Background:  Adherence to ART remains a challenge despite the advent of simpler regimens. Multi-component strategies tailored to patients and addressing diverse adherence barriers hold promise, but need formal evaluation. Managed problem solving (MAPS) is a 5-step behavioral intervention delivered through 4 face-to-face and 9 telephone contacts over 3 months followed by monthly telephone “booster sessions.”

Methods:  We conducted a randomized single-blind trial comparing MAPS with usual care. Eligible patients were:  ≥18 years old, ART naïve, restarting ART after having stopped ≥3 months prior, or failing current ART with HIV RNA ≥1000 copies/mL, and initiating an ART regimen including ≥3 drugs, at least 2 of which were active against their virus. Primary outcome was ART adherence, measured with electronic monitors (MEMS), summarized as percentage of doses taken. Secondary outcome was undetectable plasma HIV viral load dichotomized at <75 copies/mL. Adherence and viral load were compared between MAPS and usual care at month 12 using rank sum tests and χ² tests, respectively. Primary analyses were intention to treat (missing adherence data = 0% and missing viral load = not undetectable viral load). Secondary analyses included last adherence and viral load observations carried forward.

Results:  We randomized 91 to MAPS and 89 to usual care. Median age was 42 years, with 61% male, 85% black, 58% treatment-experienced, median baseline viral load of 2000 copies/mL, and median CD4 count of 250 cells/mm³. Baseline characteristics were balanced between the groups. 15 MAPS and 12 usual-care patients were lost to follow-up. In the intent-to-treat analyses, adherence was 30% higher for MAPS than for usual care (median 69% of doses taken vs 39%, p = 0.023). Secondary analyses were consistent with improved outcomes. Among individuals who remained in treatment, adherence was higher for MAPS than for usual care (77% vs 50%, p <0.002) and the proportion with undetectable plasma HIV viral load was higher for MAPS than usual care (63 of 91 [69%] MAPS vs 48 of 89 [54%] for usual care; p = 0.035). In more conservative intent-to-treat analyses, undetectable viral load was higher for MAPS than usual care at month 12 (54 of 91 [59%] MAPS vs 45 of 89 [51%] usual care), but this difference was not statistically significant (p = 0.24).

Conclusions:  MAPS is an effective ART adherence intervention, and appears to improve virologic outcome in patients remaining in treatment. MAPS should be adapted and evaluated for addressing retention in care, since the intervention’s effects were attenuated by losses to follow-up.