Comparative study on the time trends of antimicrobial resistance at animal and human hospitals in a shared community

Abstract

Antimicrobial resistance (AMR) is a significant global health threat, with increasing morbidity, mortality, and economic consequences. Human and animal health are linked in the spread and evolution of AMR, particularly due to shared antimicrobial usage in healthcare and agriculture. This study investigates the temporal trends in AMR in both a human and an animal hospital within the same community, focusing on five common drug-bug combinations that overlap between these settings. We aim to explore whether AMR patterns in these two hospitals follow similar trajectories, potentially supporting the hypothesis of cross-species transmission of AMR. Data were retrospectively collected from 2010 to 2022 from East Alabama Medical Center (EAMC) and from 2010 to 2023 from Auburn University Veterinary Teaching Hospital (AUVTH). The analysis focused on Escherichia coli and Klebsiella pneumoniae resistance to cefazolin, ceftriaxone, levofloxacin, and gentamicin, as these were overlapping pairs that had shown statistically significant time trends in preanalysis screening. Bayesian logistic regression was applied to model trends in AMR over time. Generally, the AMR trends in the two hospitals did not align, with decreasing susceptibility in the human hospital and increasing susceptibility in the animal hospital. The trends over time for K. pneumoniae susceptibility to gentamicin were the exception, with both the human and animal hospital showing little change over time. Sensitivity analysis that excluded repeated animal isolates revealed stronger alignment between settings in resistance trends, particularly a decreasing sensitivity over time for E. coli and ceftriaxone. This finding suggests possible shared environmental factors or interspecies transmission in certain instances. This study emphasizes the need to collect combined data from humans and animals to better understand AMR and create joint efforts for antimicrobial stewardship. Further research is needed to clarify the relationship between AMR in these interconnected environments and inform strategies to mitigate the spread of resistance.

Keywords: Antimicrobial resistance; Cross-species transmission; Shared community.

Fig. 1

Trend plots of estimated proportions of susceptible isolates and 95 % confidence intervals for each drug-bug combination in each hospital (animal versus human).

Fig. 2

Trend plots of estimated proportions of susceptible isolates and 95 % confidence intervals for each drug-bug combination by animal type (large versus small animals).

Fig. 3

Fitted trend plots for proportion of susceptible isolates over time with 95 % credible intervals. Logistic regression models for animal and human hospitals were fit separately. Repeated observations on the same patient from the animal hospital were treated as independent observations.

Fig. 4

Fitted trend plots for proportion of susceptible isolates over time with 95 % credible intervals. Logistic regression models for animal and human hospitals were fit separately. Only the first observation for each patient in the animal hospital was used.

Supplemental Fig. 1

CONSORT diagram for initial isolate inclusion from the animal hospital.

Supplemental Fig. 2

Fitted trend plots for proportion of susceptible isolates over time with 95 % credible intervals. Logistic regression models for animal (purple) and human (yellow) hospitals were fit separately. Repeated observations on the same patient from the animal hospital were accounted for using generalized estimating equations with an exchangeable covariance structure.