Deadly infections

Last reviewed: March 2010
Carol Bradley sitting at home with her dog
Carol Bradley of Nashville, Tenn. (with her new dog, Teyla), developed infections after surgery for stomach cancer in August 2008.
Photograph by Kristina Krug

If there's one thing that all sides agree on in the health–care–reform debate, it's the need to dramatically reduce the number of infections that patients acquire in the hospital. But infection data newly released to the public show that although some hospitals in this country are doing an excellent job of protecting patients, others are not.

Our study focuses on one of the most dreaded types of the approximately 1.7 million infections that occur each year in U.S. hospitals. They are bloodstream infections introduced through the large intravenous catheters that deliver medication, nutrition, and fluids to patients in intensive care. These so-called central-line infections account for 15 percent of all hospital infections but are responsible for at least 30 percent of the 99,000 annual hospital–infection–related deaths, according to the best estimates available.

Even for those who survive, a central–line infection means weeks or months of debilitating treatments and side effects.

"I certainly did not expect to go through the torture that happened to me," said Carol Bradley, 61, a registered nurse from Nashville, Tenn., who developed several types of infections, including a bloodstream infection, after surgery for stomach cancer in August 2008. As a result, she spent about three weeks in intensive care hooked up to feeding and breathing tubes and central–line catheters, and more than a year on antibiotics.

For decades, doctors considered central–line infections an unavoidable risk of intensive care. But in the past few years, determined reformers have shown that hospitals can cut their infection rate to zero or close to it by following a low–tech program that includes a simple checklist. In short, central–line infections are almost completely preventable.

Citizen activists across the country, including those working with Consumers Union, the nonprofit publisher of Consumer Reports, have helped to enact laws in 27 states forcing hospitals to publicly disclose their infection rates as a first step, it's hoped, toward improving them if they are less than optimum. Sixteen of the states have made that information publicly available.

For our analysis, we collected and compared central–line infection data for intensive–care units at 926 hospitals in 43 states. (Among the nation's roughly 5,000 acute–care hospitals, about 3,300 provide intensive care, but in many, there were too few patients to yield statistically meaningful data, and many are still not reporting publicly.) Our information comes from the state reports and from the Leapfrog Group, a nonprofit organization based in Washington, D.C., that for the past 10 years has worked with large employers nationwide to collect and disseminate quality information on individual hospitals. The Leapfrog information, which the hospitals submit voluntarily, includes rates of central–line infections in ICUs.

We found enormous variations within the same cities and even within the same health–care systems. For example, among Kaiser Permanente hospitals in the Los Angeles area, Harbor City Medical Center reported no infections in the 1,769 days its ICU patients were on central lines in 2008. But Woodland Hills Medical Center reported 13 infections in 1,937 central–line days in its medical–surgical ICU—more than four times the average rate for such ICUs nationwide.

Our analysis adjusts for the fact that Leapfrog and the states have data from varying mixtures of ICUs, requiring comparisons to different average infection rates. For instance, the average infection rate is two per 1,000 central–line days for a coronary ICU, so a rate of four infections would be 100 percent more infections than average.

Poorly performing hospitals included some major teaching institutions. For instance, New York University Langone Medical Center in New York City reported 39 infections in 10,119 central–line days in 2008, roughly twice the national average for its mix of ICUs. The University of Virginia Medical Center in Charlottesville didn't do much better, reporting 77 infections in 18,572 days for the 15 months ending in September 2009, also about two times the national average.

More encouragingly, nationwide, we counted 105 hospitals whose most recent public reports tallied zero central–line infections. They ranged from modest rural institutions to urban giants such as the University of Pittsburgh Medical Center Presbyterian hospital, which reported no infections among patients who were on central lines a total of 13,596 days in 2008.

Why are central lines so vulnerable to dangerous infections? And how have some hospitals managed to cut their rate to zero or near–zero?