473. Applying Validated Measurements to Assess Adherence to Antiretroviral Therapy for HIV in Puebla, Mexico: What are we missing?
Session: Poster Abstract Session: HIV Primary Care
Friday, October 21, 2011
Room: Poster Hall B1
Background: Limited data exist on the application of adherence measures validated in the U.S. to Latinos living in resource limited settings (RLS) and on barriers to adherence in Latin America. To explore adherence assessment and barriers for HIV+ Latinos receiving antiretroviral treatment (ART) at a government-sponsored site in Puebla, Mexico, we piloted three different adherence measures: a 7-day recall questionnaire; a 30-day visual analogue scale (VAS), and a qualitative, semi-structured interview.

Methods: Fifty people on ART ≥6 mos completed the questionnaire and VAS, 20 of whom underwent qualitative interviews. Non-adherence measured by the questionnaire was defined as an “almost all” response or less for any antiretroviral medication. Non-adherence measured by VAS was mean adherence calculated for all medications <100%.

Results: Nineteen (38%) respondents were women; 24 (48%) were men who have sex with men. Median age was 37y (IQR: 28.7, 45.2). Median year of HIV diagnosis was 2005. Only 32% had more than elementary school level education. Median pre-ART CD4+ cell count was 126 cells/mm3 (IQR: 47, 194), median time on ART was 34.5 mos (IQR: 25, 50), and 54% were on a PI-based regimen. Non-adherence was reported in 9 (18%) respondents by questionnaire and 19 (38%) by VAS. VAS captured all 9 reporting non-adherence by questionnaire. The kappa statistic for correlation between the two tests was 0.527. Commonly reported reasons for non-adherence on questionnaire were: forgetting (7), too busy (5), and too many pills (5). Results from the 20 qualitative interviews differed: 0 of 6 reporting non-adherence by questionnaire or VAS reported current non-adherence in the interview, though 4 of 6 reported past non-adherence. Additional adherence barriers identified in the semi-structured interviews included: depression (6), abuse (2), and lack of money (2). Six respondents mentioned the positive impact of religious faith on adherence.

Conclusion: Correlation between two standardized measures of adherence was only moderate, and semi-structured interviews captured important factors not present on a standardized questionnaire. Further research on the appropriateness of these adherence measures is needed prior to their implementation in this setting.

Subject Category: H. HIV/AIDS and other retroviruses

Rachel Rivera, MD, Infectious Diseases, UT Health Science Center San Antonio, San Antonio, TX, DeFries Triveni, M.P.H, University of California San Francisco School of Medicine, San Francisco, CA, Karen Rolston, University at Buffalo School of Medicine and Biomedical Sciences, Buffalo, NY, Sergio Saucedo, MSC, Hospital General Dr. Eduardo Vazquez , Puebla, Mexico, Mexico, Indiana Torres Escobar, MD, PhD, CAPASITS, Hospital General Dr. Eduardo Vazquez , Puebla, Mexico, Jennifer Hirsch, PhD, Mailman School of Public Health, Columbia University, New York, NY and Barbara Taylor, MD, University of Texas Health Science Center at San Antonio, San Antonio, TX; Infectious Diseases, Columbia University, New York, NY


R. Rivera, None

D. Triveni, None

K. Rolston, None

S. Saucedo, None

I. Torres Escobar, None

J. Hirsch, None

B. Taylor, None

See more of: HIV Primary Care
See more of: Poster Abstract Session

Findings in the abstracts are embargoed until 12:01 a.m. EST Thursday, Oct. 20 with the exception of research findings presented at IDSA press conferences.