Cargando…

Antiretroviral Therapy and Mortality in Rural South Africa: A Comparison of Causal Modeling Approaches

Estimation of causal effects from observational data is a primary goal of epidemiology. The use of multiple methods with different assumptions relating to exchangeability improves causal inference by demonstrating robustness across assumptions. We estimated the effect of antiretroviral therapy (ART)...

Descripción completa

Detalles Bibliográficos
Autores principales: Oldenburg, Catherine E, Seage, George R, Tanser, Frank, De Gruttola, Victor, Mayer, Kenneth H, Mimiaga, Matthew J, Bor, Jacob, Bärnighausen, Till
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Oxford University Press 2018
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6070080/
https://www.ncbi.nlm.nih.gov/pubmed/29584868
http://dx.doi.org/10.1093/aje/kwy065
_version_ 1783343618068578304
author Oldenburg, Catherine E
Seage, George R
Tanser, Frank
De Gruttola, Victor
Mayer, Kenneth H
Mimiaga, Matthew J
Bor, Jacob
Bärnighausen, Till
author_facet Oldenburg, Catherine E
Seage, George R
Tanser, Frank
De Gruttola, Victor
Mayer, Kenneth H
Mimiaga, Matthew J
Bor, Jacob
Bärnighausen, Till
author_sort Oldenburg, Catherine E
collection PubMed
description Estimation of causal effects from observational data is a primary goal of epidemiology. The use of multiple methods with different assumptions relating to exchangeability improves causal inference by demonstrating robustness across assumptions. We estimated the effect of antiretroviral therapy (ART) on mortality in rural KwaZulu-Natal, South Africa, from 2007 to 2011, using 2 methods with substantially different assumptions: the regression discontinuity design (RDD) and inverse-probability–weighted (IPW) marginal structural models (MSMs). The RDD analysis took advantage of a CD4-cell-count–based threshold for ART initiation (200 cells/μL). The 2 methods yielded consistent but nonidentical results for the effect of immediate initiation of ART (RDD intention-to-treat hazard ratio (HR) = 0.66, 95% confidence interval (CI): 0.35, 1.26; RDD complier average causal effect HR = 0.56, 95% CI: 0.41, 0.77; IPW MSM HR = 0.49, 95% CI: 0.42, 0.58). Although RDD and IPW MSM estimates have distinct identifying assumptions, strengths, and limitations in terms of internal and external validity, results in this application were similar. The differences in modeling approaches and the external validity of each method may explain the minor differences in effect estimates. The overall consistency of the results lends support for causal inference about the effect of ART on mortality from these data.
format Online
Article
Text
id pubmed-6070080
institution National Center for Biotechnology Information
language English
publishDate 2018
publisher Oxford University Press
record_format MEDLINE/PubMed
spelling pubmed-60700802018-08-09 Antiretroviral Therapy and Mortality in Rural South Africa: A Comparison of Causal Modeling Approaches Oldenburg, Catherine E Seage, George R Tanser, Frank De Gruttola, Victor Mayer, Kenneth H Mimiaga, Matthew J Bor, Jacob Bärnighausen, Till Am J Epidemiol Practice of Epidemiology Estimation of causal effects from observational data is a primary goal of epidemiology. The use of multiple methods with different assumptions relating to exchangeability improves causal inference by demonstrating robustness across assumptions. We estimated the effect of antiretroviral therapy (ART) on mortality in rural KwaZulu-Natal, South Africa, from 2007 to 2011, using 2 methods with substantially different assumptions: the regression discontinuity design (RDD) and inverse-probability–weighted (IPW) marginal structural models (MSMs). The RDD analysis took advantage of a CD4-cell-count–based threshold for ART initiation (200 cells/μL). The 2 methods yielded consistent but nonidentical results for the effect of immediate initiation of ART (RDD intention-to-treat hazard ratio (HR) = 0.66, 95% confidence interval (CI): 0.35, 1.26; RDD complier average causal effect HR = 0.56, 95% CI: 0.41, 0.77; IPW MSM HR = 0.49, 95% CI: 0.42, 0.58). Although RDD and IPW MSM estimates have distinct identifying assumptions, strengths, and limitations in terms of internal and external validity, results in this application were similar. The differences in modeling approaches and the external validity of each method may explain the minor differences in effect estimates. The overall consistency of the results lends support for causal inference about the effect of ART on mortality from these data. Oxford University Press 2018-08 2018-03-23 /pmc/articles/PMC6070080/ /pubmed/29584868 http://dx.doi.org/10.1093/aje/kwy065 Text en © The Author(s) 2018. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. http://creativecommons.org/licenses/by/4.0 This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
spellingShingle Practice of Epidemiology
Oldenburg, Catherine E
Seage, George R
Tanser, Frank
De Gruttola, Victor
Mayer, Kenneth H
Mimiaga, Matthew J
Bor, Jacob
Bärnighausen, Till
Antiretroviral Therapy and Mortality in Rural South Africa: A Comparison of Causal Modeling Approaches
title Antiretroviral Therapy and Mortality in Rural South Africa: A Comparison of Causal Modeling Approaches
title_full Antiretroviral Therapy and Mortality in Rural South Africa: A Comparison of Causal Modeling Approaches
title_fullStr Antiretroviral Therapy and Mortality in Rural South Africa: A Comparison of Causal Modeling Approaches
title_full_unstemmed Antiretroviral Therapy and Mortality in Rural South Africa: A Comparison of Causal Modeling Approaches
title_short Antiretroviral Therapy and Mortality in Rural South Africa: A Comparison of Causal Modeling Approaches
title_sort antiretroviral therapy and mortality in rural south africa: a comparison of causal modeling approaches
topic Practice of Epidemiology
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6070080/
https://www.ncbi.nlm.nih.gov/pubmed/29584868
http://dx.doi.org/10.1093/aje/kwy065
work_keys_str_mv AT oldenburgcatherinee antiretroviraltherapyandmortalityinruralsouthafricaacomparisonofcausalmodelingapproaches
AT seagegeorger antiretroviraltherapyandmortalityinruralsouthafricaacomparisonofcausalmodelingapproaches
AT tanserfrank antiretroviraltherapyandmortalityinruralsouthafricaacomparisonofcausalmodelingapproaches
AT degruttolavictor antiretroviraltherapyandmortalityinruralsouthafricaacomparisonofcausalmodelingapproaches
AT mayerkennethh antiretroviraltherapyandmortalityinruralsouthafricaacomparisonofcausalmodelingapproaches
AT mimiagamatthewj antiretroviraltherapyandmortalityinruralsouthafricaacomparisonofcausalmodelingapproaches
AT borjacob antiretroviraltherapyandmortalityinruralsouthafricaacomparisonofcausalmodelingapproaches
AT barnighausentill antiretroviraltherapyandmortalityinruralsouthafricaacomparisonofcausalmodelingapproaches