A rise in the number of Escherichia coli cases requires diligent detection efforts.
By Debby Giusti, MT(ASCP)
Page 1
Ten-year-old Brianne Kiner spent 40 days in a coma in 1993, while teams of medical personnel worked round-the-clock to keep her alive. Brianne has little memory of the 118 days she was on kidney dialysis or the 80 units of blood she received, nor does she recall the numerous times the doctors told her mother that Brianne wouldnít live through the night. What Brianne does remember is that her hospital ordeal left her with the dubious recognition of being the sickest child in the United States to survive Escherichia coli 0157:H7.1
Over a 3-month period, more than 700 children and adults in four states in the northwest became ill after eating at various Jack in the Box restaurants. They suffered severe stomach cramps and diarrhea, often bloody, and close to 200 of the ill had to be hospitalized. Fifty-five cases progressed into hemolytic uremic syndrome (HUS), a condition that can lead to kidney failure and even death. Children and the elderly are most at risk for HUS, and in the 1993 outbreak, four children died.2
Epidemiologists quickly recognized that those infected had eaten undercooked hamburgers served at more than 90 Jack in the Box restaurants in the four state area.2 The beef shipped to the restaurants was found to be contaminated with E. coli 0157, and to date, the outbreak remains the largest in U.S. history caused by the organism.

In the week of her exposure, Brianne had visited the same Jack in the Box twice, eating a hamburger each time. The double dose no doubt compounded the severity of her case. Some speculate that if the burgers had been cooked only 30 seconds longer, Brianne would not have become ill.
Spending nearly 2 months in the pediatric intensive care unit of Seattle Childrenís Hospital and Medical Center and a total of 189 days in the hospital, Brianne suffered three strokes, 10,000 seizures and every organ in her body failed. Her initial hospital bill totaled $1.2 million.1 No matter what the cost, her mother calls her daughterís survival a miracle. The medical staff believes Brianneís age helped her win the fight against the terrible pathogen.
Heightened Awareness
Brianneís story and the stories of so many affected by that first major outbreak did much to raise public awareness of the risk of E. coli 0157:H7. Consumers were advised to cook ground beef thoroughly and to use care to prevent cross-contamination of cooking surfaces and utensils. Laboratories began to test for the pathogen, and the Food Safety and Inspection Service of the United States Department of Agriculture (USDA) initiated a program to check ground beef. America seemed to breathe a sigh of relief. With proper precautions, future outbreaks could be prevented ó or so it seemed.
E. coli was first identified in 1885 by German bacteriologist Theodor Escherich.3 The gram negative rod is a normal enteric organism found in the intestine of both humans and animals. A facultative aerobic, it can be grown both aerobically and anaerobically, and in fact, when aerobic methods are used, E. coli is the dominant organism found in feces.4
The organism colonizes the human bowel shortly after birth and usually does not cause disease, but suppresses the growth of harmful bacteria and aids in the synthesis of vitamins. A relatively few strains of E. coli are pathogenic, and of those, enterohemorrhagic E. coli 0157:H7 elicits the most profound response, producing a verotoxin or shiga-like toxin identical to that of Shigella dysenteriae. The toxin attacks the epithelial cells of the intestine and damages the blood vessels, thus producing hemorrhage. The toxin is thought to be the reason some cases develop HUS.
Healthy cattle are the main reservoir of E. coli 0157, although other ruminants, such as sheep and deer, also carry the organism. Even dogs, birds, chickens and pigs have been found to harbor E. coli. The organism is excreted in the feces and can survive for long periods in fecal matter. A remarkably low dosage ó from 10 to 100 cells ó can cause disease. 5 Thus, anything that comes in contact with cattle dung or manure can be contaminated, such as farm produce, or lakes and streams by runoff water and even buildings that house cattle.
Person-to-person contact is also of concern. The organism is sloughed from the human body for up to 14 days after the onset of diarrhea.6 In 1982, an outbreak of E. coli 0157 in a nursing home in Canada killed 19 residents and proved that the organism can be deadly in institutional settings, such as day care facilities and nursing homes.7
Within 9 days of exposure, E. coli 0157:H7 causes hemorrhagic colitis with severe stomach cramps and bloody diarrhea. Patients usually have little or no fever and recover in 5 to 10 days. 6
Children under the age of 5 are highly susceptible to HUS, which is the leading cause of kidney failure in children in the United States.6 The elderly are also at risk, often developing thrombocytopenia and hemolytic anemia as well as severe renal complications. Of the children affected, 3% to 5% will die, and as many as 30% will have lasting medical problems.8 Dialysis, transfusion of packed cells and platelets may be indicated, and in extreme cases renal transplant may be required.
Antidiarrheal agents such as loperamide (Imodium) have been shown to be associated with an increased risk of HUS, perhaps by prolonging the expulsion of the organism from the intestinal tract and, therefore, increasing toxin absorption.
In 1999, evidence presented at the 35th United StatesñJapan Cholera and Related Diarrheal Diseases Conference indicated antimicrobials should not be given to children with E. coli infections. The data revealed that 15% of all North American children infected with E. coli 0157 develop HUS, and it appears that antibiotic treatment increases that risk. Bacterial lysis due to antibiotic therapy is thought to cause the release of the toxin in the intestine, making more toxin available for absorption.9
Studies are being conducted to evaluate compounds that could be taken orally by patients that would absorb the toxin from the intestine and reduce the incidence of HUS. However, the treatment would have to be given early in the disease, and since it takes 1 to 2 days to diagnose E. coli, the window of opportunity is short.
Birth of a Pathogen
E. coli 0157:H7 was first associated with a sporadic case of hemorrhagic colitis in 1975, and the first outbreaks occurred in 1982. A report in the The New England Journal of Medicine the following year described two bouts of gastrointestinal illness that caused 47 people in Oregon and Michigan to have ìsevere crampy abdominal pain, initially watery diarrhea followed by grossly bloody diarrhea and little or no fever.î Investigators narrowed their search of a cause to beef patties, rehydrated onions and pickles that the patients had eaten at a local fast-food restaurant. Three-fourths of the stool specimens collected were found to contain what the authors termed ìa rare Escherichia coli serotype, 0157:H7, that was not invasive or toxigenic by standard tests.î The organism was also found in beef from the lot shipped to the restaurant.10
That same year, a report in Lancet noted the presence of a shiga-toxin-producing E. coli in the stools of patients with HUS. Both reports led scientists to believe they were dealing with an enteric pathogen that could cause both intestinal and renal disease.
Shortly thereafter, in an attempt to determine whether E. coli 0157 was indeed a newly emerging infectious agent or one that had been previously overlooked owing to testing limitations, the Centers for Disease Control and Prevention (CDC) reviewed more than 3,000 E. coli strains serotyped from 1973 to 1983 and uncovered only one E. coli 0157:H7 isolate.8 The Public Health Laboratory in the United Kingdom and the Canadian Laboratory Centre for Disease Control conducted similar reviews. The results substantiated the theory that the medical community was dealing with a newly emerging strain of E. coli.
Foodborne Outbreaks
After the Jack in the Box outbreak, E. coli 0157 began to cause disease with surprising regularity. A total of 16 outbreaks were reported in 1993 alone.2 In August 1994, three outbreaks occurred with summer campers who grilled over the open fire and failed to adequately cook their hamburgers.11 The next year, the New Jersey Department of Health noticed sporadic cases, again due to ground beef.
In addition to undercooked hamburger, cheese curds, nonpasteurized milk and even salami were also associated with outbreaks. By 1995, 33 states had enacted legislation declaring E. coli 0157 a reportable disease.12 The same year, CDC officially instituted PulseNet, a national network of public health labs that perform pulsed-field gel electrophoresis, a type of ìfingerprinting,î on foodborne bacteria for surveillance purposes.
In 1996, the pathogen found a new vehicle for transmission. Three-year-old Anna Girand was one of many children who drank nonpasteurized Odwalla apple juice that fall and then developed diarrhea with stomach cramps. But Anna continued to drink the juice when the doctor advised her parents to ìpush liquids.î Eight days after her initial diarrhea, Anna was hospitalized with a diagnosis of E. coli 0157:H7 that progressed to HUS.
Whether the organism was introduced into the juice from dropped apples that fell onto cattle dung or manure or whether the produce had been sprayed with contaminated ground water is not known. But because the juice was not boiled or pasteurized and E. coli 0157:H7 survives in acidic environments, the organism remained viable. More than half of the 70 cases of illness in the Odwalla outbreak were children under the age of 6 years. One child died.13
That same fall, two other outbreaks involving nonpasteurized apple cider occurred in Connecticut and New York. While new regulations were slow in coming, eventually the USDA established guidelines monitoring juice production and now requires all nonpasteurized juice to be labeled.
Waterborne Outbreaks
The first reported outbreak of E. coli 0157 transmitted by water in the United States occurred from December 1989 to January 1990. Extremely cold weather caused water mains to break in Cabool, a small town in southern Missouri. When the lines were repaired, a backflow of contaminated water entered the system. Hyperchlorination at the time of repairs would have prevented the problem that led to 243 cases of illness and four deaths.14
Canadaís worst outbreak occurred in 1998 in a rural farm community where a local billboard proclaimed the area ìCanadaís Foremost Cattle County.î Nearly 2,000 residents of Walkerton ó almost half of the town ó became ill with E. coli 0157. Ninety people were hospitalized; 11 people died. People were advised to boil all water, throw out their old toothbrushes and even wipe down their doorknobs to curtail the spread of infection.
Five years earlier, a study conducted by Health Canada had compared the density of cattle in the area ó as many as 200 animals per hectare ó with the high incidence of reported E. coli cases and determined that Walkerton was a town waiting for an outbreak. Increased rain allowed runoff water that had been in contact with cattle dung to enter the municipal wells. A faulty chlorinating system failed to kill the organism, and the contaminated water was piped into the homes and consumed by unsuspecting townspeople.15
A bizarre outbreak involving recreational water stunned Georgia residents in 1998. Andy and Marisa Akin were raising their 2-year-old daughter, McCall, as a vegetarian, so they never worried about undercooked ground beef. They never imagined that a day at an Atlanta water park would be harmful to their toddler. McCall became ill with E. coli 0157, along with 26 other children. Six, including McCall, required hospitalization.
The kiddie-pool water was later identified as the source of infection with a probable fecal accident causing the contamination. A low chlorine level on the day of McCallís visit compounded the problem. After being hospitalized for 6 weeks on kidney dialysis and a respirator, McCall died on July 22, becoming the first child in the nation known to have died from E. coli transmitted in chlorinated pool water. Guidelines for both private and public pools were reviewed, and new recommendations were announced in the hopes of preventing future outbreaks from recreational water.16
Another serious water outbreak occurred in the fall of 1999 when New Yorkís Washington County Fair was in full swing. Most of the grounds at the fair were supplied with chlorinated water, but in one area vendors had access only to unchlorinated well water. Not realizing the danger, they used the water to make beverages and ice. More than 900 people experienced diarrhea, 65 were hospitalized and two died.17
Cattle Proximity
Field trips to the farm, long a favorite of both preschool and school-age children, led to 56 illnesses in Washington and Pennsylvania the following year, with 19 children hospitalized. Not only were some of the cattle on the farms colonized with E. coli 0157, but the railing in one locale was found to be positive for the pathogen. Infections occurred when children petted the animals or played on the railing and failed to wash their hands before eating. New guidelines were issued to insure the safety of children visiting farm and petting zoos.18
In 2002, a report at the Epidemic Intelligence Service Conference, CDC, documented an outbreak after a dance was held in a multi-use community facility where cattle had been previously shown. The floor was covered with sawdust that later tested positive for E. coli, as did the dust on rafters overhead. Of those who attended the event, 111 became ill and nine developed HUS.19
Another report of a pancake breakfast held in a building where cattle had been housed 4 days earlier resulted in 34 cases of E. coli. Again, the organism remained viable in the sawdust. 20
Immediately after the onset of the Jack in the Box E. coli outbreak, more than 250,000 hamburgers were recalled. The rapid response is thought to have prevented an additional 800 cases of illness.2 Since that time, food products ó most often ground beef ó have been recalled to halt the spread of disease.
In 1997, Hudson Food Inc., of Rogers, Ark., recalled 25 million pounds of beef after 17 people became ill. In July, 2002, ConAgra Beef Co. voluntarily recalled 19 million pounds of beef and beef products.21 Consumers can call the U.S.D.A. Meat and Poultry Hotline at (800) 535-4555 or visit their Web site at www.fsis.usda.gov for information on meat or poultry products.
On the Rise?
Whether ground beef, alfalfa sprouts, cantaloupe, potatoes or lettuce, it seems that any food or produce item has the potential to harbor E. coli. Mass distribution of food can cause wide-scale illness, such as an outbreak in Japan that resulted in 10,000 cases of illness in 14 separate clusters when contaminated radish sprouts were shipped throughout the country.22 Today, the CDC estimates E. coli 0157:H7 causes 73,000 cases of infection resulting in 61 deaths each year in this country.6 While some wonder if the incidence of outbreaks is on the rise, Jay Varma, MD, medical epidemiologist in the Foodborne and Diarrheal Disease Branch of the CDC, offers this reassurance: ìAfter the major outbreaks of the 1990s, the next few years saw a significant increase in the number of cases. One explanation is certainly that people in general were more aware. Another is that more labs are running tests for E. coli 0157. But for the last couple of years, the incidence has remained constant.
ìThatís a good thing and a bad thing,î continues Varma. ìItís good because there has been no increase, but bad because of all the information out there, we still havenít been able to reduce the number of cases. With the food safety measures implemented, there should be a reduction in the incidence. It may be that if we hadnít had so many interventions, we would have seen a dramatic rise in the number of cases.î
Proactive Labs
The CDC recommends the routine culture of all bloody stools and HUS patients for E. coli 0157 with Sorbitol-MacConkey (SMAC) agar. Most strains of E. coli 0157 do not ferment sorbitol, and after 18 to 24 hours of incubation at 37?C, the sorbitol-negative colonies will appear colorless on SMAC. Tests based on E. coli 0157ís inability to produce ?-glucuronidase can also be used.23
James Nataro, professor of pediatrics, pediatric infections and tropical diseases at the University of Maryland School of Medicine in Baltimore, says, ìThe biggest problem is communication to the lab that 0157 is suspected. Weíre working to educate the physician to examine the stool for blood, if possible, or at least elicit a history. They should ask the mother, ëDid you notice blood in the stool? Was it red or perhaps dark brown or black?í Whatís happening now is that the lab is asked to test for 0157 if they see blood, but more commonly they donít get the history and donít always see the blood.î
Natara challenges labs to be proactive and has this suggestion for laboratory administrators. ìAdd hemoccult analysis as a routine first step in stool testing,î he says. ìIf positive, then test for 0157:H7. It would help to fill in the cracks. And itís important to test all patients, adults as well as children. A single case may be the index case in the outbreak.î
Of note is the organismís pathogenicity. At least three cases of E. coli 0157:H7 acquired in the laboratory have progressed to hemorrhagic colitis and renal failure. The microbe is highly infectious, and adequate precautions should be taken when working with the E. coli.
Possible Solutions
Because cattle are a natural reservoir of E. coli, the slaughter house is certainly one place to look for ways to prevent the spread of infection. The U.S. Department of Agriculture has enacted a Zero Tolerance Program for fecal matter on raw beef carcasses, and a Pathogen Reduction Program has brought better regulation to areas where contamination of beef can first occur.
Meat packing plants also require constant monitoring. Beef is ground and pooled into large batches and packaged for widespread distribution. Since only a few organisms can cause infection, a single infected steer can contaminate a large quantity of ground beef.
While reducing slaughter house and packing plant contamination is essential, the scientific community recognizes the need to find other solutions to curb the spread of E. coli 0157. Irradiation of meat has been recommended, yet consumers complain that the taste is altered, especially in meat with a high fat content such as ground beef. Others are fearful that radiation may prove harmful to humans.
ìScientific evidence is fairly clear that irradiation renders food safer and poses no risk to human health that we know of,î says Varma. The CDC recognizes that public acceptance is an important question, and the agency is running national surveys to identify peopleís concerns about irradiation.
ìAlthough irradiation is one of many answers to the problem,î Varma says, ìcertain groups are concerned that if the governmentís message is irradiation, we might ignore other methods that could be beneficial. Irradiation is one answer. We donít advocate that itís the only method.î
Michael P. Doyle, PhD, regents professor of Food Microbiology, director of the Center for Food Safety and Quality Enhancement and head of the Department of Food Science and Technology at the University of Georgia, was one of the first to report on the 1993 Jack in the Box outbreak. He has continued to study E. coli 0157:H7 since that time.
ìUSDA tests reveal only 0.8% of ground beef is contaminated with 0157,î says Doyle, who believes ground beef is not the greatest risk factor to the United States. Since the organism is carried in the intestinal tract and passed in the feces of animals, it contaminates both farmland and water and poses a problem to those who work and live there, as well as to those who visit. Therefore, he says, ìIt looks like the greatest emphasis for E. coli 0157 is on the farm.
ìStudies have been done in the United States that look at the primary risk factors, and the farm comes at the top of the list, but eating uncooked ground beef is low. A similar study in England looked hard to find the risk for ground beef and found it was not as significant a problem as compared with farm exposure,î Doyle says. ìThe critical issue is to reduce the carriage or shedding in the feces in cattle.î
Probiotic bacteria seem to be one solution.
ìProbiotic means friendly microorganisms that are good for the health of humans and animals,î he explains. A subgroup are the competitive exclusion bacteria, which is how Doyle classifies the probiotic with which he has been working. The bacteria have been isolated from cattle that do not carry 0157 and are actually an antimicrobial to 0157 that lives in the rumen of cows.
ìIf we feed these bacteria to cattle,î he says, ìwithin a short time, 80% to 90% of the cattle donít shed 0157 in their feces.î The new probiotic would be inexpensive, perhaps as low as a dollar a treatment.
A veterinary pharmaceutical company is studying the probiotic. From there, the FDA would have to approve its use, but Doyle feels confident it will be on the market in the near future.
Nataro believes a vaccine for cattle looks promising, as well. The vaccine would prevent cattle from becoming colonized with E. coli 0157 and is in the testing phase.
ìSince the cattle are the main reservoir, if we can prevent cattle from harboring the organism we could prevent it from contaminating our food supply,î says Nataro. ìAnd it certainly would be more effective than a human vaccine. Safety would not be as big an issue. To administer a human vaccine to the entire population, the vaccine would have to be 100% safe, yet there still would be problems. And, of course, there would be a real concern of vaccinating the entire population against a relatively rare infection.î
With cattle, those concerns would not be a issue. ìA study of 36,000 cattle in western Canada is being done, and the data should be compiled within a year with perhaps another year before itís on the market,î he says.
Irradiation, probiotic bacteria and a vaccine for cattle are all possible ways to halt the spread of E. coli 0157 in the future. But until then, the medical community and public must be aware of the dangers of this insidious pathogen.
ìE. coli 0157:H7 is an important health problem,î says Varma, ìand the lab plays a critical role, providing information and aid in the development of new policies. But laboratorians should handle 0157 carefully. It only takes a few bugs to make a person sick. Thatís the reason itís a real challenge.î
Debby Giusti is a medical technologist and freelance writer living in the Atlanta area and an ADVANCE for Administrators of the Laboratory editorial advisory board member.