Level 2: Mid-level evidence
Healthcare-associated methicillin-resistant Staphylococcus aureus (MRSA) infections continue to be a serious problem in hospitals around the world, causing thousands of deaths every year.
Many institutions have enacted containment programs to reduce the rate of MRSA infections. A recent study reports on the efficacy of one such strategy implemented at all Veterans Affairs (VA) acute-care facilities in the United States (N Engl J Med. 2011;364:1419-1430). The VA “MRSA bundle” includes universal nasal surveillance for MRSA colonization in patients, contact precautions for patient carriers of MRSA, procedures for hand hygiene, and an institutional culture change making all personnel that come in contact with patients responsible for infection control. A total of 1,934,598 hospital admissions, transfers, or discharges (365,139 to intensive care units [ICU]) were analyzed from inception of the program in October 2007 through June 2010.
During a two-year period preceding the MRSA bundle program, there were no significant differences in monthly rates of MRSA infections in VA ICUs. Following implementation, the monthly MRSA infection rate fell significantly from the beginning of the program to the end of the study period.
Rates per 1,000 patient-days decreased from 1.64 to 0.62 for ICU patients (P <0.001) and from 0.47 to 0.26 for non-ICU patients (P <0.001). Rates of MRSA transmission also fell from 3.02 to 2.50 for ICU patients (P <0.001) and 2.54 to 2.00 for non-ICU patients (P <0.001). Over the study period, screening rates increased from 82% to 96% at admission and from 72% to 93% at transfer or discharge.
Another cluster-randomized trial evaluated a similar program including surveillance and expanded use of barrier precautions in 18 ICUs for six months (N Engl J Med. 2011;364:1407-1418). In that trial, there were no significant reductions in either MRSA colonization or infection (study also investigated vancomycin-resistant enterococcus infections).
There were a number of important differences between the two studies: The VA study was of longer duration, allowing time for changes in behavior to take full effect. Also, the VA study included hospital-wide surveillance, while the other was limited to ICUs. It is possible that infection control practices in the ICU setting were already sufficient to minimize infection rates, such that additional improvement could be hard to demonstrate.
Alan Ehrlich, MD, is an assistant clinical professor in family medicine at the University of Massachusetts Medical School in Worcester.