A losing battle: MRSA epidemic in U.S. hospitals

By the time it’s detected, other patients infected

While some drug-resistant pathogens are on the rise due to antibiotic "pressures" linked to injudicious use, an epidemic of methicillin-resistant Staphylococcus aureus (MRSA) is proceeding unchecked in the United States largely because of a failure in infection control, researchers say.

In essence, too few clinicians are identifying and isolating incoming cases in a timely manner. So after entering the hospital undetected, MRSA spreads to other patients — some considerably sicker than the original carriers — on the unwashed hands of health care workers.

"For MRSA, I think antibiotic-use practices play perhaps a less important role," says Daniel J. Diekema, MD, MSc, hospital epidemiologist at the Iowa City (IA) Veterans Affairs Medical Center. "People have demonstrated by this point that much of the increase in methicillin-resistant rates is due to patient-to-patient transmission of resistant strains."

The result is a disturbing rise in potentially lethal nosocomial bloodstream infections (BSIs) caused by MRSA. The latest data come from Diekema and colleagues participating in the Sentry surveillance program based at University of Iowa College of Medicine in Iowa City. Researchers reviewed 7,508 nosocomial BSI isolates submitted by participating hospitals throughout the United States and Canada from 1997 to 2000.

Of those infections, 1,957 were caused by S. aureus. Looking closer, the researchers found that the BSI infections caused by MRSA strains of S. aureus increased steadily at a pace of 5% to 6% from 1997 to 2000 in the United States. The percentage of MRSA among S. aureus BSIs climbed inexorably year by year: 1997 — 34%; 1998 — 39%; 1999 — 45%; and 2000 — 51%.

That means that in a medical setting in the United States today, there’s about a 50/50 chance that a staphylococcal BSI will be caused by MRSA — a pathogen that is resistant to all but a handful of antibiotics. In contrast, the percentage of MRSA among S. aureus BSIs in Canada was only 10% in 2000. However, despite widespread efforts to prevent the pathogen from getting endemically entrenched there, MRSA has more than doubled from a 4% level in Canada in 1997.

In the United States, the steady 5% annual increase is comparable to what is being found at sentinel hospitals followed by the Centers for Disease Control and Prevention. However, the Sentry surveillance reveals another interesting artifact: the rate of methicillin-resistance among "community onset" (non-nosocomial) S. aureus BSI isolates also increased in the United States from 21% in 1997 to 31% in 2000. While there has been much concern about true community-acquired MRSA developing among pediatric outpatients due to increasing antibiotic use, the more common picture appears to be people colonized or infected with MRSA simply moving between health settings.

"These patients are infected when they come into the hospital," he explains. "Now it doesn’t mean they acquired it in the community. They could have been in a long-term care facility; they could have gotten it in a previous admission to the hospital. But irrespective of that fact, when they are admitted to the hospital, no one knows that they are infected with MRSA."

Thus the patient may not be isolated for two or three days while lab work is done, creating the potential for further spread.

"We know that less than half the time that health care workers are supposed to wash their hands [that] they do," Diekema says. "Most hospitals still are waiting until they get a report from the lab of MRSA before they put a patient in isolation. With such a high proportion of Staph aureus infections coming in from the community that are methicillin-resistant, it becomes ever so much harder to control hospital spread."

One possible solution is to adopt "risk-factor" isolation, he notes, meaning, for example, that patients are admitted to contact isolation if they have been in another hospital or nursing home in the prior six months. Screening patients on admission is being tried at some hospitals, but the problem is lack of a rapid test to confirm MRSA.

"[We need] something so that when someone is admitted to the hospital, you can swab [his] nares and within a few hours know if [he is a] carrier of MRSA. Or do a PCR directly on a blood culture. Those really, as of yet, are not available."

Other strategies being tried at hospitals include using alcohol-based hand gels to improve hand washing and using mupirocin to decolonize nasal carriers.

"One of the things that we desperately need in infection control are large multicenter studies of some of these strategies to see which of them are most important," Diekema says. "A lot of different hospitals are doing different things."

As MRSA proliferates and is exposed to more antibiotics, the likelihood only increases of new resistant strains arising. The first cases of vancomycin-intermediate S. aureus have appeared over the last few years, and the new drug linezolid has run into trouble with vancomycin-resistant enterococci. Could MRSA be next?

In the Sentry data, MRSA strains were resistant to many other agents, including erythromycin (95% resistant); clindamycin (79%); ciprofloxacin (90%); gentamicin (35%); and trimethoprim-sulfa (22%). However, isolates were 100% susceptible to both vancomycin and linezolid and 99% susceptible to quinupristin/dalfopristin. Those last-line drugs are holding, but that is cold comfort when the data show that more than half of nosocomial staph BSIs are now caused by MRSA.

"We know that these were not just colonizing strains," Diekema emphasizes. "These were strains that were causing bloodstream infections in patients. One [of the reasons] is poor compliance with standard infection control measures like hand hygiene and contact precautions for those who are infected or colonized."