Longitudinal Patterns of Antimicrobial Resistance in Escherichia coli Isolated from Children <5 Years of Age Following Hospital Discharge in Kenya and the Impact of a 5-Day Course of Azithromycin

Abstract

Introduction: Antimicrobial resistance (AMR) is a growing concern worldwide, especially among gram-negative bacteria. One gram-negative species, E. coli, is responsible for most deaths attributed to AMR, especially in sub-Saharan Africa. Hospitalization is a time of increased exposure to pathogens and antibiotics. However, there is limited data on the burden of AMR post-hospital discharge and factors associated with the occurrence of extended-spectrum β-lactamase (ESBL) - producing E. coli within communities.Methods: A sequential cross-sectional data analysis from E. coli isolated from fecal samples collected from children <60 months old in Homabay and Kisii counties of Western Kenya. Fecal samples were collected at hospital discharge, and at three-month and six-month follow-ups. Fecal samples from each unique child were cultured in triplicate for E. coli. If E. coli was isolated from at least one of the samples tested in triplicate, the child was considered to have a positive E. coli culture. Any unique child with E. coli isolated at enrollment with antimicrobial susceptibility testing (AST) results was included in our analysis. We evaluated changes in the proportion of AMR to twelve selected antibiotics over six months following hospital discharge and determined AMR decline by comparing the proportion of AMR between discharge and three months and between three- and six-months post-discharge. We also determined the occurrence of ESBL-producing E. coli at six months post-hospital discharge. We conducted a univariate analysis to determine the burden of AMR post-hospital discharge and a Multivariate Poisson regression to evaluate the association between the occurrence of ESBL-producing E. coli at six months post-hospital discharge and selected risk factors. We then conducted a sensitivity analysis to examine differences in phenotypic AMR among isolates detected at discharge, three months, and six months post-hospital discharge. Results: 406 unique children were enrolled in the study, and all had E. coli isolated at the discharge time point. Most of the children (323, or 80%) had E. coli isolated at each time point (discharge, month three, and month six). E. coli isolates were predominantly from males (59.5%) and the median age of the included children was 19 months (IQR 23 months) at enrollment. Most children were hospitalized for at least three days, were of low socio-economic status (65.2%), and were HIV unexposed (84%).

There was a statistically significant decline in the proportional non-susceptibility to all antibiotics from hospital discharge to three months. The proportion of non-susceptibility isolates between three months and six months was not statistically significant for most antibiotics, except for ceftazidime 0.58 (0.36 - 0.95, p0.031), gentamicin 0.44 (0.30 - 0.63, p<0.001), and ESBL-producing E. coli 0.55 (0.32 - 0.94, p0.029). Non-susceptibility to ampicillin (AMP) and trimethoprim-sulfamethoxazole (TMP/SMX) remained highest at the end of follow-up at month six (72% and 84%, respectively). Carriage of ESBL-producing E. coli dropped from 44% at hospital discharge to 11% at six months post-hospital discharge. There were no statistically significant risk factors associated with the occurrence of ESBL-producing E. coli at six months post-hospital discharge. Conclusions: At hospital discharge, non-susceptibility to E. coli remained high, suggesting that exposure to antibiotics in the hospital is a driver of AMR in these children. Non-susceptibility to all antibiotics significantly declined up to three months post-discharge suggesting that when antibiotic pressure associated with illness and hospitalization is removed, AMR emergence also stabilizes. ESBL-producing E. coli remained stable up to six months post-hospital discharge. There were no significant correlates of ESBL-producing E. coli at six months post-hospital discharge among the risk factors examined. Overall, these findings highlight the urgent need for facility-based interventions, including increased surveillance, antibiotic stewardship, and other control measures, to reduce the spread of antimicrobial-resistant bacteria in sub-Saharan Africa.