Relationship of antiretroviral adherence measured by self-report and by rise in mean corpuscular volume to risk of virologic or clinical events in the Collaborations in HIV Outcomes Research/US (CHORUS) cohort

Laura G. Wesolowski, MPH¹, Scott R. Smith, MSPH, PhD², William Miller, MD MPH¹, Bernard M Most, PhD³, and Gregory P. Fusco, MD MPH³. (1) Department of Epidemiology, UNC School of Public Health, CB#7400 McGravran-Greenberg Hall, Chapel Hill, NC 27599-7400, 770 818 0373, gilbertlaura@yahoo.com, (2) School of Pharmacy, University of North Carolina at Chapel Hill, 205-J Beard Hall, CB #7360, Chapel Hill, NC 27599-7360, (3) CHORUS, GlaxoSmithKline, 5 Moore Dr. 17.1351B Sanders Center, Research Triangle Park, NC 27709

Background: The purpose of this study was to assess the association between antiretroviral adherence measured by self-report or mean corpuscular volume (MCV) percent rise and risk of virologic failure (>=500 copies/mL) and clinical events (new AIDS event or death). We also assessed correlation between MCV percent rise and self-reported adherence. Methods: The population included adults in the CHORUS cohort who initiated HAART and were followed for four years or until virologic failure or clinical event. Self-reported adherence (n=466) was measured using the PMAQ question: “How often in the last four weeks did you miss taking a dose of your antiretroviral medicines?” MCV percent rise (n=154) in persons on zidovudine or stavudine containing regimens was calculated as: [(Maximum MCV during first year-Pre-HAART MCV)/(Pre-HAART MCV)]*100 Results: In multivariable analyses, self-reported non-adherence (HR 1.99 95% CI 1.01, 3.92) was associated with virologic failure while MCV percent rise was not statistically associated (HR 0.98 95% CI 0.95, 1.01). MCV percent rise and self-reported non-adherence were not significantly associated with clinical event (49 clinical events). Correlation between MCV percent rise and self-reported adherence among persons on first zidovudine or stavudine containing regimen was r=0.4938, p=0.0048. Conclusions: Further research should evaluate the use of MCV percent rise as an adherence indicator. If there is not an increase in MCV during the first year on HAART or if there is an indication of non-adherence using self-report, an individualized intervention should be implemented to improve adherence before regimen changes are made to improve likelihood of virologic suppression.

Presenting author's disclosure statement:
I have a significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.
Relationship: GlaxoSmithKline funded this research