Assessment of a program for monitoring antimicrobial purchase and resistance in and on pig farms in the Midwestern United States from May 2020 through October 2023.

Introduction: Antimicrobial resistance (AMR) poses significant challenges to health and treatment options in both human and veterinary medicine. Animal AMR monitoring in the US evaluates carcasses, retail meat, live animals, and diagnostic laboratory submissions; however, there is a lack of consistent on-farm monitoring of use and resistance.

Methods: In 2020, 153 pig farms in the Midwestern United States enrolled in an antimicrobial purchase and resistance monitoring program. Intestinal samples or fecal swabs were collected biannually for 3 years from pigs and their dunging areas; antibiotic purchase data were tracked. and were isolated and underwent antibiotic susceptibility testing using either a commercial bovine/ porcine (BOPO 7F) panel (for pig samples) or the National Antimicrobial Resistance Monitoring System (NARMS) Gram-negative panel (for dunging area samples). Minimum inhibitory concentrations (MICs) of antibiotics were used to evaluate the susceptibility of pig sample isolates, while NARMS breakpoints were used to assess resistance in isolates from dunging areas.

Results: Tetracyclines were the most purchased, and penicillins were the most used antibiotic class across farm types. For pig samples, more isolates exhibited MIC values at the high end of the tested range among and isolates from wean-to-market (WTM) sites compared to breed-to-wean (BTW) sites for almost all antibiotic classes. In addition, isolates from sick pigs had higher MIC values compared to isolates from substandard but otherwise healthy pigs. Among the dunging area isolates, both bacteria had higher rates of resistance in the WTM sites compared to the BTW sites across multiple antibiotics.

Discussion: Individual ages of pigs were a likely confounder and were not controlled for, as these data were not reliably collected. A greater frequency of monitoring, along with controlling for age, recent treatments, and disease events at the individual level, would improve farm-level insights from on-farm AMR monitoring. Currently, the interpretation of phenotypic AMR data for resistance monitoring in swine medicine is limited by the lack of established veterinary breakpoints for enteric organisms. The available NARMS breakpoints are designed for humans, can be used for public health monitoring, and are likely to be applicable primarily to gastrointestinal infections involving the same bacteria in farm animals.