Low risk of attrition among adults on antiretroviral therapy in the Rwandan national program: a retrospective cohort analysis of 6, 12, and 18 month outcomes

BACKGROUND: We report levels and determinants of attrition in Rwanda, one of the few African countries with universal ART access. METHODS: We analyzed data abstracted from health facility records of a nationally representative sample of adults [≥ 18 years] who initiated ART 6, 12, and 18 months prior to data collection; and collected facility characteristics with a health facility assessment questionnaire. Weighted proportions and rates of attrition [loss to follow-up or death] were calculated, and patient- and health facility-level factors associated with attrition examined using Cox proportional hazard models. RESULTS: 1678 adults initiated ART 6, 12 and 18 months prior to data collection, with 1508 person-years [PY] on ART. Attrition was 6.8% [95% confidence interval [CI] 6.0-7.8]: 2.9% [2.4-3.5] recorded deaths and 3.9% [3.4-4.5] lost to follow-up. Population attrition rate was 7.5/100 PY [6.1-9.3]. Adjusted hazard ratio [aHR] for attrition was 4.2 [3.0-5.7] among adults enrolled from in-patient wards [vs 2.2 [1.6-3.0] from PMTCT, ref: VCT]. Compared to adults who initiated ART 18 months earlier, aHR for adults who initiated ART 12 and 6 months earlier was 1.8 [1.3-2.5] and 1.3 [0.9-1.9] respectively. Male aHR was 1.4 [1.0-1.8]. AHR of adults enrolled at urban health facilities was 1.4 [1.1-1.8, ref: rural health facilities]. AHR for adults with CD4+ ≥ 200 cells/μL vs <200 cells/μL was 0.8 [0.6-1.0]; and adults attending facilities with performance-based financing since 2004-2006 [vs. 2007-2008] had aHR 0.8 [0.6-0.9]. CONCLUSIONS: Attrition was low in the Rwandan national program. The above patient and facility correlates of attrition can be the focus of interventions to sustain high retention.