Using Whole-Genome Sequencing to Control Hospital Infection
Before the outbreak at Queen Elizabeth Hospital Birmingham, WGS was often used as a way to understand an outbreak after it had already happened. In 2010 officials at the British Columbia Centre for Disease Control asked bioinformatician Jennifer Gardy to sequence the genomes of 32 bacterial isolates from a large tuberculosis outbreak in the province. She combined the genome sequences with social-network data that tracked who had come into contact with infected people. With the help of a team including nurses and doctors on the frontline of the outbreak, she was able to reconstruct how the outbreak spread. "Our group was the first to actually use this technique to reconstruct a near-complete outbreak," Gardy says.
It wasn't long before investigators began using WGS to study outbreaks occurring within hospitals themselves. Elimination of hospital-acquired (nosocomial) infections is a priority of the U.S. Department of Health and Human Services. In 2011 an estimated 722,000 infections were acquired in U.S. hospitals, and each year hundreds of millions of infections occur worldwide. Several months after Gardy published her findings, Snitkin tackled an ongoing outbreak of carbapenem-resistant Klebsiella pneumoniae at the National Institutes of Health Clinical Center. The outbreak began with a patient in the intensive care unit. As in the A. baumannii outbreak at Queen Elizabeth Hospital Birmingham, strict infection control measures were taken, and once again, the organism was contained only temporarily before other patients became ill. Of the 18 people ultimately colonized, 11 died.
When the second patient became ill three weeks after the first patient had been discharged, Snitkin's team began routine sampling of all patients in the intensive care unit, even if they did not have the symptoms typically associated with Klebsiella. They identified several asymptomatic individuals who were colonized by the bacterium. Then investigation also found the outbreak strain in several sink drains and on a ventilator.
WGS allowed the investigators and infection control officials to reconstruct the outbreak. Because the genome of K. pneumoniae mutates so rapidly, Snitkin could track transmission by mapping the evolution of the pathogen over time. "Sequencing isolates from a set of patients can tell you that there is an outbreak, because the isolates are so closely related," Snitkin says. "In turn, finding a similarly closely related isolate on a common piece of medical equipment could suggest that this piece of equipment is a vector carrying the bacteria between patients." Having those links gave epidemiologists the evidence they needed to target cleaning and other infection control procedures to stop transmission.
The first report of researchers using WGS to stop an outbreak while it was still happening was published in 2012. A group in England, headed by University of Cambridge microbiologist Sharon Peacock, was studying an outbreak of methicillin-resistant Staphylococcus aureus (MRSA) in a neonatal intensive care unit at Cambridge's Rosie Hospital. Their investigation began after three babies tested positive for MRSA within days of each other. Preliminary tests revealed an identical pattern of antimicrobial resistance, indicating the infections were likely linked. When another baby became ill with the same strain of MRSA just days after the unit was deep-cleaned and sterilized, Peacock began collecting environmental swab samples plus blood samples from workers in the Rosie neonatal intensive care unit and families of the infants. Her team sequenced the bacteria in these samples, and she also began a retrospective analysis of MRSA samples isolated over the past six months in the unit.
The sequencing data showed that one worker in the unit carried the same strain of MRSA that was implicated in the outbreak, making her the likely source of ongoing transmission; however, since she had no symptoms, she had no idea she was infected. Treating her with appropriate antibiotics stopped transmission. In all, the group identified more than a dozen infected individuals, including six babies, who developed severe infections requiring treatment.
(Enlarge Image)
Moving Into Real Time
Before the outbreak at Queen Elizabeth Hospital Birmingham, WGS was often used as a way to understand an outbreak after it had already happened. In 2010 officials at the British Columbia Centre for Disease Control asked bioinformatician Jennifer Gardy to sequence the genomes of 32 bacterial isolates from a large tuberculosis outbreak in the province. She combined the genome sequences with social-network data that tracked who had come into contact with infected people. With the help of a team including nurses and doctors on the frontline of the outbreak, she was able to reconstruct how the outbreak spread. "Our group was the first to actually use this technique to reconstruct a near-complete outbreak," Gardy says.
It wasn't long before investigators began using WGS to study outbreaks occurring within hospitals themselves. Elimination of hospital-acquired (nosocomial) infections is a priority of the U.S. Department of Health and Human Services. In 2011 an estimated 722,000 infections were acquired in U.S. hospitals, and each year hundreds of millions of infections occur worldwide. Several months after Gardy published her findings, Snitkin tackled an ongoing outbreak of carbapenem-resistant Klebsiella pneumoniae at the National Institutes of Health Clinical Center. The outbreak began with a patient in the intensive care unit. As in the A. baumannii outbreak at Queen Elizabeth Hospital Birmingham, strict infection control measures were taken, and once again, the organism was contained only temporarily before other patients became ill. Of the 18 people ultimately colonized, 11 died.
When the second patient became ill three weeks after the first patient had been discharged, Snitkin's team began routine sampling of all patients in the intensive care unit, even if they did not have the symptoms typically associated with Klebsiella. They identified several asymptomatic individuals who were colonized by the bacterium. Then investigation also found the outbreak strain in several sink drains and on a ventilator.
WGS allowed the investigators and infection control officials to reconstruct the outbreak. Because the genome of K. pneumoniae mutates so rapidly, Snitkin could track transmission by mapping the evolution of the pathogen over time. "Sequencing isolates from a set of patients can tell you that there is an outbreak, because the isolates are so closely related," Snitkin says. "In turn, finding a similarly closely related isolate on a common piece of medical equipment could suggest that this piece of equipment is a vector carrying the bacteria between patients." Having those links gave epidemiologists the evidence they needed to target cleaning and other infection control procedures to stop transmission.
The first report of researchers using WGS to stop an outbreak while it was still happening was published in 2012. A group in England, headed by University of Cambridge microbiologist Sharon Peacock, was studying an outbreak of methicillin-resistant Staphylococcus aureus (MRSA) in a neonatal intensive care unit at Cambridge's Rosie Hospital. Their investigation began after three babies tested positive for MRSA within days of each other. Preliminary tests revealed an identical pattern of antimicrobial resistance, indicating the infections were likely linked. When another baby became ill with the same strain of MRSA just days after the unit was deep-cleaned and sterilized, Peacock began collecting environmental swab samples plus blood samples from workers in the Rosie neonatal intensive care unit and families of the infants. Her team sequenced the bacteria in these samples, and she also began a retrospective analysis of MRSA samples isolated over the past six months in the unit.
The sequencing data showed that one worker in the unit carried the same strain of MRSA that was implicated in the outbreak, making her the likely source of ongoing transmission; however, since she had no symptoms, she had no idea she was infected. Treating her with appropriate antibiotics stopped transmission. In all, the group identified more than a dozen infected individuals, including six babies, who developed severe infections requiring treatment.
(Enlarge Image)