MEDICATION ADHERENCE IN HIV-INFECTED ADULTS IN THE CURRENT ERA OF HIGHLY ACTIVE ANTIRETROVIRAL THERAPY (HAART)

Background Over the past decade, new and improved formulations of highly active antiretroviral therapy (HAART) have been introduced, which are easier to administer and may not necessitate high levels of adherence for HIV RNA suppression. At the same time, there has been an increase in the use of concomitant medications to treat chronic non-AIDS conditions (i.e., diabetes, hypertension) in HIV-infected persons. The goals of the dissertation were 1). to estimate the minimum cutoff of adherence to newer HAART needed for population HIV RNA suppression, 2). to determine whether this cutoff differed by specific regimen type, and 3). to determine if the increase in pill burden due to concomitant medication use impacted adherence to HAART. Methods We used data from three longitudinal cohort studies: the Multicenter AIDS Cohort Study (MACS), the AIDS Linked to Intravenous Experience (ALIVE), and the Veterans Aging Cohort Study Virtual Cohort (VACS) between 2001 and 2011, and analyzed them separately for this dissertation. Adherence was calculated from self-reported use in the MACS and ALIVE, and using pharmacy refill records in the VACS. In all three cohorts, the minimum needed adherence cutoff was defined as the level at which the odds of suppression was not significantly different than that observed with ≥95% adherence using repeated measures logistic regression. We controlled for confounding by indication using propensity score weighting. The effect of the number of concomitant medications on the minimum optimal adherence cutoff to HAART was also analyzed longitudinally using repeated measures logistic regression models, and we further determined if this association varied by the pharmacologic class of the concomitant medication. Results In all three cohorts, there was an increase in the proportion with ≥95% adherence, and the proportion suppressed over time. Study 1 consisted of 1,006 HAART users with 10,971 person-visits in the MACS, and 197 HAART users with 1,745 person-visits in the ALIVE. In the MACS, levels of adherence between 80-84% were sufficient for HIV RNA suppression (OR (ref ≥95%): 1.43(0.61, 3.33)). In the ALIVE, we did not observe a minimum adherence cutoff below 95%. Study 2 consisted of 21,865 HAART users who contributed 82,217 person-years of follow-up. Suppression with <95% adherence was less likely (p