AL Gifford, VA San Diego Health Care System and VA Greater Los Angeles HSR&D Center for Excellence; JE Bormann, VA San Diego Health Care System and VA Greater Los Angeles HSR&D Center for Excellence; MJ Shively, VA San Diego Health Care System and VA Greater Los Angeles HSR&D Center for Excellence; M Lee, VA Greater Los Angeles HSR&D Center for Excellence; DD Richman, VA San Diego Health Care System; SA Bozzette, VA San Diego Health Care System and VA Greater Los Angeles HSR&D Center for Excellence

Objectives: HIV patients who do not adhere to their multi-drug antiretroviral regimens (ARV) risk high plasma HIV levels, ARV resistance, inability to suppress HIV even with potent regimens, disease progression, and transmission of resistant HIV to others. We developed a program to educate and support HIV patients and evaluated program effects on ARV adherence.

Methods: HIV-infected adults (N=209) at VA and non-VA care centers were randomized in 40/30/30 proportion to Self-Management Education and Support (SME), Social Support control (SS), or Printed Materials control (PM). Patients on multi-drug ARV treatment (N=168) were blocked for equal representation across groups. SME is a group patient education program co-led by a trained nurse and a trained HIV+ peer-educator which teaches patient medication management/adherence skills and other self-care skills, and is delivered in 6 1.5-2 hour sessions. Medication adherence was assessed using a validated ARV adherence self-report instrument that classifies adherence as Perfect (100%), Fair (80-99%), or Poor (<80%). Multivariate ordered logit models and intention-to-treat analysis were used to adjust for baseline adherence differences and assess program effects on adherence immediately post-intervention, and after a 6-month follow-up period during which no education or support were provided.

Results: Adjusting for baseline adherence, SME patients were significantly more likely to have high post-intervention adherence than PM controls (Odds Ratio [OR] 3.1; p<0.01); differences with SS controls were not significant (OR 1.8; p=0.17). Adjusted post-intervention probabilities of perfect adherence depended on both baseline adherence and intervention group, and were: perfect at baseline - SME 85%, SS 76%, PM 64%; fair at baseline - SME 47%, SS 34%, PM 22%; poor at baseline - SME 13%, SS 8%, PM 5% (p<0.01). Post-intervention probabilities of poor adherence were: for those perfect at baseline - SME 2%, SS 3%, PM 5%; fair at baseline - SME 9%, SS 14%, PM 23%; Poor at baseline - SME 36%, SS 50%, PM 64% (p<0.01). Adherence at 6-month follow-up was similar between the three groups, with adjusted probabilities of perfect adherence 72% (SME), 67% (SS), and 78% (PM) for those perfect at baseline; 41% (SME), 35% (SS), and 48% (PM) for those fair at baseline; 15% (SME), 12% (SS), 20% (PM) for those poor at baseline (all p>0.2). Results are pending for program effects on serum drug levels, plasma HIV concentrations, and emergence of drug resistant strains.

Conclusions: A group patient education and support program using a self-management approach to increase medication adherence (SME) led to better post-intervention adherence than distributing printed educational materials, however the difference did not persist. Future work should seek ways to maintain medication adherence effects over time.

Impact: With further improvements, effective and long-lasting programs to teach HIV patients skills for correct and consistent medication use could lead to better HIV suppression and contribute to long-term disease remission.