Infection Killed Almost 19,000 in 2005, Study Says
by Kevin Sack
The New York Times
October 16, 2007
ATLANTA — Nearly 19,000 people died in the United States in 2005 after being infected with a virulent drug-resistant bacterium that has spread rampantly through hospitals and nursing homes, according to the most thorough study to be conducted of the disease’s prevalence.
The study, which was published today in The Journal of the American Medical Association, suggests that invasive infections with methicillin-resistant Staphylococcus aureus, or M.R.S.A., may be twice as common as previously thought, according to its lead author, Dr. R. Monina Klevens. If the mortality estimates are correct, the number of deaths associated with M.R.S.A. each year would exceed those attributed to HIV/AIDS, Parkinson’s disease, emphysema or homicide.
By extrapolating data collected in nine locations, the researchers established the first true baseline for M.R.S.A. in the United States, projecting that 94,360 patients developed an invasive infection from the pathogen in 2005 and that nearly one of every five, or 18,650 of them, died.
The authors, who work for the Centers for Disease Control and Prevention, cautioned that their methodology differed significantly from previous studies and that direct comparisons were therefore risky. But they said they were surprised by the prevalence of the serious infections they found, which they calculated as 3.2 cases per 1,000 people.
In an accompanying editorial in the medical journal, Dr. Elizabeth A. Bancroft, an epidemiologist with the Los Angeles County Department of Public Health, characterized that finding as “astounding.” She wrote that the prevalence of invasive M.R.S.A. — when the bacteria has not merely colonized on the skin, but has attacked a normally sterile part of the body, like the organs or bloodstream — is greater than the combined rates for other conditions caused by invasive bacteria, including bloodstream infections, meningitis and flesh-eating disease.
The study also concluded that 85 percent of invasive M.R.S.A. infections are associated with health-care treatment. Previous research had indicated that many hospitals and long-term care centers have become breeding grounds for M.R.S.A. because bacteria may be transported from patient to patient by doctors, nurses and unsterile equipment.
“This confirms in a very rigorous way that this is a huge health problem,” said Dr. John A. Jernigan, the deputy chief of prevention and response in the C.D.C.’s Division of Healthcare Quality Promotion. “And it drives home that what we do in health care will have a lot to do with how we control it.”
The findings are likely to further stimulate an already active debate about whether hospitals and other medical centers should test all patients for M.R.S.A. upon admission. Some hospitals have had notable success in reducing their infection rates by isolating infected patients and then taking extra precautions, like requiring workers to wear gloves and gowns.
But other research has suggested that such techniques may be excessive, and may have the unintended consequence of diminishing medical care for sequestered patients. The C.D.C., in guidelines released last year, recommended that hospitals attempt to reduce their infection rates by first improving hygiene procedures and that they resort to screening high-risk patients only if other methods fail.
Dr. Lance R. Peterson, an epidemiologist with Evanston Northwestern Healthcare, said the Chicago-area hospital system reduced its rate of invasive M.R.S.A. infections by 60 percent after it began screening all patients in 2005.
“This study puts more onus on organizations that don’t do active surveillance to demonstrate that they’re reducing their M.R.S.A. infections,” he said. “Other things can work, but nothing else has been demonstrated to have this kind of impact. M.R.S.A. is theoretically a totally preventable disease.”
Numerous studies have shown that busy hospital workers disregard basic standards of hand-washing more than half the time. This week, Consumers Union, the nonprofit publisher of Consumer Reports, called for hospitals to begin publishing their hand-washing compliance rates.
“This study just accentuates that the hospital is ground zero, that this is where dangerous infections are occurring that are killing people every day,” said Lisa A. McGiffert, manager of the group’s “Stop Hospital Infections” campaign.
Though the C.D.C. estimates that M.R.S.A. represents only 10 percent to 20 percent of all infections acquired in health-care settings, the bacterium is feared for its opportunism and deadliness.
First isolated in the United States in 1968, it is resistant to a number of antibiotics and can cause infections of surgical sites, the urinary tract, the bloodstream and the lungs, leading to extensive and expensive hospital stays. The bacteria can be brought unknowingly into hospitals and nursing homes by patients who show no symptoms, and then takes advantage of weakened immune systems, incisions and wounds.
Of the infections studied by Dr. Klevens and her colleagues, 27 percent were considered to have originated during a patient’s current hospital stay. Another 58 percent were deemed to be associated with a previous hospitalization, nursing home stay, surgery or dialysis. Only 14 percent were cases without a defined health care risk factor, meaning the infection likely originated in the community.
A major difference with previous analyses is that the new study compiled actual confirmed cases of M.R.S.A. infection, rather than relying on coded patient records that sometimes lack precision. In the new study, higher prevalence rates and death rates were found for the elderly, blacks and men. The figures also varied greatly by geography, with Baltimore’s incidence rates far exceeding those of the eight other locations studied.
Dr. Klevens said further research would be needed to understand the racial and geographic disparities.
The C.D.C.’s latest estimate of all infections associated with health care, also taken from a study by Dr. Klevens, was 1.7 million cases and 99,000 associated deaths in 2002.