Background:  HIV⁺ prisoners have poor HIV treatment outcomes after release from prison.

Methods: A multi-site, community-based randomized controlled trial of directly administered ART (DAART) versus self-administered therapy (SAT) among 154 HIV⁺ prisoners on ART was conducted for 6 months after prison release. Subjects were randomized 2:1 to DAART:SAT and DAART was provided daily. Primary outcome is percent with viral suppression (viral load <400) at 6 months; secondary outcomes are viral load <50 and DCD4. For viral load, missing = non-suppression and missing CD4 = no change from baseline.

Results:  Between 2004 and 2009, 154 subjects randomized 2:1 to DAART (N = 103) and SAT (N = 51), respectively. The mean age was 45.6 years and comprised a high proportion of people of color (87%), men (81.2%), homelessness on the day of release (43.1%), Axis I psychiatric illnesses (43.7%), use of buprenorphine upon release (33.1%), pre-incarceration hazardous drinking (39.7%), and prescribed once-daily (84%), boosted protease inhibitor (PI) (50.7%) or non-nucleoside reverse transcriptase inhibitor (NNRTI)-based (34.7%) regimens. Baseline labs were viral load <400 = 78.8%, viral load <50 = 57%, and mean CD4 = 390.8 cells/mL. DAART and SAT groups did not differ significantly on any baseline characteristic aside from higher CES-D score (17.4 vs 19.8, p = 0.027) among SAT subjects, but no difference in percent with Major Depression (CES-D ³16); 52.7% vs 57.1%, p = ns. Most (88.3%) subjects were retained for 6 months. At the end of 6 months, the DAART group had significantly more subjects with viral load <400 (78.6% vs 52.9%, p <0.001), viral load <50 (59.2 vs 43.1, p = 0.05), but no significant difference in change in CD4 (+8.3 vs –10.1, p = 0.20). Using multiple regression examining the correlates of achieving viral load <400, only receipt of DAART (AOR = 3.90 [1.76 to 8.67]) and having a baseline viral load <50 (AOR = 2.43 [1.10 to 5.36]) were associated with persistent viral suppression at 6 months. 

Conclusions:  DAART, compared to SAT, maintains HIV treatment outcomes among released prisoners. This trial is the first to confirm the superiority of DAART among this population. DAART, however, is expensive and may not be necessary for all released HIV⁺ prisoners since more than half of subjects on SAT maintained viral suppression.