Randomized controlled trial of an intervention to prevent adherence failure among HIV-infected patients initiating antiretroviral therapy.

OBJECTIVE: Compare the efficacy of a multicomponent social support intervention to standard-of-care counseling on medication adherence among HIV-infected patients initiating antiretroviral therapy. DESIGN: Randomized controlled trial. Generalized estimating equations tested for differences in the percentage of participants achieving 90% adherence. MAIN OUTCOME MEASURES: Pill-taking, electronically monitored over 6 consecutive months; plasma viral load (VL), assessed at 3 and 6 months following initiation of therapy. RESULTS: Of 226 participants who were randomized and began the trial, 87 (38%) were lost to the study by 6 months. The proportion of adherent participants declined steadily over time, with no time by group interaction. Sustained adherence was associated with increased odds of achieving an undetectable VL (OR=1.78; 95% CI=1.01, 3.13). In intention-to-treat analyses, a larger proportion of the intervention group than the control group was adherent (40.15% vs. 27.59%, p=.02) and achieved an undetectable VL p=.04). However, the majority of participants who remained on study experienced some reduction in VL (>or=1-log drop or undetectable), regardless of experimental condition. CONCLUSION: The multicomponent social support intervention significantly improved medication adherence over standard-of-care counseling; evidence for improved virologic outcomes was inconsistent. Early discontinuation of care and treatment may be a greater threat to the health of HIV patients than imperfect medication-taking.

The menstrual cycle does not affect human immunodeficiency virus type 1 levels in vaginal secretions.

To determine whether the menstrual cycle affects human immunodeficiency virus (HIV) type 1 levels in vaginal secretions, vaginal lavage samples were collected at 7, 14, and 21 days after initiation of menses, to compare virus levels during the follicular, ovulatory, and luteal phases. During 33 menstrual cycles in 25 women, HIV-1 RNA levels in vaginal secretions ranged from <1000 to 5.3x10(7) copies per lavage, and weekly changes ranged from <0.5 to 2.5 log(10) copies per lavage. HIV-1 RNA levels in vaginal lavage samples from days 7, 14, and 21 were not significantly different. No discernible pattern was found in changes of vaginal virus loads (VVLs) during the menstrual cycle. VVLs were not correlated with plasma estradiol or progesterone levels (P>.05). These results suggest that hormonal changes during the menstrual cycle do not have a significant effect on HIV-1 RNA levels in vaginal secretions.