2 - (SYM 16) the Salience of Structural Barriers and Behavioral Health to Antiretroviral Therapy Adherence in People Receiving HIV Primary Care in South Africa

South Africa (SA) is home to the largest number of people living with HIV (PLWH) in the world, and both structural and individual level variables affect HIV care engagement. Two of the most frequently identified individual variables that affect adherence to antiretroviral therapy (ART) are depression and substance use, with alcohol being the most used substance in SA. As part of planning relevant cognitive-behavioral interventions, the present study sought to examine the relative association of structural barriers with ART adherence while accounting for individual-level behavioral health concerns, depression and alcohol use, in Khayelitsha, a peri-urban settlement outside Cape Town. PLWH (N=194) from six primary care clinics completed an interviewer-administered psychosocial assessment (in isiXhosa) which included a self-report qualitative rating of past-two-week adherence, the Center for Epidemiologic Studies Depression Scale, the Alcohol Use Disorders Identification Test, and the Structural Barriers to Medication Taking questionnaire (e.g., “I do not take my ART pills because I cannot afford the food I need to eat when I take them”).  In the first model, linear regression was used to measure associations between depression and alcohol use with structural barriers to ART use. In the second model, linear regression was used to measure associations of both individual- and structural-level factors with self-reported ART adherence via sequential stepwise entry. Most participants (83%) were women, 98.5% were Black, and the average age was 41.25 years (SD=9.96). In the first model, depression (b=.06, p< .001) and alcohol use (b=.14, p< .001) together accounted for 22% of the variance (p< .001), and both were significant unique predictors of structural barriers. In step 1 of the second model, depression was a significant predictor of ART adherence (b=-.016, p=.033). In step 2, alcohol use, but not depression, was uniquely significant (-.032, p=.002). In step 3, structural barriers to medication taking were added, and were the only significant unique predictor of adherence (b=-1.58, p< .001), with this step accounting for 19% of the variance (p< .001). Although depression and alcohol use are consistently identified as barriers to adherence, in this sample of PLWH in SA, structural barriers played a unique role over and above those individual concerns. This study points to the need to place behavioral health interventions in the context of structural vulnerabilities in PLWH in SA, and emphasizes the need for multilevel interventions.