Background Adherence to antiretroviral treatment is critical to the success of the therapy.

Purpose The authors conducted a study to assess adherence in patients included into a cohort from 2001 until 2008.

Materials and methods The authors performed a retrospective cohort analysis of adherence data from each new patient enrolled between 2001 and 2008. Pharmacy refill records from all medication in the therapy were used to measure mean annual adherence. The primary outcome was optimal adherence (considered as ≥ 95%). Multivariate logistic regression and survival analysis for repeated measurement was applied. Gender, age at the moment of the recruitment and being immigrant were also collected.

Results There were 241 HIV-positive adults eligible for analysis (68.5% male; mean age: 39.1±8.3). In our cohort, 137 (56.9%) were antiretroviral- naïve and 104 (43.1%) antiretroviral- experienced patients. 8.3% were immigrants and the median of follow-up was 4 years (1- 6). Naïve patients showed statistically better mean adherence in the first year and also higher rate of patients with optimal adherence (p < 0.001 in both cases). In the immigrant population the rate of non-adherence was higher (p =0.07). Regarding the multivariate analysis, non-immigrant patients (OR 5.4; 95% CI 1.9 to 15.4) and starting treatment after 2005 (OR 2.93; 95% CI 1.4 to 6.1) showed to be predictors of optimal adherence. For every five- year increase in age, being non-immigrant had 14% higher probability to be adherent (OR 1.14; 95% CI 1.05 to 1.23). During the follow-up, being naïve was the unique variable to maintain optimal adherence.

Conclusions In our cohort, antiretroviral naïve and non-immigrant patients who started treatment after 2005 had higher probability to achieve optimal adherence during the first year. But the only predictor of maintain good adherence levels was being naive.