Trust and the acceptance of and adherence to antiretroviral therapy

Background: Antiretroviral therapy (ART) has resulted in reduced AIDS incid ence and mortality. Socially marginalized individuals with HIV infection, p articularly injection drug users (IDUs), have received less ART and derived less benefit than others. Little is known about the therapeutic process ne cessary to promote acceptance of and adherence to ART among marginalized HI V-infected populations. We report on the correlates of both acceptance of a nd adherence to ART among HIV infected prisoners, most of whom are IDUs. Design: Using a cross-sectional survey design within four ambulatory prison HIV clinics, 205 HIV-infected prisoners eligible for ART were recruited be tween March and October 1996. Measurements: Detailed interviews were conducted that included personal cha racteristics, health status and beliefs, and validated standardized scales measuring depression, health locus of control, social desirability and trus t in physician, medical institutions and society. Acceptance and adherence were documented by self-report and validated for a subset by pharmacy revie w. Clinical information was obtained from standardized chart review. Adhere nce was defined as having taken greater than or equal to 80% of ART. Results: The acceptance of (80%) and adherence to (84%) ART among this grou p of prisoners was high. Multiple regression models demonstrated that corre lates of acceptance of and adherence to ART differed. Acceptance was associ ated with trust in physician (8% increase for each unit increase with trust in physician scale) and trust in HIV medications (threefold reduction for those mistrustful of medication). Side effects (OR = 0.09), social isolatio n (OR = 0.08), and complexity of the antiretroviral regimen (OR = 0.33) wer e associated with decreased adherence. The prevalence of health beliefs sug gesting an adverse relationship between ART and drugs of abuse was high (ra nge 59 to 77%). Adherence did not differ among those receiving directly obs erved therapy (82%) or self-administration (85%). Conclusions: ART can be successfully administered within a correctional set ting. Trust and the therapeutic relationship between patient and physician remain central in the ART initiation process. Characteristics of the therap eutic agents and the degree of social isolation predict adherence. These re sults may inform the design of interventions to improve both acceptance of and adherence to ART particularly among marginalized populations who have n ot derived full benefit from these potent new therapies.