When a 43-year-old female lung-transplant patient arrived at the National Institutes of Health's (NIH) Clinical Center in June 2011, the hospital's infection-control team was on high alert. The woman carried a "superbug" resistant to all but two antibiotics, and the NIH staff wanted to keep the dangerous bacteria contained.

The patient was isolated in the intensive-care unit. Staff members donned gowns and gloves before entering her room. Her nurses cared for no other patients.

When she was discharged in July to return to New York, the NIH thought these measures had worked. There were no signs the bug — Klebsiella pneumoniae — had spread.

But a few weeks later, the hospital staff was "horrified," said Tara Palmore, an NIH infection-control specialist, when a second patient tested positive for the bacteria. A third and fourth soon followed. Those three patients eventually died as the bacteria grew impervious to every known antibiotic, even new experimental drugs.

The outbreak, not made public until now, was a minor disaster for the NIH, as its experts fought for months to control a superbug that was killing patients.

Over six months — and despite building literal and figurative walls to stop it — the bacteria hopped to 17 patients, 11 of whom died. Six of those deaths were attributed to the superbug, NIH staff reported Wednesday in the journal Science Translational Medicine.

The outbreak provides a case study of the dangers of the latest wave of hospital-bred bacteria and the extreme measures hospitals must adopt to stem the rising superbug tide. The research hospital now tests every patient transferring in from another facility for superbugs, tests every ICU patient twice a week and screens every patient in the 234-bed facility monthly for antibiotic-resistant bacteria.

"This was our introduction to antibiotic-resistant Klebisella," Palmore said. "We absolutely knew what it was, and we hoped we would never see it."

Nationwide, about 6 percent of hospitals are battling outbreaks of the class of superbugs known as carbapenem-resistant bacteria, which includes the NIH Klebsiella, said Alexander Kallen of the Centers for Disease Control and Prevention (CDC). These bacteria usually live harmlessly in our intestinal tracts. But in patients with immune systems weakened by, say, anti-rejection transplant drugs or chemotherapy, the bacteria can turn dangerous, gaining resistance to the most powerful antibiotics and seeping into the blood. That's what happened at NIH.

The CDC detected this bacteria in 2000. "Since then, we've seen it spread across the country," Kallen said, to 41 states.

Early in the NIH outbreak, hospital staff turned to advanced genetic technology to show that the outbreak had originated with the New York patient. By reading the DNA of the bacteria, the NIH scientists saw that the bugs from patients No. 2 and 3 were so closely related — differing at just one or two genetic letters out of 6 million — they must have come from the first patient, although from two places on her body.

"We eventually learned that they were all connected," said Palmore of the 17 patients.

In late August, the staff built a wall in the ICU and moved the Klebsiella-positive patients into a new, six-bed unit. Blood-pressure cuffs and other reusable gear were tossed after one use.

The hospital hired monitors to ensure doctors and nurses were donning gowns, gloves and masks and scrubbing their hands after examining the patients. If monitors fell asleep or otherwise shirked, Palmore fired them and found new ones. At one point, nine monitors were on duty.

Staff members took throat and rectal swabs from every ICU patient to sweep for the bug. When the bacteria were found in two patients outside the ICU, the sweep was broadened to every patient in the hospital.

They swabbed equipment, sinks, toilets and furniture. They hired a contractor to fumigate patient rooms with vaporized hydrogen peroxide, a potent disinfectant.

Still, they found Klebsiella on a ventilator that had been bleached. They found it in a sink drain in a patient's room. And they kept finding it in patients.

By January, the strict measures had paid off. No new cases have occurred since, Palmore said, although two Klebsiella-positive patients remain at the hospital.

NIH staff treated the patients who survived with colistin, a decades-old antibiotic that fell out of favor after it became apparent it can severely damage the kidneys. But as the superbug spread, it became resistant even to colistin. So the staff tried experimental antibiotics. None of those worked, either. Six patients eventually died from Klebsiella infections that spread to their blood.

Later this year, the CDC is launching a program in 10 cities to watch for hospital-borne outbreaks of Klebsiella and related superbugs.