Loading…

Sensitivity analysis for unobserved confounding in causal mediation analysis allowing for effect modification, censoring and truncation

Causal mediation analysis is used to decompose the total effect of an exposure on an outcome into an indirect effect, taking the path through an intermediate variable, and a direct effect. To estimate these effects, strong assumptions are made about unconfoundedness of the relationships between the...

Full description

Saved in:
Bibliographic Details
Published in:Statistical methods & applications 2022-10, Vol.31 (4), p.785-814
Main Author: Lindmark, Anita
Format: Article
Language:English
Subjects:
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by cdi_FETCH-LOGICAL-c434t-e64c60b29709b413acc8d3676e9a9c508fb61a9b8a9f33801dffc0c9cf61b7f3
cites cdi_FETCH-LOGICAL-c434t-e64c60b29709b413acc8d3676e9a9c508fb61a9b8a9f33801dffc0c9cf61b7f3
container_end_page 814
container_issue 4
container_start_page 785
container_title Statistical methods & applications
container_volume 31
creator Lindmark, Anita
description Causal mediation analysis is used to decompose the total effect of an exposure on an outcome into an indirect effect, taking the path through an intermediate variable, and a direct effect. To estimate these effects, strong assumptions are made about unconfoundedness of the relationships between the exposure, mediator and outcome. These assumptions are difficult to verify in a given situation and therefore a mediation analysis should be complemented with a sensitivity analysis to assess the possible impact of violations. In this paper we present a method for sensitivity analysis to not only unobserved mediator-outcome confounding, which has largely been the focus of previous literature, but also unobserved confounding involving the exposure. The setting is estimation of natural direct and indirect effects based on parametric regression models. We present results for combinations of binary and continuous mediators and outcomes and extend the sensitivity analysis for mediator-outcome confounding to cases where the continuous outcome variable is censored or truncated. The proposed methods perform well also in the presence of interactions between the exposure, mediator and observed confounders, allowing for modeling flexibility as well as exploration of effect modification. The performance of the method is illustrated through simulations and an empirical example.
doi_str_mv 10.1007/s10260-021-00611-4
format article
fullrecord <record><control><sourceid>proquest_swepu</sourceid><recordid>TN_cdi_swepub_primary_oai_DiVA_org_umu_190068</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2711890810</sourcerecordid><originalsourceid>FETCH-LOGICAL-c434t-e64c60b29709b413acc8d3676e9a9c508fb61a9b8a9f33801dffc0c9cf61b7f3</originalsourceid><addsrcrecordid>eNp9kcFu3CAURa2qkTpN8gNdIXVbmveMg2E5mjZNpJG6SBRlhzCGiMgDUzCJ5gv627XHUbLr6iG9cy9wb1V9QfiOAO1FRqg5UKiRAnBE2nyoVsiRUSnw4ePxLGh9ifCp-pzzEwBjrGGr6u-tDdmP_tmPB6KDHg7ZZ-JiIiXELtv0bHtiYnCxhN6HR-IDMbpkPZCd7b0efQzvOj0M8WWmZgPrnDUj2cXeO2-O5DdiputimhEdejKmEpbNWXXi9JDt-es8re6uft5trun296-bzXpLTcOakVreGA5dLVuQXYNMGyN6xltupZbmEoTrOGrZCS0dYwKwd86AkcZx7FrHTiu62OYXuy-d2ie_0-mgovbqh79fq5geVdkVhXLKUUz814Xfp_in2Dyqp1jS9N2s6hZRSBAIE1UvlEkx52Tdmy-CmvtRSz9q6kcd-1HNJGKvT9nPedj0bv0f1T9CVpdP</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2711890810</pqid></control><display><type>article</type><title>Sensitivity analysis for unobserved confounding in causal mediation analysis allowing for effect modification, censoring and truncation</title><source>EconLit s plnými texty</source><source>EBSCOhost Business Source Ultimate</source><source>Springer Nature</source><creator>Lindmark, Anita</creator><creatorcontrib>Lindmark, Anita</creatorcontrib><description>Causal mediation analysis is used to decompose the total effect of an exposure on an outcome into an indirect effect, taking the path through an intermediate variable, and a direct effect. To estimate these effects, strong assumptions are made about unconfoundedness of the relationships between the exposure, mediator and outcome. These assumptions are difficult to verify in a given situation and therefore a mediation analysis should be complemented with a sensitivity analysis to assess the possible impact of violations. In this paper we present a method for sensitivity analysis to not only unobserved mediator-outcome confounding, which has largely been the focus of previous literature, but also unobserved confounding involving the exposure. The setting is estimation of natural direct and indirect effects based on parametric regression models. We present results for combinations of binary and continuous mediators and outcomes and extend the sensitivity analysis for mediator-outcome confounding to cases where the continuous outcome variable is censored or truncated. The proposed methods perform well also in the presence of interactions between the exposure, mediator and observed confounders, allowing for modeling flexibility as well as exploration of effect modification. The performance of the method is illustrated through simulations and an empirical example.</description><identifier>ISSN: 1618-2510</identifier><identifier>ISSN: 1613-981X</identifier><identifier>EISSN: 1613-981X</identifier><identifier>DOI: 10.1007/s10260-021-00611-4</identifier><language>eng</language><publisher>Berlin/Heidelberg: Springer Berlin Heidelberg</publisher><subject>Chemistry and Earth Sciences ; Computer Science ; Continuity (mathematics) ; Economics ; Empirical analysis ; Exposure ; Finance ; Health Sciences ; Humanities ; Insurance ; Law ; Management ; Mathematics and Statistics ; Medicine ; Original Paper ; Physics ; Probability and Uncertainty ; Regression models ; Sensitivity analysis ; Statistical Theory and Methods ; Statistics ; Statistics and Probability ; Statistics for Business ; Statistics for Engineering ; Statistics for Life Sciences ; Statistics for Social Sciences ; statistik</subject><ispartof>Statistical methods &amp; applications, 2022-10, Vol.31 (4), p.785-814</ispartof><rights>The Author(s) 2021</rights><rights>The Author(s) 2021. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c434t-e64c60b29709b413acc8d3676e9a9c508fb61a9b8a9f33801dffc0c9cf61b7f3</citedby><cites>FETCH-LOGICAL-c434t-e64c60b29709b413acc8d3676e9a9c508fb61a9b8a9f33801dffc0c9cf61b7f3</cites><orcidid>0000-0002-4600-0060</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>230,314,780,784,885,27924,27925</link.rule.ids><backlink>$$Uhttps://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-190068$$DView record from Swedish Publication Index$$Hfree_for_read</backlink></links><search><creatorcontrib>Lindmark, Anita</creatorcontrib><title>Sensitivity analysis for unobserved confounding in causal mediation analysis allowing for effect modification, censoring and truncation</title><title>Statistical methods &amp; applications</title><addtitle>Stat Methods Appl</addtitle><description>Causal mediation analysis is used to decompose the total effect of an exposure on an outcome into an indirect effect, taking the path through an intermediate variable, and a direct effect. To estimate these effects, strong assumptions are made about unconfoundedness of the relationships between the exposure, mediator and outcome. These assumptions are difficult to verify in a given situation and therefore a mediation analysis should be complemented with a sensitivity analysis to assess the possible impact of violations. In this paper we present a method for sensitivity analysis to not only unobserved mediator-outcome confounding, which has largely been the focus of previous literature, but also unobserved confounding involving the exposure. The setting is estimation of natural direct and indirect effects based on parametric regression models. We present results for combinations of binary and continuous mediators and outcomes and extend the sensitivity analysis for mediator-outcome confounding to cases where the continuous outcome variable is censored or truncated. The proposed methods perform well also in the presence of interactions between the exposure, mediator and observed confounders, allowing for modeling flexibility as well as exploration of effect modification. The performance of the method is illustrated through simulations and an empirical example.</description><subject>Chemistry and Earth Sciences</subject><subject>Computer Science</subject><subject>Continuity (mathematics)</subject><subject>Economics</subject><subject>Empirical analysis</subject><subject>Exposure</subject><subject>Finance</subject><subject>Health Sciences</subject><subject>Humanities</subject><subject>Insurance</subject><subject>Law</subject><subject>Management</subject><subject>Mathematics and Statistics</subject><subject>Medicine</subject><subject>Original Paper</subject><subject>Physics</subject><subject>Probability and Uncertainty</subject><subject>Regression models</subject><subject>Sensitivity analysis</subject><subject>Statistical Theory and Methods</subject><subject>Statistics</subject><subject>Statistics and Probability</subject><subject>Statistics for Business</subject><subject>Statistics for Engineering</subject><subject>Statistics for Life Sciences</subject><subject>Statistics for Social Sciences</subject><subject>statistik</subject><issn>1618-2510</issn><issn>1613-981X</issn><issn>1613-981X</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><recordid>eNp9kcFu3CAURa2qkTpN8gNdIXVbmveMg2E5mjZNpJG6SBRlhzCGiMgDUzCJ5gv627XHUbLr6iG9cy9wb1V9QfiOAO1FRqg5UKiRAnBE2nyoVsiRUSnw4ePxLGh9ifCp-pzzEwBjrGGr6u-tDdmP_tmPB6KDHg7ZZ-JiIiXELtv0bHtiYnCxhN6HR-IDMbpkPZCd7b0efQzvOj0M8WWmZgPrnDUj2cXeO2-O5DdiputimhEdejKmEpbNWXXi9JDt-es8re6uft5trun296-bzXpLTcOakVreGA5dLVuQXYNMGyN6xltupZbmEoTrOGrZCS0dYwKwd86AkcZx7FrHTiu62OYXuy-d2ie_0-mgovbqh79fq5geVdkVhXLKUUz814Xfp_in2Dyqp1jS9N2s6hZRSBAIE1UvlEkx52Tdmy-CmvtRSz9q6kcd-1HNJGKvT9nPedj0bv0f1T9CVpdP</recordid><startdate>20221001</startdate><enddate>20221001</enddate><creator>Lindmark, Anita</creator><general>Springer Berlin Heidelberg</general><general>Springer Nature B.V</general><scope>C6C</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>8FD</scope><scope>H8D</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope><scope>ADHXS</scope><scope>ADTPV</scope><scope>AOWAS</scope><scope>D8T</scope><scope>D93</scope><scope>ZZAVC</scope><orcidid>https://orcid.org/0000-0002-4600-0060</orcidid></search><sort><creationdate>20221001</creationdate><title>Sensitivity analysis for unobserved confounding in causal mediation analysis allowing for effect modification, censoring and truncation</title><author>Lindmark, Anita</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c434t-e64c60b29709b413acc8d3676e9a9c508fb61a9b8a9f33801dffc0c9cf61b7f3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>Chemistry and Earth Sciences</topic><topic>Computer Science</topic><topic>Continuity (mathematics)</topic><topic>Economics</topic><topic>Empirical analysis</topic><topic>Exposure</topic><topic>Finance</topic><topic>Health Sciences</topic><topic>Humanities</topic><topic>Insurance</topic><topic>Law</topic><topic>Management</topic><topic>Mathematics and Statistics</topic><topic>Medicine</topic><topic>Original Paper</topic><topic>Physics</topic><topic>Probability and Uncertainty</topic><topic>Regression models</topic><topic>Sensitivity analysis</topic><topic>Statistical Theory and Methods</topic><topic>Statistics</topic><topic>Statistics and Probability</topic><topic>Statistics for Business</topic><topic>Statistics for Engineering</topic><topic>Statistics for Life Sciences</topic><topic>Statistics for Social Sciences</topic><topic>statistik</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Lindmark, Anita</creatorcontrib><collection>Springer Open Access</collection><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Technology Research Database</collection><collection>Aerospace Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts – Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><collection>SWEPUB Umeå universitet full text</collection><collection>SwePub</collection><collection>SwePub Articles</collection><collection>SWEPUB Freely available online</collection><collection>SWEPUB Umeå universitet</collection><collection>SwePub Articles full text</collection><jtitle>Statistical methods &amp; applications</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Lindmark, Anita</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Sensitivity analysis for unobserved confounding in causal mediation analysis allowing for effect modification, censoring and truncation</atitle><jtitle>Statistical methods &amp; applications</jtitle><stitle>Stat Methods Appl</stitle><date>2022-10-01</date><risdate>2022</risdate><volume>31</volume><issue>4</issue><spage>785</spage><epage>814</epage><pages>785-814</pages><issn>1618-2510</issn><issn>1613-981X</issn><eissn>1613-981X</eissn><abstract>Causal mediation analysis is used to decompose the total effect of an exposure on an outcome into an indirect effect, taking the path through an intermediate variable, and a direct effect. To estimate these effects, strong assumptions are made about unconfoundedness of the relationships between the exposure, mediator and outcome. These assumptions are difficult to verify in a given situation and therefore a mediation analysis should be complemented with a sensitivity analysis to assess the possible impact of violations. In this paper we present a method for sensitivity analysis to not only unobserved mediator-outcome confounding, which has largely been the focus of previous literature, but also unobserved confounding involving the exposure. The setting is estimation of natural direct and indirect effects based on parametric regression models. We present results for combinations of binary and continuous mediators and outcomes and extend the sensitivity analysis for mediator-outcome confounding to cases where the continuous outcome variable is censored or truncated. The proposed methods perform well also in the presence of interactions between the exposure, mediator and observed confounders, allowing for modeling flexibility as well as exploration of effect modification. The performance of the method is illustrated through simulations and an empirical example.</abstract><cop>Berlin/Heidelberg</cop><pub>Springer Berlin Heidelberg</pub><doi>10.1007/s10260-021-00611-4</doi><tpages>30</tpages><orcidid>https://orcid.org/0000-0002-4600-0060</orcidid><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier ISSN: 1618-2510
ispartof Statistical methods & applications, 2022-10, Vol.31 (4), p.785-814
issn 1618-2510
1613-981X
1613-981X
language eng
recordid cdi_swepub_primary_oai_DiVA_org_umu_190068
source EconLit s plnými texty; EBSCOhost Business Source Ultimate; Springer Nature
subjects Chemistry and Earth Sciences
Computer Science
Continuity (mathematics)
Economics
Empirical analysis
Exposure
Finance
Health Sciences
Humanities
Insurance
Law
Management
Mathematics and Statistics
Medicine
Original Paper
Physics
Probability and Uncertainty
Regression models
Sensitivity analysis
Statistical Theory and Methods
Statistics
Statistics and Probability
Statistics for Business
Statistics for Engineering
Statistics for Life Sciences
Statistics for Social Sciences
statistik
title Sensitivity analysis for unobserved confounding in causal mediation analysis allowing for effect modification, censoring and truncation
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-24T17%3A46%3A30IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_swepu&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Sensitivity%20analysis%20for%20unobserved%20confounding%20in%20causal%20mediation%20analysis%20allowing%20for%20effect%20modification,%20censoring%20and%20truncation&rft.jtitle=Statistical%20methods%20&%20applications&rft.au=Lindmark,%20Anita&rft.date=2022-10-01&rft.volume=31&rft.issue=4&rft.spage=785&rft.epage=814&rft.pages=785-814&rft.issn=1618-2510&rft.eissn=1613-981X&rft_id=info:doi/10.1007/s10260-021-00611-4&rft_dat=%3Cproquest_swepu%3E2711890810%3C/proquest_swepu%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-c434t-e64c60b29709b413acc8d3676e9a9c508fb61a9b8a9f33801dffc0c9cf61b7f3%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_pqid=2711890810&rft_id=info:pmid/&rfr_iscdi=true