Genome Detectives Solve a Hospital’s Deadly Outbreak
By GINA KOLATA
Published: August 22, 2012
The ambulance sped up to the red brick federal research hospital on June 13, 2011, and paramedics rushed a gravely ill 43-year-old woman straight to intensive care. She had a rare lung disease and was gasping for breath. And, just hours before, the hospital learned she had been infected with a deadly strain of bacteria resistant to nearly all antibiotics.
The hospital employed the most stringent and severe form of isolation, but soon the bacterium, Klebsiella pneumoniae, was spreading through the hospital. Seventeen patients got it, and six of them died. Had they been infected by the woman? And, if so, how did the bacteria escape strict controls in one of the nation’s most sophisticated hospitals, the Clinical Center of the National Institutes of Health in Bethesda, Md.?
What followed was a medical detective story that involved the rare use of rapid genetic sequencing to map the entire genome of a bacterium as it spread and to use that information to detect its origins and trace its route.
“We had never done this type of research in real time,” said Julie Segre, the researcher who led the effort.
The results, published online Wednesday in the journal Science Translational Medicine, revealed a totally unexpected chain of transmission and an organism that can lurk undetected for much longer than anyone had known. The method used may eventually revolutionize how hospitals deal with hospital-acquired infections, which contribute to more than 99,000 deaths a year.
“It could transform infection control in hospitals and in the community,” said Dr. Sharon J. Peacock, a clinical microbiologist at the University of Cambridge in England, who was not involved in the study. But she added a cautionary note: The challenge now is interpretation of the genetic data. Most hospitals do not have the expertise. They need tools, perhaps Web-based, to do the data analysis. At first, the hospital was confident that it could contain bacteria that could easily kill other patients whose immune systems were weakened by disease, said Dr. Tara Palmore, deputy epidemiologist at the Clinical Center. The doctors knew the bacteria would be almost impossible to stop once they got into patients’ bloodstreams.
“It really is the proverbial superbug,” Dr. Palmore said.
So the hospital used an approach called enhanced contact isolation. The patient was kept in a single room in intensive care. Everyone who entered had to wear a mask and gloves. Every piece of equipment that touched the patient had to be disinfected. And items like blood pressure cuffs and stethoscopes that could not be disinfected were thrown away.
After 24 hours, the woman was moved to a regular private room. For her entire stay, she could walk in the hallway only if no one else was around and if she wore a gown and gloves. For physical therapy, she could use only equipment that was dedicated to her.
A month after she arrived, she was discharged. It seemed no one had picked up the bacteria. Everyone breathed a sigh of relief.
But on Aug. 5, lab technicians found the bacteria in the trachea of a man who had never been in the same area as the infected woman.
“We were worried that he could have gotten it from that first patient, but we just didn’t see how that was possible,” Dr. Palmore said. “That was when we first realized the limit of traditional culturing methods.”
The lab results could not tell the hospital whether the man had been infected by the woman or had an unrelated superbug.
But another patient tested positive for the micro-organism on Aug. 15, and another on Aug. 23. About a patient a week was turning up positive for K. pneumoniae.
Dr. Segre, a genome researcher, proposed sequencing the entire genome of the first patient’s bacteria and comparing it with the genome sequences of bacteria from other infected patients. That could enable scientists to detect minute genetic changes that were the bacterium’s fingerprints. And they could use that knowledge to track the chain of infection.
When the first bacterial genome was sequenced, in 1995, it took three years. This time, researchers did it in just a couple of days.
Sequencing revealed that all the K. pneumoniae originated from the first patient, who transmitted the bacteria from her lung and throat on three occasions.
The woman’s lung bacteria differed from those in her throat by seven DNA base pairs out of six million — a chance occurrence that allowed the researchers not only to identify her bacteria in other patients but also to know where they came from.
It showed the chain of transmission was more complex than anyone had anticipated. Patients were not infected in the order that they appeared to have been. The bacterium had smoldered in many of them, below the level of detection with the usual smears from the groin and throat.
The most surprising was Patient 4. He tested positive six weeks after the first patient left the hospital and died soon after, though not directly because of the infection. But this man had lymphoma, and it was thought that someone with a disease like that, which weakens the immune system, would have become ill within days.
“Suddenly seeing this long latency period was very worrisome,” Dr. Segre said.
The doctors were left with a mystery. How did the bacteria travel from the first patient to the others? The hospital staff had been scrupulous about hand-washing; it had isolated patients with the bacteria in a separate intensive care unit, and a staff member had been there 24 hours a day to watch as health workers and other visitors washed their hands and put on gowns and gloves. When researchers looked for the bacteria on staff members’ hands, they found none.
But they discovered the bacteria in a respirator that had been used by a patient who had the bacteria in his body but had not gotten ill. The respirator had been cleaned, but the disinfecting procedure had failed. The bacteria were also in the sink drains after the rooms had been cleaned. The hospital ended up removing plumbing to get rid of the bacteria.
“We didn’t understand the environmental stability of this organism,” Dr. Segre said.
The hospital finally controlled the outbreak by doing periodic rectal swabs of all patients and looking for the bacteria, a method that requires special equipment but that finds the bacteria even when they are undetectable in swabs from the groin and throat. The doctors also undertook the difficult task of telling patients about the outbreak, including the first woman whose infection ultimately killed six other people,
“They were understandably upset,” Dr. Palmore said. She apologized to one man.
“He in some way was still not satisfied,” Dr. Palmore said. “How could you be?”