Tuesday, March 18, 2008

The Ups and Downs of Universal MRSA Screening

By Crystal Phend
EVANSTON, Ill., March 17 -- No sooner had Swiss researchers given an unshakable thumbs down to universal screening for methicillin-resistant Staphylococcus aureus when a study here came to the opposite conclusion.Introduction of a protocol for active surveillance at admission reduced the rate of hospital-associated MRSA by 69.6% in three affiliated hospitals, reported Lance R. Peterson, M.D., of Evanston Northwestern Healthcare here, and colleagues in the March 18 issue of the Archives of Internal Medicine. These observational findings contrasted sharply with those of the similar Swiss study that found no benefit to surveillance for surgical patients, a finding reported in the March 12 issue of the Journal of the American Medical Association (See: Hospital Admission Screening Fails to Stop MRSA Infections).Yet it would be premature to recommend that all hospitals institute universal screening, commented Ebbing Lautenbach, M.D., M.P.H., of the University of Pennsylvania, in an accompanying editorial.
"How MRSA behaves in one hospital can be very different from how it behaves in another hospital," he said. "You really need more information at the hospital level to figure out what makes sense for that given hospital."
The need for interventions to halt increasing MRSA rates at hospitals is universally recognized, but working out the details has been tricky, he said.
Several states have legislated uniform surveillance for MRSA in healthcare settings, and Congress recently began deliberations on the Strategies to Address Antimicrobial Resistance (STAAR) Act to boost federal efforts at surveillance, prevention, and control. However, the CDC and other organizations have recommended against routine or mandated surveillance.
Surveillance can be an important tool against MRSA, but "mandating screening limits a hospital's flexibility to design infection prevention programs that will best protect their patient population," according to a statement from the Association for Professionals in Infection Control and Epidemiology.
Part of the reason for the controversy has been lack of solid data, Dr. Lautenbach said. Aside from the Swiss study, studies looking at active surveillance were small, focused on only select hospital units or specific patient populations, or were limited to outbreak control.
Dr. Peterson and colleagues examined outcomes at their three-hospital healthcare system with sequential surveillance protocols.
In the baseline year, there was no routine surveillance for MRSA colonization or attempt at decolonization.
In the second year, all patients admitted to the ICU were tested for MRSA nasal colonization using culture-based assays with a turnaround time of 2.5 days. The 8.3% (277 of 3,334) of patients positive for MRSA were put in contact isolation but not decolonized.
In the third year, all hospitalized patients were tested on day one of admission using polymerase chain reaction-based assays with a turnaround time of 0.67 days. The 6.3% (3,926 of 62,035) of patients with MRSA underwent topical decolonization therapy as well as contact isolation. The hospitals achieved 90% adherence to the surveillance program.
The five-day decolonization regimen consisted of 2% mupirocin (Bactroban) twice daily and a 4% chlorhexidine (Peridex, Periogard) wash or shower every two days.
The results were impressive, Dr. Lautenbach said.
The number of hospital-associated clinical MRSA infections in the bloodstream, respiratory, urinary tract, or surgical site fell from 8.9 at baseline to 7.4 during ICU surveillance and 3.9 per 10,000 patient-days during universal surveillance.
This represented a nonsignificant 36.2% decrease from baseline to ICU surveillance (P=0.17) and 69.6% (P=0.03) from baseline to universal surveillance.
While universal surveillance improved MRSA infection rates by more than half during the hospitalization and on 30-day follow-up, the effect did not extend further. The rate of methicillin-susceptible S. aureus infection was unchanged by the interventions.
The researchers said their findings would likely be generalizable to most U.S. hospitals because of the intermediate size and community-based nature of those included in the study.
However, "MRSA epidemiology and patient populations differ substantially across institutions and regions," Dr. Lautenbach said. "Each institution may need to tailor its intervention to its unique needs and resources."
Targeting high-risk populations such as nursing home residents for screening may be more cost-effective, he noted.
Furthermore, it's not clear that universal surveillance was responsible for the benefits, he said. "We need to know precisely which component of the intervention was most important."
At the same time as surveillance expanded to all admitted patients, hospitals in the study also started providing feedback on screening adherence, reduced turnaround time for MRSA detection, and started recommending treatment to reduce MRSA colonization.
These factors likely help explain the difference in findings compared with the Swiss study, Dr. Lautenbach said.
"While it's important to find people who are colonized with MRSA," he said. "It's not just finding it; it's what you do with that information."
Dr. Peterson reported receiving honoraria from Becton Dickinson; grants from Becton Dickinson, Cepheid, 3M, Roche, and Nanosphere; and pending grants from Becton Dickinson, Cepheid, 3M, and Roche. Co-authors reported consultancies for GlaxoSmithKline and Roche, honoraria from Becton Dickinson and Roche, grants from Roche, and pending grants from Becton Dickinson.
Dr. Lautenbach reported no conflicts of interest.
Primary source: Annals of Internal MedicineSource reference:Robicsek A, et al "Universal surveillance for methicillin-resistant Staphylococcus aureus in 3 affiliated hospitals" Ann Intern Med 2008; 148: 409-418. Additional source: Annals of Internal MedicineSource reference: Lautenbach E "Expanding the universe of methicillin-resistant Staphylococcus aureus prevention" Ann Intern Med 2008; 148: 474-476.

No comments: