Reducing Transmission of Resistant Bacteria in ICUs

Published - Written by Ishani Ganguli

Health care-associated infections harm patients and consume limited resources. Such infections with methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococcus (VRE) have become a popular target for lawmakers and regulators to improve quality, and some prevention efforts have been codified. However, we still don’t have a good sense of which efforts are worthwhile. Two studies in this week’s NEJM attempt to tackle this question.

In 2007, after a promising four-year pilot study of their multi-pronged effort to decrease the rate of health-care associated MRSA infections, the Veterans Healthcare System in Pittsburgh directed acute care VA hospitals across the country to launch the same program. Over three years and across 150 hospitals, providers checked all new admissions and transfers for nasal MRSA, used contact precautions and hand hygiene, and promoted the idea that infection control was the responsibility of everyone who had contact with patients. Each month, they recorded the adherence to and impact of their efforts into a central database.


When Jain et al evaluated this program, with its reach of almost 2 million patients, they found that more were screened for MRSA at admission (82% to 96%) and more were screened at transfer or discharge (72% to 93%) during this period compared to pre-implementation. The rates of health care-associated MRSA infections in the ICU, stagnant for the two years before the trial, dropped from 1.64 infections per 1000 patient-days in October 2007 to 0.62 per 1000 patient-days in June 2010. Outside the ICU, rates fell from 0.47 to 0.26 per 1000 patient days.

In contrast to Jain et al’s real-world observational study, Huskins et al did a cluster-randomized, controlled trial of more than 9000 patients entering 18 ICUs. The intervention unit adopted universal gloving and contact precautions for patients known to be colonized or infected with MRSA or VRE, while the control unit continued with current practice.

Huskins and his team found that in the intervention units, providers were more likely to place infected or colonized patients  on barrier precautions and were more likely to wash their hands and use clean gloves than their control unit counterparts (a median of 47% of contacts vs 25%), but still not nearly as often as required by protocol. The researchers found no difference between the groups in the rates of MRSA and VRE colonization or health care-associated infections.

NEJM deputy editor Lindsey Baden, MD notes the difficulty in conducting a randomized controlled trial in the “frenetic environment” of an ICU. “How do you make sure everybody gowns and gloves appropriately? The high traffic in and out of patient rooms makes tracking and enforcing infection control practices quite difficult.”

And though the VA study intervention yielded a significant decrease in infection rates, he says, its design leaves question about just how this came about.  “Scientifically, I want to know which intervention made a difference. Was it the MRSA bundle? If so, which components were most important? Perhaps other practices that emerged over time significantly contributed to the decline in infections” Baden says. Unpacking this question would help hospitals best deploy their limited resources. At the same time Baden says, knowing that you’ve “markedly diminish a bad outcome” using basic techniques like good line management is a great start.

In an accompanying editorial, Richard Platt, MD of Harvard Medical School acknowledges the difference in results: “Taken together, these studies leave considerable uncertainty about whether the MRSA bundle, including routine surveillance cultures, is worthwhile in all settings. These studies do, however, underscore the importance of carefully evaluating the effect of existing state mandates to perform surveillance testing.”

What factors unique to the VA may have contributed to this program’s success? To what extent might these results be reproduced outside the context of the VA?  Which of these interventions could work in your health care workplace?