Community-based progress indicators for prevention of mother-to-child transmission and mortality rates in HIV-exposed children in rural Mozambique

Abstract

Background: Eliminating mother-to-child HIV-transmission (EMTCT) implies a case rate target of new pediatric HIV-infections< 50/100,000 live-births and a transmission rate < 5%. We assessed these indicators at community-level in Mozambique, where MTCT is the second highest globally. Methods: A cross-sectional household survey was conducted within the Manhiça Health Demographic Surveillance System in Mozambique (October 2017–April 2018). Live births in the previous 4 years were randomly selected, and mother/child HIV-status was ascertained through documentation or age-appropriate testing. Estimates on prevalence and transmission were adjusted by multiple imputation chained equation (MICE) for participants with missing HIV-status. Retrospective cumulative mortality rate and risk factors were estimate by Fine-Gray model. Results: Among 5000 selected mother-child pairs, 3486 consented participate. Community HIV-prevalence estimate in mothers after MICE adjustment was 37.6% (95%CI:35.8–39.4%). Estimates doubled in adolescents aged < 19 years (from 8.0 to 19.1%) and increased 1.5-times in mothers aged < 25 years. Overall adjusted vertical HIV-transmission at the time of the study were 4.4% (95% CI:3.1–5.7%) in HIV-exposed children (HEC). Pediatric case rate-infection was estimated at 1654/100,000 live-births. Testing coverage in HEC was close to 96.0%; however, only 69.1% of them were tested early(< 2 months of age). Cumulative child mortality rate was 41.6/1000 live-births. HIV-positive status and later birth order were significantly associated with death. Neonatal complications, HIV and pneumonia were main pediatric causes of death. Conclusions: In Mozambique, SPECTRUM modeling estimated 15% MTCT, higher than our district-level community-based estimates of MTCT among HIV-exposed children. Community-based subnational assessments of progress towards EMTCT are needed to complement clinic-based and modeling estimates.