A Mom Suspected SpinachAs public health officials looked for the source of lethal E. coli, the mother of two sick children began to focus on their salad.
By Maria L. La Ganga
Times Staff Writer
October 9, 2006
MILWAUKEE — On a Sunday afternoon early last month, 3-year-old Caroline Zientek began to complain. Her stomach hurt, she said, and she had diarrhea.
She didn't get better. By late Tuesday night, the normally happy toddler was shrieking in pain. And on Wednesday at 5 a.m., her mother, a nurse, was worried enough that she packed her daughter into the family's blue Nissan and sped through the darkness to Children's Hospital of Wisconsin.
"I think it's food poisoning," Ana Maria Zientek told the emergency room nurse, describing the waves of cramping that had her daughter doubled over in pain every 20 minutes.
When Dr. Jane Machi examined Caroline at 7:30 a.m., she found her lethargic, clingy, seriously dehydrated and unable to urinate. She had dark circles under her eyes.
Machi ordered a stool sample and a series of blood tests. Without waiting for the lab results, she started Caroline on intravenous fluids and admitted the child to the hospital.
Machi had seen these symptoms before. The last time, the diagnosis had been Escherichia coli O157:H7.
E. coli bacteria ordinarily live harmlessly in the digestive systems of animals that chew their cuds. But after the pathogen is excreted in feces, it occasionally ends up in human food. It can be spread from a cow's intestines to meat during processing. Water and passing birds can spread it to fields where crops are grown.
E. coli has many forms, and O157:H7 is one of the most virulent. Particularly in young children and the elderly, it can sometimes cause hemolytic uremic syndrome, in which red blood cells break down and platelets clump up in the kidneys, causing them to fail. Many patients with the syndrome end up on dialysis. Some die.
When doctors told her their suspicions, Ana Maria Zientek knew her daughter was in grave danger. What neither Zientek nor the doctors knew Sept. 6 was that Caroline was not alone. Physicians around the country had begun seeing patients with bloody diarrhea and abdominal cramps.
But no one knew where the E. coli was coming from. For Ana Maria Zientek, that became a galvanizing question.
A day after Caroline Zientek was admitted to Children's Hospital, the medical director of a nearby blood center called Dr. Jeffrey Davis, Wisconsin's chief medical officer for communicable diseases.
He told Davis that five adults in separate hospitals in the area had undergone a blood-cleansing procedure called plasma exchange over the previous few days. They all had hemolytic uremic syndrome, with its signature kidney failure and anemia.
Davis was immediately concerned. It had been 20 years since the veteran scientist had last seen so many patients with hemolytic uremic syndrome in so short a time in such a small region.
E. coli infections had been confirmed in three of the patients. "I asked if it was O157:H7. He said yes," Davis recalled. "This was a very powerful sentinel event."
By itself, a small outbreak of E. coli O157:H7 is neither unusual nor alarming. Wisconsin sees about 200 cases each year. Summer months bring the most infections, because people swim in lakes and rivers, where the microbes can be found, and they eat undercooked hamburgers at backyard barbecues.
But because of the hemolytic uremic syndrome and other reports of O157:H7 from around the state, Davis had a bad feeling.
He set his staff to work, calling hospitals across the state for more information. From Children's Hospital of Wisconsin, they learned about Caroline Zientek and another child, both now confirmed to have the bacterium.
By the end of the day, microbiologists at the state hygiene lab had teased out the DNA fingerprint in stool samples given by each of the patients. More than 3,000 unique strains of E. coli O157:H7 have been identified. In Wisconsin, eight patients had identical forms, suggesting a common — but unknown — source of infection.
Wisconsin public health officials knew they had a serious problem. They also had a responsibility to alert health officials in other states in case the outbreak was larger than they knew.
So on Friday, Sept. 8, microbiologist Linda Machmueller sat at her computer in the Wisconsin State Laboratory of Hygiene in Madison and posted a terse message on PulseNet, a federal Web board run by the Centers for Disease Control and Prevention that allows scientists around the country to communicate about possible disease outbreaks.
"Wisconsin has a cluster of 8 E. coli O157:H7," she typed, including seven local cases and one from Illinois. They all appeared to "match the pattern" for a strain of the organism that the U.S. Department of Agriculture had isolated earlier in hamburger patties from Texas. The microbiologist attached a copy of the deadly organism's DNA fingerprint.
She had no idea what would happen next. "You never know when you post these things if it's going to amount to anything," she said.
Spreading the Word
The Wisconsin posting landed on PulseNet at 5:14 p.m. EDT, after everyone at the Web board's Atlanta headquarters had gone home for the weekend. So it wasn't until Monday, Sept. 11, that database manager Molly Joyner read the brief note, checked the DNA fingerprint and began trying to figure out what was going on.
PulseNet, as Wisconsin's Davis puts it, is a little "like a dating service for bacteria." It allows public health labs throughout the country to compare the organisms they're seeing with those being found in other states.
The bacterium isolated in the Wisconsin outbreaks was not a highly unusual strain. Two or three cases a week are commonly posted on PulseNet.
But by the end of the day on Sept. 11, Joyner had discovered that nine states had posted single matching E. coli samples to PulseNet in the weeks leading up to the Wisconsin cluster, although it was unclear if they were connected. And Minnesota e-mailed that afternoon with yet another match.
This was clearly a national outbreak.
'It's the Spinach'
At Children's Hospital in Milwaukee, the Zientek family had moved in for the long haul, to a double room on the eighth floor.
Three days after Caroline was admitted, her older brother David, 6, was rushed to the hospital with kidney failure from a severe case of E. coli. The next day, Caroline's condition worsened and her kidneys also began to fail. There was little the doctors could do but watch and wait for the children to get better — or not.
"As a parent, it was the most agonizing thing," Zientek said. "I asked the doctor, 'When will the time be that the body decides what it's going to do? How long are we going to sit on our hands?' "
During long hours at the hospital, Zientek tried to puzzle out how her children had become ill.
Health officials kept asking her about beef, a logical question given that it is the source of many E. coli outbreaks.
But the salad-loving self-proclaimed "fanatic about kitchen cleanliness" had all but banished beef from the family diet years ago. She'd decided it was too risky after watching a documentary about "deadly bugs" that profiled an especially virulent strain of E. coli O157:H7 and its common link to ground beef.
She wondered if her kids could have picked up the bug hunting for frogs in the creek near their house. She also wondered about something else. Her brother in New Mexico had found an article on the Internet linking E. coli and leafy greens. Zientek couldn't stop thinking about a bag of Dole baby spinach in her refrigerator, which she had purchased Aug. 25 and served in a salad.
On that Monday, Zientek spoke to the mother of another child with O157:H7 on the same floor of the hospital.
The woman had found the same article on the Internet. "We looked at each other and said, 'It's the spinach,' " Zientek recalled.
'Green Leafy Stuff'
At 5:30 p.m. on Sept. 12, Melissa Plantenga picked up the phone at the Oregon Public Health Division and canceled dinner plans with a friend: "I've just started working on this outbreak," she said, "and I think it's going to be big."
Although Oregon public health officials were unaware of the E. coli chatter starting to bubble up over PulseNet because of computer problems, what they'd seen in their own state over the last few days was enough to start them thinking.
Four days earlier, the state public health lab had reported three matching cases of E. coli O157:H7. Now there were three more. It was time for Plantenga to get to work.
Plantenga's job was to pinpoint the source of the E. coli. Although she often describes her duties as monitoring "all of the stool samples that come into the Oregon state public health division," the 29-year-old research analyst is actually a disease detective.
During five hours of telephone interviewing that night, Plantenga probed the minutiae of the Oregon victims' diets: Did you eat at any fast-food restaurant? Airport kiosk? Where did you buy the food you ate in the seven days before you got sick? Which bakery? Any ethnic specialty stores? Any eggs, dairy, meats, poultry, vegetables, fresh produce?
By the time she went home, she had talked to four of the six patients, including one who answered her questions from a hospital bed. Back in her office Sept. 13, Plantenga reached the other two patients.
A pattern had begun to form. Five of the six had eaten fresh spinach in the week before E. coli struck.
On learning the news, William E. Keene, an epidemiologist who directs Oregon's outbreak investigations, spent his lunch hour sending e-mails to other scientists in the close, clubby world of epidemiology. The first one went out at 1:14 p.m.: We have a cluster of six E. coli O157:H7 cases with matching DNA fingerprints. Pursuing possible leads. Do you have anything similar?
Dr. Christopher R. Braden, who leads food-borne outbreak investigations at the CDC, responded in less than 30 minutes.
He had discussed the Wisconsin cluster with officials there earlier in the day. "The pattern you have looks similar to my untrained eye," Braden wrote. "No hypothesis yet, but WI cases tend to be female, in 20s to over 40 … thinking green leafy stuff."
Thirteen minutes later, Keene responded with a one-sentence note: "We think it might be bagged spinach."
No Smoking Gun
By Sept. 14, federal officials had a tough call to make. The CDC had confirmed that identical strains of E. coli O157:H7 had sickened 50 in eight states, including Wisconsin, New Mexico, Utah and Oregon, and had killed an elderly woman.
During the first major conference call about the outbreak, public health officials from the CDC and afflicted states laid out their evidence to the Food and Drug Administration that spinach was the source of the deadly bacterium.
But no one had discovered a smoking gun: a bag of the leafy green vegetable from a patient's refrigerator from which microbiologists had actually grown the correct strain of E. coli.
Jeffrey Davis in Wisconsin reminded his fellow health officials of the strawberry scare of 1996. In that case, hundreds of people in nine states and Toronto were sickened by an outbreak of cyclospora, a parasite that causes intense diarrhea, weight loss and fatigue.
Epidemiologists in Houston, where more than 100 had fallen ill, fingered California strawberries. The real culprit, it later turned out, was raspberries from Guatemala. But by the time scientists realized their mistake, Golden State growers had lost between $20 million and $40 million.
"The strawberry industry has never forgiven us," Braden said.
The stakes were high.
If they waited to warn the public until they were absolutely certain that spinach was the culprit, more people could get sick and die. But if they went public and were wrong, an innocent industry could be devastated.
When the conference call ended, no decision had been made about how to move forward.
Dr. David Acheson hung up the phone convinced that his agency, the FDA, needed to move swiftly. The conference call had been the FDA's first chance to collect data on the widening outbreak, and the news was grim. There was no time to waste.
Hours later, the FDA called an after-hours telephonic news conference. "We're not advising people to cook product that may have E. coli," Acheson told reporters across the country. "What we're advising people is, don't eat it."
Davis and his colleagues in Wisconsin knew nothing about Acheson's announcement until it was over, and they spent the next several hours answering phone calls from the media.
"We didn't know that they were going to really do that and come out as strongly as they did," said the CDC's Braden. "We thought they might want another day or so of information."
Back in Milwaukee, things were looking up. Although Caroline Zientek was still being cared for at Children's Hospital, David had been released and was recovering from kidney shutdown at home.
A few days earlier, Ana Maria Zientek had turned over a bag of Dole baby spinach to the Milwaukee Health Department.
On Sept. 14, as state and federal officials conferred about the growing outbreak, she was calling the department to find out what had happened to the produce.
No one at the department had any information. But three hours later, her suspicions were confirmed.
"The story broke on the national news," she said. "I about fell off my chair."
On Sunday, Sept. 17, during what had become a daily conference call of health officials across the country, New Mexico epidemiologist Chad Smelser announced a breakthrough.
With the number of infected people at 109 and spinach a distant memory on supermarket shelves, things had been moving quickly. Based on the packaging of suspect spinach, health officials believed that at least some of the contaminated crop had been packaged by Natural Selection Foods in San Juan Bautista, Calif. But researchers still needed that smoking gun: a spinach sample with the strain of E. coli implicated in the outbreak.
Health workers around the nation had collected dozens of bags of spinach from the refrigerators of the victims. Growing E. coli O157:H7 in a lab, however, is a tricky business.
A food sample is processed in a soup of antibiotics and mold suppressants to stop other organisms from growing and allow the E. coli O157:H7 to flourish. It is then placed in a petri dish or test tube to grow. But one contaminated leaf doesn't mean the entire bag is contaminated. And the best tests often come back with false negatives.
Microbiologists at the state lab in Albuquerque had been trying since Friday to coax E. coli O157:H7 to grow in spinach retrieved from a patient's refrigerator. At last, Smelser was able to share some good news: "It looks like we're growing something that's going to be E. coli."
Cheers went up from those listening to the call.
Two days later, PulseNet scientists confirmed that the E. coli grown in Albuquerque matched the strain at the heart of the outbreak. Ultimately, the organism was found in 13 bags of spinach from 10 states, including two from Wisconsin, the hardest hit state. All had the same brand.
"The circle was complete, said Patricia Fox, head of Wisconsin's communicable disease epidemiology unit. "It doesn't always happen in epidemiology."
As of Friday, three people had died from eating tainted spinach. One hundred and ninety-nine in 26 states were sickened during the outbreak. Officials still have not discovered how the spinach packaged by Natural Selection came to be contaminated, and they may never know. But they are confident the outbreak has been contained.
Caroline and David Zientek are still anemic from battling kidney failure, but they are home and their health is improving. Their doctor recently cleared them to fly to Florida for a family wedding.
But there has been emotional fallout from their battles with E. coli. Caroline is back in diapers and wants to sleep in the same room as her mother. Getting either child to the doctor these days is "a war," Ana Maria Zientek said, because "they were so poked and prodded."
As for the nurse who suspected what made her children sick before the nation even knew there was an outbreak? She's feeling largely vindicated. And she no longer serves her family raw vegetables.
"No one would believe the motherly instinct," said her husband, Paul, a tall, lanky Milwaukee police officer. "My friends are all calling her Columbo."
Tracking a lethal bug
Public health sleuths sought the cause of E. coli infections as the number of cases rose. By Friday there had been 199 confirmed cases and three deaths.
Key points in seeking the source-
Sept. 5 —Wisconsin public pealth officials discover a cluster of fiveE. coli O157:H7 cases. A livestock barn at a county fair is suspected.Sept. 8 —Wisconsin State Laboratory of Hygiene posts the DNAfingerprint of the E. coli strain on PulseNet along with a messageabout eight cases. Hamburger patties from Texas are suspected.Sept. 13—Health officials from Wisconsin, Oregon and the Centers forDisease Control and Prevention talk for the first time about apossible national outbreak. Spinach is believed to be the culprit.Sept. 14—After a wider conference call, U.S. Food and DrugAdministration advises consumers not to eat bagged fresh spinach.Sept. 15—National Selection Foods of San Juan Bautista, Calif.,recalls all products containing spinach that have use-by dates fromAug. 17 through Oct. 1.Sept. 19—Microbiologists in New Mexico confirm that E. coli grownfrom an open bag of spinach retrieved from a patient's refrigerator isidentical to the strain that had sickened 131 people and killed onewoman up to then.Sept. 20—A 2-year-old boy dies in Salt Lake City, becoming thesecond fatality.Oct. 5—Nebraska health officials report that an elderly woman'sdeath was caused by E. coli.
--Sources: Centers for Disease Control and Prevention, Times reporting
More on this outbreak: Dole and Natural Selections Spinach E. coli Outbreak