Premium
Association between Self‐Reported Adherence to HIV Therapy and Disease Markers
Author(s) -
Yin Connie,
Rayner Craig R,
Read Tim,
Fairley Christopher K
Publication year - 2002
Publication title -
journal of pharmacy practice and research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.222
H-Index - 22
eISSN - 2055-2335
pISSN - 1445-937X
DOI - 10.1002/jppr2002324315
Subject(s) - medicine , cart , human immunodeficiency virus (hiv) , viral load , immunology , mechanical engineering , engineering
Non‐adherence to highly active antiretroviral therapy (HAART) in HIV is widely accepted as a major cause of treatment failure. As a measure of adherence, all patients prescribed HAART at the Melbourne Sexual Health Clinic fill in a simple questionnaire documenting the number of doses per day and the number of doses missed in the last 4, 7 and 28 days. Objective: To investigate the association between mean HIV‐RNA viral load (VL), CD4 antigen count, the number of doses per day, and the 4‐day, 7‐day and 28‐day self‐reported adherence (SRA), and reasons for missed doses. Method: Data were collected for patients attending the clinic from May 2001 until January 2002. SRA [100‐ 100 x (number of doses missed + number of prescribed doses for that period)], number of doses per day, CD4 counts and VL results were averaged within patients. Statistical analyses were performed using classification and regression tree (CART) analysis, a multi variable non‐parametric tree‐based modelling technique. Results: 139 patients were evaluable for analysis of VL and 195 for analysis of CD4 counts. Mean (range) SRAs were: 4 days, 97.5% (45.8‐100%); 7 days, 97.7% (57.1‐100%); and 28 days, 98.1% (62.5‐100% ). The mean number of doses per day was 2.2 (range 1‐5). The average 28‐day SRA and number of doses/day were significantly associated with VL. Conclusion: Average 28‐day SRA and the number of doses per day were associated with VL. Reasons for non‐adherence were consistent with reasons cited in the literature. Further exploration of issues, including recall bias, are required before these results may be applied routinely in the clinical setting.