06 March 2008

"Leaky" hospitals: Is CA-MRSA really HA? Or v.v?

Within the community of scientists researching MRSA — which must be getting bigger all the time, since MedLine records about 25 new papers every week — there is some tension over whether hospital-acquired or community-associated MRSA causes the most disease and early death (morbidity and mortality, in epidemiologist-speak).

The latest paper to stoke this fire was a much-reported, CDC-authored Journal of the American Medical Association paper and an accompanying commentary that were published last Oct. 17. The paper found that invasive MRSA (the most serious cases, of bloodstream infections and endocarditis for instance) causes an estimated 94,360 infections and 18,650 deaths per year. The commentary famously and accurately said that MRSA accounts for more deaths in the United States each year than AIDS does. (Paper: Klevens RM, Morrison MA, Nadle J, et al. Invasive methicillin-resistant Staphylococcus aureus infections in the United States. JAMA 2007;298:1763-71. Abstract here. Commentary: Bancroft EA. Antimicrobial resistance: it's not just for hospitals. JAMA 2007;298:1803-4. MedLine listing, but no abstract.)

An interesting feature of the Klevens paper, which was based on data from the CDC's Active Bacterial Core surveillance system, is that it re-sorts MRSA cases using a new set of definitions: community-associated, hospital-onset, and a new category they called "hospital-acquired, community-onset" — that is, colonized in the hospital but not symptomatic until after the person was discharged. Using these new definitions, the CDC group found that the vast majority of invasive MRSA cases were healthcare associated: 26.6% of the 8987 cases on which they based their study occurred while the patients were in hospital, and 58.4% after they left and were living at home again. Only 13.7 percent of the invasive MRSA cases turned out to be community-associated.

In other words, the most serious cases of MRSA look like they are CA cases, but actually are not.

That finding prompted the team to call for better infection control: "If, in fact, these infections represent acquisition during transitions of care from acute care, it follows that strategies to prevent and control MRSA among in-patients, if properly applied, may have an impact on these infections as well as on the traditional hospital-onset infections."

What's interesting about this is that it appears to return us to the "leaky hospitals" hypothesis of CA-MRSA. This view, which reigned from the emergence of MRSA in the 1960s to well into the 1990s, held that hospitals are the source of almost all MRSA cases if you just look hard enough. It was dislodged by microbiological analysis in several key papers in the late 90s, which showed that the strains causing CA-MRSA cases really were genetically different from the hospital strains.

The CDC paper seems to swing opinion back the other way, toward HA-MRSA as the source of the most serious cases of MRSA disease at least. But, here's an important point: The CDC paper bases its resorting on an assessment of risk factors, such as whether patients had prior exposure to the health care system. It does not sort the cases by microbiology: Of the 8987 cases used in the analysis, genetic-fingerprint (PFGE) results were available for only 864.

Now another group of researchers has arisen to say: That's not good enough. In a letter published in February, Michael David and colleagues of the University of Chicago Children's Hospital chide the CDC group for not considering that someone may have been in a healthcare environment, been discharged, and afterward picked up a community strain, In other words: Unless you do the microbiology, you won't know which cases are which. (David MZ, Siegel JD, Chambers HF, Daum RS. Determining whether methicillin-resistant Staphylococcus aureus is associated with health care. JAMA 2008;299:519. MedLine listing, but no abstract.)

Important note: The University of Chicago group, led by Dr. Robert Daum, were the first to challenge the "leaky hospitals" hypothesis and raise the alarm about CA-MRSA in a pathbreaking paper — really the first significant CA-MRSA paper — in 1998. (Herold BC, Immergluck LC, Maranan MC, et al. Community-acquired methicillin-resistant Staphylococcus aureus in children with no identified predisposing risk. JAMA 1998;279:593-8. Abstract here.)

So: Are healthcare strains, with their unique resistance patterns, responsible for the worst community cases? Are community-staph cases really healthcare cases in disguise? And what if community staph was causing healthcare cases as well?

More on that to come.


Medifix said...

Well argued and critically analyzed, but where do we go from here?

Maryn McKenna said...

First comment! Thank you. And an excellent question, because it gets immediately to real-world resource questions. It seems inarguable that, if we want to know how the epidemiology of MRSA is changing, we have to have more microbio. But microbio is costly — in direct spending, time and person-hours. (There was a presentation at the IDSA mtg last fall — and of course I can't find the slides — about the cost to hospitals just to do the microbio to respond to some states' new mandatory-reporting laws. IIRC it was more than 1 new employee.)