Hospital wastewater-based antimicrobial resistance epidemiology: a scoping review

Abstract Introduction: Wastewater-based epidemiology (WBE) offers pooled biological samples of defined populations, complementing clinical surveillance for infectious diseases. Hospitals are hotspots for antimicrobial resistance (AMR). We performed a scoping review assessing hospital wastewater for AMR monitoring. Methods: We searched Medline, Embase, Biosis, Web of Science, and biorxiv.org from inception to May 5, 2026, for studies using hospital wastewater to monitor AMR. The primary outcome was AMR detection and concordance with traditional surveillance. Concordance was reported by resistance class, and reported as strain- or gene-level, as discordant, or as not assessed. Secondary outcomes included sampling methods, and laboratory techniques. Results: Of 4,695 screened studies, 36 from 22 countries met inclusion criteria (high-income 53%, upper-middle-income 42%, and lower-middle-income 5%). Traditional surveillance used mostly routine diagnostics (58%); only 39% of studies aligned temporally with wastewater sampling. Extended-Spectrum Beta-Lactamase-producing organisms showed strain-level concordance in 2/6 studies, and gene-level concordance in 4/6; carbapenem resistance showed strain-level concordance in 2/9 and gene-level concordance in 4/9. Concordance was present for vancomycin-resistant enterococci (2/3), while inconsistent for Methicillin-resistant Staphylococcus aureus (0/2) and Candida auris (1/2) . One study quantified correlation statistically, finding strong association (Spearman’s rho = .88). Wastewater sampling methods included grab sampling (39%), composite sampling (36%), and was unreported (25%). Most studies combined culture-based (89%) and molecular methods (PCR 75%, WGS 42%, metagenomics 17%), mostly targeting Gram-negatives (83%). Conclusion: Hospital WBE showed variable concordance with traditional surveillance, but methodological heterogeneity and limited temporal alignment constrain comparisons. Standardized frameworks and patient-level studies, particularly from low- and middle-income countries, are needed.