Combining machine learning with whole-genome sequencing, University of Pittsburgh School of Medicine scientists and Carnegie Mellon University researchers greatly improved the speed at the rate at which outbreaks of infectious diseases are detected in a hospital setting. This is in contrast to the traditional methods of tracking them.
The results were published in the journal Clinic Infectious Diseases today. They show a way for health professionals to identify and stop hospital-based infectious disease outbreaks, thereby saving lives and reducing costs.
Hospitals are using an outdated method of stopping the spread of infectious diseases in patients. The methods haven’t been changed much over the last century. Our process detects important outbreaks that would otherwise go unnoticed by traditional infection prevention monitoring.
Lee Harrison, M.D. Senior author and professor of infectious diseases at Pitt’s School of Medicine and epidemiology at Pitt’s Graduate School of Public Health
The Enhanced Detection System for Healthcare-Associated Transmission (EDS-HAT) couples the recent development of affordable genomic sequencing with computer algorithms connected to the vast trove of data in electronic health records. When the sequencing determines that two or more patients in a hospital have near-identical strains of an infection, machine learning quickly searches those patients’ electronic health records for similarities which could be due to close proximity to hospital beds, a procedure using the same equipment or shared health care provider – thereby alerting health professionals to investigate and stop any further transmission.
Ordinarily, this process requires that healthcare professionals notice that two or more patients are suffering from a similar infection and alert their infection prevention team, who will then go through patient records to attempt to determine how the infection was transmitted.
Alexander Sundermann (M.P.H.), C.I.C.F.A.P.I.C. clinical research coordinator, doctoral student at Pitt Public Health, said, “This is an extremely labor-intensive procedure that is dependent on health professionals being able to detect an outbreak that is shared between patients at the beginning.” While this may work if patients are located within the same hospital unit, it’s unlikely that patients will be infected even if they are located in separate units that have different health care teams.
From November 2016 until November 2018, UPMC Presbyterian hospital conducted EDS-HAT. There was a delay of six months on a few selected infectious pathogens that are often associated with healthcare-associated infections. However the hospital continued utilize traditional methods of prevention for infections. The team looked into how EDS-HAT performed.
EDS-HAT detected 99 clusters of similar infections over the two-year period and identified at least one potential transmission route in 65.7% of those clusters. In the same time frame, infection prevention utilized whole genome sequencing to aid in the investigation of fifteen suspected outbreaks, two of which revealed genetically related infections.
The team estimates that EDS-HAT could have prevented as many as 63 transmissions from one patient to the next of an infectious disease. It could also have saved the hospital an estimated $692,500.
EDS-HAT identified an outbreak of vancomycin-resistant Enterococcus foecium in one case-study. It traced the outbreak back to an interventional radiology procedure that involved the injection of sterile contrast. This was done in accordance with the manufacturer’s instructions. EDS-HAT detected the outbreak and UPMC alerted the manufacturer to improper sterilization practices.
“In that case, EDS-HAT linked the dots between seemingly unconnected patient infections occurring in different hospital units, preventing this outbreak but potentially preventing similar outbreaks at other hospitals,” Harrison said. “That instance demonstrates the value of EDS-HAT.”
UPMC will soon introduce EDS-HAT at UPMC Presbyterian Hospital. The hospital also expects that this innovation will benefit other programs to prevent infections. The original EDS-HAT was primarily focused on drug-resistant bacterial pathogens. However it is set to expand to include sequencing of respiratory viruses including COVID-19.
Sundermann, A.J. and. al. (2021). Whole Genome Sequencing Surveillance, and Machine Learning of an Electronic Health Record for Enhanced Healthcare Detection. Clinical Infectious Diseases. doi.org/10.1093/cid/ciab946.
Content Source: https://www.news-medical.net/news/20211117/AI-and-whole-genome-sequencing-improve-quick-detection-of-infectious-disease-outbreaks.aspx