Infectious-cause hospitalisation in a province-wide cohort of children with antenatal HIV exposure compared to children without HIV exposure.
Shani T de Beer, Florence Phelanyane, Suzanne M Ingle, Amy L Slogrove, Brian S Eley, Emma Kalk, Alexa Heekes, Kim Anderson, Andrew Boulle, Mary-Ann Davies, Hayley E Jones
Abstract
Open AccessOBJECTIVES: We described and compared infectious-cause hospitalisation outcomes among children born without HIV in the Western Cape (WC), South Africa, during the WHO Option B+ (2013-2015) and universal ART (2016-2018) eras by exposure to maternal HIV and ART. DESIGN: Retrospective cohort. METHODS: Using data from the WC Provincial Health Data Centre, we described rates, causes and risk factors of infectious-cause hospitalisations, up to age 3 years, among children born at a public WC health facility. We compared rates of and risk factors for admission, in children exposed to maternal HIV and uninfected (HEU) and children HIV unexposed and uninfected (HUU), in the neonatal, post-neonatal (age >28 days to ≤12 months), and age >12-36 month periods using mixed-effects Poisson regression. Regression models were adjusted for maternal age and suburb of residence. RESULTS: We included 398,334 mother-child pairs, 17.2% children HEU and 82.8% HUU. Infectious-cause hospitalisation, between birth and age 3 years, occurred in 11.5% vs. 10.9% of children HEU and HUU respectively. Children HEU experienced higher rates of hospitalisation than children HUU, irrespective of maternal ART history, during the neonatal period (adjusted incidence rate ratios, aIRRs: 1.34-1.66) and post-neonatal period (aIRRs: 1.13-1.42), but not during the >12-36 month period. Among children HEU, maternal VL ≥1000/mL vs. <1000/mL during pregnancy was associated with higher admission rates during the post-neonatal period (aIRR = 1.15; 95% CI:1.06-1.25). CONCLUSIONS: Irrespective of timing of maternal ART start, children HEU vs. HUU had higher rates of infectious-cause hospitalisation during the first year of life, but not thereafter.