High rates of unplanned interruptions from HIV care early after antiretroviral therapy initiation in Nigeria

Background: Unplanned care interruption (UCI) challenges effective HIV treatment. We determined the frequency and risk factors for UCI in Nigeria. Methods: We conducted a retrospective-cohort study of adults initiating antiretroviral therapy (ART) between January 2009 and December 2011. At censor, patients were defined as in care, UCI, or inactive. Associations between baseline factors and UCI rates were quantified using Poisson regression. Results: Among 2,496 patients, 44 % remained in care, 35 % had ≥1 UCI, and 21 % became inactive. UCI rates were higher in the first year on ART (39/100PY), than the second (19/100PY), third (16/100PY), and fourth (14/100PY) years (p<0.001). In multivariate analysis, baseline CD4>350/uL (IRR 3.21, p<0.0001), being a student (IRR 1.95, p<0.0001), and less education (IRR 1.58, p=0.001) increased risk for UCI. Fifty-five percent of patients with UCI and viral load data had HIV viral load>1,000 copies/ml upon return to care. Discussion: UCI were observed in over one-third of patients treated, and were most common in the first year on ART. High baseline CD4 count at ART initiation was the greatest predictor of subsequent UCI. Conclusions: Interventions focused on the first year on ART are needed to improve continuity of HIV care.