Genomic surveillance detects, helps stop hospital outbreaks
Click Here to Manage Email Alerts
Key takeaways:
- Real-time whole-genome sequencing detected genetically related clusters and identified potential sources.
- Interventions were then able to prevent further transmission.
SEATTLE — A hospital used real-time whole-genome sequencing to detect genetically related clusters of infection, identify potential sources and stopped further transmission, researchers reported.
“We have previously researched the impact of our tool called the Enhanced Detection System for Healthcare-associated Transmission (EDS-HAT), which combines both whole-genome sequencing (WGS) surveillance and machine learning of the electronic health record to detect and investigate outbreaks of bacterial pathogens,” Alexander J. Sundermann, DrPH, MPH, an epidemiologist and assistant professor of infectious diseases at the University of Pittsburgh, told Healio.
EDS-HAT was able to detect major outbreaks that otherwise go undetected, which could ultimately prevent serious infections and save costs, Sundermann said.
“Given these findings, we pivoted to performing real-time WGS at our hospital in November 2021 to detect and investigate outbreaks,” he said. “The purpose of this study was to share the impactful outbreaks we detected after a whole year of performing WGS surveillance, as well as the infection prevention implications after interventions were made.”
Sundermann and colleagues performed weekly surveillance between Nov. 1, 2021, and Oct. 31, 2022, during which they collected and sequenced cultured isolates of bacterial pathogens from patients who were hospitalized for 3 days or longer or those who had a recent health care exposure in the prior 30 days.
According to findings presented by Sundermann at the Society for Healthcare Epidemiology of America Spring Conference, isolates with 15 or fewer single nucleotide polymorphisms (SNPs) were considered genetically related clusters, except for Clostridioides difficile isolates, which were considered related with 2 SNPs or fewer.
The researchers investigated isolates considered to be related clusters for epidemiological links and implemented interventions to stop transmission.
Of 1,633 unique patient isolates sequences over the course of a year, Sundermann and colleagues identified 74 clusters that comprised 12.9% of patients, with a median cluster size of two patients. In all, 56.2% of isolates had an epidemiological link to an earlier isolate, which the researchers said indicated a potential transmission, and infection prevention interventions for initiated for 89.2% of clusters after notification.
A total of 69 interventions, including unit education (n = 28), hand hygiene observations (n = 16), enhanced cleaning (n = 16), environmental cultures or removal of endoscope (n = 7), and enhanced microbiology surveillance (n = 2), were then performed.
Following these interventions, 59 subsequent infections were identified, including 17 (28.8%) with no clear epidemiological link and 41 (69.5%) with a link either to a new transmission route (n = 37) or the same route before intervention (n = 4). During the study, there was only one subsequent infection within a cluster occurring after intervention from the same route, which the researchers suspected was unit-based transmission of vancomycin-resistant Enterococcus faecium.
According to Sundermann, these results overall demonstrated that they were able to rapidly detect bacterial outbreaks with as few as two patients using WGS surveillance and were able to then direct infection prevention interventions where epidemiological commonalities were found.
“The bottom line is that we show EDS-HAT, when run in real time, can effectively detect important outbreaks that are otherwise missed and guide infection prevention interventions to prevent future infections,” Sundermann said.
Collapse
Sundermann A, et al. Real-time whole genome sequencing surveillance for outbreak detection and intervention. Presented at: Society for Healthcare Epidemiology of America Spring Meeting; April 11-14, 2023; Seattle.
Read more about
Click Here to Manage Email Alerts