OBJECTIVE

This systematic review aimed to (i) determine the risk of antimicrobial resistance (AMR) development associated with antimicrobial use (AMU) and other exposure factors in broilers, and (ii) identify best management practices to mitigate preharvest AMR development of enteric bacteria alongside broiler production.

METHODS

Study selection criteria comprised the population, exposure or intervention, comparator, and outcome framework and included broiler (population), AMU or other management practices (exposure or intervention), organic or antibiotic-free production (comparator), and the presence of AMR-enteric bacteria/genes (outcome). Peer-reviewed primary research studies were searched in PubMed on 19 December 2022, and AGRICOLA, Embase, Scopus, and Web of Science on 10 February 2023. The risk of bias in studies was assessed using the modified ROBIS-E risk of bias assessment tool. The results were synthesised and presented narratively according to PRISMA 2020 guidelines.

RESULTS

In total, 205/2699 studies were subjected to full-text review, with 15 included in the final synthesis. Enteric bacteria Escherichia coli, Salmonella and Campylobacter exhibited AMR and multidrug resistance against several critically important antimicrobials (aminoglycoside, cephalosporin, chloramphenicol, macrolide, penicillin, quinolone, tetracycline, and sulfonamide) for human health. The risk of AMR development in bacteria was shown to be potentially higher with AMU in broiler production. Substandard farm management practices, poor biosecurity measures, and conventional production systems have also been associated with the dissemination of AMR in bacteria.

CONCLUSIONS

Our findings indicate that AMU exposure is associated with considerably higher risk of AMR development in enteric bacteria. Antimicrobial stewardship, organic/antibiotic-free broiler production, good farm management practices, and high-level biosecurity measures are able to substantially mitigate preharvest AMR development in enteric bacteria. However, most of studies were cross-sectional, and therefore causal inference cannot be established.