RESUMO
BACKGROUND: The introduction of Antiretroviral Therapy (ART) has brought a remarkable reduction in HIV-related mortality and morbidity both in adults and children living with HIV/AIDS. Adherence to ART is the key to the successful treatment of patients as well as containment of drug resistance. Studies based on caregivers' report have shown that adherence to ART among children is generally good. However, subjective methods such as caregivers' report are known to overestimate the level of adherence. This study determined the rate of adherence and its predictors using unannounced home-based pill count and compared the result with caregivers' report in a tertiary referral hospital in Ethiopia. METHODS: A cross-sectional study was conducted between December 1, 2011 and January 30, 2012. The study participants were 210 children on ART and their caregivers attending pediatric ART clinic of Tikur Anbessa Hospital (TAH), Addis Ababa University. Caregivers were interviewed at the ART clinic using a structured questionnaire. Then, unannounced home-based pill count was done 7 days after the interview. RESULTS: Caregiver-reported adherence in the past 7 days prior to interview was 93.3%. Estimated adherence using unannounced home-based pill count was found, however, to be 34.8%. On multivariate logistic regression model, children with married [aOR = 7.85 (95% CI: 2.11,29.13)] and widowed/divorced [aOR = 7.14 (95% CI: 2.00,25.46)] caregivers, those who were not aware of their HIV sero-status [aOR = 2.35 (95% CI:1.09, 5.06)], and those with baseline WHO clinical stage III/IV [OR = 3.18 (95% CI: 1.21, 8.40] were more likely to adhere to their ART treatment. On the other hand, children on d4T/3Tc/EFV combination [OR = 0.10 (95% CI: 0.02, 0.53)] were less likely to adhere to their treatment. Caregivers' forgetfulness and child refusal to take medication were reported as the major reasons for missing doses. CONCLUSION: The level of adherence based on unannounced home-based pill count was unacceptably low. Interventions are urgently needed to improve adherence to ART among children at TAH. Besides, a longitudinal study measuring adherence combined with clinical parameters (viral load and CD4 count) is needed to identify a simple and reliable measure of adherence in the study area.