May 2012
Feature Story

Anne Ford

Is universal surveillance for methicillin-resistant Staphylococcus aureus in hospitals a crucial tool in the war against MRSA? Or is it a misguided, needlessly expensive approach?

To Lance Peterson, MD, the answer is easy. “In reading the data, it’s clear that universal surveillance works,” he says. Dr. Peterson is director of microbiology and infectious diseases research and associate epidemiologist at NorthShore University HealthSystem, Evanston, Ill., and clinical professor in the Departments of Pathology and Medicine, University of Chicago Pritzker School of Medicine.

To Michael Edmond, MD—professor of internal medicine and chair of the Division of Infectious Diseases at Virginia Commonwealth University, Richmond, as well as hospital epidemiologist for the university’s health system—the answer is equally clear. “The argument has been that the only way to control MRSA is to do active surveillance,” he says, “and I think our data disproves that. At VCU, we don’t do active surveillance, and yet our MRSA rates are very low.”

And to Loren Miller, MD, investigator at LA BioMed and director of the Infection Prevention and Control Program at Harbor-UCLA Medical Center, Torrance, Calif., the answer is: We don’t know. “The simple answer is that the data are mixed,” he says. “Here’s what seems to happen every few years: Major studies come out within a few months of each other, and one shows a benefit [to universal MRSA testing and decolonization] and the other doesn’t.”

All this to say: The debate over the advisability of universal MRSA surveillance in hospitals is far from over. Those for it say that screening all patients for MRSA upon admission (and putting anyone found to be colonized under contact precautions) results in substantially reduced infection rates. Those against it say it’s a mistake to target MRSA in particular, rather than health-care–acquired infections in general, and that it’s possible to reduce MRSA infections without the expense and effort of universal surveillance.

And those who haven’t yet decided? Well, they have other MRSA-related things on their minds, such as which test to use and how to pay for it. Even as the jury on universal MRSA testing remains out, new rapid MRSA tests have come to market in the last few years. Of course, these tests, which yield results much more quickly than the traditional culture-based testing, are also pricier than culture.

At the same time, desired turnaround times haven’t gotten any shorter. A survey conducted by the American Society for Microbiology and the Association for Professionals in Infection Control and Epidemiology recently found that 51 percent of the infection preventionists surveyed want MRSA and other health-care–associated infection results within 12 hours—more quickly than culture can provide.

Why such a tight time frame?

“In a word, it’s to prevent transmission of pathogenic microorganisms to other patients,” says Lillian A. Burns, MT, MPH, administrative director of infection control and epidemiology at Staten Island University Hospital, New York, and clinical advisor to APIC. The need to prevent health-care–associated infections among patients is something on which everyone can agree.

Let’s momentarily abandon the debate about universal surveillance in favor of focusing on some good news: The cost of rapid MRSA tests has fallen in recent years as more of them have entered the marketplace, Dr. Peterson says. And, he points out, their price seems less expensive when you realize that “the agar-based systems are not cheap either. If you want to get sensitivity anywhere near [that of] the PCR, you have to do an enriched test, which takes two plates and gets you close to the cost of some of the commercial PCR assays.”

The bad news: “The manufacturer suggested retail price [for rapid tests] runs from $25 to $45 or $50,” he says, although “some large-volume labs are able to get their prices lower than that.”

Of the several rapid assays on the market, the one most consistently characterized by sources as the priciest is Cepheid’s Xpert MRSA test. “The Cepheid test, although it is very easy to perform, is very expensive,” says Rodney Arcenas, PhD, ABMM, clinical scientist for microbiology and molecular testing at Pathology Consultants of South Broward, Hollywood, Fla.

Still, some users assert, the Xpert test, which is performed on Cepheid’s GeneXpert instrument, is worth it. In 2007, Loyola University Medical Center in Maywood, Ill., began using the Xpert in its universal MRSA screening initiative. Since then, “we’ve seen a sustained two-thirds drop in nosocomial MRSA infections,” says Paul C. Schreckenberger, PhD, professor of pathology, director of clinical microbiology, and associate director of molecular pathology.

“We like the fact that it’s very sensitive,” he says of the Xpert test. “We like the fact that it has about one minute of tech time involved. All we have to do is add the sample and close the lid and place it in the instrument, and we’re done. We like the simplicity of it, we like the sensitivity of it, we like the fact that we can get the results turned around rapidly. Cost of the test has not been an issue. We have gotten the cost down over the years to something that, I guess, is still an expensive test, but we feel like there’s a big return on investment.”

The cost was brought down in two ways: “One, we’ve negotiated,” he says. “Also, Loyola joined a 65-hospital network, and so we were able to renegotiate all our reagent costs and get it down even lower. We keep seeing a decline in price of doing the test, so that’s a good thing.”

Denise Uettwiller-Geiger, PhD, counts herself among the Xpert’s satisfied users. Dr. Geiger is director of laboratory and clinical trials at John T. Mather Memorial Hospital, Port Jefferson, NY. In 2008, the hospital began putting ICU and CCU patients under active MRSA surveillance, using the Xpert test to turn MRSA results around in less than two hours and assigning patients to beds based on their MRSA status. “Prior to doing real-time PCR 24/7, the infection rate in 2007 was 0.90 per 1,000 discharges, and I’m happy to say that in 2011, the infection rate was 0.17 per 1,000 discharges,” she says. “So we had an 82 percent reduction.”

The Xpert was a big part of the program’s success, she says. “First of all, it’s a single-cartridge reagent system. It’s very easy to use, very stable,” she says. “That’s why we’re able to provide testing 24/7.” The cost per test is about $63, including labor and equipment lease. But, she points out, “by decreasing the number of infections by 82 percent, we’ve seen an almost 86 percent reduction in costs associated with infection. Even though there’s a larger per-test cost, the downstream benefits to the organization certainly outweigh the cost of us performing the test. We feel that we chose the right technology for the type of program that we wanted to put in place.”

Some laboratories use a different test for the bulk of their MRSA testing but keep the Cepheid test around for occasions when they need to get a result out as rapidly as possible. “We use the Xpert for stat requests for MRSA screening,” says Raymond Podzorski, PhD, clinical microbiologist with ProHealth Laboratories, Waukesha, Wis. “These are mostly pa-tients whose screening was somehow miss-ed at their initial presurgical workup, and then they come into the facility and they haven’t had their MRSA screening yet.”

Dr. Podzorski’s laboratory once used the Cepheid test for all MRSA testing, but switched to the Roche LightCycler MRSA Advanced assay in August of last year. “We only have staffing to run the MRSA screening on the day shift,” he explains. “We typically test 40 to 60 MRSA screening swabs a day, and the GeneXpert has eight modules. In the morning there’s 40 to 60 samples sitting there waiting to be tested, and we’ve got eight spots to foot them. So we were basically on a batch system anyway, even though the GeneXpert is a random-access instrument.”

Once the Roche assay came to market, the laboratory tried it out, finding that “timewise, it really didn’t take significantly longer hands-on time to perform,” Dr. Podzorski says. “And it didn’t have a major impact on the turnaround time because we could only do them on the day shift anyway. There was a savings in cost per test, so we switched.” He found, too, that after purchasing the instrumentation necessary to run the Roche test (an analyzer, capillary centrifuge, and shaker device), “the cost savings easily paid for that in a very short time.”

Dr. Arcenas, too, keeps the Cepheid assay on hand for stat purposes but uses the Roche LightCycler for most MRSA testing. “It’s our workhorse assay,” he says of the latter. “We run maybe two to three batches a day, depending on our volume. It’s performed well for us.” Before adopting the LightCycler assay, he did consider BD’s GeneOhm MRSA test, but found it too labor-intensive. Of the Roche test, he says, “It’s pretty straightforward. Our techs like it.”

He does have one small complaint, however. “As far as the analysis part, there’s LightCycler software that analyzes the data, but we have to export it to software where it makes the final interpretation and calls the results ‘positive’ or ‘negative’ or ‘invalid.’ We’re used to it now, but it would be nice to have the LightCycler software make the calls for us, so that we only have to deal with one software. It could improve workflow, but it’s not a huge limitation.”

Dr. Peterson and several of his colleagues at NorthShore presented a poster at the Mar. 31–April 3 meeting of the European Congress of Clinical Microbiology and Infectious Diseases in London. The poster summarized the results of a study they performed on four tests: the BD GeneOhm, the Cepheid Xpert MRSA, the Cepheid Xpert SA Nasal Complete, and the Roche LightCycler MRSA Advanced. While all the tests were found to have similarly high specificities, ranging from 98.7 (BD) to 99.5 (Cepheid Xpert SA), the Roche assay outperformed the others in sensitivity, with a score of 98.3—well above the next-highest performer, the BD test (which was found to have a sensitivity of 92.1).

In Dr. Peterson’s view, that variation in sensitivity may stem from differing methods of specimen preparation. “With the Roche assay, the way the nasal swab is prepared collects all the DNA within the swab,” he says, “whereas the others require a type of vortexing procedure off the swab.” (The research used in the study was supported by an investigator-initiated grant from Roche, as well as by a grant from the Agency for Healthcare Research and Quality.)

Yet another MRSA assay now has FDA clearance, though it’s not for nasal screening but for identifying Staphylococcus aureus and methicillin resistance from positive blood cultures. It’s the MicroPhage KeyPath MRSA/MSSA blood culture test, which, says the company’s literature, uses “bacteriophage amplification technology” to “provide the performance of molecular results at a price that is more in line with chromogenic culture media.” Evaluating it now for adoption is Mike Loeffelholz, PhD, ABMM, associate professor in the Department of Pathology and director of the clinical microbiology laboratory at the University of Texas Medical Branch, Galveston.

The MicroPhage test’s predicted lower price point is especially attractive for his laboratory, Dr. Loeffelholz says, thanks to the damage UTMB suffered from Hurricane Ike in 2008. “Ike nearly shut down UTMB for good,” he says. “It didn’t, but it has set us back a long time. It delayed a lot of things we wanted to implement, such as turnaround time improvements. If it weren’t for Ike, we would have been doing something more state-of-the-art by now for MRSA. We are still using what many labs have been doing for many years, which is a tube coagulase test for a rapid and presumptive identification of staph, and then we do culture on standard blood agar plates. That means you’re looking often at 24 hours for a confirmed Staph aureus identification, and 48 hours for MRSA versus MSSA. This is the state that I think everybody was at five or 10 years ago.”

Dr. Loeffelholz was drawn to the MicroPhage test not only because of its price but also its walkaway time and ease of use. “After the hurricane, we were substantially downsized,” he says. “We simply don’t have the resources to run PCR on positive blood cultures. What attracted us to this test to evaluate it was its apparent simplicity and minimum hands-on time, which is crucial for our lean staffing.”

“So we are in evaluation mode right now,” he says. “We’re running it prospectively on positive blood cultures, and then based upon that performance, we may decide to prepare a business plan to bring that test in. So far it’s performing well. We’ve done a dozen or so specimens, and we’re looking at overall pretty good correlation with the traditional methods.”

Ask Harbor-UCLA’s Dr. Miller which rapid MRSA test he uses, and you’ll get some chuckles in return. “I’m laughing because I work for the county, and we’re always struggling to pay for any new technology,” he says. “We don’t use rapid tests for MRSA at my hospital because of the cost. We use culture. There are data to suggest that if you use a rapid test, you can save money, because you can get people without MRSA out of their private rooms; you don’t have to use gowns and gloves for people you know don’t have MRSA. These rapid tests may have some financial and practical benefits if you use them—and can afford them.”

As per California law, Harbor-UCLA performs MRSA screening on patients who are admitted to the ICU, are on hemodialysis, have been hospitalized within the past 30 days, are coming from a nursing home, or are being admitted for surgery. Harbor-UCLA’s limited resources mean that patients found to be colonized with MRSA are put in contact precautions, but not decolonized.

“The data’s not clear that decolonizing them helps the individual patient or even helps prevent MRSA from spreading,” Dr. Miller explains. “There are no convincing data to show it helps. On a limited budget, we pay for things that evidence shows clearly matter, that will improve patient outcomes. Most experts don’t feel there’s convincing data that it [decolonization] works; nonetheless, a lot of people think you should do it because it looks promising. Things that are evidence-based, we absolutely do, regardless of the cost. But in our hospital, we can’t afford the promising but unproven.”

Speaking of “unproven”: That’s how Virginia Commonwealth’s Dr. Edmond characterizes the value of universal MRSA surveillance. In his view, screening all patients for MRSA is a misguided approach. “First, the expense of it” is prohibitive, he says. “Not just what it costs to do the test, but the opportunity cost, because there’s a lot of work associated with making sure the cultures are sent, tracking them, making sure patients who have positive results are moved to contact precautions.”

Second, he says, “putting people into contact precautions is not necessarily a benign thing to do.” And third, “even if you eliminated all MRSA infections, it would have no impact on your other infections.” Those include multidrug-resistant gram-negative rods and other organisms “for which we have very few therapies.” He would rather see hospitals devoting resources to interventions that reduce all infections, not just MRSA. Certainly his own institution has had success going that route; in the past 12 months, he says, VCU has seen one device-associated MRSA infection in its eight ICUs.

Then, too, universal MRSA screening and decolonization can result in unintentionally giving aid to the enemy, Dr. Miller says: “One of the key antibiotics used for decolonization is mupirocin. Anecdotally, in our area, hospitals that use a lot of it are beginning to see increases in mupirocin-resistant MRSA. One hospital, I’ve been told, has about a 20 percent high-level mupirocin resistance rate. Which is pretty staggering.”

On the other side of the aisle is Dr. Schreckenberger, who argues that universal MRSA surveillance can result not only in better patient care and lower infection rates but also in increased hospital revenue. “It used to be that if somebody got an infection you could add an additional ICD-9 code to the bill,” he says. “Well, you can’t do that anymore. It makes a lot more sense to prevent it.”

The first time he asked his hospital’s CEO to support universal MRSA surveillance, he recalls, the proposal was rejected. Then a new CEO came in and immediately gave his assent. “When I asked him why it was so easy for him to support this when his predecessor wasn’t going to,” says Dr. Schreckenberger, “he said, ‘The way I look at it is we need more revenue. And I can’t build revenue if my rooms are tied up with patients with infections. The No. 1 thing Loyola does in terms of revenue is our surgical department. I have more than enough ORs, more than enough surgeons, and I can’t keep the ORs full. Why not? Because we don’t have enough acute-care beds. Anything we do to prevent a person from leaving that bed on time means another surgery that gets postponed or delayed.’”

Supporters of universal MRSA surveillance point to a 2011 study in the New England Journal of Medicine that examined the results of the VA health care system’s implementation of universal MRSA screening (using Cepheid’s Xpert test). Between October 2007 and June 2010, the study found, the VA’s MRSA transmission rate fell by 17 percent in ICUs and 21 percent in non-ICUs, while the incidence of health-care–associated MRSA infections in ICUs dropped by 62 percent (Jain R, et al. N Engl J Med. 2011;364:1419–1430).

Dr. Peterson calls it “the most important paper that’s come out in the last couple years.”

“They showed in their 153 hospitals, with well over 1 million patients, that their universal surveillance program has been highly effective, just like ours,” he says. (In 2005, NorthShore implemented universal MRSA surveillance and decolonization, resulting in a 70 percent drop in infection in the first year. NorthShore still does universal screening but has switched to a policy of testing only those patients who have certain risk factors such as recent use of antibiotics. “So we’re screening everybody but testing only half of them,” he says.)

Of the VA results, Dr. Edmond, in response, points to a paper that argues that no more than six percent of the VA’s drop in MRSA rates was due to the universal surveillance policy (Gurieva T, et al. Clin Infect Dis. 2012; April 4 [Epub ahead of print]).

“The literature is unfortunately relatively murky,” Dr. Peterson concedes. “The same issue of the New England Journal of Medicine that showed the VA’s dramatic response published another study that had been done by the NIH in ICUs, and they found no benefit from doing surveillance” (Huskins WC, et al. N Engl J Med. 2011;364:1407–1418).

“My own personal belief is that the [Huskins] study was suboptimal regarding the MRSA testing design,” he continues. “They sent all their swabs to a centralized laboratory, and they did broth-enriched culture testing so the turnaround time was very slow, and by the time the results got back to the ICU, 60 percent of the patient days had already been discharged. If I had seen that design, I would have predicted no reduction in MRSA transmission or disease. Inconsistent literature makes it very confusing for most of the public to understand the issue.”