Mobility and Clinic Switching Among Postpartum Women Considered Lost to HIV Care in South Africa.

This version is the Accepted Manuscript, and was published in final edited form as: J Acquir Immune Defic Syndr. 2017 April 01; 74(4): 383–389. doi:10.1097/QAI.0000000000001284OBJECTIVE: Retention in HIV care, particularly among postpartum women, is a challenge to national antiretroviral therapy programs. Retention estimates may be underestimated because of unreported transfers. We explored mobility and clinic switching among patients considered lost to follow-up (LTFU). DESIGN: Observational cohort study. METHODS: Of 788 women initiating antiretroviral therapy during pregnancy at 6 public clinics in Johannesburg, South Africa, 300 (38.1%) were LTFU (no visit ≥3 months). We manually searched for these women in the South African National Health Laboratory Services database to assess continuity of HIV care. We used geographic information system tools to map mobility to new facilities. RESULTS: Over one-third (37.6%) of women showed evidence of continued HIV care after LTFU. Of these, 67.0% continued care in the same province as the origin clinic. Compared with those who traveled outside of the province for care, these same-province "clinic shoppers" stayed out-of-care longer {median 373 days [interquartile range (IQR): 175-790] vs. 175.5 days (IQR: 74-371)} and had a lower CD4 cell count on re-entry [median 327 cells/μL (IQR: 196-576) vs. 493 cells/μL (IQR: 213-557). When considering all women with additional evidence of care as engaged in care, cohort LTFU dropped from 38.1% to 25.0%. CONCLUSION: We found evidence of continued care after LTFU and identified local and national clinic mobility among postpartum women. Laboratory records do not show all clinic visits and manual matching may have been under- or overestimated. A national health database linked to a unique identifier is necessary to improve reporting and patient care among highly mobile populations