"Testing high-risk groups isn't enough," he said. "You stop some [of the bacteria's] spread, but not enough" to really make a difference in hospital-acquired infections.
The Evanston Northwestern study describes a three-stage rollout of MRSA screening across Evanston Hospital, Glenbrook Hospital and Highland Park Hospital. First, epidemiologists tested all hospital patients to estimate the number of people who carry the drug-resistant bacteria on their bodies. That was 8.5 percent.
Then, the hospitals began testing every patient being admitted to intensive care units for MRSA. Those with the bacteria were placed in isolation, and special precautions were taken, including gowns and gloves for providers and rubdowns with disinfectants.
This is the regimen now required in Illinois for all hospitals under legislation passed last year.
To their surprise, scientists at Evanston Northwestern found this level of intervention didn't make a dent in the rate of hospital-acquired MRSA infections.
Peterson suspected the screening had been insufficient. Epidemiologists had found that patients colonized by MRSA were dispersed across the hospitals, not concentrated in a few units. And MRSA had begun circulating widely in the community, making it difficult for a targeted hospital-based effort to snuff it out.
To Peterson, the trends argued that testing should be expanded to all patients, not just intensive-care patients, and Evanston Northwestern became the first hospital group in the country to screen universally for MRSA in August 2005.
This time, the intervention worked, and hospital-acquired MRSA infections plummeted by 70 percent.
Given those findings, Peterson argues that the results of the JAMA study are not surprising. In that case, only half of the patients in a Swiss hospital's surgical wards were tested for colonization with MRSA. The other half were subject to normal infection control practices, and there's no evidence that patients elsewhere in the hospital were tested.
A reasonable conclusion is that perhaps 30 percent of the Swiss hospital's patients were screened for bacteria, Peterson said.
This wasn't universal screening, in other words. "I think their data says you can't just test certain units in the hospital and make an impact," Peterson said.
Proponents of universal screening will seize on the new article's findings to support their cause. But an accompanying editorial in the Annals of Internal Medicine sounds a note of caution.
It's still not clear which part of Evanston Northwestern's universal screening intervention worked, said Dr. Ebbing Lautenbach of the University of Pennsylvania School of Medicine. Was it the isolation of colonized patients and other precautions? Was it the use of antibiotic ointments to eliminate MRSA in patients who tested positive? Was it ongoing feedback to medical units with elevated infection rates?
Until further evidence comes in, "each institution may need to tailor its intervention to its unique needs and resources," Lautenbach concludes.