outcomes Research
License:
This work is licensed under a CC BY 4.0 License. Read Full License
View the published version at BMC Medical Research Methodology.
Background: Causal effect estimation with observational data is subject to bias due to confounding, which is often controlled for using propensity scores. One unresolved issue in propensity score estimation is how to handle missing values in covariates.
Method: Several approaches have been proposed for handling covariate missingness, including multiple imputation (MI), multiple imputation with missingness pattern (MIMP), and treatment mean imputation. However, there are other potentially useful approaches that have not been evaluated, including single imputation (SI) + prediction error (PE), SI+PE + parameter uncertainty (PU), and Generalized Boosted Modeling (GBM), which is a nonparametric approach for estimating propensity scores in which missing values are automatically handled in the estimation using a surrogate split method. To evaluate the performance of these approaches, a simulation study was conducted.
Results: Results suggested that SI+PE, SI+PE+PU, MI, and MIMP perform almost equally well and better than treatment mean imputation and GBM in terms of bias; however, MI and MIMP account for the additional uncertainty of imputing the missingness.
Conclusions: Applying GBM to the incomplete data and relying on the surrogate split approach resulted in substantial bias. Imputation prior to implementing GBM is recommended.
Figure 1
Due to technical limitations, full-text HTML conversion of this manuscript could not be completed. However, the manuscript can be downloaded and accessed as a PDF.