Invasive carbapenem-resistant enterobacterales infection at a South African paediatric hospital

Bockarie, Yemah

Invasive carbapenem-resistant enterobacterales infection at a South African paediatric hospital

Thesis / Dissertation

2026

Publisher

University of Cape Town

Department

Department of Paediatrics and Child Health

Faculty

Faculty of Health Sciences

Abstract

Background: Carbapenem-resistant Enterobacterales (CRE) cause significant morbidity and mortality. The global dissemination of CRE is a public health concern, and its impact on children is increasingly being recognised in low-and middle-income countries (LMIC). Studies describing paediatric CRE infections in LMIC are limited. Therefore, this study describes the incidence risk, clinical and microbiological characteristics, treatment, and outcome of children with invasive CRE infections at Red Cross War Memorial Children's Hospital (RCWMCH) in Cape Town, South Africa. Methods: A retrospective description was completed on invasive CRE infections diagnosed between January 2016 and December 2021. Clinical and microbiological data were extracted from hospital records and the National Health Laboratory Service Microbiology database. All invasive CRE infections over the study period were used to estimate the incidence risk per 10,000 hospital admissions. Further analysis was completed on infections with complete datasets using STATA 15.0. Categorical variables were summarised using frequencies and percentages, while continuous variables were presented as means and standard deviations or median and interquartile ranges as appropriate. Student T-test or Mann-Whitney U test was used for comparison of continuous variables depending on their distribution, while Pearson's chi-squared and Fisher's exact test were used to compare categorical variables. To explore factors associated with 30-day mortality, univariable analysis was performed to estimate unadjusted risk ratios (RR) and corresponding 95% confidence intervals for the association between each exposure variable and 30-day mortality. A 2-sided p-value of <0.05 was considered statistically significant. Results: The overall incidence risk was 8.4/10,000 admissions. Of 85 infections with sufficient clinical and/or antibiotic information, the median age at CRE diagnosis was 3.9 months (IQR 0.8-44.4) and 80% (68/85) were healthcare-associated. Sites of CRE infection included bloodstream 39% (33/85), urinary tract 22% (19/85) and respiratory tract 21% (18/85). Klebsiella species (68%, 58/85) of which Klebsiella pneumoniae (57/58) predominated, and Serratia marcescens (22%, 19/85) were most isolated. Carbapenemase detection in 63 of 72 isolates tested were OXA-48 (46/63, 73%) and NDM (17/63, 27%). CRE isolates were most susceptible to amikacin 79% (67/85) and tigecycline 78% (66/85). Mortality within 30 days of diagnosis was 33% (28/85). Factors significantly associated with 30-day mortality on univariable analysis were: any organ dysfunction, cardiovascular dysfunction or shock, bladder catheterisation, TPN, and ceftazidime-avibactam and/or colistin-based therapy while non-severe anaemia lowered risk. Conclusion: This study has provided insight into CRE epidemiology at RCWMCH, adding to the limited number of paediatric studies from Africa. Potential 30-day mortality risks were identified on univariable analysis. In resource-limited settings, cost-related barriers usually limit access, availability, and the timely initiation of ceftazidime-avibactam and colistin in paediatric clinical use. Additionally, its use in the sickest patients may negatively affect treatment outcomes, hence the association with mortality seen in our setting. Larger prospective hospital-based paediatric studies are needed to fully evaluate risk factors for 30-day mortality in resource-constrained settings.