Spatial pattern and risk factors of resistance to important antibiotics among E. coli from veterans in seven U.S. Midwest states

TL;DR

The study maps antibiotic resistance in E. coli among veterans in seven U.S. Midwest states and identifies risk factors to guide better antibiotic use.

Contribution

The study provides a large-scale analysis of E. coli resistance patterns and risk factors across seven U.S. Midwest states using Bayesian and multilevel regression models.

Findings

High resistance rates were observed for fluoroquinolone (29%) and TMP-SMX (22%).

Resistance rates varied geographically, with higher levels in southern urban regions of Illinois, Indiana, and Ohio.

Male gender, diabetes, and prior antibiotic exposure were significant risk factors across all antibiotic classes.

Abstract

Effective antibiotic stewardship programing in clinical settings necessitates a good understanding of local prevalences of antimicrobial resistance and important patient and community risk factors. However, most studies are limited in sample size and geographic coverage.

This study utilized phenotypic resistance data of Escherichia coli from the Veteran’s Health Administration of the United States (U.S.), incorporating 126,777 unique cultures from veteran outpatients from seven Midwest states from 2010 to 2023, to examine the spatial pattern and important individual- and county-level risk factors for resistance to four important classes of antibiotics. We utilized Bayesian conditional autoregressive zero-inflated Poisson regression models to generate smoothed rates of resistance in each county and multilevel logistic regression models to detect risk factors for resistance.

High overall rates of resistance were seen for fluoroquinolone (29%) and TMP-SMX (22%). Geographic variation was seen among and between antibiotic classes. Certain urban regions in the southern parts of Illinois, Indiana, and Ohio had higher local resistance rates for fluoroquinolone and TMP-SMX. Being male, having diabetes, and previous exposure to antibiotics are significant risk factors for all classes of antibiotics while the significance of other risk factors varied across classes.

Diverse geographic patterns of resistance level may reflect differences in local prescribing practices, while the differential correlations with risk factors likely reflect their clinical indications and prescribing patterns in clinical settings. The local resistance rates and risk factors for different classes of antibiotics should provide important guidance in practicing empirical prescribing and antibiotic stewardship in clinical settings.

Figures