-
Tytuł:
-
Common Methods for Handling Missing Data in Marginal Structural Models: What Works and Why.
-
Autorzy:
-
Leyrat C
Carpenter JR
Bailly S
Williamson EJ
-
Źródło:
-
American journal of epidemiology [Am J Epidemiol] 2021 Apr 06; Vol. 190 (4), pp. 663-672.
-
Typ publikacji:
-
Journal Article; Research Support, Non-U.S. Gov't; Review
-
Język:
-
English
-
Imprint Name(s):
-
Publication: Cary, NC : Oxford University Press
Original Publication: Baltimore, School of Hygiene and Public Health of Johns Hopkins Univ.
-
MeSH Terms:
-
Computer Simulation*
Models, Statistical*
Electronic Health Records/*statistics & numerical data
Data Interpretation, Statistical ; Humans
-
References:
-
Am J Epidemiol. 2009 Sep 15;170(6):687-94. (PMID: 19675141)
Biom J. 2015 Mar;57(2):254-70. (PMID: 25352223)
Stat Med. 2013 Apr 30;32(9):1584-618. (PMID: 23208861)
Am J Epidemiol. 2008 Sep 15;168(6):656-64. (PMID: 18682488)
BMJ. 2009 Jun 29;338:b2393. (PMID: 19564179)
J Clin Epidemiol. 2008 Apr;61(4):344-9. (PMID: 18313558)
Biometrics. 2012 Mar;68(1):129-37. (PMID: 22050039)
Sleep. 1991 Dec;14(6):540-5. (PMID: 1798888)
Stat Med. 2020 May 20;39(11):1641-1657. (PMID: 32103533)
Biostatistics. 2004 Jul;5(3):445-64. (PMID: 15208205)
Stat Med. 2009 Dec 20;28(29):3657-69. (PMID: 19757484)
Stat Med. 2010 Feb 20;29(4):431-43. (PMID: 20025082)
Clin Trials. 2007;4(2):125-39. (PMID: 17456512)
PLoS One. 2016 Jun 17;11(6):e0157318. (PMID: 27314230)
Med Care. 2019 Mar;57(3):237-243. (PMID: 30664611)
Stat Med. 2007 Jan 15;26(1):20-36. (PMID: 17072897)
Int J Biostat. 2008;4(1):Article 13. (PMID: 22462119)
J Am Stat Assoc. 2018;113(521):369-379. (PMID: 30034062)
Stat Methods Med Res. 2019 Jan;28(1):3-19. (PMID: 28573919)
Stat Methods Med Res. 2006 Jun;15(3):213-34. (PMID: 16768297)
Int J Epidemiol. 2019 Feb 1;48(1):254-265. (PMID: 30358847)
Am J Epidemiol. 2015 Dec 15;182(12):1047-55. (PMID: 26589708)
BMC Med Res Methodol. 2012 Jul 11;12:96. (PMID: 22784200)
Stat Med. 2009 Apr 30;28(9):1402-14. (PMID: 19222021)
Epidemiology. 2000 Sep;11(5):550-60. (PMID: 10955408)
-
Grant Information:
-
MC_UU_12023/21 United Kingdom MRC_ Medical Research Council; MR/M013278/1 United Kingdom MRC_ Medical Research Council; MR/S01442X/1 United Kingdom MRC_ Medical Research Council; MR/T032448/1 United Kingdom MRC_ Medical Research Council
-
Contributed Indexing:
-
Keywords: complete cases; inverse probability weighting; last observation carried forward; missingness pattern approach; multiple imputation; propensity score; time-varying confounding
-
Entry Date(s):
-
Date Created: 20201015 Date Completed: 20210420 Latest Revision: 20230222
-
Update Code:
-
20240105
-
PubMed Central ID:
-
PMC8631064
-
DOI:
-
10.1093/aje/kwaa225
-
PMID:
-
33057574
-
Marginal structural models (MSMs) are commonly used to estimate causal intervention effects in longitudinal nonrandomized studies. A common challenge when using MSMs to analyze observational studies is incomplete confounder data, where a poorly informed analysis method will lead to biased estimates of intervention effects. Despite a number of approaches described in the literature for handling missing data in MSMs, there is little guidance on what works in practice and why. We reviewed existing missing-data methods for MSMs and discussed the plausibility of their underlying assumptions. We also performed realistic simulations to quantify the bias of 5 methods used in practice: complete-case analysis, last observation carried forward, the missingness pattern approach, multiple imputation, and inverse-probability-of-missingness weighting. We considered 3 mechanisms for nonmonotone missing data encountered in research based on electronic health record data. Further illustration of the strengths and limitations of these analysis methods is provided through an application using a cohort of persons with sleep apnea: the research database of the French Observatoire Sommeil de la Fédération de Pneumologie. We recommend careful consideration of 1) the reasons for missingness, 2) whether missingness modifies the existing relationships among observed data, and 3) the scientific context and data source, to inform the choice of the appropriate method(s) for handling partially observed confounders in MSMs.
(© The Author(s) 2020. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health.)