E-mail Abstract Author Session Search Abstracts Program

Session 84 Poster Presentations
Neuropathogenesis: Surrogate Markers
Session Day and Time: Wednesday 1:30 - 3:30 pm
Room: Hall B


695
Intervention Trial Using a Novel Electronic Device in HAART Initiators with Cognitive Dysfunction
A. Andrade*1, H. Davis2, R. L. Skolasky1, S. Celano1, A. W. Wu1, A. M. Letzt3, J. McArthur1
1Johns Hopkins Univ, Baltimore, MD; 2CDC, Atlanta, GA; and 3Adherence Tech Corp, Burke, VA

Background: Patients (pts) with HIV infection and cognitive deficits may miss HAART doses. We examined whether the Disease Management Assistance System (DMAS)—a prototypic programmable verbal prompting device that provides verbal reminders and records dosing times—enhanced adherence to HAART in HIV+ subjects with cognitive deficits.

Methods: Prospective, trial for pts with a history of < 3 HAART or naïve, randomized to control (monthly ARV education alone) or intervention (DMAS + monthly ARV education). Adherence was recorded monthly in all subjects using MEMScap. CD4 count and plasma viral load (VL) were obtained at baseline, wk 12 and 24; CSF VL at baseline and wk 24. Neuropsychological tests assessed psychomotor performance, memory, and working memory at baseline and week 24 (impaired defined as ³ 1.5 SD below mean).

Results: Fifty-eight (58) subjects completed the 24-week study: 52% were ARV-naïve (41% intervention and 62% control) and most had memory impairment. The difference in the distribution of ARV naïve pts was not significant across the two groups (p = 0.115). The table shows overall median adherence scores for all DMAS users compared to controls.

                                    Wk4     Wk8     Wk12   Wk16   Wk20   Wk24   Overall

MEMS (Control)          68%     69%     69%     71%     66%     67%     66%

MEMS (Intervention)    78%     69%     84%     79%     75%     80%     78% (absolute +12%)

Adherence measured by MEMScap and DMAS were highly correlated (0.86, p = 0.09). Mean CD4 changes were similar for the two groups (p = 0.176); 11/29 intervention pts and 4/29 controls had undetectable VL (p = 0.014) at week 24. At wk 24 CSF VL decreased 1.15 and 0.85 log in the intervention and control groups, respectively (p = 0.088). For DMAS users with cognitive impairment, overall adherence was 76% vs 64% for controls (p = 0.002) and the absolute difference was 12%. The difference in cognitively normal subjects was 8% (p = 0.157).

Conclusions: Adherence to HAART was significantly improved in DMAS users with cognitive impairment compared to controls receiving standard education. The rate of virologic suppression to undetectable levels was nearly 3 times higher among DMAS users. This trial demonstrates the feasibility of using a programmable verbal prompting device, and suggests that adherence to HAART can be improved even in those with cognitive impairment.