Identifying associations in Escherichia coli antimicrobial resistance patterns using additive Bayesian networks

Area of Science:

• Veterinary Microbiology
• Epidemiology
• Antimicrobial Resistance

Background:

• Antimicrobial resistance (AMR) in animal production is complex and not fully understood.
• Multiple resistance patterns are common, complicating risk factor analysis.
• Identifying systematic associations between resistances is essential for effective AMR prevention strategies.

Purpose of the Study:

• To identify systematic associations between antimicrobial resistance patterns in Escherichia coli from Canadian pigs.
• To lay the groundwork for comprehensive, farm-environment-focused AMR risk factor analyses.

Main Methods:

• Bayesian network analysis was employed.
• Examined resistance patterns to 13 antimicrobials in pig fecal E. coli isolates.
• Data sourced from Canada's five major pork-producing provinces.

Main Results:

• Identified numerous systematic associations between individual antimicrobial resistances.
• Sixteen resistance associations were corroborated by existing literature.
• Associations involved key antimicrobial classes: β-lactams, folate inhibitors, tetracyclines, aminoglycosides, and quinolones.

Conclusions:

• This study provides a foundational understanding of AMR epidemiology in pigs.
• Findings support the development of holistic, multivariate risk factor analyses for on-farm AMR.
• Further research can integrate these findings into livestock production environment causal factors for a complete AMR picture.