CHARLOTTESVILLE, Va. -- The doctors tried one antibiotic afteranother, racing to stop the infection as it tore through the man's body,but nothing worked.

In a matter of days after the middle-agedpatient arrived at University of Virginia Medical Center, the stubbornbacteria in his blood had fought off even what doctors consider "drugsof last resort."

"It was very alarming; it was the first timewe'd seen that kind of resistance," says Amy Mathers, one of thehospital's infectious-disease specialists. "We didn't know what to offerthe patient."

The man died three months later, but the bacteriawasn't done. In the months that followed, it struck again and again inthe same hospital, in various forms, as doctors raced to decipher thesecret to its spread.

The superbug that hit UVA four years ago --and remains a threat -- belongs to a once-obscure family ofdrug-resistant bacteria that has stalked U.S. hospitals and nursinghomes for over a decade. Now, it's attacking in hundreds of thoseinstitutions, a USA TODAY examination shows, and it's a fight themedical community is not well positioned to win.

The bacteria,known as Carbapenem-Resistant Enterobacteriaceae, or CRE, are named fortheir ability to fight off carbapenem antibiotics -- the last line ofdefense in the medical toolbox. And so far, they've emerged almostexclusively in health care facilities, picking off the weakest ofpatients.

The bacteria made headlines this summer after a CRE strain of Klebsiella pneumoniaebattered the National Institutes of Health Clinical Center outsideWashington, D.C. Seven died, including a 16-year-old boy. (Hospitalsdon't reveal victims' names in keeping with medical privacy rules.) Butthat case was neither the first nor the worst of the CRE attacks.

USA TODAY's research shows there have been thousands of CRE casesthroughout the country in recent years -- they show up as everythingfrom pneumonia to intestinal and urinary tract infections. Yet evenlarger outbreaks like the UVA episode, in which seven patients alsodied, have received little or no national attention until now.

Thebacteria's ability to defeat even the most potent antibiotics hasconjured fears of illnesses that can't be stopped. Death rates amongpatients with CRE infections can be about 40%, far worse than other,better-known health care infections such as MRSA or C-Diff, which haveplagued hospitals and nursing homes for decades. And there are growingconcerns that CRE could make its way beyond health facilities and intothe general community.

"From the perspective of drug-resistantorganisms, (CRE) is the most serious threat, the most serious challengewe face to patient safety," says Arjun Srinivasan, associate directorfor prevention of health care-associated infections at the Centers forDisease Control and Prevention.

Since the first known case, at aNorth Carolina hospital, was reported in 2001, CREs have spread to atleast 41 other states, according to the CDC. And many cases still gounrecognized, because it can be tough to do the proper laboratoryanalysis, particularly at smaller hospitals or nursing homes.

Toassess the threat and what's being done to stop it, USA TODAYinterviewed dozens of health care authorities and reviewed hundreds ofpages of journal articles, clinical reports, and state and federalhealth care data. The examination shows:

• CRE infections alreadyare endemic in several major U.S. population centers, including NewYork, Los Angeles and Chicago, which account for hundreds of confirmedcases. Smaller pockets of cases have been reported across much of thecountry, including Oregon, Wisconsin, Minnesota, Pennsylvania, Maryland,Virginia and South Carolina.

• There is no reliable national dataon the scope of the CRE problem. The CDC has urged states to trackcases, but only a handful do so -- and they're just getting going. "Wedon't have enough ... data to tell what the trend looks like," saysStephen Ostroff, director of epidemiology at the Pennsylvania Departmentof Health. "All we know is that it is here."

• There is littlechance that an effective drug to kill CRE bacteria will be produced inthe coming years. Manufacturers have no new antibiotics in developmentthat show promise, according to federal officials and industry experts,and there's little financial incentive because the bacteria adaptquickly to resist new drugs.

• Many hospitals -- and an evengreater percentage of nursing homes -- lack the capacity, such as labcapability, to identify CRE, or the resources to effectively screen andisolate patients carrying the bacteria. And even when screening ispossible, there's a lack of consensus on whom to target.

"We'reworking with state health departments to try to figure out how big aproblem this is," says the CDC's Srinivasan, noting that his agency canpool whatever incidence data states collect. "We're still at a pointwhere we can stop this thing. You can never eradicate CRE, but we canprevent the spread. ... It's a matter of summoning the will."

Other experts are less optimistic.

"Myconcern is that there aren't a lot of methods in our tool kit that aresignificantly effective in curbing the spread of these infections," saysEli Perencevich, a professor and infectious-disease doctor at theUniversity of Iowa's Carver College of Medicine.

The spread ofCRE threatens to change the face of health care, crippling hospitalunits that specialize in treatments such as organ transplants andchemotherapy, which rely on the ability to control infections inpatients with weak immune systems.

If unchecked, "these (bacteria)are going to greatly impact the kind of surgeries (and) treatments wecan have," Perencevich says. "We're entering the post-antibiotic era;that's a very big problem."

Tracking an elusive killer

TheUVA epidemiologists knew their CRE outbreak would be tough to contain-- they'd read about other cases in medical literature and knew that thebacteria spread fast, with frighteningly high death rates.

But it quickly became clear that this case would be even more difficult than most.

Whenthe doctors began analyzing the bacteria in their first patient, who'dtransferred from a hospital in Pennsylvania, they found not one, but twodifferent strains of CRE bacteria. And as more patients turned up sick,lab tests showed that some carried yet another.

"We were reallyfrustrated; we hadn't seen anything like this in the literature," saysCosti Sifri, the hospital epidemiologist. "The fact that we haddifferent bacteria told us these cases were not related, but the shoeleather epidemiology suggested to us that all these (infections) camefrom the same patient. ... We realized we might be seeing a mobilegenetic event."

In other words, it looked like a single resistance gene was jumping among different bacteria from the Enterobacteriaceae family, creating new bugs before their eyes.

Thedoctors went back to the lab with even more urgency. It was January2008, five months after the first case turned up, and they'd identifiedfive patients harboring three distinct species of CRE.

Three of those patients already were dead.

Mobile patients; mobile bugs

Thereare many challenges to containing the spread of CRE, but one of themost daunting -- and immediate -- is figuring out where it's showing up.

There is no billing code for CRE infections under Medicare orMedicaid, the health care programs for the elderly, poor and disabled,and there's no federal reporting requirement for the infections. Sogetting a reliable national picture of prevalence or where cases areconcentrated is a challenge.

Based on academic studies and datafrom the handful of states and counties that require at least somereporting, it's clear that CRE is spreading fast. USA TODAY surveyedthose states and counties, and every one of them has found cases.

InLos Angeles County alone, a year of surveillance through mid-2011turned up 675 cases at hospitals, nursing homes and clinics. InMaryland, a 2011 survey by the state health department identified 269patients carrying CRE and estimated that up to 80% of the state'shospitals had seen at least one case during the year.

But the data are so isolated, and the reporting methodologies so varied, that the reports are of little practical use.

"Ifwe don't know the scope and we don't know the distribution -- how bigis the problem and where is the problem -- it's hard to know the nextpiece, which is what (prevention strategies) are you going to implementand where?" says Claudia Steiner, a physician and research officer atthe U.S. Agency for Healthcare Research and Quality.

It'sespecially important to know where CRE bacteria are emerging becausethey spread among patients who bounce between or among clinics, surgicalcenters, rehabilitation facilities, nursing homes and, of course,hospitals.

In the Chicago area, where scores of CRE infectionshave been found since 2008, studies show that about 3% of hospitalpatients in intensive care carry the bacteria, says Mary Hayden,director of clinical microbiology and an infectious-disease doctor atRush University Medical Center. Those same studies have found CREs beingcarried by about 30% of patients in long-term care facilities.

Notall of those patients are symptomatic: The bacteria can lurk, unseen,until a carrier's immune system is compromised or until the bug finds apath into the body and infection sets in. And as those patients movefrom one facility to another, the bacteria move with them, oftenclinging to caregivers' hands -- and moving to new victims.

"We have to think about a new approach, a regional approach, tocontrolling these organisms, because ... no facility is an island," Haydensays. If a nursing home patient is carrying CRE and gets sick in thenight, "the staff there just want to get him to a hospital," she adds."They may not know much about his (history), so that information doesn'tcome with him."

But the bacteria do.

The problem underscores the need for some sort of universal patientrecord system that can allow clinicians to see key aspects of apatient's medical history as that person moves among facilities, Haydensays. The technical hurdles and privacy concerns are challenging, sheadds, but some Chicago-area hospitals are working with public healthagencies to develop a model.

Meanwhile, the bacteria cycle from one facility to the next -- and back.

"It is continually reintroduced; I don't think it is going away,"says David Landman, an infectious-disease doctor at the State Universityof New York's Downstate Medical Center in Brooklyn. "You need extremecontrol efforts."

A new tracking plan

Back at UVA, the doctors' theory was proving correct: Theyidentified a common resistance gene among the different CRE bacteriaattacking the hospital, and it matched what they found in the initialpatient from Pennsylvania. The gene was jumping, one by one, to otherspecies of Enterobacteriaceae bacteria, creating new carbapenem-defying bugs.

The doctors were seeing, in real time, a phenomenon that had worriedresearchers for years: the ability of CRE to share resistance genesacross different members of the Enterobacteriaceae family.

The big fear is that the genes may start to convey resistance tomore common strains of the bacteria, turning routine illnesses, such asurinary tract infections, into untreatable nightmares. Worst-casescenario: Resistance could move to bacteria outside of health care, sopeople could pick it up in the community through something as simple as ahandshake.

The UVA doctors were in uncharted waters. Medical literature on CREssaid to look for resistance in certain types of Enterobacteriaceaebacteria, "but we were seeing it in all kinds of bacteria," saysMathers, the infectious disease specialist.

The doctors sent out an urgent set of new instructions: Patients sickened with any form of Enterobacteriaceae bacteriashould be checked immediately to see whether it iscarbapenem-resistant, even if it's a strain not normally associated withCRE infections.

"We told the lab to look at anything that has a possible link withthis (resistance) gene" that the hospital had identified, Mathers says."Any hint of resistance, then we need to know about it."

Stopping the untreatable

There's not much hope for a new treatment of CRE infections.

A few drugs show marginal effectiveness, including an old antibioticshelved decades ago because of high toxicity. And there's littleincentive for drug companies to invest in developing alternatives.Effective medications would be taken only until a patient recovered,making them far less profitable than life-long drugs for chronicillnesses. Plus, CREs develop new resistance quickly, so any newantibiotic isn't likely to last.

"If you look at the current pipeline of antibiotics (in development)... none of them really is going to be active against these bacteria,"says Gary Roselle, director of the Infectious Diseases Service for theDepartment of Veterans Affairs health care system.

"The reality is, (CRE infections) are remarkably difficult to treat,they often have bad outcomes ... and they're increasing nationally," addsRoselle, a doctor who oversees infection control for the VA's hundredsof hospitals, clinics and nursing homes. "I'm assuming this is going toget worse, and there likely won't be new antibiotics to treat it in thenear future, so the focus has to be on prevention."

CDC guidance for controlling CRE rests on traditional infectioncontrol strategy: rigorous hand cleaning by staff and visitors;isolating infected patients and requiring gowns and gloves for anyonecontacting them; cutting antibiotic use to slow the development ofresistant bacteria; and limiting use of invasive medical devices, suchas catheters, that give bacteria a path into the body.

But the measure that may hold the most promise is contentious:screening patients for the bacteria so carriers can be isolated. There'sdisparate opinion over who should be screened. Every patient? Onlythose whose history puts them at high risk for infection? Only thoseshowing symptoms?

Because many hospitals and nursing homes lack the resources to domuch screening, some patient advocates say the priority should belooking for more common bugs, such as MRSA (Methicillin-resistantStaphylococcus aureus), which is more treatable than CRE.

"Why cause hospitals to use resources for a pathogen with unknown(prevalence)?" says Michael Bennett, president of the Coalition forPatients' Rights. "Doesn't it make sense to attack the biggest problem?"

But screening has proved effective at facilities that have cut high CRE rates.

InNew York City, where CRE cases are endemic at many facilities,Bronx-based Montefiore Medical Center cut prevalence rates in halfacross its nine intensive care units with a program that relied heavilyon screening. The initiative tested all intensive-care patients using anexperimental, high-speed assay for the bacteria, and carriers wereisolated immediately.

The initiative, which grew to include sampling of patients acrossall units of Montefiore's three-hospital network, revealed that 40% ofMontefiore's CRE cases involved patients who had arrived with thebacteria when transferred from nursing homes and other institutions.

"So even if I had a perfect program to stop all patient-to-patienttransmission in the hospital, the maximum impact I could have would be a60% reduction in prevalence," says Brian Currie, the hospital's vicepresident for research and an assistant dean at the affiliated AlbertEinstein College of Medicine.

Currie sees the cut in Montefiore'sCRE rates as "a significant achievement," but he notes that theinitiative also underscored the trials ahead. The doctor and his staffidentified 11 nursing homes and several hospitals that regularly -- andunwittingly -- send CRE-infected patients to his facility.

"It's amazing how little awareness many of the providers have," he says.

The challenge at nursing homes, which typically have no labs andlimited ability to test patients for infectious bacteria, is even moredaunting.

"Personnel working in long-term care facilities may be unaware of'new' resistance (bacteria)," researchers concluded in a 2008 study ofCRE infections in New York nursing homes, published in Clinical Infectious Diseases. The risk of CRE in nursing home patients "should be of great concern."

New tools in the fight

Once the UVA doctors figured out that a single gene was driving thespread of CRE through the hospital, they still needed to figure out away to find it -- and stop it. And the clock was ticking.

By April 2008, eight months after they'd identified their firstinfection, 13 additional patients had been infected with related strainsof the bacteria. Seven were dead.

Back in the lab, the doctors figured out that the gene was hitching aride among bacteria on mobile pieces of DNA, called plasmids, that canmove from one cell to another. With more work, they developed a genetictest that could identify those plasmids -- and the bacteria they'daffected -- in days.

Traditional tests to identify the resistance-carrying plasmids can take months.

"Half the story is the outbreak and half the story is how we figuredit out," says Sifri, the epidemiologist. "We had to understand what washappening before we could attack the problem."

The lessons learned at UVA have helped them target CRE screening ofat-risk patients across the hospital, as well as those checking in. Andwith rapid identification and isolation of carriers, heavy vigilance onhand washing, and other infection-control measures, the hospital hasbeen able to control its outbreak, Sifri says.

But the bacteria are there to stay, lurking somewhere, invisible and always a threat.

"We have continued to have patients with CREs that are related tothis (first) event," Sifri says. "We haven't been able to close the dooron this. ... I'm not sure you ever can."