I mentioned in my last post that I’ve been away on assignment and have been trying to catch up to an onslaught of news. One of the things that broke while I was gone was a new paper in Science Translational Medicine describing the ferocious impact on a hospital at the National Institutes of Health of the arrival of carbapenem-resistant Klebsiella pneumoniae, known for short as KPC or CRKP.
Even though the news is now several days old — the paper went live at noon Wednesday and has been covered in most major media since — I think it’s worth doubling back to take a closer look. Because, with all respect to my media colleagues, I think some of this week’s stories have omitted the larger context. So, a different kind of post for me — less news, more analysis, based on this book, this magazine story, and these past posts on antibiotic resistance. Here we go:
The outbreak described in the paper involved 18 patients: the index patient, a 43-year-old woman who had a serious lung disorder and who arrived at NIH already colonized with CRKP, and and 17 other people who were being treated at NIH who acquired the bug in the following 6 months. Of those 17, patients, 11 died — six of them as a direct result of the infection, and five because of their underlying diseases (lymphomas, tumors and aplastic anemia), though the resistant infection no doubt contributed to their decline. The index patient survived.
A word about the bug: CRKP (covered particularly in this post) is a gut-dwelling bacterium which has acquired the ability to make enzymes that foil the attack of almost every antibiotic used in medicine. In most cases, CRKP responds to only two drugs, neither of which are ideal treatments. One is an old (1949) compound called colistin, which is still effective after decades largely because it is destructive to the kidneys and physicians have been reluctant to use it. The other is a new (2005) antibiotic called tigecycline that doesn’t work well in all tissues. Beyond those two drugs, there is nothing to treat this infection: no additional compounds in the medicine cabinet, no potential compounds under development. When CRKP and similar infections (NDM-1, OXA, VIM, all broadly known as the CREs) acquire resistance to those two remaining drugs — and several of the infections in this outbreak did — they become untreatable. There is nothing to be done.
A word about the hospital: It was the Clinical Center of the National Institutes of Health, which is not like any other hospital in the United States. No one goes to the Clinical Center except, effectively, by invitation. Everyone there is seriously ill and involved in some kind of clinical trial; the hospital doesn’t do primary care, labor and delivery, or elective surgeries, and has no emergency room. That makes a difference to this story in two ways: On the one hand, a higher proportion of the patients were probably vulnerable to this bug, which tends to attack debilitated ICU patients. On the other hand, because it is a hospital that is entirely dedicated to critical care — and also entirely federally funded — it has a deeper bench of knowledge, and much greater resources, than an average community or even academic hospital that is caring for a wider array of patients.
Patients in the NIH outbreak. (Source: Snitkin et al.)
And, quickly, the outlines of the outbreak: The index patient was admitted in June 2011. Because she was already known to be carrying CRKP — possibly picked up in the New York City hospital from which she had been transferred — the Clinical Center immediately put her under what is called “enhanced contact” isolation: gowns and gloves for anyone entering her room, reserving equipment such as a stethoscope for just her use, rigorously cleaning equipment that couldn’t be reserved for her, and double-cleaning the room after she checked out in July. Nevertheless, starting in August, other patients began to acquire CRKP at the rate of about once a week; the last patient in the outbreak, a 37-year-old woman with sickle cell who survived, was recognized Dec. 14. As the outbreak grew in size, the hospital took extreme measures, from grouping both patients and staff so no one with or caring for the bug had any contact with anyone who did not, to hiring additional personnel to watch the staff, to repeatedly checking every patient for colonization at different points on their bodies, to peroxide-bombing rooms, to ripping out plumbing, to literally rebuilding the ICU. Eventually the outbreak was shut down, and whole-genome sequencing of isolates from all the patients revealed that despite the infection-control efforts, and the time lag, the entire outbreak could be traced back to Patient No. 1. (For an explanation of how the sequencing worked, read Ed Yong’s post.)
In the coverage of this paper, I see several things that trouble me: the sense that the outbreak was extraordinary, and the claims that the outbreak should have been disclosed and controlled. In a very narrow sense, those observations are correct — and yet I think they are missing context. Which is simply this:
There was nothing unusual about this outbreak, other than the resources that the NIH infection preventionists were able to marshal to attack it by means of their unique funding. Outbreaks of CRKP and other CREs are happening in health care across the United States — at NIH, at academic medical centers, at community hospitals, in nursing homes — all the time.
source: cdc.gov/hai
That is not to say that the occurrence of CRKP is not an emergency. It is. The point is that it is an emergency happening, not just in a single hospital, but across the entire country, and that we are completely failing to address.
To the right is the CDC’s most recent, state-level map of CRKP and CRE occurrence in the U.S. The organism or one of its close cousins has been found in patients in healthcare facilities in 39 states and in Puerto Rico. Notably, this has all occurred in approximately a decade: The first KPC isolate was spotted in North Carolina in 1996, the first outbreaks occurred in ICUs in New York in the early 2000s, and since then, CRKP has spread not just across the United States but to more than a dozen countries across the world.
Once you start to understand how broadly this bug has spread, you have to reframe (I think) two other aspects of the coverage this week.
One is the claim, advanced by the government of the county where NIH sits, that the Clinical Center should have disclosed its ongoing outbreak. Technically, that may — or may not — be correct: Maryland does have a law mandating disclosure of hospital infections, but NIH is a federal agency and not under Maryland’s jurisdiction. Emotionally, of course one would want to know, and as a matter of patients’ rights, one should. And yet: CRKP has been detected in many, many healthcare institutions, and because it is carried in the intestines, is likely to be in many more institutions that have not detected it. Disclosure is laudable, but by itself it does nothing to protect. And because it has spread so widely, there may effectively be nowhere safe to go.
Another is the tinge, in some stories, that the NIH preventionists should somehow have done more to shut this outbreak down. We are telling an incomplete story if we do not acknowledge that, when it comes to the prevention of healthcare infections, we don’t know what “more” is. Under federal healthcare-payment rules, infection prevention is an unreimbursed activity, and infection-prevention research is by far the lowest of NIH’s research-funding streams. As a result, when the NIH preventionists piled intervention on intervention, they were not following a widely recommended protocol, backed by repeated pieces of science, that told them what would work to control their outbreak. They were simply doing everything they could think of. When other hospitals face outbreaks, they think of other things, possibly conducted in a different order. Despite the size and obvious threat of the CRE epidemic, they are all working in the dark.
If I had to distill what bothers me most about what I read this week, it is that much of the coverage was either a villain story (hospital permits deadly infection!) or a hero story (genomic detectives save the day!) But in fact, the story of CRKP is neither. It is a story of systems breaking down. As a topic, systems are deeply undramatic. But unless we start paying more attention to systems — for research funding, for disease surveillance, for drug development; for addressing, in an organized way, the accelerating loss of the antibiotic miracle — the result will be deeply dramatic: more outbreaks, and more deaths.
Cite: E. S. Snitkin, A. M. Zelazny, P. J. Thomas, F. Stock, N. C. Program, D. K. Henderson, T. N. Palmore, J. A. Segre, Tracking a Hospital Outbreak of Carbapenem-Resistant Klebsiella pneumoniae with Whole-Genome Sequencing. Sci. Transl. Med. 4, 148ra116 (2012).
Update: For more important context, from people actually down in the trenches of combatting this bug, read Dr. Judy Stone at Scientific American Blogs and Dr. Eli Perencevich at Controversies in HAI Prevention.
