Drinking water chlorination impact on fecal carriage of extended-spectrum beta-lactamase-producing Enterobacteriaceae in Bangladeshi children in a double-blind, cluster-randomized controlled trial

Abstract

Background Water, sanitation, and hygiene (WASH) services have the potential to interrupt transmission of antimicrobial-resistant bacteria and reduce the need for antibiotics, thereby reducing selection for resistance. However, evidence of WASH impacts on antimicrobial resistance (AMR) is lacking. Methods We evaluated extended-spectrum beta-lactamase (ESBL)-producing Escherichia coli and ESBL-KESC (Klebsiella spp., Enterobacter spp., Serratia spp., and Citrobacter spp.) carriage in the feces of 479 Bangladeshi children under 5 years of age enrolled in a double-blind, cluster-randomized controlled trial of in-line drinking water chlorination in two low-income urban communities in Bangladesh. We additionally assessed the interventions impact on circulating beta-lactamase genes in fecal metagenomes and in genomes of fecal ESBL-E. coli isolates. Findings We detected ESBL-E. coli in 65% (n = 309) and ESBL-KESC in 12% (n = 56) of enrolled children. We observed no effect of the intervention on the prevalence of ESBL-E. coli (relative risk [95% confidence interval] = 0.98 [0.78, 1.23]) when controlling for study site and age. Although ESBL-KESC ( 0.76 [0.44, 1.29]) was lower among children in the intervention group, the relative risk was not significant. Concentrations of ESBL-E. coli (log10 CFU/g-wet) were on average [95% confidence interval] 0.13 [-0.16, 0.42] higher in the intervention group and ESBL-KESC (log10 CFU/g-wet) were 0.10 [-0.22, 0.02], lower in the intervention group, when controlling for study site and age. Furthermore, the distribution of ESBL-E.coli sequence types, type of beta-lactamase-encoding genes in ESBL-E. coli isolates, and the presence and relative abundance of beta-lactamase-encoding genes in childrens fecal metagenomes did not differ significantly between the intervention and control children. Interpretation One year of in-line drinking water chlorination in communities did not meaningfully impact the carriage of ESBL-E. coli among children in an area of high ESBL-E. coli carriage. While ESBL-KESC was at lower prevalence than ESBL-E. coli, in the intervention group, limited study power prevented a clear interpretation of treatment effect. Development and evaluation of effective interventions to reduce AMR carriage are needed to support calls for WASH embedded in current National and Global AMR Action Plans.