Escherichia coli 0157:H7 Infections: Epidemiology

Roberts, James R. MD

In Focus: Part 2 in a Series

A compendium of Dr. James Roberts' InFocus columns is available in book form. The 302-page volume, InFocus: Roberts' Practical Guide to Common Medical Emergencies, is available from Lippincott Williams & Wilkins for $59.95 by calling (800)638–3030.

Article Outline

Author Credentials and Financial Disclosure: James R. Roberts, MD, is the Chairman of the Department of Emergency Medicine and the Director of the Division of Toxicology at Mercy Health Systems, and a Professor of Emergency Medicine and Toxicology at the Drexel University College of Medicine, both in Philadelphia, PA. Dr. Roberts has disclosed that he has no significant relationships with or financial interests in any commercial companies that pertain to this educational activity.

Learning Objectives: After reading this article, the physician should be able to:

1. Explain the environmental issues involved in the transmission of E. coli 0157:H7.

2. Describe the incidence of colitis in children compared with adults.

3. Recall the principles of diagnosis and treatment of colitis.

Release Date: January 2004

Acute diarrhea, often times bloody and associated with severe cramping and constitutional symptoms, is a common presentation to the emergency department. Such infections, even when they are from common and well known bacteria such as salmonella, campylobacter, or shigella, are annoying and distressful to the patient although severe consequences are rare.

Even without treatment, almost all forms of acute bacterial diarrhea are self-limited. Stool cultures, fecal leukocytes, the physical exam, and the history are often nonspecific, and many cases in outpatients are never diagnosed with certainty. Therapy is usually supportive, but empiric antibiotic therapy and the judicious use of antimotility agents are now considered reasonable. Usually antibiotics are prescribed while awaiting culture results, but in some cases clinicians will give a short course of cipro or bactrim and Immodium, and the majority of patients will get better. However, one type of diarrhea that's sure to spell problems for both the emergency physician and patient is colonization with one specific type of Escherichia coli, the 0157:H7 serotype.

Last month's column provided a clinical overview of E. coli 0157:H7 infections. This month's column highlights some of the epidemiological issues involved with E. coli 0157:H7 infections. Many physicians will never diagnose even a single case of this severe infectious disease in their careers, but that's simply due to not looking for it. We probably all see it more than we think. Unless we obtain follow-up on hospitalized patients, this disease will continue to stay under our clinical radar screen.

Statistically, E. coli 0157:H7 colitis is less common than shigella, salmonella, and campylobacter, but it is much more common than one would intuit. We don't diagnose this infection because it often fails to present with the same GI gusto as other infectious diarrheas. Even when you send a routine stool culture to the lab, most hospitals will fail to find this culprit. Because the clinical course in most healthy adults is usually a self-limited one, nobody is the wiser when the patient spontaneously gets better. A bout of “culture negative” yet miserable diarrhea is dismissed as viral when the labs are normal and the patient finally gets better. Importantly, if you look for E. coli 0157:H7, you will certainly find it in an unexpectedly high proportion of otherwise clandestine and “routine” cases. Most cases are not associated with epidemics that make the 6 o'clock news, so don't expect the Centers for Disease Control and Prevention to descend on your hospital in most cases. This month's column highlights the fact the children are so susceptible they can contract E. coli 0157:H7 merely by petting a cow.

Contrary to adults who generally fair well after even a rocky course with E. coli 0157:H7, children and the elderly are more prone to complications, even fatal ones.

This is a recent article that highlights an epidemiological evaluation of an E. coli 0157:H7 outbreak of 151 patients with confirmed infections. It is unusual because it involves a petting zoo and a large number of children. The authors note that E. coli 0157:H7 causes an estimated 60 deaths per year in the United States, with 73,000 annual cases. Healthy cows are the main recognized animal reservoir for this bacterium, and often these animals harbor the organisms as part of their normal bowel flora. The animals do not get sick, but they excrete the bacterium into the environment via their feces. The bacterium can remain viable on the ground or on fence posts for weeks.

Most prior outbreaks were due to exposure to contaminated food (almost exclusively undercooked ground beef) or water and person-to-person contact. However, direct transmission of E. coli 0157:H7 from incidental exposure to live animals and their environment is a growing concern. Most reports in the past have been single cases or small clusters, so this 151-patient outbreak is quite unusual. The epidemiological investigation that followed was quite extraordinary.

In September 2000, an unexpectedly large number of E. coli 0157:H7 infections occurred in a few counties in Pennsylvania. The outbreak was recognized by county and state health departments, and the organizations instituted active case investigations shortly after the first case was reported. The majority of patients were visitors to a local petting farm. The establishment was a traditional dairy farm that had for several decades operated a small domestic animal petting zoo for children. Approximately 2000 people visited the farm each day, and food and beverages were available on site. An epidemiological investigation was performed to determine the magnitude of the outbreak, to identify risk factors for infection, to interrupt the transmission, and to describe the bacteriology in this farm environment.

A special reporting agency was established for physicians who had ill patients and for laboratories that had tested stool that was positive for the E. coli 0157:H7. By the time the official investigation began, six patients had developed the hemolytic uremic syndrome (HUS).

For clinical purposes, a probable case was defined as acute diarrhea (three or more loose stools in a 24-hour period) in any patient who developed such symptoms within 10 days after visiting the farm. A confirmed case was one in which the E. coli 0157:H7 was isolated from stool or the hemolytic uremic syndrome was identified. Controls were individuals who had visited the farm but did not have diarrhea within 10 days. All individuals were asked to fill out a questionnaire that identified exposure, food or beverages consumed at the farm, and handwashing practices.

The entire farm population of domestic animals was subjected to stool collection with analysis for E. coli 0157:H7. Swabs also were obtained from the animal's environments, including railings around enclosures and standing water. Samples were placed on sorbitol-MacConkey agar, the medium commonly used to isolate the E. coli subgroup 0157:H7. Isolates also were tested for shiga toxin production.

Fifty-one confirmed and probable cases were found in a five-county area. No resident or employee of the farm was infected. The ages of the infected patients ranged from 1 to 50, with a medium age of 4. Of those infected, 92 percent were 10 or younger. About one-third of the patients had bloody diarrhea; almost half had fever and vomiting. Eight individuals (16%), all 10 or younger, developed the hemolytic uremic syndrome. Renal failure resolved in all but one child. In this patient, end-stage renal failure developed, and he required transplantation. There were no deaths. Once the farm was closed, no additional cases were reported.

Those infected were more likely than controls to have direct contact with calves (such as petting), touching the railings, or having contact with manure in the stalls at the farm. Those who bit their nails or had purchased food or drink at the farm also were more likely to develop the infection than controls. Visitors who washed their hands in sinks were less likely to become ill, supporting a protective benefit from handwashing.

Of the 216 cattle at the farm, 15 percent were colonized with E. coli 0157:H7. Overall, calves and heifers were more often colonized than older cattle. E. coli 0157:H7 was found in one sample from a watering trough and one sample from a fence railing.

The authors conclude that this large outbreak of E. coli 0157:H7 infections was due to direct transmission of bacteria from animals and their environment to people, mainly children. Close contact with the animals was associated with the illness. Although typically only about two percent of all cow herds are colonized with E. coli 0157:H7, a high percentage (15%) of the animals in this farm exhibited that characteristic.

The authors note that E. coli 0157:H7 can survive in the environment for several months, therefore posing a threat of infection to humans even without direct contact to animals. This report also highlights the predilection for children via fecal-oral contamination to contract clinical infection with this bacterium. Eating in the contaminated environment (hand-to-mouth transmission) and lack of handwashing were identified behaviors that contributed to this outbreak.

E. coli 0157:H7 – Piecing Together a Jigsaw Puzzle New Engl J Med 2002;347:608

This is an editorial accompanying the above report that provides an overall review of this potentially deadly bacterium. Hemorrhagic colitis linked to E. coli 0157:H7 associated with ground beef was first reported approximately 20 years ago. Prior outbreaks have been almost solely associated with undercooked ground beef consumption, yet recent epidemiology has provided further evidence that the disease can be transmitted by environmental (non-food) exposure. Outbreaks also have been noted with such remote contact as camping on the land on which farm animals had previously grazed.

Because E. coli 0157:H7 can survive for a long periods in the environments, one does not have to have direct contact with an animal that is actively excreting the organism to become infected. Finally, person-to-person transmission has been well demonstrated. It is well established that infections with E. coli 0157:H7 are potentially fatal, and unfortunately children bear the greatest incidence of complications.

Therapy remains supportive, correcting and maintaining food and electrolyte imbalance, avoiding the use of antimicrobial and antimotility agents, and managing complications. The two most serious complications are thrombotic microangiopathic thrombotic thrombocytopenic purpura (TTP), and the hemolytic uremic syndrome. Because there are few therapeutic options once the colitis has occurred, prevention and control measures are essential for reducing morbidity and mortality.

Comment: The importance of this report is its conclusions about how E. coli 0157:H7 colitis can be spread in the environment and the tremendous susceptibility of children to this infection. One can readily recognize how a single business, such as a restaurant, can reek havoc on a population, and this petting zoo had an amazing 2000 visitors a day who were potentially exposed. No one ate undercooked ground beef in this report, those affected merely visited the contaminated environment. E. coli 0157:H7 is an amazingly contagious organism that can be transmitted by simply eating a sandwich while visiting a cattle farm.

E. coli 0157:H7 is probably a much more common cause of bloody diarrhea in the United States than previously considered. It's probably less common than campylobacter or salmonella enteritis, but in some states, its incidence is about the same as shigella infections. As with most infectious diseases, serious infections and complications are more common in children and the elderly. Mortality rates among nursing home patients can be as high as 35 percent. The most common source has always been inadequately cooked ground beef. Because the bacterium is present in the intestines of healthy cattle, contamination of the meat during slaughter and the grinding process transfers the organism to the products sold in supermarkets.

Properly cooked ground beef can eliminate one's exposure, and the USDA requires that commercially-sold ground beef be cooked thoroughly (internal temperature of 160 degrees F) so that juices are no longer pink. This effectively kills the E. coli 0157:H7 organism. Note that the contaminated meat does not contain the toxin; it contains the bacterium that produces the toxin once the bacterium is in the human intestine. Rarer exposures have occurred from consuming apple cider, raw milk, or contaminated water.

Person-to-person transmission is well known in day care centers and nursing homes. Food handlers may be responsible for some institutional outbreaks, hence the omnipresent sign in restrooms: “Employees must wash their hands.” As an aside, emergency physicians and nurses are at risk for becoming infected from their patients, and should not allow parents to bring their children to visit a sick relative in the hospital or ED.

E. coli bacteria have at least 30 serotypes, of which the 0157:H7 is one of the more toxic ones. This serotype produces a toxin that is cytotoxic to intestinal vascular endothelial cells, thus producing a platelet-thrombin complex, vascular occlusion, hemorrhagic colitis, and renal failure. Importantly, E. coli 0157:H7 does not cause an invasive bacteremia, but produces a toxemia and a cytotoxic event secondary to the toxin's effect on blood vessels. Blood cultures will not identify these bacteria nor their toxemia.

The incubation period is two to eight days (usually three to five), with watery diarrhea being the initial symptom. Diarrhea becomes bloody hours or days later. Most patients have some sort of bloody diarrhea, but frank GI bleeding sometimes is mistaken for hemorrhage from a noninfectious cause. Severe abdominal cramping and vomiting typically accompany the diarrhea. Fever may be present, but it's not a uniform finding and its absence raises suspicion for the disease. Curiously, fecal leukocytes are found in relatively small numbers, compared with the striking voluminous WBC seen in shigella dysentery. Endoscopic and radiographic findings are nonspecific and consistent with only inflammatory changes, sometimes mistaken for C. difficile, inflammatory bowel disease, or bowel ischemia.

There are two serious complications of E. coli 0157:H7 hemorrhagic colitis: HUS and TTP. HUS is common in children, and may occur in as many as 20 to 25 percent of cases. Even without HUS, E. coli 0157:H7 hemorrhagic colitis in a child is a very serious process. TTP occurs much less commonly, and is usually the counterpart of HUS in adults. Curiously, these complications typically occur five to 20 days after the onset of diarrhea, and the colitis may actually be totally resolved and stool cultures now negative.

An episode of diarrhea may be unappreciated by the time the diagnosis of HUS or TTP is made. Included in the differential are other types of colitis or inflammatory bowel disease. I would likely consider C. difficile or other forms of diarrhea long before I thought of E. coli 0157:H7. Confirming the diagnosis requires specific stool culturing technique with serotyping of the E. coli 0157:H7. Sending a routine stool sample to the lab won't make the diagnosis. When E. coli 0157:H7 is present, the lab will report “normal flora.” Blood cultures will be negative.

Emergency physicians have become somewhat cavalier in their approach to the routine performance of stool cultures in ED patients. I am still quite liberal with this test, and order a stool culture in all but the most benign cases. It seems reasonable to send cultures from all patients given antibiotics. After studying this topic, I will specifically look for E. coli 015:H7 more often.

The management of possible E. coli 0157:H7 hemorrhagic colitis presents somewhat of a conundrum. While the empiric use of antibiotics is scientifically supported when one strongly suspects salmonella, shigella, or campylobacter, antibiotic therapy does not make a major impact on the clinical scenario caused by these common organisms. Despite the observation that E. coli 0157:H7 is sensitive to most common antibiotics, their use will not shorten the course. In fact, there is a trend toward a more serious infection when antibiotics are chosen. Normal competing bowel flora may be killed in a misguided attempt to cure salmonella, shigella, or campylobacter. This theoretically may allow E. coli 0157:H7 to flourish.

Lysis of the bacteria also may increase toxin release. However, E. coli 0157:H7 can be eliminated by most antibiotics, at least theoretically decreasing complications. Antimotility agents, such as Immodium, probably won't hurt anyone for a few days, although this also has been related to more serious disease. This disease must be treated unknowingly with antibiotics and Immodium many times, without sequelae, so the entire antibiotic antimotility issue is poorly defined.

One other key that might prompt suspicion of E. coli 0157:H7 is that stools are usually very bloody, if not totally bloody, suggesting diverticulitis or some other noninfectious process. Fever is usually absent, and even if one does a stool leukocyte swab, one will be misled because they just aren't there. I would not expect to find E. coli 0157:H7 in run-of-the-mill watery brownish diarrhea or in the slightly heme positive specimen. The pro sends a stool culture in the person with frank rectal bleeding and with symptoms not thought to be due to acute bacterial infection. One should always culture bloody diarrhea or bloody stools in children by using techniques to isolate E. coli 0157:H7.

Having further studied this fascinating disease, I can think of many patients that I have admitted with acute lower GI bleed that probably were never even considered for the possibility of E. coli 0157:H7 colitis. Undoubtedly, I have treated a mild case of E. coli 1157:H7 in an outpatient without knowing it. Also in that rare patient who has signs and symptoms suggestive of TTP, I can never remember having asked about a distant bout of bloody diarrhea that may have been the cause.

Back to Top | Article Outline

Conferences Listed on EM-

Listings of meetings in emergency medicine and related fields are now listed on Emergency Medicine News web site, To view the listings, click on Events.

To list a meeting in Emergency Medicine News and on the web site, send the name of the meeting, date, location, sponsor, and contact information, including a phone number to: Emergency Medicine News, 333 Seventh Ave., 19th Fl., New York, NY 10001; (201)358-8596 (fax);

© 2004 Lippincott Williams & Wilkins, Inc.