Effect of highly active antiretroviral therapy on time to acquired immunodeficiency syndrome or death using marginal structural models

Stephen R. Cole, Miguel A. Hernán, James M. Robins, Kathryn Anastos, Joan Chmiel, Roger Detels, Carolyn Ervin, Joseph Feldman, Ruth Greenblatt, Lawrence Kingsley, Shenghan Lai, Mary Young, Mardge Cohen, Alvaro Muñoz

Research output: Contribution to journalArticle

204 Citations (Scopus)

Abstract

To estimate the net (i.e., overall) effect of highly active antiretroviral therapy (HAART) on time to acquired immunodeficiency syndrome (AIDS) or death, the authors used inverse probability-of-treatment weighted estimation of a marginal structural model, which can appropriately adjust for time-varying confounders affected by prior treatment or exposure. Human immunodeficiency virus (HIV)-positive men and women (n = 1,498) were followed in two ongoing cohort studies between 1995 and 2002. Sixty-one percent (n = 918) of the participants initiated HAART during 6,763 person-years of follow-up, and 382 developed AIDS or died. Strong confounding by indication for HAART was apparent; the unadjusted hazard ratio for AIDS or death was 0.98. The hazard ratio from a standard time-dependent Cox model that included time-varying CD4 cell count, HIV RNA level, and other time-varying and fixed covariates as regressors was 0.81 (95% confidence interval: 0.61, 1.07). In contrast, the hazard ratio from a marginal structural survival model was 0.54 (robust 95% confidence interval: 0.38, 0.78), suggesting a clinically meaningful net benefit of HAART. Standard Cox analysis failed to detect a clear net benefit, because it does not appropriately adjust for time-dependent covariates, such as HIV RNA level and CD4 cell count, that are simultaneously confounders and intermediate variables.

Original languageEnglish (US)
Pages (from-to)687-694
Number of pages8
JournalAmerican Journal of Epidemiology
Volume158
Issue number7
DOIs
StatePublished - Oct 1 2003
Externally publishedYes

Fingerprint

Structural Models
Highly Active Antiretroviral Therapy
Acquired Immunodeficiency Syndrome
HIV
CD4 Lymphocyte Count
RNA
Confidence Intervals
Proportional Hazards Models
Cohort Studies
Survival
Therapeutics

Keywords

  • Acquired immunodeficiency syndrome
  • Antiretroviral therapy
  • Causality
  • Confounding factors (epidemiology)
  • Highly active

ASJC Scopus subject areas

  • Epidemiology

Cite this

Effect of highly active antiretroviral therapy on time to acquired immunodeficiency syndrome or death using marginal structural models. / Cole, Stephen R.; Hernán, Miguel A.; Robins, James M.; Anastos, Kathryn; Chmiel, Joan; Detels, Roger; Ervin, Carolyn; Feldman, Joseph; Greenblatt, Ruth; Kingsley, Lawrence; Lai, Shenghan; Young, Mary; Cohen, Mardge; Muñoz, Alvaro.

In: American Journal of Epidemiology, Vol. 158, No. 7, 01.10.2003, p. 687-694.

Research output: Contribution to journalArticle

Cole, SR, Hernán, MA, Robins, JM, Anastos, K, Chmiel, J, Detels, R, Ervin, C, Feldman, J, Greenblatt, R, Kingsley, L, Lai, S, Young, M, Cohen, M & Muñoz, A 2003, 'Effect of highly active antiretroviral therapy on time to acquired immunodeficiency syndrome or death using marginal structural models', American Journal of Epidemiology, vol. 158, no. 7, pp. 687-694. https://doi.org/10.1093/aje/kwg206
Cole, Stephen R. ; Hernán, Miguel A. ; Robins, James M. ; Anastos, Kathryn ; Chmiel, Joan ; Detels, Roger ; Ervin, Carolyn ; Feldman, Joseph ; Greenblatt, Ruth ; Kingsley, Lawrence ; Lai, Shenghan ; Young, Mary ; Cohen, Mardge ; Muñoz, Alvaro. / Effect of highly active antiretroviral therapy on time to acquired immunodeficiency syndrome or death using marginal structural models. In: American Journal of Epidemiology. 2003 ; Vol. 158, No. 7. pp. 687-694.
@article{908c65b4e6e849b5b89ea4ef4794407c,
title = "Effect of highly active antiretroviral therapy on time to acquired immunodeficiency syndrome or death using marginal structural models",
abstract = "To estimate the net (i.e., overall) effect of highly active antiretroviral therapy (HAART) on time to acquired immunodeficiency syndrome (AIDS) or death, the authors used inverse probability-of-treatment weighted estimation of a marginal structural model, which can appropriately adjust for time-varying confounders affected by prior treatment or exposure. Human immunodeficiency virus (HIV)-positive men and women (n = 1,498) were followed in two ongoing cohort studies between 1995 and 2002. Sixty-one percent (n = 918) of the participants initiated HAART during 6,763 person-years of follow-up, and 382 developed AIDS or died. Strong confounding by indication for HAART was apparent; the unadjusted hazard ratio for AIDS or death was 0.98. The hazard ratio from a standard time-dependent Cox model that included time-varying CD4 cell count, HIV RNA level, and other time-varying and fixed covariates as regressors was 0.81 (95{\%} confidence interval: 0.61, 1.07). In contrast, the hazard ratio from a marginal structural survival model was 0.54 (robust 95{\%} confidence interval: 0.38, 0.78), suggesting a clinically meaningful net benefit of HAART. Standard Cox analysis failed to detect a clear net benefit, because it does not appropriately adjust for time-dependent covariates, such as HIV RNA level and CD4 cell count, that are simultaneously confounders and intermediate variables.",
keywords = "Acquired immunodeficiency syndrome, Antiretroviral therapy, Causality, Confounding factors (epidemiology), Highly active",
author = "Cole, {Stephen R.} and Hern{\'a}n, {Miguel A.} and Robins, {James M.} and Kathryn Anastos and Joan Chmiel and Roger Detels and Carolyn Ervin and Joseph Feldman and Ruth Greenblatt and Lawrence Kingsley and Shenghan Lai and Mary Young and Mardge Cohen and Alvaro Mu{\~n}oz",
year = "2003",
month = "10",
day = "1",
doi = "10.1093/aje/kwg206",
language = "English (US)",
volume = "158",
pages = "687--694",
journal = "American Journal of Epidemiology",
issn = "0002-9262",
publisher = "Oxford University Press",
number = "7",

}

TY - JOUR

T1 - Effect of highly active antiretroviral therapy on time to acquired immunodeficiency syndrome or death using marginal structural models

AU - Cole, Stephen R.

AU - Hernán, Miguel A.

AU - Robins, James M.

AU - Anastos, Kathryn

AU - Chmiel, Joan

AU - Detels, Roger

AU - Ervin, Carolyn

AU - Feldman, Joseph

AU - Greenblatt, Ruth

AU - Kingsley, Lawrence

AU - Lai, Shenghan

AU - Young, Mary

AU - Cohen, Mardge

AU - Muñoz, Alvaro

PY - 2003/10/1

Y1 - 2003/10/1

N2 - To estimate the net (i.e., overall) effect of highly active antiretroviral therapy (HAART) on time to acquired immunodeficiency syndrome (AIDS) or death, the authors used inverse probability-of-treatment weighted estimation of a marginal structural model, which can appropriately adjust for time-varying confounders affected by prior treatment or exposure. Human immunodeficiency virus (HIV)-positive men and women (n = 1,498) were followed in two ongoing cohort studies between 1995 and 2002. Sixty-one percent (n = 918) of the participants initiated HAART during 6,763 person-years of follow-up, and 382 developed AIDS or died. Strong confounding by indication for HAART was apparent; the unadjusted hazard ratio for AIDS or death was 0.98. The hazard ratio from a standard time-dependent Cox model that included time-varying CD4 cell count, HIV RNA level, and other time-varying and fixed covariates as regressors was 0.81 (95% confidence interval: 0.61, 1.07). In contrast, the hazard ratio from a marginal structural survival model was 0.54 (robust 95% confidence interval: 0.38, 0.78), suggesting a clinically meaningful net benefit of HAART. Standard Cox analysis failed to detect a clear net benefit, because it does not appropriately adjust for time-dependent covariates, such as HIV RNA level and CD4 cell count, that are simultaneously confounders and intermediate variables.

AB - To estimate the net (i.e., overall) effect of highly active antiretroviral therapy (HAART) on time to acquired immunodeficiency syndrome (AIDS) or death, the authors used inverse probability-of-treatment weighted estimation of a marginal structural model, which can appropriately adjust for time-varying confounders affected by prior treatment or exposure. Human immunodeficiency virus (HIV)-positive men and women (n = 1,498) were followed in two ongoing cohort studies between 1995 and 2002. Sixty-one percent (n = 918) of the participants initiated HAART during 6,763 person-years of follow-up, and 382 developed AIDS or died. Strong confounding by indication for HAART was apparent; the unadjusted hazard ratio for AIDS or death was 0.98. The hazard ratio from a standard time-dependent Cox model that included time-varying CD4 cell count, HIV RNA level, and other time-varying and fixed covariates as regressors was 0.81 (95% confidence interval: 0.61, 1.07). In contrast, the hazard ratio from a marginal structural survival model was 0.54 (robust 95% confidence interval: 0.38, 0.78), suggesting a clinically meaningful net benefit of HAART. Standard Cox analysis failed to detect a clear net benefit, because it does not appropriately adjust for time-dependent covariates, such as HIV RNA level and CD4 cell count, that are simultaneously confounders and intermediate variables.

KW - Acquired immunodeficiency syndrome

KW - Antiretroviral therapy

KW - Causality

KW - Confounding factors (epidemiology)

KW - Highly active

UR - http://www.scopus.com/inward/record.url?scp=0141620176&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0141620176&partnerID=8YFLogxK

U2 - 10.1093/aje/kwg206

DO - 10.1093/aje/kwg206

M3 - Article

C2 - 14507605

AN - SCOPUS:0141620176

VL - 158

SP - 687

EP - 694

JO - American Journal of Epidemiology

JF - American Journal of Epidemiology

SN - 0002-9262

IS - 7

ER -