FACTOR	SPECIFIC PATHOGEN(S) AND OTHER CONSIDERATIONS
Presentation to a health care facility	Degree of illness overall greater in patients seeking evaluation; increased probability of “not norovirus” cause to 50%
Travel history	Especially foreign travel and to endemic areas of dysenteric disease
Recent hospitalization	Clostridium difficile from antibiotic exposure
Day care attendance	Rotavirus, Shigella, Giardia
Nursing home residence	C. difficile, medication side effects, tube feedings, ischemic colitis, fecal impaction, and overflow diarrhea
Wilderness exposure	Giardia or Cryptosporidium
Antibiotic therapy	C. difficile, antibiotic side effects
Raw shellfish, farm animals and fair livestock, pet reptiles or amphibians, petting zoos	Salmonella species, Escherichia coli O157:H7 and non-O157 Shiga toxin–producing E. coli, Vibrio species
Epidemic of multiple patients with a short time of onset	Norovirus; less commonly, Campylobacter jejuni, Salmonella species, Cryptosporidium
Acute vomiting and diarrhea after eating suspected contaminated food	Bacillus cereus, Clostridium botulinum, Staphylococcus aureus
Epidemic of severe gastroenteritis traced to eggs, poultry, meat, or dairy products	C. jejuni, Salmonella species
Homosexuality (males)	Giardia lamblia, Entamoeba histolytica
Abdominal pain Nausea, vomiting Bloody stool Fever Rectal pain Tenesmus	Severe bacterial infections: Salmonella, Campylobacter, Shigella, EPEC, Yersinia or Vibrio species Also consider surgical abdomen, GI bleeding Inflammatory bowel disease
Diarrhea >7-14 days’ duration	Protozoa and microsporidia, C. difficile, Campylobacter, Shiga toxin–producing E. coli
Hemolytic uremic syndrome	E. coli O157:H7 or other species
Stool WBC count	Not reliable for diagnosis of bacterial cause
Colonic ulcerations	Inflammatory bowel disease
Proctitis	Bacterial cause highly probable
Pseudomembranes	Toxic megacolon, C. difficile
Chronic disease (e.g., cirrhosis, DM)	Complicated course expected with any form of diarrheal illness
Organ transplantation	Abnormally severe illness from rotavirus and adenovirus Increased frequency of cytomegalovirus Severe illness from dysenteric diarrhea Spore-forming protozoa and microsporidia
HIV infection, other immunodeficiency disorders	Severe illness from common bacteria, spore-forming protozoa, and microsporidia Increased frequency of cytomegalovirus and Mycobacterium avium complex

DM, diabetes mellitus; EPEC, enteropathogenic E. coli; GI, gastrointestinal; HIV, human immunodeficiency virus; WBC, white blood cell.

Persistent diarrhea is defined as that lasting for more than 14 days, and chronic diarrhea, more than 30 days.⁷ Infectious agents of persistent and chronic diarrhea include bacteria, parasites, and rarely viruses.¹³ Common bacterial pathogens include Aeromonas, Plesiomonas, Campylobacter, Clostridium difficile, Salmonella, and Mycobacterium tuberculosis. Parasites causing chronic diarrhea are colonic forms such as Amoeba, Trichuris, Yersinia, and Schistosoma species or small intestinal pathogens such as Giardia, Cryptosporidium, Cyclospora, Isospora, and Strongyloides.¹³ In developing countries, chronic diarrhea is more likely to have a bacterial cause. In developed countries, chronic diarrhea is caused by noninfectious disorders such as irritable bowel syndrome, malabsorption syndromes, laxative abuse, and inflammatory bowel disease.¹² Categorization of the stool type as watery, inflammatory, or fatty may assist in proper classification of the chronic diarrhea syndrome as infectious or noninfectious. Testing for HIV or immunodeficiency is important because patients with these conditions commonly have chronic diarrhea. Evaluation should include testing for cryptosporidia, microsporidia, mycobacteria (i.e., Mycobacterium avium complex), herpes simplex virus (HSV), Isospora, Cyclospora, and cytomegalovirus. In addition, parasitic and helminthic infections should be ruled out.

Noninfectious Diarrhea.: Noninfectious causes of diarrhea (see Box 31-2) are responsible for approximately 15% of all cases of diarrheal illness. The distinction between infectious and noninfectious causes may not be clinically apparent. A complete evaluation should consider possible surgical pathology of the abdomen, including gastrointestinal bleeding, ischemic bowel, acute appendicitis, intussusception, ectopic pregnancy, and partial bowel obstruction.¹¹ The differential diagnosis also includes possible toxic exposures or ingestions, such as heavy metal poisoning, or ingestion of plant-borne or fish-borne toxins. Endocrine pathology, such as adrenal insufficiency, hyperthyroidism, diabetic enteropathy, and hormone-secreting tumors, and other systemic illnesses should be considered, and special attention should be directed at underlying medical conditions, medication use, and past surgical history.

Ancillary Testing

Most cases of acute diarrhea are self-limited, and laboratory and diagnostic tests should be kept to a minimum unless required for epidemiologic studies. Testing is indicated in patients who have a high probability of a “non-norovirus” clinical picture and have worrisome historical data, signs, and symptoms associated with an increased probability of those causes. Ancillary testing should never compromise empirical treatment when indicated (as discussed later). Fever with a toxic appearance and volume depletion, blood- or mucus-containing stools, frequent voluminous stools, and other risk factors (see Table 31-1) should prompt a diligent search for a specific causative disorder in order to guide appropriate therapy.^12,14 A white blood cell count is rarely helpful and not sensitive or specific enough to aid in diagnostic decision-making, although hemoglobin determination is useful to screen for anemia from blood loss, and abnormalities in platelet and coagulation parameters may contribute to identification of a cause for gastrointestinal bleeding. A comprehensive chemistry panel, including renal function tests, can be important when significant volume loss is suspected or when significant diarrhea has been present for 48 to 72 hours. Liver function studies, thyroid tests, serum lipase assay, and a pregnancy test may be helpful in selected cases.

Hemoccult and fecal cell count: The presence of fecal leukocytes is not specific or sensitive enough to use as the sole criterion to decide which patients with presumed bacterial gastroenteritis should be treated empirically with antibiotics. With inflammatory diarrhea of various causes, red and white blood cells are seen on stool examination. Included are bacterial, parasitic, and noninfectious causes, such as chemotherapy, radiation therapy, hypersensitivity reactions, autoimmune disorders, and inflammatory bowel disease. The presence of fecal leukocytes does not delineate which patients would benefit from empirical antimicrobial therapy. The presence of blood does not always correlate with the presence of fecal leukocytes, so reliance on positive stool guaiac test result alone as a rationale for antibiotic therapy is not recommended. The presence of blood without fecal leukocytes may indicate amebiasis, malignancies, heavy metal poisoning, fissures, hemorrhoids, bowel ischemia, or primary gastrointestinal bleeding.

Assays for calprotectin and lactoferrin, produced by leukocytes, are sensitive and specific and may be more useful than microscopic examination of the stool, but these tests are rarely, if ever, of use in the ED.⁸

C. difficile toxin assay: This test is indicated if the patient reports recent antibiotic use. C. difficile–associated diarrhea most commonly occurs during or shortly after the antibiotic course. In 25% to 40% of cases, however, onset of the diarrhea may be delayed as long as 12 weeks after antibiotic therapy. The most commonly implicated antibiotics are cephalosporins, penicillins, and clindamycin. Although C. difficile accounts for only 10 to 20% of antibiotic-associated diarrhea, an assay for C. difficile toxin gives a positive result in nearly all cases of antibiotic-associated pseudomembranous colitis.¹⁵ Approximately 3% of adult patients and 65% of newborns may be colonized with C. difficile.

Escherichia coli O157:H7 toxin assay: This test is considered in endemic areas and in patients with suspected hemolytic-uremic syndrome.¹⁶

Stool culture for bacteria: Stool cultures may be warranted in patients who are febrile, toxic appearing, immunocompromised, at the extremes of age, experiencing a prolonged course, or not responding to conventional treatment. Studies have shown a 2% positive rate, thus proving that routine cultures are of limited value.¹⁶

Stool examination for ova and parasites: The assessment of stool for ova and parasites is not routinely recommended. This study is used in patients with chronic diarrhea (Entamoeba histolytica, Cryptosporidium); patients with a history of travel to developing countries, particularly to Nepal or areas of Russia (Cryptosporidium, Giardia, Cyclospora)¹⁷; patients with exposure to infants in daycare centers (Cryptosporidium, Giardia); and patients with HIV infection (E. histolytica, Giardia).¹⁸

Giardia antigen assay and serologic testing for amebiasis may be considered in patients exposed to poor sanitation, HIV-infected patients, patients with a history of travel to developing countries, patients with a history of backpacking, and patients with daycare exposures.

Urinalysis: A urinalysis and a urine pregnancy test should be obtained only when urinary tract infection is a possibility, a gastrointestinal origin for the symptoms is not clear, or pregnancy is suspected.

Radiographic studies: Plain radiographs and contrast computed tomography (CT) may be indicated for patients thought to have a surgical abdomen and to identify pathologic abnormalities, such as tumor, obstruction, free air, fistulae, blind loops, and those associated with Crohn’s disease.

Gastrointestinal referral: Referral may be indicated in the evaluation of chronic diarrhea and for workup beyond the scope of the ED (e.g., endoscopy, further stool studies, biopsy).

Empirical Management

Oral rehydration is the treatment of choice for mild to moderate fluid losses (Fig. 31-1). Oral rehydration can be accomplished with sports beverages, commercial rehydration solutions, or a balanced clear liquid diet in the home (e.g., consisting of water, salt-containing liquids such as canned soups, and potassium from oranges or bananas). The WHO has defined an oral rehydration solution (WHO-ORS) that can be made by dissolving the following in 1 L of clean water:

Figure 31-1 Approach to the patient with acute diarrhea. Abx, antibiotics; CT, computed tomography; ED, emergency department; GE, gastroenterologist; GI, gastrointestinal; IV, intravenous; temp., temperature; US, ultrasonography; VSs, vital signs.

3.5 g of sodium chloride

2.9 g of trisodium citrate or 2.5 g of sodium bicarbonate

1.5 g of potassium chloride

20 g of glucose or 40 g of sucrose

Replacement of micronutrients, particularly copper and zinc, has been recommended, especially in developing countries.¹⁹ The concept of bowel rest has been abandoned because it may worsen diarrhea and lead to more severe dehydration. The choice of oral rehydration fluids depends on the extent of dehydration and the underlying health of the patient. In otherwise healthy patients with mild to moderate dehydration, fluids such as sports drinks, diluted fruit juices, and soft drinks supplemented with soups, broths, or crackers may be sufficient to replace the fluid and sodium losses associated with acute diarrhea. Such frequently used “clear liquids” may contain excess sugars and insufficient sodium content, however, leading to an osmotic diarrhea. Beverages containing caffeine should be avoided because caffeine increases cyclic adenosine monophosphate levels and may cause a secretory diarrhea. Milk and other products containing lactose also should be avoided because viral and bacterial pathogens, responsible for many cases of diarrhea, may cause a transient lactase deficiency, leading to malabsorption and osmotic diarrhea. Food intake is encouraged, but foods high in simple sugars should be avoided because the osmotic effect is counterproductive. Foods with a high fat content may delay gastric emptying and should be avoided. The BRAT (bananas, rice, apples, and toast) diet has long been recommended, particularly with pediatric patients. Although no controlled studies have examined the efficacy of the BRAT diet, it remains a commonly recommended strategy. The pectin in the peel of apples is constipating (pectin, found in fruit peel, is the “pectate” in Kaopectate), and bananas provide potassium. If this diet is used for extended periods, adequate provision of protein and energy needs of the patient becomes a concern.

In patients with evidence of more severe dehydration, intravenous fluid resuscitation with normal saline or lactated Ringer’s solution is the preferred treatment. Pediatric patients should receive a bolus of 20 mL/kg of normal saline, which may be repeated as indicated. Specific treatment for diarrhea should be directed toward the suspected cause. In patients with suspected surgical pathology, further diagnostic testing and surgical consultation may be required. With toxic exposures, treatment consists of early decontamination, supportive care, and, if appropriate, administration of specific antidotes. Other noninfectious causes of diarrhea are treated as indicated.

Because the specific pathogen causing infectious diarrhea is rarely identified in the ED and the results of cultures are usually unavailable, any antimicrobial treatment is empirical and guided by knowledge of the common causes of infectious diarrhea (see Box 31-1). Viral and noninvasive bacterial gastroenteritis tend to be self-limiting and require only supportive therapy. Empirical antibiotic treatment is directed against invasive bacterial and parasitic organisms that cause the greatest harm. Antibiotic treatment is initiated in patients with a suspected invasive process and severe diarrhea, systemic symptoms, fever, or abdominal pain and in patients who appear toxic. The current recommendation for empirical treatment of a systemically ill–appearing adult is ciprofloxacin, 500 mg orally twice a day, or levofloxacin, 500 mg orally every 24 hours for 3 to 5 days.²⁰ Fluoroquinolones are efficacious against most organisms that cause dysenteric illnesses and have been shown to be more effective than trimethoprim-sulfamethoxazole.^17,21 Fluoroquinolones should not be administered to pregnant patients or children younger than 18 years. The antibiotic treatment of severe gastroenteritis in children has been associated with the development of hemolytic-uremic syndrome and thrombotic thrombocytopenic purpura if the bacterial cause is enterohemorrhagic E. coli O157:H7, although Salmonella, Shigella, and Campylobacter also have been implicated. If possible, treatment for pediatric patients should be based on culture results with supportive care initially.

If amebic dysentery is of concern in high-risk patients (see Table 31-1), treatment with metronidazole after stool analysis for ova and parasites is recommended. In patients with a history of recent antibiotic use suspected of having C. difficile colitis, a C. difficile toxin assay followed by vancomycin or metronidazole is appropriate.²²

The use of antimotility agents in the treatment of acute enteritis has been controversial, with the literature divided over this issue. Patients with simple, acute viral gastroenteritis benefit from antimotility agents and often obtain significant relief of symptoms, with less fluid loss and without significant complications. Loperamide is the safest and most effective medication. Relief of symptoms is achieved much more rapidly than with bismuth subsalicylate (Pepto-Bismol) in patients with inflammatory diarrhea or antibiotic-associated colitis. In pediatric age groups, the use of opioids, loperamide, or diphenoxylate with atropine rarely has been associated with the precipitation of toxic megacolon and hemolytic-uremic syndrome. Because the beneficial effects of these medications are modest, they should be avoided or used with extreme caution in these high-risk patients.

Probiotics have been used as an alternative to traditional antibiotic therapy for diarrhea. Lactobacillus and other bacteria have proved to be effective in restoring the normal gastrointestinal flora that is disrupted during diarrhea illness. This approach has been most effective with traveler’s diarrhea and nonspecific diarrhea in children.