Use of directed acyclic graphs (DAGs) to identify confounders in applied health research: review and recommendations.
Directed acyclic graphs
causal diagrams
causal inference
confounding
covariate adjustment
graphical model theory
observational studies
reporting practices
Journal
International journal of epidemiology
ISSN: 1464-3685
Titre abrégé: Int J Epidemiol
Pays: England
ID NLM: 7802871
Informations de publication
Date de publication:
17 05 2021
17 05 2021
Historique:
accepted:
12
10
2020
pubmed:
18
12
2020
medline:
8
7
2021
entrez:
17
12
2020
Statut:
ppublish
Résumé
Directed acyclic graphs (DAGs) are an increasingly popular approach for identifying confounding variables that require conditioning when estimating causal effects. This review examined the use of DAGs in applied health research to inform recommendations for improving their transparency and utility in future research. Original health research articles published during 1999-2017 mentioning 'directed acyclic graphs' (or similar) or citing DAGitty were identified from Scopus, Web of Science, Medline and Embase. Data were extracted on the reporting of: estimands, DAGs and adjustment sets, alongside the characteristics of each article's largest DAG. A total of 234 articles were identified that reported using DAGs. A fifth (n = 48, 21%) reported their target estimand(s) and half (n = 115, 48%) reported the adjustment set(s) implied by their DAG(s). Two-thirds of the articles (n = 144, 62%) made at least one DAG available. DAGs varied in size but averaged 12 nodes [interquartile range (IQR): 9-16, range: 3-28] and 29 arcs (IQR: 19-42, range: 3-99). The median saturation (i.e. percentage of total possible arcs) was 46% (IQR: 31-67, range: 12-100). 37% (n = 53) of the DAGs included unobserved variables, 17% (n = 25) included 'super-nodes' (i.e. nodes containing more than one variable) and 34% (n = 49) were visually arranged so that the constituent arcs flowed in the same direction (e.g. top-to-bottom). There is substantial variation in the use and reporting of DAGs in applied health research. Although this partly reflects their flexibility, it also highlights some potential areas for improvement. This review hence offers several recommendations to improve the reporting and use of DAGs in future research.
Sections du résumé
BACKGROUND
Directed acyclic graphs (DAGs) are an increasingly popular approach for identifying confounding variables that require conditioning when estimating causal effects. This review examined the use of DAGs in applied health research to inform recommendations for improving their transparency and utility in future research.
METHODS
Original health research articles published during 1999-2017 mentioning 'directed acyclic graphs' (or similar) or citing DAGitty were identified from Scopus, Web of Science, Medline and Embase. Data were extracted on the reporting of: estimands, DAGs and adjustment sets, alongside the characteristics of each article's largest DAG.
RESULTS
A total of 234 articles were identified that reported using DAGs. A fifth (n = 48, 21%) reported their target estimand(s) and half (n = 115, 48%) reported the adjustment set(s) implied by their DAG(s). Two-thirds of the articles (n = 144, 62%) made at least one DAG available. DAGs varied in size but averaged 12 nodes [interquartile range (IQR): 9-16, range: 3-28] and 29 arcs (IQR: 19-42, range: 3-99). The median saturation (i.e. percentage of total possible arcs) was 46% (IQR: 31-67, range: 12-100). 37% (n = 53) of the DAGs included unobserved variables, 17% (n = 25) included 'super-nodes' (i.e. nodes containing more than one variable) and 34% (n = 49) were visually arranged so that the constituent arcs flowed in the same direction (e.g. top-to-bottom).
CONCLUSION
There is substantial variation in the use and reporting of DAGs in applied health research. Although this partly reflects their flexibility, it also highlights some potential areas for improvement. This review hence offers several recommendations to improve the reporting and use of DAGs in future research.
Identifiants
pubmed: 33330936
pii: 6012812
doi: 10.1093/ije/dyaa213
pmc: PMC8128477
doi:
Types de publication
Journal Article
Research Support, Non-U.S. Gov't
Review
Langues
eng
Sous-ensembles de citation
IM
Pagination
620-632Subventions
Organisme : Medical Research Council
ID : MR/K501402/1
Pays : United Kingdom
Informations de copyright
© The Author(s) 2020. Published by Oxford University Press on behalf of the International Epidemiological Association.
Références
Eur J Epidemiol. 2019 Mar;34(3):211-219
pubmed: 30840181
Environ Health. 2017 Oct 23;16(1):110
pubmed: 29078795
Epidemiology. 1999 Jan;10(1):37-48
pubmed: 9888278
Obesity (Silver Spring). 2017 Aug;25(8):1451-1459
pubmed: 28660652
Epidemiology. 2011 Sep;22(5):745
pubmed: 21811114
Int J Epidemiol. 2020 Feb 1;49(1):322-329
pubmed: 31325312
Int J Epidemiol. 2016 Dec 1;45(6):1887-1894
pubmed: 28089956
Am J Public Health. 2018 May;108(5):616-619
pubmed: 29565659
Ann Stat. 2013 Feb;41(1):196-220
pubmed: 25544784
Environ Res. 2013 Oct;126:91-7
pubmed: 23890968
Int J Epidemiol. 2016 Dec 1;45(6):1787-1808
pubmed: 27694566
Am J Epidemiol. 2013 Feb 15;177(4):292-8
pubmed: 23371353
Epidemiology. 2010 Jan;21(1):159
pubmed: 20010223
Arthritis Rheumatol. 2016 May;68(5):1156-64
pubmed: 26714126
BJOG. 2019 Jul;126(8):973-982
pubmed: 30891907
Int J Epidemiol. 2021 Jan 23;49(6):2058-2064
pubmed: 31298274
PLoS One. 2015 Jun 04;10(6):e0128933
pubmed: 26042802
Int J Epidemiol. 2021 Jan 23;49(6):2074-2082
pubmed: 32380551
BMC Med Res Methodol. 2008 Oct 30;8:70
pubmed: 18973665
Int J Colorectal Dis. 2016 Mar;31(3):579-85
pubmed: 26670673
Environ Health. 2017 Nov 21;16(1):127
pubmed: 29162109
Int J Epidemiol. 2016 Dec 1;45(6):1776-1786
pubmed: 26800751
Biom J. 2018 May;60(3):431-449
pubmed: 29292533
J Acquir Immune Defic Syndr. 2016 Jan 1;71(1):87-93
pubmed: 26334742
Soc Sci Med. 2018 Aug;210:2-21
pubmed: 29331519
Epidemiology. 2010 Jul;21(4):586-7
pubmed: 20539116