Preventing a Hospital “Horror Show”

In the ongoing battle between man and microbe, the antibiotic vancomycin is a last line of defense, the drug to turn to when few others will work.

The existence of a vancomycin-resistant strain of enterococci (VRE), therefore, spreads fear among hospital administrators. An outbreak of this “superbug” in one facility quickly can become a problem for other hospitals. 

“A VRE outbreak is a horror show for a hospital,” says Bruce Y. Lee, MD, MBA, an associate professor in International Health.  “Infections can be very stubborn, taking weeks, months or even years to eliminate.”

Lee’s team developed a sophisticated computer model that demonstrates how VRE can spread across a county and calls for new approaches to VRE management. The results were published in the August 2013 issue of the American Journal of Infection Control.

“The culprit behind VRE spread may be how frequently two facilities share patients, not how close they are geographically,” Lee says. “This makes it difficult to track the original source of an outbreak.” 

Lee’s analysis drew upon real patient data from 29 hospitals in Orange County, California. His study shows that a mere 10 percent increase in VRE at a single hospital can produce a nearly 3 percent increase in every hospital countywide. There are an estimated 20,000 to 85,000 cases of VRE infection each year in U.S. hospitals.

“The health care community needs to realize how hospitals are interconnected via sharing patients and, in the event of an outbreak, to expand vigilance and control efforts.” —Bruce Y. Lee

Hospitals and health care facilities often transfer patients for reasons that range from specialized care to insurance considerations. Lee’s study even took into account patients discharged at one facility before being readmitted by another. More than half of the patient transfers in his study fell in this largely overlooked category.

The researchers found additionally that in an outbreak, infection control specialists declare victory too soon, cutting short important control measures—such as heightened surveillance, contact isolation and better cleaning—that could halt the cascade of the infection surging through a region. “The health care community needs to realize how hospitals are interconnected via sharing patients and, in the event of an outbreak, to expand vigilance and control efforts much wider and for potentially longer periods of time,” Lee says.

One easily overlooked weapon against VRE is better inter-hospital communication, he says. Currently, hospitals know too little about the facilities they share patients with, and region-wide computer databases do not exist. Incentive programs for cooperation could help, as well.

“VRE control is every hospital’s concern. As long as one hospital in your region is struggling with VRE control, your hospital is at risk,” Lee says.

While Lee was at the University of Pittsburgh, he established a collaborative team at the Pittsburgh Supercomputing Center and University of California, Irvine, to develop the computational modeling software platform known as Regional Healthcare Ecosystem Analyst (RHEA). RHEA simulates health care facilities in a region and patients flowing among them and surrounding communities, allowing researchers to infect virtual patients with VRE.

“While this sort of computational modeling is fundamental in finance, meteorology and other fields, it is relatively new in public health,” says Lee, who joined the School’s International Vaccine Access Center in July as director of Operations Research.

“Improving the way health care products and services are administered can improve the lives of millions of people.”