Neonatal and infant diagnostic HIV‐PCR uptake and associations during three sequential policy periods in Cape Town, South Africa: a longitudinal analysis

INTRODUCTION: To strengthen the early infant diagnosis (EID) programmes and timeously identify and treat HIV‐infected infants, birth HIV‐PCR for some/all infants has been recommended in the Western Cape, South Africa since 2014. Operational data on the implementation of such programmes in low‐ and middle‐income countries are limited. METHODS: Utilizing the electronic records platform at primary care facilities, we developed an electronic register which consolidated obstetric and HIV‐related data, allowing us to track a cohort of HIV‐infected/exposed mother/infant dyads longitudinally from antenatal care through delivery to infant HIV‐PCR. We assessed guideline implementation and impact on EID of three sequential EID policies in a referral chain of facilities in Cape Town (primary‐tertiary care). Birth HIV‐PCR was indicated in period 1 if symptomatic; period 2 if meeting high‐risk criteria for transmission; and period 3 for all HIV‐exposed neonates. RESULTS: We enrolled 2012 HIV‐exposed infants; 89.2% had at least one HIV‐PCR at any point. The majority of birth tests were performed in hospital versus primary care regardless of policy period. Almost half of all infants (47.9%) had at least one high‐risk criterion for vertical infection; of these, 39.7% had a birth test. Infants with more risk factors were more likely to have birth EID. Receipt of a birth HIV‐PCR significantly reduced the likelihood of receiving a follow‐up test at six to ten weeks, even after adjusting for potential confounders (aOR 0.18 (0.12 to 0.26)). The proportion of infants tested at six to ten weeks old dropped from 92.9% (period 1) to 80.2% in period 3 and those receiving birth HIV‐PCR increased, peaking at 67.4% during period 3. The proportion of positive birth tests was highest (2.9%) when birth tests were restricted to infants meeting high‐risk criteria, with a low proportion positive for the first time at six to ten weeks. During period 3, the proportion positive at six to ten weeks was high (2.4%), highlighting the importance of follow‐up to detect intrapartum and early postpartum infections. CONCLUSIONS: Over all policy periods, EID guidelines were incompletely implemented across all levels of care but especially in primary care. Birth HIV‐PCR reduced return for follow‐up testing, such follow‐up testing is critical for the effectiveness of the programme.