Background: In Côte d’Ivoire a retrospective national-level cohort study conducted in 2009 indicated that 77%, 72% and 75% of children living with HIV (CLHIV) remained in care at 12, 36, and 48 months after starting ART, respectively. The CDC-sponsored Improving HIV Surveillance and Program Evaluation in Côte d’Ivoire project assessed retention progress since that study.

Methods and materials: A retrospective review of medical records was conducted using 2-stage cluster sampling. HIV care and treatment centers (CTC) were sampled based on patient caseload, then records of patients under 15 years of age who initiated ART between 2012–2016 were randomly selected. Time-to-event analysis was performed to estimate the cumulative attrition rates per total number of person-years (PYs) of observation. Cox proportional hazard regressions were conducted to identify factors associated with attrition.

Results: 1198 patient records from 33 CTC were reviewed. At time of ART initiation, CLHIVs’ mean age was 5.8 years, 609 (51%) were male, median CD4 count was 529 cells/μL (interquartile range: 270-910), and 210 (21%) were severely undernourished. Retention was 91%, 84%, 74%, 72%, and 70% at 12, 24, 36, 48 and 60 months after ART initiation, respectively. 309 attrition events occurred over 3,169 PYs of follow-up, with 266 lost to follow-up (LTFU), 29 transfers and 14 deaths. Multivariate analyses showed determinants for LTFU included attending a CTC with a mixed (public-private) status [adjusted hazard ratio (aHR) 6.05, 95% confidence interval (CI): 4.23–8.65], a CTC with no on-site laboratory (aHR 4.01, 95% CI: 1.70–9.46), and a CTC without an electronic medical record system (aHR 2.22, 95% CI: 1.59–3.12). Age, clinical and immunological status at ART initiation, and parent's HIV status were not related to attrition.

Conclusion: Attrition rates were high, with no substantial improvement since 2009 for time periods longer than 12 months since ART initiation. Retention improvements in Cote d’Ivoire would require a focus on scaling up innovative service delivery models such as decentralization of services to the community, coordination with local or regional labs for facilities without an on-site laboratory, and rigorous tracking systems for patients in care.