Hospitals' dirty secret

Patient safety at the nation's hospitals took a hit again as three more studies connected poor practices at hospitals with costly and often lethal infections in hospitalized patients.

Three groups of researchers studied hospital-acquired infections in reports released together last week that reaffirmed not only the deadly, costly and wasteful nature of these medical errors, but also the validity of a 2-year-old Pennsylvania public reporting program on hospital-acquired infections.

"It's 1992 New York all over again," said David Nash, editor in chief of the American Journal of Medical Quality, where the studies appeared in a 40-page special section. Nash was referring to the New York launch of the first-in-the-nation, statewide open-heart-surgery report-card program. "Look how far we've come," he said. "That was front-page news for a week, and Pennsylvania was right behind New York." Pennsylvania started its heart surgery reporting program in 1994.

The patient-safety "trifecta," as Nash called the three studies, were presented as "Hospital-Acquired Infections: Meeting the Challenge," a supplement to the November-December 2006 issue of the AJMQ, published by the American College of Medical Quality.

The studies showed:

• On average, central line-associated bloodstream infections at one Pittsburgh hospital cost more to treat than what was covered by hospital-acquired-infection-related payments, regardless of payer.

• The likelihood of surgical wound infections, as reported in the statewide Pennsylvania hospital-acquired-infections initiative, can be predicted to a statistically significant degree by patient acuity, but to an even greater degree by variance between hospitals.

• Across multiple diagnoses at admission, variance in mortality, length of stay and charges between hospital-acquired-infection cases and nonhospital-acquired-infection cases statewide in Pennsylvania cannot be explained by the commonly heard defense that "my patients are sicker."

Nash, the physician chairman of the Department of Health Policy at Jefferson Medical College of Thomas Jefferson University in Philadelphia, wrote in an introductory commentary that the AJMQ special section "provides some intriguing, albeit unsettling, information regarding the link between (hospital-acquired infections) and increased hospital costs, and makes it clear that it is the process of care, not the underlying clinical condition of the patient, that drives the current epidemic of" hospital-acquired infections. The reports, Nash wrote, "will stir emotion" and likely "will be met with skepticism by some, shock by others and incredulity by the remainder."

But only by bucking the conventional wisdom "that infection in the hospital is a byproduct of our day-to-day business, we will go a long way toward restoring the faith of the public in what we do, reducing the cost of our actions, and improving the overall quality of medical care in the United States."

The supplement is one of many recent efforts to tackle the problem. The Centers for Disease Control and Prevention released guidelines in October designed to help hospital staff better manage methicillin-resistant Staphylococcus aureus -- considered to be a "superbug."

The National Quality Forum, the standard-bearer for hospital and physician quality measures, said it is studying whether to add hospital-acquired infections to its list of "never" events. A report on the topic is expected sometime next year.

And the CMS proposed to eliminate reimbursement payments for infections that it deems preventable beginning in October 2008.

Key to the three studies was data from Pennsylvania. Beginning in January 2004, acute-care hospitals in Pennsylvania were required to report cases of hospital-acquired infections from orthopedic, neurosurgery and circulatory-system surgeries; central line-associated infections of the bloodstream; and urinary-tract infections from Foley catheters. The reports were gathered in a database maintained by the Pennsylvania Health Care Cost Containment Council. Reporting on 2004 hospital-acquired-infection data began last year.

Two weeks ago, 2005 data from 168 hospitals went online showing that 19,154 patients had acquired an infection that year while hospitalized in Pennsylvania. Of them, 2,478 died in the hospital, a mortality rate of 12.9% compared with a rate of 2.3% for all patients.

Hospital-acquired-infection patients' average length of stay was 20.6 days compared with 4.5 days for all patients, and their average charge was $185,260 compared with $31,389.

Marc Volavka, executive director of the council, said at a news conference last week in Washington that many physicians and hospital administrators who fought the Pennsylvania council's hospital-acquired-infection public-reporting program did so because they thought it would have little impact on the infection rate and would be too complicated for the public to understand.

But with the release, "no hospital went out of business," Volavka said. "No doctor went out of business."

Of the three AJMQ studies, the most dramatic and poignant was produced by researchers looking into the cost of 54 central line-associated bloodstream infections in the 28-bed medical intensive-care and coronary-care units at Allegheny General Hospital in Pittsburgh between July 2002 and June 2005.

Of the 54 central-line cases studied, only four of the least-complicated produced a positive gross operating margin for the hospital. Those results were more than offset by the often soaring costs of the remaining 50 cases. Payments for all 54 cases ranged wildly, from $4,546 to $299,318—costs even more so, from $15,565 to $353,205. Bottom line, a central-line case yielded an operating loss for the hospital of $26,839 on average and $1,449,306 in total.

The toll on patients was even more grim. Twenty-two of the 54 central-line-infection patients died. Only nine went home, with the bulk of the survivors being discharged to long-term-care or other facilities.

To humanize the numbers, the report included five vignettes, including a case study of a 37-year-old video programmer who was admitted with acute pancreatitis and hypertriglyceridemia. He had a femoral venous catheter placed for treatment, the tip of which later was found to be colonized with a drug-resistant strain of staph bacteria. On his fifth day in the hospital, he started running a fever, his blood pressure dropped and his breathing began to labor and fail, landing him in the ICU on a ventilator. Then his kidneys faltered, requiring dialysis; he developed abdominal abscesses, and after 19 days on the ventilator, underwent a tracheostomy.

Eighty-six days after admission, he was discharged to a nursing home. Even though he was insured by "a large commercial payer," the hospital received $200,031 for his care against $241,844 in costs, an operating loss of $41,813.

Data from the Allegheny Hospital study were "in keeping" with those gathered by the Pennsylvania council statewide, showing that hospital-acquired infections were more costly "by a factor of six," the report said. In addition, "we did not find any significant relationship between different payers and the magnitude of operational losses." Physician researcher Richard Shannon, lead author of the Allegheny study, said going in, "There was a tacit assumption that hospitals were making money on these cases because of outlier payments." "I think no one realized … that a large amount of money gets consumed in these infections, and that money gets tied up that we could use to do other things that we cannot now do," said Shannon, who is now senior vice chairman of the Department of Medicine at the University of Pennsylvania Health System, and was previously chairman of the department of medicine at Allegheny General Hospital.

Another of the studies, "Factors Associated with Risk of Surgical Wound Infections," attempted to determine to what degree patient acuity played in the matter. Looking at risk-adjusted surgery data for selected procedures from the Pennsylvania council, the study found that generally there was a statistically significant but relatively "poor" association between the condition of the patient at admission and the incidence of hospital-acquired infections. But when specific hospitals were added as a variable, "the prediction of infection was improved by 23% to 33%" depending on type of surgery, the study found.

"The patient factors are statistically significant, and there is no way to get around that," said Christopher Hollenbeak, lead author of the report. "Our data show those factors that are attributable to hospitals are much more significant," said Hollenbeak, a health economist and associate professor of surgery and health evaluation sciences at Penn State College of Medicine, Hershey, Pa.

The last of the three studies similarly looked at patient-level data from the Pennsylvania council and from data-mining firm MediQual, Marlborough, Mass., to try and determine if sicker patients were skewing the data. "The question that was first raised with Pennsylvania's report was the people (with hospital-acquired infections) are very ill, so of course they had higher mortality. It comes free with the territory," said lead author Richard Johannes, vice president of clinical affairs at MediQual, a division of pharmaceutical wholesaler Cardinal Health. What the MediQual study tried to answer was whether this was so, he said. The study's thoroughly descriptive title, "Adverse Outcomes From Hospital-Acquired Infection in Pennsylvania Cannot Be Attributed to Increased Risk," gives the answer. "It tells you that even after you select the sickest of the sick, you still have this increase in the hospital-acquired infections," Johannes said. The report concluded that "Reducing factors that contribute to the development of (hospital-acquired infections) can save lives, decrease costs and save valuable healthcare resources."

The Association for Professionals in Infection Control and Epidemiology said it was pleased that the studies looked at the economics of hospital-acquired infections -- especially how much could be reinvested back into a quality infection -- control program if they were greatly reduced or eliminated.

Brent James, director of the Institute for Health Care Delivery Research at Intermountain Healthcare in Salt Lake City, said he was not surprised Shannon's research showed a financial benefit would accrue to Allegheny from reducing central-line infections, but based on more than two decades of experience in various quality improvement projects at the institute, unfortunately, that isn't the rule.

At the institute, "What we see consistently is that about three-quarters of the time that when you save money, your costs drop, but your revenues drop even more," James said.

-- with Matthew DoBias

This article initially appeared in the Nov. 27 edition of Modern Healthcare magazine.

What do you think? Write us with your comments at hitsdaily@crain.com. Please include your name, title and hometown.