Ten-year trends in anti-retroviral therapy persistence among US medicaid beneficiaries, 2001-2010.

OBJECTIVE: Whether the rate of HIV antiretroviral therapy (ART) persistence has improved over time in the U.S. is unknown. We examined ART persistence trends between 2001 and 2010, using non-HIV medications as a comparator.

METHODS: We conducted a retrospective cohort study using Medicaid claims. We defined persistence as the duration of treatment from the first to the last fill date before a 90-day permissible gap, and used Kaplan-Meier curves and Cox proportional hazard models to assess crude and adjusted non-persistence. The secular trends of ART persistence in 43 598 HIV patients were compared with the secular trends of persistence with angiotensin-converting enzyme inhibitors (ACE) or angiotensin receptor blockers (ARB), statins, and metformin in (1) non-HIV-infected patients and (2) subgroups of HIV patients who started these control medications while using ART.

RESULTS: Median time to ART non-persistence increased from 23.9 months in 2001-2003 to 35.4 months in 2004-2006, and was not reached for those starting ART in 2007-2010. In adjusted models, ART initiators in 2007-2010 had 11% decreased hazards of non-persistence compared with those who initiated in 2001-2003 (p < 0.001). For non-HIV patients initiating ACE/ARB, statins, and metformin, the hazard ratios (HR) for non-persistence comparing 2007-2010 to 2001-2003 were 1.07, 0.94, and 1.02, respectively (all p < 0.001). For HIV patients initiating the three control medications, the HRs of non-persistence comparing 2007-2010 to 2001-2003 were 0.71, 0.65, and 0.63, respectively (all p < 0.001).

CONCLUSIONS: Persistence with ART improved between 2001 and 2010. Persistence with control medications improved at a higher rate among HIV patients using ART than HIV-negative controls.