Adherence over 48 weeks in an antiretroviral clinical trial: variable within patients, affected by toxicities and independently predictive of virological response

Objectives: To investigate adherence to antiretroviral therapy over 48 week s, to investigate the association between adherence and treatment-related s ymptoms and to investigate the impact of adherence on virological response over 48 weeks among established predictors of treatment success. Methods: One-hundred-and-sixty HIV-1 infected protease inhibitor- and stavu dine-naive patients participating in a trial of ritonavir/saquinavir versus ritonavir/saquinavir/ stavudine completed an adherence questionnaire and a symptom checklist at weeks 12, 24, 36 and 48. We calculated odds ratios be tween experienced symptoms and non-adherence. Regression models were used t o determine predictors of HIV-1 RNA below 400 copies/ml at week 48, and of the area about the change from baseline over 48 weeks (ACFB) in serum HIV-1 RNA. Results: The percentage of patients reporting missing medication, deviation from time schedule, and dietary prescriptions at separate time-points rang ed from 12 to 15%, 32 to 35% and 17 to 22%, respectively. The percentage th at changed their level of adherence during 48 weeks ranged from 29% for ski pping medication to 48% for deviation from time-schedule. Experienced side- effects were associated with an increased likelihood of non-adherence. Not skipping medication was an independent predictor of both having a serum HIV -1 RNA below 400 copies/ml at week 48 and the ACFB over 48 weeks in serum H IV-1 RNA. Conclusions: Adherence was an independent predictor of virological response over 48 weeks. The level of adherence is variable within patients over tim e. This suggests the need for continued adherence monitoring in all patient s as part of standard medical practice.