Background
Scale up of antiretroviral therapy (ART) has led to substantial declines in HIV related morbidity and mortality. However, attrition from ART care remains a major public health concern and has been identified as one of the key reportable indicators in assessing the success of ART programs. This study describes the incidence and predictors of attrition among adults initiating ART in a rural HIV clinic in Coastal Kenya.

Methods
A retrospective cohort study design was used. Adults (≥15 years) initiated ART between January 2008 and December 2010 were followed up for two years. Attrition was defined as individuals who were either reported dead or lost to follow up (LFU, ≥ 180 days late since the last clinic visit). Kaplan Meier survival probabilities and Weibull baseline hazard regression analyses were used to model the incidence and predictors of time to attrition.

Results
Of the 928 eligible participants, 308 (33.2% [95% CI, 30.2 – 36.3]) underwent attrition at an incident rate of 23.1 (95% CI, 20.6 – 25.8)/100 pyo. Attrition at 6 and 12 months was 18.4% (95% CI, 16.0 – 21.1) and 23.2% (95% CI, 19.9 – 25.3) respectively. Gender (male vs. female, adjusted hazard ratio [95% CI], p-value: 1.5 [1.1 – 2.0], p = 0.014), age (15 – 24 vs. ≥ 45 years, 2.2 [1.3 – 3.7], p = 0.034) and baseline CD4 T-cell count (100 – 350 cells/uL vs. <100 cells/uL, 0.5 [0.3 – 0.7], p = 0.002) were independent predictors of time to attrition.

Conclusions
A third of individuals initiating ART were either reported dead or LFU during two years of care, with more than a half of these occurring within six months of treatment initiation. Practical and sustainable biomedical interventions and psychosocial support systems are warranted to improve ART retention in this setting.