Serotype and Commodity Contributions to Quinolone Resistance Trends Among FSIS NARMS Salmonella Isolates, 2013-2023.

Food-animal surveillance can identify antimicrobial-resistance patterns that may affect foodborne disease risk, but pooled summaries can obscure commodity and serotype drivers. This study analyzed public FSIS NARMS cecal and HACCP/slaughter-source Salmonella records accessed through FDA NARMS Now to examine quinolone-resistance trends by calendar period, commodity, and serotype. Records were filtered to Salmonella isolates with reported growth collected from 2013 to 2023. Calendar periods were defined descriptively as 2013-2016, 2017-2022, and 2023 to bracket major U.S. antimicrobial-stewardship milestones, including completion of GFI #213 implementation in 2017 and implementation of GFI #263 in 2023; these periods were treated as temporal landmarks rather than causal policy exposures. Outcomes were ciprofloxacin non-susceptibility, ciprofloxacin resistance only, and nalidixic acid resistance. Analyses included commodity-stratified percentages with Wilson 95% confidence intervals, serotype contribution summaries, and binomial models with standard errors clustered by year. The dataset included 42,232 Salmonella isolates; 41,608 were tested for ciprofloxacin and 41,606 for nalidixic acid. Overall, ciprofloxacin non-susceptibility increased from 1.96% in 2013-2016 to 17.84% in 2017-2022 and 25.71% in 2023. Nalidixic acid resistance increased from 1.36% to 17.02% and 24.01%, respectively. These increases were concentrated in chickens, where ciprofloxacin non-susceptibility rose from 1.7% to 28.7% and 42.4%. Salmonella Infantis accounted for 75.0% of all ciprofloxacin non-susceptible isolates, and Infantis plus Enteritidis accounted for 85.6%. After excluding both serotypes, adjusted period odds ratios were substantially attenuated. Findings indicate that quinolone-resistance trends in FSIS NARMS food-animal Salmonella data were strongly shaped by chicken-associated serotype composition, supporting serotype-aware surveillance.