Researchers in Kuwait have documented near-universal resistance to first-line antibiotics in Salmonella enterica isolated from retail chicken products, with 97.1% of isolates resistant to ampicillin, 95.7% to tetracycline and 89.9% to cefotaxime — a third-generation cephalosporin that sits in the front line of human salmonellosis treatment.

The study, published in Poultry Science (Volume 105, Issue 12, December 2026; DOI 10.1016/j.psj.2026.107844) as an open-access full-length article, characterized 72 Salmonella enterica isolates recovered from chicken products sold at retail in Kuwait. Ahmad Al-Dhumair of Kuwait University and colleagues at the Ministry of Health tested each isolate against eleven antibiotics using the disk-diffusion method, then subjected all isolates to whole-genome sequencing to identify resistance genes, plasmid replicons and integrons, and to assign sequence types by multilocus sequence typing (MLST).

What the genomes showed

Sequencing identified 32 antimicrobial resistance genes across the collection. Six dominated: tetA in 98.6% of isolates, qacE in 87.5%, sul1 in 86.1%, floR in 81.7%, aac(3)-IV in 80.3% and blaCTX-M-65 in 79.2%. The presence of blaCTX-M-65 in roughly four of five isolates is the genotypic counterpart to the high cefotaxime resistance observed phenotypically, and it signals extended-spectrum beta-lactamase activity circulating in the food supply.

Mobile genetic elements appeared at similarly high frequencies. 79.1% of isolates carried the IncFIB(pN55391) plasmid replicon, and 83.3% carried integrons — the attC/intI elements that capture and disseminate resistance cassettes. In practical terms, the resistance in these bacteria is not only widespread but portable: the machinery for horizontal transfer is present in most isolates.

MLST resolved seven sequence types, with two lineages dominating: ST32 at 33.3% and ST314 at 31.9% of isolates. In silico pathogenicity prediction classified every isolate as a human pathogen, at confidence levels of 95.5% to 98.5%.

Where genotype and phenotype disagreed

For clinicians and diagnostic laboratories, the study's concordance analysis carries the most immediate weight. Agreement between genetic markers and phenotypic susceptibility varied sharply by drug.

Concordance ran high for meropenem (100%; kappa = 1.0), tetracycline (95.5%; k = 0.02), chloramphenicol (78.3%; k = 0.35) and gentamicin (79.7%; k = 0.43). It collapsed for cefotaxime (10.1%; k = 0.0) and nalidixic acid (30.4%; k = 0.0).

The divergence cuts both ways and matters at the bench. For tetracycline, raw percentage agreement looks excellent but the kappa statistic of 0.02 shows that agreement is driven almost entirely by the prevalence of resistance rather than by predictive power of the marker. For cefotaxime and nalidixic acid, genotype alone would have badly misestimated susceptibility. The authors' take-away: whole-genome sequencing and conventional susceptibility testing are complementary, not interchangeable, and surveillance programs should run both.

Implications by sector

Public health. Non-typhoidal Salmonella is a leading foodborne pathogen, and the study found multidrug resistance to be widespread across the isolates, including resistance to clinically important drug classes — beta-lactams, phenicols, aminoglycosides, tetracyclines, sulfonamides and quinolones. The prediction that all isolates are human pathogens at 95.5–98.5% confidence, combined with the high load of transferable resistance elements, frames retail chicken in this market as a reservoir from which resistance genes can move into human-associated bacteria, not just a source of infection.

Food animal. The resistance profile reads as a signature of selection pressure along the poultry production chain. The dominance of tetA, sul1, floR and aac(3)-IV — markers classically linked to tetracycline, sulfonamide, florfenicol/chloramphenicol and aminoglycoside use in animal agriculture — plus the near-fixation of the CTX-M-65 beta-lactamase, gives poultry producers and veterinarians in the region a concrete target list for antimicrobial stewardship review.

Companion animal. The study did not sample companion animals, but its findings are relevant to small-animal practice in the region: dogs and cats fed raw chicken diets sourced from the same retail supply would be exposed to the same multidrug-resistant strains and their mobile resistance elements.

The surveillance argument

The authors stop short of policy prescriptions beyond one clear recommendation: Kuwait needs routine surveillance that pairs whole-genome sequencing with conventional susceptibility testing, embedded in a broader One Health framework. The kappa data supply the technical justification — sequencing alone would have missed the cefotaxime and nalidixic acid resistance phenotypes that disk diffusion caught, while phenotyping alone would miss the plasmid and integron context that predicts spread.

The study, funded under an open-access Creative Commons license, provides a baseline that future retail sampling in the Gulf region can be measured against, and its ST32/ST314 lineage data give outbreak investigators reference profiles for source attribution.