Real-Time Surveillance of Antimicrobial Resistance

Abstract

Antimicrobial resistance is among the most consequential public health threats of the twenty-first century, contributing to an estimated 1.3 million deaths globally in 2019 and projected to cause 10 million annual deaths by 2050 if current trends continue. Effective response depends on real-time surveillance that detects resistance emergence early, tracks transmission within and between healthcare facilities, and informs antimicrobial stewardship interventions before resistance becomes entrenched. Conventional AMR surveillance based on periodic laboratory summary reports provides a retrospective view that is typically weeks to months out of date and lacks the spatial resolution needed for targeted response. We evaluated five AMR surveillance frameworks across 68 European hospitals and regional laboratory networks covering 14.8 million bacterial isolates from 2018 to 2024. Frameworks ranged from passive aggregate reporting through real-time whole-genome-sequencing-enabled surveillance with ML-based outbreak detection. Performance was benchmarked using our AMR Surveillance Effectiveness Score (AMSES), integrating outbreak detection speed, transmission chain reconstruction accuracy, antimicrobial stewardship feedback impact, cross-border signal sharing, and resource-setting equity. The WGS-integrated real-time framework achieved the highest AMSES (0.919), detecting AMR outbreaks 14.2 days earlier than passive reporting and correctly reconstructing 86.4% of transmission chains against epidemiologically-confirmed reference. Carbapenemase-producing Enterobacterales outbreaks and ESBL-producing organisms showed the strongest real-time surveillance advantage, while MRSA surveillance benefited most from pooled cross-border data sharing