Hospital Enterococcus faecium demonstrates distinct environmental and patient reservoirs: a genomic point prevalence survey
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Objectives
Enterococcus faecium is a common cause of hospital-acquired infections and is associated with vancomycin resistance. The hospital environment is often a reservoir for E. faecium due to its ability to survive on surfaces. In 2019 we noted increasing prevalence of vanA E. faecium causing bloodstream infections. We aimed to assess the hospital environment as a reservoir of vanA E. faecium .
Methods
We conducted a point prevalence survey of the environment at the Alfred Hospital, Melbourne, Australia. We sampled all wards (n=12), 10 surface types and 357 surfaces. Resulting E. faecium isolates underwent short-read whole genome sequencing and were compared to invasive infection (bloodstream) isolates. The van operon, multi-locus sequence type (MLST) and core genome MLST (cgMLST) were determined.
Results
Twenty percent (73/357) of surfaces isolated E. faecium with a median of 20% positivity per ward (IQR 11.6-30.6%). vanA and vanB were detected in 12/73 (16%) and 58/73 (79%) environmental genomes, respectively. In contrast, a higher proportion of bloodstream isolates carried vanA (32/72 [44%], P <0.001). Environmental genomes belonged to 10 MLSTs and 10 cgMLSTs, with ST796/ST78 accounting for 54/73 (74%) genomes. Clinical STs and cgMLSTs overlapped with environmental but were more diverse. We noted 13 putative genomic transmission clusters. One cluster contained 25/44 (57%) vanA genomes, with the majority being clinical (22/26 genomes, 85%). Environmental-clinical links were observed in 292/895 (33%) putative genomic transmission links. In completed assemblies, the vanA and vanB operons were located on pRE25-like plasmids and the chromosome, respectively.
Conclusion
We noted a 20% prevalence of E. faecium environmental colonisation but the vanA operon was detected in only 12/73 genomes. We identified distinct environmental and patient reservoirs, with most environmental genomes carrying the vanB operon and having distinct STs/cgMLST clusters. Environment-clinical reservoir spillover was detected but accounted for a minority of putative transmission links.
