William P. Ciesla Jr. MD
Richard L. Guerrant MD
Essentials of Diagnosis
Diarrheal illnesses caused by bacterial, protozoal, or viral pathogens remain a major cause of morbidity and mortality throughout the world. In developing countries, poor sanitation and substandard living conditions create an environment in which diarrheal pathogens exact a terrible toll, especially among children. In the industrialized world, infectious diarrhea remains a common problem among all age groups. Moreover, epidemiologic events such as the AIDS epidemic, the globalization of the food supply, and emergence of new pathogens are increasingly factors in diarrheal illnesses. Chlorine-resistant Cryptosporidium parvum now threatens our drinking water, Cyclospora cayetanensis and EHEC menace our food supply, and toxigenic C difficile increasingly is a problem in hospitalized and institutionalized patients. Given the ubiquitous nature of infectious diarrhea, it is imperative that primary care physicians and infectious disease specialists have a firm understanding of the diagnosis and treatment of these diverse and potentially devastating illnesses.
In industrialized countries, a century of achievements in public health and sanitation have markedly diminished the widespread carnage of diarrheal pathogens common in the developing world. Nevertheless, infectious diarrhea remains a significant problem, one that appears to be increasing with the emergence of new epidemiologic factors and pathogens. Every year in the United States, there are an estimated 25 million cases and 5,000–10,000 deaths caused by infectious diarrhea. As elsewhere in the world, children experience the highest incidence of diarrhea; between 2.0 and 3.2 illnesses per child per year (up to 5.0 illnesses per child per year for children in day care). However, adults and especially the elderly are not spared. The elderly population especially is at risk for the more serious sequelae of infectious diarrhea, and most of the mortality related to gastroenteritis in adults occurs in this group.
A number of different epidemiologic factors are important to consider when approaching a patient who may have diarrhea caused by an infectious agent. Contaminated food or water ingestion, recent travel, antibiotic use, HIV positivity or AIDS, and institutional or day care all are important clues to identify the patient who may be at increased risk for developing infectious diarrhea.
Contaminated Food. The acquisition of diarrheal pathogens through the food supply is a major problem both in the developing world and in the United States, where 400–600 documented food-borne outbreaks and millions of cases are estimated to occur each year. A number of foods and pathogens have been associated with food-borne diarrheal illnesses: raw or poorly cooked shellfish (Vibrio parahemolyticus, Norwalk virus), chicken (Campylobacter jejuni; see Chapter 57), eggs (Salmonella enteriditis; see Chapter 53), chocolate milk (Listeria monocytogenes; see Chapter 51), reheated fried rice (Bacillus cereus; see Chapter 51), and undercooked beef, unpasteurized apple cider, lettuce, and alfalfa sprouts (EHEC; see Chapter 53). The mode of food contamination is varied; an infected food handler with poor hygiene, untreated human or animal waste used as fertilizer, unclean water used to wash or prepare food, and animal feces mixing with meat products during or after slaughter all have been implicated. In addition, evolving social and economic factors are emerging to further facilitate the spread of food-borne pathogens in the United States. First, over one-third of the produce now consumed in the United States is imported, often from developing regions with varied agricultural practices and abundant exotic pathogens. This threat manifested itself recently in several outbreaks associated with imported Guatemalan raspberries contaminated with the parasite Cyclospora sp. Second, an increasing proportion of meals are now prepared outside the home (eg, in restaurants), exposing large numbers of people to common source outbreaks related to, for example, an infected food handler or contaminated foodstuffs. Finally, the intentional contamination of the food supply by diarrheal pathogens has emerged as a form of biological terrorism. One incident in the United States involved > 700 cases of Salmonella gastroenteritis related to the intentional contamination of restaurant salad bars with Salmonella typhimurium. This criminal act was committed by a religious commune in an apparent attempt to influence voter turnout for a local election.
Contaminated Water. Drinking water, often contaminated by human or animal fecal waste, is a common mode of acquisition of diarrheal pathogens in the developing world and when natural disasters disrupt the normal sanitation infrastructure in industrialized countries. In addition, many lakes, rivers, and streams contain myriad organisms such as Giardia lamblia and EHEC. Hikers and swimmers are at risk when they drink unfiltered water from these sources. In the past, municipal water supplies in the United States were generally considered free from diarrheal pathogens because of modern chlorination and filtering procedures.This notion was challenged in 1993 when a large water-borne outbreak of infectious diarrhea in Milwaukee was caused by the parasite Cryptosporidium parvum. This outbreak affected > 400,000 people despite the use of standard purification procedures; further outbreaks have occurred despite even more stringent filtration standards.
HIV and AIDS. HIV-positive and AIDS patients are at significant risk for diarrheal illnesses. More than 30 million patients are believed to be infected with HIV worldwide, and it has been estimated that 50–90% of these patients will develop a significant diarrheal illness at some point, especially as the CD4 count drops below 200. Depending on the pathogen involved, these patients can develop debilitating chronic diarrhea with few therapeutic options. Infectious agents and drug side effects are the two major causes of diarrhea in this population.
Hospitalization or Institutional Care. Hospitalization or institutional care is another important and often overlooked risk factor. In the United States, 1–2% of all hospitalized patients will develop an episode of nosocomial diarrhea, leading to volume and electrolyte losses as well as predisposing the patient to other nosocomial infections. Large populations of compromised patients, frequent antibiotic use, and C difficile spores in the hospital environment all contribute to the risk in this population. Children in day care are also at increased risk, given the poor personal hygiene and clustering of large numbers of children together in this setting. Outbreaks are often caused by organisms such as Shigella or Giardia spp. that require a small infectious dose and are readily spread from child to child by direct contact.
Antibiotic use places patients at risk for the development of antibiotic-associated diarrhea, especially in the hospital setting. The onset of diarrhea usually occurs while the antibiotic is being administered, but onset can be delayed for several weeks after the antibiotic has been stopped. Virtually all antibiotics have been implicated as a cause; clindamycin has the highest risk with 10–20% of patients who receive antibiotics developing antibiotic-associated diarrhea, but penicillins and cephalosporins are the most important culprits given their widespread use. C difficile infection accounts for ~20% of all antibiotic-associated diarrhea and virtually all cases of pseudomembranous colitis secondary to antibiotic use. The etiology of the remaining cases of antibiotic-associated diarrhea remains unknown.
Travel to Developing Countries. Travel from an industrialized country to the developing world places large numbers of people at risk for the acquisition of infectious diarrhea. An estimated 8 million Americans travel to developing regions each year, and up to 50% will develop diarrhea during their stay. Although these illnesses are rarely life-threatening, they have a significant impact on the tourist industry and on business travelers. Traveler's diarrhea has also been a major problem in military deployments, leading to the incapacitation of large numbers of soldiers. Early in operation Desert Storm during the Gulf War in 1990, for example, the attack rate for infectious diarrhea among American soldiers in Saudi Arabia was as high as 10%/week.
Documenting a careful and directed history is absolutely critical when approaching the patient who may have infectious diarrhea. Clues in the history are very helpful in assessing the severity of the patient's illness, narrowing the differential diagnosis of potential pathogens, and choosing an appropriate therapeutic course.
In general, two different clinical syndromes are described in patients with infectious diarrhea, each with its own set of causative organisms. Noninflammatory diarrhea occurs with pathogens that act primarily in the small intestine to induce fluid secretion by various mechanisms and produce minimal, if any, intestinal inflammation. These patients usually present with voluminous watery diarrhea, nausea and vomiting, abdominal cramps, and minimal or low-grade fevers. The second syndrome, inflammatory diarrhea, is typically seen with pathogens that induce inflammation in the colon via invasion or cytotoxins. These patients present with fever, diarrhea with smaller volumes of stool (often < 1 L), blood or mucus in the stool (dysentery), tenesmus, and lower quadrant cramping. A history of grossly bloody stools should suggest EHEC infection, now the major infectious cause of bloody diarrhea in the United States. The duration of the diarrhea should be determined, because this information is also helpful in narrowing the differential.
Whether inflammatory or noninflammatory, diarrhea may be acute (lasting < 2 weeks) or persistent (lasting > 2 weeks); some refer to diarrhea lasting > 1 month as “chronic.” Acute diarrhea is usually caused by infectious agents (viruses and bacteria), the ingestion of preformed bacterial toxins, or drugs. Persistent diarrhea is more often caused by parasitic pathogens or noninfectious causes.
The hydration status of the patient should be assessed by asking about orthostatic symptoms (dizziness or lightheadedness upon standing), decreased urine output, increased thirst, and the ability of the patient to take oral fluids. For viral infections and certain types of bacterial food poisoning, prominent nausea and vomiting may be a clue to the diagnosis. High-grade fever is more common with pathogens that invade or disrupt the colonic epithelium, leading to intestinal inflammation. Mild abdominal cramping or pain is often present. More severe abdominal pain, localized to the right lower quadrant and mimicking acute appendicitis, can be seen with Yersinia enterocolitica and C jejuni infections.
The patient should be asked about similar illnesses in family members or close contacts, suggesting a common source outbreak related to a specific pathogen. The clinician must also be aware of any potential outbreaks that may be occurring in the community and always be alert to potential clustering of similar cases, which may point to an unrecognized outbreak. A careful history of recent food ingestion, including any deviations from the normal diet, should be obtained. HIV risk factors or a history of HIV positivity or AIDS is important, as is a sexual history for anal intercourse-related sexually transmitted diseases (STDs) (proctitis caused by gonorrhea, syphilis, herpes, or Chlamydia infection).
The key to examining the patient with diarrhea is to assess the hydration status and determine if there has been significant volume loss. Supine and standing blood pressure and pulse rates should be determined to assess for orthostatic hypotension. Supine hypotension and resting tachycardia can indicate life-threatening hypovolemia and the need for aggressive fluid resuscitation. The mucous membranes should be assessed for moisture and the skin for tenting. In infants, lethargy, dry mucous membranes, flat fontanelles, sunken eyes, poor capillary refill, and poor skin turgor may all be clues of significant hypovolemia.
The abdominal exam usually is unrevealing in most cases of infectious diarrhea. There may be diffuse, mild tenderness and hyperactive bowel sounds, but usually no rebound or guarding. However, severe cases of C difficile may result in peritoneal signs and Y enterocolitica and C jejuni can mimic an acute appendicitis as mentioned. A rectal examination and stool guaiac test should be performed to look for occult blood. A fresh stool sample should be obtained in a specimen cup for gross examination to give an objective assessment of the patient's complaints. The appearance of the stool—watery, mucoid, or grossly bloody—can be very helpful in narrowing the differential diagnosis and focusing the laboratory examination.
Another, more stable marker for the presence of intestinal inflammation is lactoferrin. Lactoferrin is an iron-binding glycoprotein released from neutrophil secondary granules; its presence in stool is indicative of colon inflammation. False positives are known to occur in breast-fed infants, so fecal lactoferrin should be used with caution in this population. In one study, fecal lactoferrin, detected using a commercially available latex agglutination assay, was 83–93% sensitive and 61–100% specific for patients with Salmonella, Campylobacter, or Shigella spp. detected by stool culture. Lactoferrin tests have the advantage of being sensitive in cases in which the pathogen can lyse the fecal leukocytes as a result of cytotoxic toxins (C difficile), via adherence mechanisms (Entamoeba histolytica) or if the stool sample is not examined immediately.
Testing for fecal leukocytes or lactoferrin is advocated as a way to help identify infections by diarrheal pathogens that result in intestinal inflammation, because the patients may benefit from antimicrobial therapy and tend to have more severe sequelae. The clinical presentation can provide clues to identify these patients, but there can be considerable overlap in symptoms between the inflammatory and noninflammatory syndromes. The presence of fecal leukocytes or lactoferrin provides information in addition to the history to decide whether a stool culture and antibiotic therapy are indicated. The limitations of using fecal leukocyte or lactoferrin testing for this purpose are obvious from looking at their respective sensitivities and specificities. However, as long as the clinician recognizes these limitations and interprets the results in the context of the clinical setting, fecal leukocyte or lactoferrin tests should remain important tools in approaching the patient with infectious diarrhea.
Stool cultures should be considered for all patients suspected of having inflammatory diarrhea based on the clinical and epidemiologic history, a positive fecal leukocyte or lactoferrin test, or both. In the United States, the most common infectious causes of inflammatory enteritis are C jejuni, Salmonella spp., and Shigella spp. Most laboratories will automatically culture for these organisms if stool is sent for routine bacterial culture. Stool culture for other bacterial pathogens often will require special media or handling. In these cases, communication with the microbiology laboratory about the clinically suspected pathogens is very helpful to ensure that the specimens are cultured properly. Patients with grossly bloody diarrhea or a suggestive epidemiologic history should have their stool cultured for EHEC O157:H7. However, an increasing portion of EHEC infections are now caused by non-O157 serotypes (≤ 50%), so if EHEC is suspected, non-O157 serotypes must be considered. A commercially available enzyme immunoassay can now identify the presence of Shiga bacillus toxin in E coli isolates or in the stool. It is sensitive, specific, and often will remain positive after the stool culture becomes negative.
For hospitalized patients, the use of routine bacterial stool cultures and parasite examinations should be governed by the “3-day rule” for bacterial cultures and the “4-day rule” for parasite exams recently advocated by the College of American Pathologists. These rules are based on several studies showing that the yield of such tests is extremely low in patients who develop diarrhea after being hospitalized for > 3–4 days and such tests are not cost effective. The pathogens that are detected by these tests are uncommonly acquired in the hospital setting in the United States. However, the application of these rules should be flexible and allow for those cases where, after careful clinical consideration, potential pathogens likely to be detected by these tests are still in the differential irrespective of the length of hospitalization.
If the diarrhea persists for > 7 days or if the history raises the possibility of a parasitic infection, then the stool should be examined by the appropriate studies. Helpful tests include the classic ova and parasite exam, the funnel gauze test for Strongyloides spp., direct fluorescence staining for Giardia and Cryptosporidium spp., modified acid fast stain of the stool for Cryptosporidium or Cyclospora spp., or trichrome staining for microsporidia. The ova and parasite (O+P) exam involves concentrating the stool specimen and then preparing a smear to look for parasite cysts or trophozoites and intestinal helminths. Among the diarrheal pathogens that can be detected by the O+P smear are E histolytica and G lamblia. Unfortunately, the O+P exam can have a low yield, so consideration should be given to using one of the pathogen-specific tests noted above.
A history of recent antibiotic use or hospitalization in a patient with diarrhea should raise the concern of C difficile, and the stool should be tested for the presence of C difficile toxins. The gold standard assay is the stool-cytotoxin test, looking at the ability of toxin B to induce cytopathic effects in cultured cells. It is 94–100% sensitive and 99% specific. Several new immunoassays are now available that are cheaper and faster than the stool-cytotoxin assay, but also less sensitive and specific. Several stool samples should be tested over sequential days before completely excluding C difficile infection.
In addition to stool studies, patients with severe diarrhea, fever, abdominal pain, or volume loss should have serum chemistries, including sodium, potassium, chloride, blood urea nitrogen, and creatinine, and have a complete blood count checked.
Direct imaging such as sigmoidoscopy, colonscopy, or esophagogastroduodenoscopy can play a role in cases of chronic diarrhea in which the diagnosis remains unclear despite use of appropriate studies. Esophagogastroduodenoscopy with duodenal aspiration or biopsy can be used to detect G lamblia. Sigmoidoscopy or colonoscopy may show mucosal changes like erythema, ulcers, or pseudomembrane formation (C difficile). Biopsy may reveal adherent bacteria (enteroaggregrative E coli), viral inclusions (cytomegalovirus colitis) or pathology consistent with noninfectious diseases.
The most important question in managing the patient with suspected infectious diarrhea is to decide who will benefit from a directed laboratory evaluation and specific antimicrobial therapy. For most cases of noninflammatory diarrhea, the illness will be self-limited and require only supportive care. In these cases, extensive laboratory evaluation will be of no benefit except from an epidemiologic standpoint and will waste considerable health care resources. As a general rule, patients with a duration of illness longer than 2 days, high fever, systemic toxicity, bloody stools, fecal leukocytes or lactoferrin, severe volume depletion or abdominal pain, and those who are immunocompromised (ie, with immunosuppressive chemotherapy after organ transplant or with AIDS) should have a more extensive evaluation initiated at presentation. Other patients can be followed for resolution of their illness and evaluated only if one of the above criteria develops or as part of a public health investigation.
Figure 20-1 outlines a general approach to the patient with infectious diarrhea, using the history and the presence or absence of fecal leukocytes or lactoferrin to classify the diarrhea as noninflammatory or inflammatory. Unfortunately, there are no prospective studies that validate this algorithm by showing a decrease in morbidity or mortality. However, keeping in mind the limitations of fecal leukocytes or lactoferrin as previously discussed, it is useful to begin to formulate the differential diagnosis and to guide therapy. Box 20-1 lists the common pathogens causing inflammatory and noninflammatory diarrhea, bloody diarrhea, and diarrhea in AIDS patients and in travelers to developing countries. In the following section, we discuss some of the more common pathogens, briefly outlining general epidemiologic, clinical, and diagnostic information that will be useful in narrowing the differential diagnosis. The reader is referred to the specific chapters elsewhere in this book for more detailed and complete information about the potential pathogens. Finally, in considering the differential diagnosis in a patient with diarrhea, the clinician must always remember noninfectious causes and obtain a careful history of medication use, enteral feedings, and symptoms suggestive of inflammatory bowel disease, endocrinopathies, ischemic colitis, or malabsorption.
Bacterial Species.C jejuni (Chapter 57) is probably the major cause of community-acquired inflammatory enteritis in the United States, causing an estimated 2 million cases each year. The organism is transmitted via poorly cooked chicken and contaminated milk or water or on unwashed cooking utensils. The infection usually results in an acute enteritis with watery or bloody stools, fever, and abdominal pain that usually resolves within 1 week. The diagnosis is made by isolating the organism by stool culture.
Salmonella (Chapter 53) gastroenteritis is another major cause of inflammatory diarrhea acquired via contaminated food, milk, or water. Within 48 h of ingestion, most patients will develop fever, watery diarrhea, and abdominal cramping that resolves within 10 days. The stool contains moderate numbers of fecal leukocytes, and the diagnosis is made by stool culture. S enteriditidis (associated with undercooked eggs) and S typhimurium are the most common Salmonella isolates implicated in the United States. A small percentage of patients with enteric fever syndromes from Salmonella typhi or Salmonella paratyphi, characterized by fever and abdominal pain, may present with constipation, diarrhea, or both, early in the course. Bacteremia and subsequent metastatic foci of infection are serious complications that can occur with Salmonella gastroenteritis and enteric fever.
Shigella (Chapter 53) species should be suspected in patients who present with acute diarrhea combined with fever and blood or mucus in the stool. As compared with Salmonella or Campylobacter spp., transmission occurs more often via direct patient-to-patient spread. In the United States, Shigella infection affects mainly children and is especially important in the day care setting. Shigella sonnei accounts for 60–80% of cases in the United States and less commonly causes bloody stools. Shigella flexerni is most common in the developing world and is more frequently associated with grossly bloody stools. Other pathogenic species include Shigella dysenteriae and Shigella boydii. Patients usually complain of fever and watery diarrhea early in the course. Later, the stool volume may decrease and tenesmus with bloody stools (40% of cases) may develop. The duration usually is ≤ 5 days, but occasionally can last weeks. The stool classically has sheets of fecal leukocytes, and the diagnosis is confirmed by stool culture.
Several E coli species (Chapter 53) can lead to diarrhea.
EHEC are Shiga toxin-producing E coli that are an emerging diarrheal pathogen. The intestines of cattle appear to be the major reservoir of the organism. Infection is associated with the ingestion of contaminated meats, vegetables, and water. The meat, especially ground beef, is contaminated by feces with slaughter. Vegetables and water are contaminated by cattle waste used as fertilizer or present in runoff. After an incubation period of 3–4 days, patients develop cramping, abdominal pain, and diarrhea. The stool may then become bloody over several days. Interestingly, EHEC infection may have no associated fever or fecal leukocytes or lactoferrin. Thus, acute bloody diarrhea with no fecal leukocytes or fever should not exclude the consideration of EHEC, which is now the major infectious cause of grossly bloody diarrhea in the United States. Of those infected, ≤ 6% may develop hemolytic uremic syndrome (HUS) 2–14 days after disease onset. The diagnosis is confirmed by isolating E coli O157:H7 by culture or via the demonstration of Shiga toxin from E coli isolates or in the stool (non-O157 serotypes). Enteroinvasive E coli strains cause diarrhea via invasion and induction of an inflammatory reaction in the intestine. Enteroinvasive E coli strains cause a disease similar to that caused by Shigella spp. and are rare in the United States. Enteroaggregative E coli is a recently described pathogen that can cause a persistent, mildly inflammatory diarrhea and malnutrition among people in the developing world and in patients with AIDS in the United States.
Figure 20-1. General approach to the patient with infectious diarrhea. (Reprinted, with permission, from Guerrant R, Bobak D: Bacterial and protozoal gastroenteritis. N Engl J Med 1991;325:327.)
C difficile (Chapter 53) is a major cause of antibiotic-associated and nosocomial diarrhea. The disease normally occurs in the hospital setting, but community-acquired cases are being increasingly recognized. The organism overgrows the colon when the normal flora is altered by antibiotics, certain chemotherapeutic agents, or after abdominal surgery. Two protein toxins, toxins A and B, are released into the colon and lead to the development of diarrhea. The disease ranges from asymptomatic carriage to mild diarrhea to more severe forms of colitis with or without pseudomembrane formation. Pseudomembranes are mucosal plaques comprised of neutrophils and cellular debris, reflecting the inflammatory nature of the toxins. The most severe forms can lead to toxic megacolon and colonic perforation. The diagnosis is made by detecting the toxins in the stool via the methods discussed previously.
V parahaemolyticus and other Vibrio species (Chapter 57) are the major causes of seafood-related diarrhea in the United States, Japan, and the developing world. In the United States, the disease is most common along the Atlantic and Gulf coasts. The consumption of improperly cooked crab and shrimp meat and exposure to contaminated seawater (including the use of seawater to clean cooking utensils on cruise ships) are the main vehicles for disease transmission. In the United States, acute self-limited outbreaks are the rule, with explosive watery diarrhea and cramping abdominal pain developing within 24 h after exposure and lasting for up to several days. The organism will not grow on routine stool cultures; the diagnosis requires that the organism be isolated on thiosulfate citrate bile salts sucrose agar.
Parasites.E histolytica (Chapter 81) is a major protozoan cause of invasive diarrhea, especially in the developing world. It is typically found in lower socioeconomic areas with poor sanitation, in institutionalized patients, and in immigrants from endemic areas. The intestinal disease caused by E histolytica includes asymptomatic infection, noninvasive diarrhea, acute rectocolitis (dysentery), and fulminant colitis with the risk of toxic megacolon and colonic perforation. Amebic colitis is usually associated with heme-positive stools, but fecal leukocytes can be absent owing to the ability of the pathogen to lyse human polymorphonuclear neutrophil leukocytes. Male patients are at greater risk for developing amebic liver abscesses. The diagnosis is made by identifying the parasite in the stool, by commercially available antigen detection assays, or by serology.
Viruses. Cytomegalovirus (Chapter 33) is an important cause of inflammatory diarrhea in compromised patients, especially those with AIDS. AIDS patients with cytomegalovirus enteritis present with watery diarrhea and can develop hematochezia and fever. On endoscopy, the mucosa may show plaquelike pseudomembranes, erosions, or ulcers, and biopsy should reveal the characteristic intranuclear inclusions.
Bacteria. Enterotoxigenic E coli strains (Chapter 53) producing heat-labile enterotoxin (LT) or heat-stable enterotoxin (ST) toxins are a major cause of diarrhea in the developing world and among travelers. Transmission occurs via contaminated food or water. After an incubation period of 1–2 days, the patient develops watery diarrhea (3–10 unformed stools per day) and cramps. The symptoms usually resolve over several days or < 1–2 days if treated early with effective antimicrobial agents. An unfortunate minority will have disease that lasts longer than 1 week. Fever and bloody stools are uncommon.
Vibrio cholerae (Chapter 57) is a well characterized diarrheal pathogen that has caused seven major pandemics since the 1800s. The seventh pandemic spread to Latin America in the early 1990s, and cases have been imported into the United States. The organism is ingested with contaminated food or water, and the infection that results ranges from asymptomatic to the full-blown cholera syndrome with voluminous watery diarrhea. Severe cases can rapidly progress to hypovolemic shock and death in < 12 h. Fever or abdominal pain is usually absent. The stool is often described as similar to rice water in appearance and is rarely bloody. Diagnosis is made clinically and confirmed by culturing the organism from the stool.
Mycobacterium avium complex (Chapter 62) is a common cause of diarrhea in patients with advanced HIV infection. Disseminated M avium complex presents with fever, sweats, weight loss, and, in 40% of cases, watery diarrhea. Diagnosis is made by isolating M avium complex from the stool and blood.
Protozoa.G lamblia (Chapter 84) causes prolonged diarrhea in both immunocompetent and compromised patients. The incubation period is 1–2 weeks after exposure to contaminated water or direct contact with an infected patient. The infection may be asymptomatic or loose, and foul-smelling stools can develop, often associated with cramping, flatulence, nausea, and cramps. The symptoms may persist for weeks to months if untreated. The diagnosis is confirmed by identifying the trophozoites or cysts in the stool or duodenal aspirates or by detecting Giardia antigens in the stool via a commercially available enzyme-linked immunosorbent assay.
Cryptosporidium parvum (Chapter 83) is another highly infectious protozoa that is endemic throughout much of the world, including the United States. The organism is spread via contaminated water and direct patient contact. After an incubation period of 1 week, immunocompetent hosts develop diarrhea, cramps, and weight loss lasting for days to weeks before resolving. Low-grade fever occurs in 50% of infected individuals. In AIDS patients, the organism can cause a severe chronic diarrhea and biliary tract infections. The diagnosis is made by identifying the cysts by a modified acid-fast stain of the stool (often missed on routine O+P) or via a commercially available fluorescence antibody stain.
Cyclospora cayetanensis (Chapter 83) is an emerging protozoal cause of diarrhea associated with contaminated drinking water or foods (raspberries and lettuce). Unlike Cryptosporidium spp., the life cycle of C cayetanesis requires a maturation phase outside the host and is thus less likely to be spread directly by person to person contact. In immunocompetent patients, after an incubation period of 1 week, a self-limited illness with watery diarrhea, cramps, and prominent fatigue ensues. It is a rare cause of diarrhea in AIDS patients in the United States, likely owing to the widespread use of trimethoprim-sulfamethoxazole (TMP-SMX) prophylaxis in these patients. It is, however, a common cause of diarrhea in this population elsewhere in the world. The diagnosis is made by identification of the organism on a modified acid-fast stain of the stool (not identified on routine O+P) or by detecting the oocytes in the stool by fluorescence microscopy (the autofluorescent oocysts without special staining).
Isospora belli (Chapter 83) is endemic in tropical countries, and infection is associated with travel or residence in these regions. Like C cayetenesis, it causes self-limited infections in normal hosts and chronic diarrhea in AIDS patients except in the United States, where it is uncommon probably owing to TMP-SMX use. It is diagnosed by modified acid-fast staining of stool to detect the cysts.
Microsporidia infections (Chapter 83) caused by Enterocytozoon bieneusi and Septata intestinalis are endemic in the United States and worldwide. Infection is commonly encountered in patients with advanced AIDS and causes a prolonged watery diarrhea with frequent relapses. The cysts are not identified by acid-fast stains, but rather require a modified trichrome stain of the stool or small-bowel biopsy.
Viral. Rotaviruses (Chapter 37) are a major cause of gastroenteritis worldwide and are estimated to be responsible for 10–20% of all diarrhea-related deaths. In the United States, millions of infections occur annually leading to > 100,000 pediatric admissions and 200 deaths. The peak incidence of disease is in the 6- to 24-month-old age group, and virtually all children have been infected by age 3. Adults remain susceptible to infection, but tend to have milder symptoms. Spread of the virus appears to be by the fecal-oral route and, for unknown reasons, the infection occurs mainly in the winter months. The disease is characterized by mild fever, watery diarrhea, nausea, and vomiting over 2–5 days. Immunodeficient children and adults are predisposed to developing chronic rotavirus infections with diarrhea that can last for months. The diagnosis is confirmed by rotavirus antigen enzyme-linked immunosorbent assay of the stool or by a less sensitive, more rapid latex agglutination test.
Caliciviruses (Chapter 38), including Norwalk or Norwalk-like viruses, are very common causes of viral gastroenteritis that are spread via the fecal-oral route. Outbreaks have occurred in relation to the consumption of contaminated water, food, or shellfish. Most outbreaks are characterized by an incubation period of 1–2 days, a duration of illness of 1–3 days, prominent vomiting, and a negative stool workup. Watery diarrhea with 4–8 stools/24 h is common, and low-grade fever occurs in 50% of cases. The secondary attack rate in affected households can be very high. Diagnosis is based on clinical features and a lack of identifiable pathogens on laboratory workup. Serologies can be helpful to confirm the diagnosis retrospectively.
BOX 20-1 Microbiology of Infectious Diarrhea
Volume depletion and electrolyte abnormalities remain the most serious sequelae of diarrheal illnesses worldwide. It is especially a problem in the very young and the elderly, owing to a lack of ready access to fluids for rehydration or to underlying diseases that may limit the ability of elderly people to cope with the volume changes. In some diseases, such as cholera, the volume loss may be massive, and the patient can progress to hypovolemic shock within 24 h. Electrolyte losses in the stool are also common with diarrhea, potentially leading to hypokalemia and metabolic acidosis. Profound hypoglycemia, leading to lethargy, coma, and seizures, can occur rarely with severe cholera, especially in children.
Malnutrition is being increasingly recognized as a major complication of repeated or chronic diarrheal illnesses in the developing world. There is growing evidence that children with such infections exhibit growth retardation and possibly cognitive dysfunction compared with matched controls. If future studies confirm this relationship, then malnutrition may well emerge as the most devastating long-term sequela of infectious diarrhea.
Nosocomial infections are increased in hospitalized patients with preceding or ongoing episodes of nosocomial diarrhea. In one study, patients with nosocomial diarrhea had a 10-fold higher risk of nosocomial urinary tract infections as compared with matched controls without diarrhea.
Hemolytic uremic syndrome (HUS) is another complication of enteric pathogens, most notably EHEC. Patients with HUS typically present with renal failure, microangiopathic hemolytic anemia, and thrombocytopenia 2–14 days after the onset of the diarrhea. In prior outbreaks, 6% of infected patients with E coliO157:H7 developed HUS, 50% of whom required dialysis and had mortality as high as 1.2% (> 250 deaths per year in the United States). The risk of developing HUS after EHEC infection appears to be increased with the use of antimotility drugs, and these agents are contraindicated in suspected cases. Antimicrobial therapy may increase the risk of HUS as well, but this remains controversial.
Guillain-Barrésyndrome, an acute demyelinating polyneuropathy, is another potential complication of enteric infections, especially after C jejuni infection. Of patients with Guillain–Barré, 20–40% had a documented C jejuni infection in the prior several weeks. Patients usually present with ascending motor weakness and may require mechanical ventilation for respiratory muscle involvement. The mechanism of how this infection may trigger this immune-mediated syndrome remains unknown.
Postinfectious arthritis has been described several weeks after an invasive diarrheal illness with Campylobacter, Shigella, Salmonella, or Yersinia spp., especially in patients with a human leukocyte antigen-B27 phenotype.
As with the diagnostic approach, optimal therapy for a given diarrheal illness should be guided by a number of different factors. The age and health of the patient; volume status; ability to take oral fluids; presence of fever, blood, leukocytes, or lactoferrin in the stool; and suspected pathogen all need to be considered. For example, an adult in the United States with < 1 day of noninflammatory diarrhea can be treated with oral fluids and monitored. A child with voluminous diarrhea in a cholera-endemic region and significant volume loss will require more aggressive fluid resuscitation and antimicrobial therapy. Finally, a child with bloody diarrhea after consumption of an undercooked hamburger will require the consideration of different sequelae and treatment concerns than those in the first two cases. Given the huge range of presentations, sequelae, and therapies available with the various diarrheal pathogens, it is important to emphasize that therapy must be tailored to each patient based on the available clinical and laboratory data.
Antibacterial therapy should be considered in all patients with diarrhea in whom certain pathogens are suspected, because antibiotics in these cases can decrease the duration of symptoms and prevent serious sequelae. Which organisms should be covered in an individual patient will depend on the available clinical and epidemiologic data. In general, patients with suspected Shigella, Campylobacter, or severe C difficile infection, traveler's diarrhea, or V cholerae infection should be started on empiric therapy once the diagnosis is entertained, because the antimicrobial agents will influence the course of the illness. For Shigella and Campylobacter spp., antimicrobial therapy has been shown to decrease the duration of the diarrhea. Fecal excretion of these organisms is also decreased with antibiotic therapy and in settings where the potential for spread is present; for example, day care is another indication for therapy.
For nontyphoidal Salmonella gastroenteritis, antibiotic therapy may decrease the duration of illness, but paradoxically may prolong fecal excretion of the organism or predispose to systemic invasion. Antimicrobial therapy for Salmonella gastroenteritis should be considered for those patients who may be at increased risk for bacteremic dissemination, including the very young, the elderly, or immunocompromised patients, and for those with an enteric fever syndrome. Otherwise, nontyphoidal Salmonella gastroenteritis is generally self-limited and antimicrobial agents are best avoided. As mentioned, antibiotic therapy is not currently recommended for suspected or confirmed EHEC infections, especially in children. To date, studies are contradictory on whether the incidence of HUS is decreased or increased in patients who had received antibiotics, and further studies are needed to clarify this issue.
C difficile infections should be treated initially by stopping the antimicrobial therapy. If the symptoms persist or if the patient is severely ill and has underlying medical conditions, then specific therapy should be started. Although antibiotic therapy is generally not indicated in noninflammatory diarrhea, antimicrobial agents can shorten the duration of the symptoms and attenuate the volume loss in patients with cholera or traveler's diarrhea. Boxes 20-2 and 20-3 list the suggested therapies (for adults and children) that should be started once the respective pathogens are suspected.
The issue of antimicrobial resistance must be considered when choosing antibiotic therapy for enteric pathogens. For example, Shigella and Salmonella spp. are increasingly resistant to ampicillin and TMP-SMX, whereas Campylobacter resistance to fluoroquinolones is emerging (especially in Thailand). Stool cultures are therefore not only valuable to confirm suspected bacterial pathogens, but also to supply susceptibility patterns for a given isolate. Knowledge of the local resistance patterns, adequate culture specimens before antibiotic use, and monitoring of the patient's clinical response are all required to ensure that the correct antimicrobial therapy has been given.
Specific antiparasitic therapy is currently available for patients with confirmed or suspected G lamblia, E histolytica, I belli, and C cayentanensis infections. No specific therapy has proven consistently effective against Cryptosporidium spp. in AIDS patients; paromomycin has had limited success. In general, resistance has not been a problem with the protozoal pathogens, but relapses can occur with pathogens like Giardia spp.
Traveler's diarrhea, most commonly caused by enterotoxigenic E coli, can be reduced from a duration of 3–5 days to 2 days or less with effective antimicrobial therapy. TMP-SMX and ciprofloxacin hydrochloride are the two most commonly used antibiotics, but resistance to TMP-SMX is now fairly widespread. As with other forms of diarrhea, oral rehydration is critical and should be initiated as soon as the illness begins regardless of the use of antimicrobial agents. Antimotility agents can be used once effective antimicrobial agents have been initiated, especially in patients with dysentery symptoms or bloody stools.
With adequate volume replacement, good supportive care, and antimicrobial therapy if indicated, the prognosis of infectious diarrhea is excellent with minimal morbidity and mortality. As with many illnesses, the morbidity and mortality are concentrated in the very young and the elderly. In the United States, the mortality now associated with infectious diarrhea is significantly < 1.0%. There are exceptions such as EHEC infections, which had an associated mortality of 1.2% in most studies related to the development of hemolytic uremic syndrome. Oral rehydration therapy and improvements in sanitation have begun to have an impact on mortality in the developing world. Nevertheless, considerable work remains to reduce the levels of mortality in these regions to that of the developed world. The biggest improvements likely will come with changes in economic and social conditions rather than from any further improvements in medical therapy.
Prevention & Control
Because most diarrheal pathogens are spread via the fecal-oral route, their acquisition can be prevented by maintaining good personal hygiene. This includes frequent handwashing after toilet use and especially during food preparation. The value of good hygiene by food handlers and cooks to avoid the spread of food-borne pathogens can not be overstated. Diapers of children with diarrhea should be disposed of as soon as possible with minimal contamination of the living environment. Human sewage should be removed from living areas, and farm animals should be kept separate from human dwellings. Nosocomial diarrhea can be reduced by using antibiotics judiciously, by avoiding broad-spectrum antibiotics if possible, and by careful handwashing, which may potentially reduce the spread of C difficile spores within the hospital environment.
BOX 20-2 Empiric Antimicrobial Therapy for Infectious Diarrhea in Adults
BOX 20-3 Empiric Antimicrobial Therapy for Infectious Diarrhea in Children
Because food and water are common vectors for pathogens, these areas should be given special attention. Drinking water, water used to clean food, or water that is used in cooking should be filtered and chlorinated. If there is any concern about water safety or if unpurified water is taken from lakes or rivers, then it should be boiled for several minutes before consumption. When swimming in lakes or rivers, people should take care not to ingest the water. Sea water is often contaminated with Vibrio species, and caution should be exercised when using this water for cooking or washing or when swimming.
All fruits and vegetables should be washed thoroughly with clean (boiled, filtered, or treated) water before consumption, especially given the quantity of produce now imported into the United States. Untreated human or animal waste should not be used to fertilize fruits and vegetables. All meats and seafood should be cooked adequately. Chicken and ground beef should be cooked until the red or pink is gone. Only pasteurized dairy products and juices should be consumed. A recent outbreak of EHEC was associated with drinking unpasteurized apple cider made with apples contaminated, after falling to the ground, by cattle feces.
The prevention of traveler's diarrhea begins with the avoidance of potentially contaminated food or water and with the maintenance of good personal hygiene. All drinking water should be boiled, or carbonated beverages should be consumed. Ice cubes should be avoided, unless made with boiled water. Food should be well cooked, and unpeeled fruits and food from salad bars and street vendors should be avoided. These measures can reduce, but not eliminate, the risk of acquiring traveler's diarrhea and are difficult to strictly adhere to. Bismuth subsalicylate tablets taken 4 times a day for up to 3 weeks and antibiotic prophylaxis with fluoroquinolones are effective in preventing many cases of traveler's diarrhea. However, given the cost, the risk of contributing to resistance, drug side effects, and the fact that a short course of antimicrobial agents can significantly ameliorate traveler's diarrhea if it occurs, such prophylaxis is not currently recommended by most experts.
Vaccines are a promising area for the prevention of infectious diarrhea, but the effectiveness and scope of the currently available vaccines are limited. At present, vaccines are available for V cholerae, typhoid fever, and rotavirus. The current parenteral cholera vaccine is not very effective and not recommended for use. Newer oral cholera vaccines based on the toxin B-subunit or inactivated whole cells are more effective, and the immunity is of longer duration. However, neither is currently licensed in the United States. The older parenteral typhoid vaccine was only 70% effective and had frequent side effects. A newer parenteral vaccine based on the capsular polysaccharide Vi-antigen is now available in the United States. It is also 70% protective, but requires only 1 dose and has fewer side effects. An oral, live-attenuated typhoid vaccine is available, requiring 1 capsule every other day for 4 doses and having an efficacy similar to that of the other two vaccines. The recently approved oral rotavirus vaccine covers serotypes 1–4 and is given to infants at 2, 4, and 6 months of age. In several international trials, the vaccine decreased the incidence of gastroenteritis by 50%, severe disease by 75%, and hospitalization for dehydration by almost 100%. Fever was the most common side effect, especially in the first week after the administration, but overall the vaccine was very safe. More recent data, however, have shown an increased risk of intussusception in infants vaccinated in the prio 1–2 weeks. Vaccination in infants is no longer recommended.
Choi SW et al: To culture or not to culture: fecal lactoferrin screening for inflammatory bacterial diarrhea. J Clin Microbiol 1996;34:928.
Dupont H et al: Guidelines on acute infectious diarrhea in adults. Am J Gastroenterol 1997;92:1962.
Griffin P: Escherichia coli O157:H7 and other enterohemorrhagic E coli. In Blaser M et al: Infections of the Gastrointestinal Tract. Raven, 1995.
Guerrant R: Lessons from diarrheal illnesses: demography to molecular pharmacology. J Infect Dis 1994;169:1206.
Guerrant R, Bobak D: Bacterial and protozoal gastroenteritis. N Engl J Med 1991;325:327.
Gueirent RL,0 von Gilder T, et al: Practice guidelines for the management of infectious diarrhea. Clin Infect Dis 2001; 32:331.
Fekety R: Guidelines for the diagnosis and management of Clostridium difficile-associated diarrhea and colitis. Am J Gastroenterol 1997;92:739.
Hines J, Nachamkin I: Effective use of the clinical microbiology laboratory for diagnosing diarrheal diseases. Clin Infect Dis 1996;23:1292.
Mackenzie W et al: A massive outbreak in Milwaukee of Cryptosporidium infection transmitted through the public water supply. N Engl J Med 1994;331:161.