Self-transfers and factors associated with successful tracing among persons lost to follow-up from HIV care, Sheema District, Southwestern Uganda: retrospective medical records review, 2017-2021

Background Due to improved coverage and scale-up of antiretroviral therapy (ART), patients are increasingly transferring between ART-providing sites. Self-transfers may constitute a high proportion of patients considered lost to follow-up (LTFU), and if overlooked when reporting patients who have dropped out of HIV care, may result in an incorrect estimation of retention. We determined the prevalence of self-transfers, and successful tracing, and identified associated factors among people living with HIV (PLHIV) LTFU from care at public health facilities in Sheema District, Southwestern Uganda. Methods We conducted a cross-sectional retrospective medical records review during February and March 2022. We included records of all PLHIV who were LTFU from 2017 to 2021, and who were registered at government-owned ART clinics in Sheema District. LTFU was considered for those who were not taking ART refills for a period of >= 3 months. We abstracted demographic and clinical data from medical records at the selected clinics. Participants were traced via phone calls or in-person to ascertain the outcomes of LTFU. We performed multivariate modified Poisson regression to identify factors associated with self-transfer, and successful tracing. Results Overall, 740 patients were identified as LTFU from three ART-providing clinics; of these, 560 (76%) were self-transfers. The mean age was 30 (SD +/- 10) years, and most (69%, n = 514) were female; the majority (87%, 641/740) were successfully traced. Age (adjusted prevalence ratio [aPR] = 1.13, 95% CI 1.01-1.25, P = 0.026 for those aged 18-30 years compared to > 30 years), female sex (aPR = 1.18, 95% CI 1.11-1.25, P < 0.001), and having WHO clinical stage 1-2 (aPR = 2.34, 95% CI 1.89-3.91, P < 0.001) were significantly associated with self-transfer. Presence of a phone contact in the patient's file (aPR = 1.10, 95% CI 1.01-1.90, P = 0.026) was associated with successful tracing of the patients considered LTFU. Conclusion Self-transfers accounted for the majority of patients recorded as LTFU, highlighting the need to account for self-transfers among patients considered LTFU, to accurately estimate retention in care. ART-providing facilities should regularly update contact information for PLHIV to enable successful tracing, in the event that the patients are LTFU. This calls for a health-tracking system that easily identifies self-transfers across ART-providing clinics using unique patient identifiers.