The causal effect of opioid substitution treatment on HAART medication refill adherence. (15th May 2015)

• Record Type: Journal Article
• Authors: Nosyk, Bohdan; Min, Jeong E.; Colley, Guillaume; Lima, Viviane D.; Yip, Benita; Milloy, M.-J.S.; Wood, Evan; Montaner, Julio S.G.
• Is Part Of: AIDS. Volume 29:Number 8(2015)
• Publication Date: 2015-05-15
• Subjects: health administrative data -- HAART -- HIV/AIDS -- marginal structural modelling -- opioid dependence -- opioid substitution treatment; AIDS (Disease) -- Periodicals; Acquired Immunodeficiency Syndrome; AIDS (Disease); Periodicals; Periodicals; 616.9792005
• DOI: 10.1097/QAD.0000000000000642
• Languages: English
• ISSNs: 0269-9370

Abstract

Abstract : Background: People who inject drugs (PWID) account for roughly 13% of the prevalent HIV/AIDS population outside of sub-Saharan Africa, and access to opioid substitution treatment (OST) is limited in many settings globally. OST likely facilitates access to HAART, yet sparse evidence is available to support this hypothesis. Our objective was to determine the causal impact of OST exposure on HAART adherence among HIV-positive PWID in a Canadian setting. Methods: We executed a retrospective cohort study using linked population-level data for British Columbia, Canada (January 1996–March 2010). We considered HIV-positive PWID after meeting HAART initiation criteria. A marginal structural model was estimated on a monthly timescale using inverse probability of treatment weights. The primary outcome was 95% HAART adherence, according to pharmacy refill compliance. Exposure to OST was defined as 95% of OST receipt, and we controlled for a range of fixed and time-varying covariates. Results: Our study included 1852 (63.3%) HIV-positive PWID with a median follow-up of 5.5 years; 34% were female and 39% had previously accessed OST. The baseline covariate-adjusted odds of HAART adherence following OST exposure was 1.96 (95% confidence interval: 1.72–2.24), although the adjusted odds estimated within the marginal structural model was 1.68 (1.48–1.92). Findings were robust to sensitivity analyses on model specification. Conclusion: In a setting characterized by universal healthcare and widespread access to both office-based OST and HAART, OST substantially increased the odds of HAART adherence. This underlines the need to address barriers to OST globally to reduce the disease burden of both opioid dependence and HIV/AIDS. Abstract : Supplemental Digital Content is available in the text