Background:  Adherence to medication is critical to the success of ART. A number of adherence interventions have been employed, but to date there is little evidence supporting specific interventions to improve adherence to ART. This pilot study, ACTG 731, was conducted to determine whether an intervention delivered by telephone would improve adherence outcomes of persons starting ART.

Methods:  The randomized clinical trial was conducted with ART-naïve subjects (n = 109) enrolling in ACTG 384 at 5 U.S. sites. Subjects (85% male, 51% white, median HIV RNA of 5.0 log₁₀) who consented to substudy 731 were randomly assigned to receive standard care or standard care plus 12 structured telephone calls. The calls, delivered over the first 16 weeks of ART by a nurse at a central site (in accordance with HIPPA), were structured to address common barriers to ART adherence and recommend self-management strategies. Outcome measures were collected over 64 weeks and included an ACTG adherence questionnaire, Medication Event Monitoring System (MEMS), and 384 study end-points. Data were analyzed using descriptive, mixed model for repeated measures, Kaplan-Meier, and Cox PH regression techniques.

Results:  The rate of self-reported adherence was high in both treatment groups (98%, mean weeks 4 to 64) with >64% reporting perfect adherence. Even with a sizeable ceiling effect, a significantly better overall treatment effect was observed in the telephone group (p = 0.023). In a post-hoc analysis, the difference in overall treatment effect was strengthened (p <0.001) when the comparison was limited to subjects reporting <100% early adherence (mean week 64 = 96% [telephone]/91% (standard of care)]. Self-reported adherence was significantly associated with adherence as measured with MEMS. Comparing time to primary regimen failure, the KM survival curve for the telephone group remained above the standard of care across weeks 20 to 64; a Cox PH model that controlled for baseline RNA stratification, baseline CD4, gender, age, race/ethnicity, and randomized ART treatment arm, showed that telephone calls were associated with a lower relative hazard (HR = 0.68; 95%CI 0.38 to 1.23) for regimen failure, but the difference was not significant (p = 0.21).

Conclusions:  Findings indicate that proactive telephone calls delivered from a central site, unaffiliated with subjects’ trial sites, can improve adherence. The treatment effects appear durable, however, the intervention must be tested in a larger population with greater variance in rates of adherence to fully establish clinical benefits.