HIV exposure as a risk factor for death among neonates with healthcare-associated bloodstream infections in South Africa, 2016-2017
No Thumbnail Available
Date
2021
Authors
Zwane, Thembekile Buhle Christina
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Background: Hospitalised neonates (aged 0-28 days) are at high risk of developing bloodstream
infections (BSI). Bacterial infections account for 26% of deaths during the neonatal period in low and middle-income countries. HIV-exposed neonates may have lower specific antibody levels at
birth compared to HIV-unexposed neonates, which makes them vulnerable to various infectious
agents. We hypothesized that HIV exposure was associated with death among neonates with
Staphylococcus aureus, carbapenem-resistant Enterobacterales (CRE) or Candida BSI.
Methods: A cross-sectional study was conducted at five hospitals participating in GERMS-SA
surveillance for culture-confirmed BSIs, 2016-2017. A case was defined as a neonate with
isolation of S. aureus, CRE or Candida from a blood culture specimen. Demographic, clinical and
in-hospital outcome was collected by chart review/interview. We defined neonatal exposure to
maternal HIV infection as any neonate with an HIV polymerase chain reaction (PCR) test result,
assuming that this test would be ordered only if the mother was known to be HIV-seropositive.
We used classical and multivariable logistic regression analyses to determine the association
between HIV exposure and in-hospital mortality. We adjusted for age, sex, birthweight, gestational
age, a central venous catheter (CVC) in situ, antimicrobial use and looked for effect modification
of HIV exposure on mortality by birth weight in the final adjusted model.
Results: There were 380 cases and 86.8% (330/380) of the cases had a known outcome and HIV
status. The percentage of missing information varied across cases and therefore the total for each
variable was different. The median age was 11 days (IQR 7-18) and males accounted for 52.3%
(195/373). Cases with CRE accounted for 5.2% (20/380), Candida 37.4% (142/380) and S. aureus
cases 57.4% (218/380). The percentage of HIV-exposed neonates was 32.7% (108/330), of these
108 had a known outcome. The overall crude case-fatality ratio (CFR) was 25.4% (84/330). The
CFR among HIV-exposed neonates was 26.8% (29/108). HIV-exposed neonates had 14% higher
crude odds of death compared to HIV-unexposed neonates (odds ratio [OR] 1.14, 95% CI: 0.68-
1.93). After adjustment, the OR for death was 1.19 (95% CI: 0.60-2.37, p=0.62). Among babies
with birthweight ≥2.5 kg, the adjusted OR for death was 4.01 (95% CI: 0.14-113.86, p=0.42)
among HIV-exposed versus HIV-unexposed neonates. Among neonates with a low birth weight,
the stratum-specific OR for the effect of HIV exposure on mortality was 1.12 (95% CI: 0.55-2.28,
p=0.74).
Conclusion: HIV-exposed neonates with a BSI had 19% higher adjusted odds of death than HIV unexposed neonates, though the 95% CI spanned one. Birth weight modified the effect of HIV
exposure on mortality and this was stronger among neonates with normal birthweight (≥2.5 kg).
Our total sample size may not have been sufficiently large. Missing data probably resulted in
selection bias. However, we recommend using a prospective cohort study design for the analysis
to accurately measure 30-day in hospital mortality from admission date.
Description
A research report submitted in partial fulfilment of the requirements for the degree of Master of Science in Epidemiology to the Faculty of Health Sciences, School of Public Health, University of the Witwatersrand, Johannesburg, 2021