Uganda’s “EID Systems Strengthening” model produces significant gains in testing, linkage, and retention of HIV-exposed and infected infants: An impact evaluation

Introduction
A review of Uganda’s HIV Early Infant Diagnosis (EID) program in 2010 revealed poor retention outcomes for HIV-exposed infants (HEI) after testing. The review informed development of the ‘EID Systems Strengthening’ model: a set of integrated initiatives at health facilities to improve testing, retention, and clinical care of HIV-exposed and infected infants. The program model was piloted at several facilities and later scaled countrywide. This mixed-methods study evaluates the program’s impact and assesses its implementation.

Methods
We conducted a retrospective cohort study at 12 health facilities in Uganda, comprising all HEI tested by DNA PCR from June 2011 to May 2014 (n = 707). Cohort data were collected manually at the health facilities and analyzed. To assess impact, retention outcomes were statistically compared to the baseline study’s cohort outcomes. We conducted a cross-sectional qualitative assessment of program implementation through 1) structured clinic observation and 2) key informant interviews with health workers, district officials, NGO technical managers, and EID trainers (n = 51).

Results
The evaluation cohort comprised 707 HEI (67 HIV+). The baseline study cohort contained 1268 HEI (244 HIV+). Among infants testing HIV+, retention in care at an ART clinic increased from 23% (57/244) to 66% (44/67) (p < .0001). Initiation of HIV+ infants on ART increased from 36% (27/75) to 92% (46/50) (p < .0001). HEI receiving 1st PCR results increased from 57% (718/1268) to 73% (518/707) (p < .0001). Among breastfeeding HEI with negative 1st PCR, 55% (192/352) received a confirmatory PCR test, a substantial increase from baseline period. Testing coverage improved significantly: HIV+ pregnant women who brought their infants for testing after birth increased from 18% (67/367) to 52% (175/334) (p < .0001). HEI were tested younger: mean age at DBS test decreased from 6.96 to 4.21 months (p < .0001). Clinical care for HEI was provided more consistently. Implementation fidelity was strong for most program components. The strongest contributory interventions were establishment of ‘EID Care Points’, integration of clinical care, longitudinal patient tracking, and regular health worker mentorship. Gaps included limited follow up of lost infants, inconsistent buy-in/ownership of health facility management, and challenges sustaining health worker motivation.

Discussion
Uganda’s ‘EID Systems Strengthening’ model has produced significant gains in testing and retention of HEI and HIV+ infants, yet the country still faces major challenges. The 3 core concepts of Uganda’s model are applicable to any country: establish a central service point for HEI, equip it to provide high-quality care and tracking, and develop systems to link HEI to the service point. Uganda’s experience has shown the importance of intensively targeting systemic bottlenecks to HEI retention at facility level, a necessary complement to deploying rapidly scalable technologies and other higher-level initiatives.