Using intent-to-treat comparisons, it has been shown that the integration of antiretroviral therapy (ART) and tuberculosis (TB) treatment improves survival. Because the magnitude of the effect of ART initiation during TB treatment on mortality is less well understood owing to noncompliance, we used instrumental variables (IV) analyses.
We studied 642 HIV-TB co-infected patients from the Starting Antiretroviral Therapy at Three Points in Tuberculosis trial. Patients were assigned to start ART either early or late during TB treatment or after TB treatment completion. We used 2-stage predictor substitution and 2-stage residuals inclusion methods under additive and proportional hazards regressions with a time-fixed measure of compliance defined as the fraction of time on ART during TB treatment. We moreover developed novel IV methods for additive hazards regression with a time-varying measure of compliance.
Intent-to-treat results from additive hazards models showed that patients in the early integrated arms had a reduced hazard of -0.05 (95% confidence interval [CI]: -0.09, -0.01) when compared with the sequential arm. Adjustment for noncompliance changed this effect to -0.07 (95% CI: -0.12, -0.01). An additional time-varying IV analysis on the overall effect of ART exposure suggested an effect of -0.29 (95 % CI: -0.54, -0.03).
IV analyses enable assessment of the effectiveness of TB and ART integration, corrected for noncompliance, and thereby enable a better public health evaluation of the potential impact of this intervention.
From the aCentre for the AIDS Programme of Research in South Africa (CAPRISA), University of KwaZulu-Natal, Durban, South Africa
bMRC-CAPRISA HIV-TB Pathogenesis and Treatment Research Unit, Durban, South Africa
cThe South African DST/NRF Centre of Excellence in Epidemiological Modelling and Analysis (SACEMA), University of Stellenbosch, Stellenbosch, South Africa
dSchool of Mathematics, Statistics and Computer Science, University of KwaZulu-Natal, Pietermaritzburg, South Africa
eDepartment of Applied Mathematics, Computer Science and Statistics, Ghent University, Gent, Belgium
fMedical Statistics Unit, the London School of Hygiene and Tropical Medicine, London, United Kingdom.
Submitted September 10, 2016; accepted September 17, 2018.
Nonhlanhla Yende-Zuma is supported by CAPRISA, which was established as part of the Comprehensive International Program of Research on AIDS (CIPRA) (grant # AI51794) from the US National Institutes of Health. The U.S. President’s Emergency Plan for AIDS Relief (PEPfAR) funded the care of all the participants in the trial. The Global Fund to fight AIDS, Tuberculosis and Malaria funded the cost of the drugs used in the trial. The research infrastructure to conduct this trial, including the data management, laboratory and pharmacy cores were established through the CIPRA grant. Nonhlanhla Yende-Zuma is also supported by the South African Medical Research Council (SAMRC). This publication was made possible by grant number 5R24TW008863 from the Office of Global AIDS Coordinator and the U. S. Department of Health and Human Services, National Institutes of Health (NIH OAR and NIH OWAR). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the government. The support of the DST-NRF Centre of Excellence in Epidemiological Modelling and Analysis (SACEMA), University of Stellenbosch toward this research is hereby acknowledged. Opinions expressed and conclusions arrived at, are those of the author and are not necessarily to be attributed to SACEMA.
The authors report no conflicts of interest.
Availability of data and code: Dataset and code will be made available upon request but access will be controlled and each request will be considered on a case by case basis.
Supplemental digital content is available through direct URL citations in the HTML and PDF versions of this article (www.epidem.com).
Correspondence: Nonhlanhla Yende-Zuma, Centre for the AIDS Programme of Research in South Africa (CAPRISA), 2nd floor K-RITH tower, University of KwaZulu-Natal, 719 Umbilo Road, Congella, Durban, South Africa, 4013. E-mail: email@example.com.