Rédiger un PREreview

Infections caused by carbapenem-producing Enterobacterales (CPEs) are a persistent and growing threat in healthcare settings. Yet, current infection prevention and control (IPC) surveillance methods, which largely rely on the spatial and temporal proximity of patients, often misattribute or miss infection transmission events. Here, we develop and retrospectively evaluate an integrated methodology that combines analyses of ward-level patient movement data and whole-genome sequencing (WGS) data analyses, which provide measures of bacterial and plasmid similarity. Specifically, we evaluate this methodology across two datasets: a CPE outbreak of diverse carbapenem types (103 genomes, January 2021–March 2021) and an Imipenem-Hydrolysing β-lactamase-positive CPE outbreak (82 genomes, June 2016–October 2019), using standard clinical criteria and conservative genomic thresholds to quantify how often current IPC surveillance methods correctly identify genomically confirmed transmission events. Findings show that, across 3,423 patient contact–genome pairs, current IPC surveillance methods detected only 20.5% of genomically confirmed transmission events whilst maintaining 98.5% specificity, with missed events arising from temporal, spatial, and cross-species, mechanistic blindspots. In contrast, WGS-enabled IPC surveillance methods provided a 25–47-day earlier detection window and, in a linked economic evaluation, delivered annualised savings of up to £3.6 million, as well as a return on investment exceeding 2-fold in 7 of 8 cost scenarios. By operationalising high-throughput WGS data analysis with clinically relevant patient movement data, we evidence that it may be possible to disrupt and thereby mitigate the effects of AMR-driven CPE outbreaks, supporting investigations into the adoption of WGS-enabled IPC surveillance as a standard-of-care tool.