Carbapenem-resistant Enterobacterales (CRE) colonisation as a predictor for subsequent CRE infection: A retrospective...

Carbapenem-resistant Enterobacterales (CRE) are associated with significant morbidity and mortality. Carbapenem-resistant Enterobacterales colonisation is an important prerequisite for infection, and its surveillance is crucial to reduce spread. However, data from South Africa are limited.

We aimed to determine CRE colonisation prevalence, the incidence of subsequent CRE infections and the risk factors associated with each.

We retrospectively reviewed hospital records from 686 patients admitted to a medical high-care ward at a tertiary hospital in Gauteng, South Africa, between October 2019 and May 2022. Patients were grouped by CRE colonisation status on arrival and discharge. Data on comorbidities, indwelling devices and antibiotic exposure were collected.

The prevalence of CRE colonisation was 12.4% (95% confidence interval [CI]: 10.1-15.1), with Klebsiella pneumoniae (81.2%) being the most common CRE isolated and OXA-48-like enzymes (94.5%) being the most frequent carbapenemase detected. Risk factors for CRE colonisation on the univariate analysis included exposure to antibiotics (odds ratio [OR]: 2.21; 95% CI: 0.98-4.96, P = 0.048) and presence of a central venous line (OR: 6.33; 95% CI: 1.78-22.46, P = 0.001). Of patients colonised with a CRE, 21.2% subsequently developed a culture-positive infection within 180 days from the initial colonisation result and the majority within 30 days. These infections were mostly CREs (OR: 4.0, 95% CI: 1.3-12.7), and where the infections were CREs, the causative CRE organism and carbapenemase subtype were identical in each case.

Our study documented higher CRE prevalence rates than those previously reported from South Africa. Given the association between CRE colonisation and subsequent infection, urgent measures are required to reduce CRE colonisation rates. As the organism and carbapenemase detected in the initial colonisation and subsequent CRE infection were closely related, knowledge of prior CRE colonisation may assist clinicians with antibiotic choice if patients present with an infection within 30 days of CRE colonisation.

This study reports higher CRE colonization rates in South Africa than previously documented, highlighting the urgent need to reduce colonization. The close genetic link between CRE colonization and subsequent infection suggests that knowledge of prior colonization can guide clinicians in selecting effective antibiotics, particularly for infections occurring within 30 days. These findings support targeted interventions to address the rising CRE threat.