Sewage surveillance for assessing clinical antibiotic resistance prevalence: Combining metagenomic and phenotypic data.
JSON
Flach CF, Berglund F, Osena G, Huijbers PMC, Larsson DGJ
One Health 23 (-) 101485 [2026-12-00; online 2026-06-18]
Surveillance of antibiotic resistance in clinical isolates is a cornerstone for the management of bacterial infections but is limited in large parts of the world, often due to lack of resources. Sewage surveillance has been proposed as a promising, resource-efficient complement to the traditional surveillance approach based on samples from many individual patients. Both phenotypic data on resistance in sewage isolates and abundance of antibiotic resistance genes in sewage have been shown to correlate with resistance prevalence in clinical isolates. Here, we aimed to directly compare and combine an isolate-based and a gene-based sewage surveillance approach to evaluate what best can reflect clinical resistance rates. The two approaches, based on susceptibility testing of collected E. coli isolates and metagenomic sequencing, respectively, were applied to municipal sewage samples collected in ten European countries. The data generated was related to available data on resistance to aminopenicillins, fluoroquinolones, third generation cephalosporins and aminoglycosides prevalence in clinical E. coli isolates using beta regression models. None of the tested individual predictors were superior across all four investigated classes of antibiotics. For modelling of aminopenicillin resistance, a clearly higher R2 value was obtained when isolate-based and gene-based data was combined as predictors, also after adjusting for the number of included variables. We conclude that there could be a value of including both isolate- and gene-based sewage data for predictions of resistance rates in clinical isolates, while emphasizing the value of linking predictors to specific species and classes of antibiotics.
NGI Stockholm (Genomics Production) [Service]
National Genomics Infrastructure [Service]
PubMed 42381665
DOI 10.1016/j.onehlt.2026.101485
Crossref 10.1016/j.onehlt.2026.101485
pmc: PMC13316215
pii: S2352-7714(26)00169-2