Host and environmental correlates of multi-drug resistance in Kenyan children with acute bacterial diarrhea

Bacterial diarrhea results in significant morbidity and mortality in children in sub-Saharan Africa. Antibiotic treatment can be a life-saving intervention, but the emergence of antibiotic resistance limits its clinical efficacy. Data on the burden and risk factors for antibiotic resistance in enteric pathogens are needed to inform diarrhea management recommendations and resistance control interventions. Stool/rectal swab samples of children aged 6 mos – 15 yrs presenting with acute diarrhea in western Kenya were cultured for bacterial pathogens. HIV-uninfected children with identified Shigella or Salmonella species, or enteropathogenic [EPEC], enterotoxigenic [ETEC], enteroaggregative [EAEC], or enteroinvasive Escherichia coli [EIEC] were included in this substudy. Resistance to ampicillin, ceftriaxone, ciprofloxacin, cotrimoxazole, and tetracycline was determined using MicroScan Walkaway40 Plus. To evaluate correlates of multi- !3 drug resistance (MDR [resistance to ≥ 3 classes of antibiotics]), we used multivariable log- binomial regression to estimate prevalence ratios (PRs) and 95% confidence intervals (CIs). Of 292 children in the analysis, median age was 22.5 mos (interquartile range: 10.5-41.5 mos), 60.6% used pit latrines and 8.6% were HIV-exposed. Resistance to cotrimoxazole (96.2%) was most common among all pathogens, followed by ampicillin (79.1%) and tetracycline (73.0%). Phenotypic MDR was identified in 60.3% of children; and in 38.2% of Shigella, 40.0% of Salmonella, 73.0% of EPEC, 54.1% of ETEC, 76.0% of EAEC, and 72.2% of EIEC isolates. Children 6-24 mos were more likely to have MDR infections identified than those 24-59 mos (PR = 1.51 [95% CI: 1.19, 1.90]) whereas there was no difference in MDR prevalence between children in the two older age categories, >59m vs. 24-59m (PR = 1.30 [95% CI: 0.91, 1.87]). Children in households with a shared pit latrine were more likely to have MDR (aPR = 1.92 [95% CI: 1.08, 3.38]), than those with flush toilets, as were children in households that practiced open defecation (aPR = 1.91 [95% CI: 1.11, 3.30]). Children living in a household with 2 or more persons per room were 22% more likely to have an MDR pathogen than children living with fewer than 2 persons per room (PR = 1.22 [95% CI: 1.04, 1.43]). Duration of exclusive breastfeeding, malnutrition, maternal HIV, and water source were not associated with MDR infections in this study. Young children and those living in contaminated environments may be at higher risk for infection by antibiotic resistant enteric pathogens.