Effect of antibiotic and healthcare exposure on long-term dynamics of extended-spectrum beta-lactamase producing (ESBL) Enterobacterales colonisation in Cambodian children

Background: The gastrointestinal tract is a major reservoir of potentially pathogenic and antimicrobial resistant (AMR) bacteria, such as extended-spectrum-beta-lactamase-producing Enterobacterales (ESBL-E). Healthcare and antibiotic exposure are key determinants for ESBL-E colonisation, but little is known about the long-term temporal dynamics of carriage after exposure. Methods: COMRU-META was a clinical cohort of 1–59-month-olds seen at Angkor Hospital for Children, Cambodia, in 2021-22. Metadata (sociodemographic, clinical and environmental) and rectal swabs (RS) were collected at presentation to healthcare and at 1, 3 and 6-month follow-up visits. To detect ESBL and carbapenem resistant Enterobacterales (CRE), we cultured RS on selective chromogenic media, performed MALDI-TOF mass-spectrometry for isolate identification and determined antimicrobial susceptibilities by disk diffusion and automated minimum inhibitory concentration testing. We performed multivariable logistic regression to assess factors associated with ESBL-E and CRE colonisation at baseline and Cox regression to assess the effect of healthcare and antibiotic exposure on the time-to-gain and time-to-loss of ESBL-E colonisation during follow-up. Results: Among 605 children (median age 1.4 years, 47% female), colonisation at each visit over the 6 months by ESBL Escherichia coli was 85-88% and by ESBL Klebsiella pneumoniae, 27-29%. CRE colonisation proportion range was 1-2%. At baseline, most common risk factors for ESBL-E and CRE colonisation were previous healthcare or antibiotic exposure, while ESBL and CRE K. pneumoniae carriers were also younger and malnourished. For ESBL E. coli, there were 160 colonisation ‘gain’ episodes and 145 ‘loss’ episodes during the 6-month follow-up, not associated with antibiotic and healthcare exposure. For ESBL K. pneumoniae, children with any antibiotic (HR: 1.40, 95% CI: 1.05-1.89), 3rd generation-cephalosporin (HR: 2.06, 95% CI: 1.12-3.79) and inpatient exposures (HR: 1.42, 95% CI: 1.04-1.96) were more likely to have colonisation ‘gain’ episodes (N=250); children with healthcare exposure (HR: 0.70, 95% CI: 0.49-1.00) and who were inpatients (HR: 0.66, 95% CI: 0.43-1.00) were less likely to have colonisation ‘loss’ episodes (N=235). Conclusion: ESBL-E carriage is prevalent in Cambodian children. Persistent healthcare and antibiotic exposure were associated with changes in colonisation dynamics for ESBL K. pneumoniae, but not for ESBL E. coli. These factors should be prioritised in AMR prevention strategies