Retention in Care and Patient-Reported Reasons for Undocumented Transfer or Stopping Care Among HIV-Infected Patients on Antiretroviral Therapy in Eastern Africa: Application of a Sampling-Based Approach

Background. Improving the implementation of the global response to human immunodeficiency virus requires understanding retention after starting antiretroviral therapy (ART), but loss to follow-up undermines assessment of the magnitude of and reasons for stopping care. Methods. We evaluated adults starting ART over 2.5 years in 14 clinics in Uganda, Tanzania, and Kenya. We traced a random sample of patients lost to follow-up and incorporated updated information in weighted competing risks estimates of retention. Reasons for nonreturn were surveyed. Results. Among 18 081 patients, 3150 (18%) were lost to follow-up and 579 (18%) were traced. Of 497 (86%) with ascertained vital status, 340 (69%) were alive and, in 278 (82%) cases, updated care status was obtained. Among all patients initiating ART, weighted estimates incorporating tracing outcomes found that 2 years after ART, 69% were in care at their original clinic, 14% transferred (4% official and 10% unofficial), 6% were alive but out of care, 6% died in care (< 60 days after last visit), and 6% died out of care (>= 60 days after last visit). Among lost patients found in care elsewhere, structural barriers (eg, transportation) were most prevalent (65%), followed by clinic-based (eg, waiting times) (33%) and psychosocial (eg, stigma) (27%). Among patients not in care elsewhere, psychosocial barriers were most prevalent (76%), followed by structural (51%) and clinic based (15%). Conclusions. Accounting for outcomes among those lost to follow-up yields a more informative assessment of retention. Structural barriers contribute most to silent transfers, whereas psychological and social barriers tend to result in longer-term care discontinuation.