Regional prospective whole-genome sequencing surveillance of ESBL-producing Escherichia coli and Klebsiella pneumoniae in the Netherlands: a multicentre study on nosocomial and interhospital transmission

Objectives Whole-genome sequencing (WGS) can identify genomic clusters of multidrug-resistant organisms, but distinguishing nosocomial transmission from community-associated spread requires detailed epidemiological context and often remains challenging. The present study aimed to quantify nosocomial transmission of ESBL-producing Escherichia coli and Klebsiella pneumoniae (ESBL-EcKp), to identify wards with possible outbreaks, and to explore interhospital spread through prospective regional surveillance. Methods A prospective genomic surveillance study was conducted across five Dutch hospitals (2,641 beds, serving ~ 1.7 million inhabitants). Phenotypically suspected ESBL-EcKp isolates from routine diagnostics (June 2022–May 2023) were sequenced. One isolate per patient per species, and per resistance mechanism was analysed. Genomic relatedness was determined using whole-genome multilocus sequence typing (≤ 43 alleles for E. coli , ≤ 22 for K. pneumoniae ). Pairwise comparisons were combined with complete admission histories to assess epidemiological links. Results Among 480 E. coli and 117 K. pneumoniae isolates, 51.9% and 27.4% respectively were genomically related to at least one other isolate. Hospital links were identified in 33.6% of genomically related versus 28.5% of unrelated E. coli comparisons (p < 0.001), and in 70.0% versus 31.7% for K. pneumoniae (p < 0.001). For E. coli , six wards were identified as outbreak-prone, and for K. pneumoniae , two wards. Eleven E. coli index isolates formed cross-hospital genomic links with 53 related isolates across all hospitals, one K. pneumoniae index isolate linked cases across two hospitals. Conclusions Prospective regional WGS surveillance revealed both nosocomial and possible interhospital transmission of ESBL-producing Escherichia coli and Klebsiella pneumoniae . This approach can identify outbreak-prone wards without additional sampling and may support more efficient, targeted infection prevention strategies.