Lonny M. Hecker, M.D., David R. Saunders, M.D., and David Losh, M.D.

1.0 Introduction

Diarrhea affects most individuals at some time during their lives. The occasional loose or watery stool is so common that few individuals seek medical advice, unless the diarrhea is persistent. In this chapter diarrhea is divided into acute and chronic categories based on whether it has been present for less than or greater than four weeks.

This chapter will present evidenced based recommendations for the evaluation and treatment of diarrhea in adults. The information provided is not necessarily applicable to children and should not be used to guide pediatric practice. It includes important papers published prior to 1990 and those accessed via a MEDLINE search, 1990 - 1998, using the MeSH headings Diarrhea, acute, and diarrhea, chronic. Other headings searched included malabsorptive syndrome, and Fordtran, JS, whose research group has made so many contributions to the subject.

2.0 Physiology and Pathophysiology

Diarrhea results when the remarkable efficiency of the gut for absorbing water, electrolyte, and nutrients is impaired. About 9-10 liters of water and electrolyte enter the upper jejunum daily, of which one liter is delivered to the cecum, and one-tenth of a liter is delivered to the outside world. Decreasing this efficiency from 99% to 98% would double fecal water to produce potentially a wetter stool. A great variety of drugs, toxins, pathogens, and food stuffs can impair the efficiency of salt and water absorption.

The chief contribution of the stomach to digestion and absorption is metered delivery of food and drink to the small intestine so that the absorptive capacity of the upper small intestine is not overwhelmed.

Carbohydrate and protein in the small and large intestines are especially important in increasing the efficiency of sodium and water absorption. Soluble starches are digested by pancreatic amylase into small chains of glucose molecules which, together with the ingested disaccharides (lactose and sucrose), are hydrolyzed to monomers by brush border enzymes.

Absorption of sodium (and water) is coupled to the absorption of glucose and galactose, especially in the duodenum and jejunum. Much of the available sugars have been absorbed when chyme arrives in the ileum, where sodium absorption relies on sodium/hydrogen and chloride/bicarbonate exhangers. Carbohydrate which escapes absorption in the small intestine is fermented by colonic bacteria to short-chain fatty acids whose colonic absorption enhances sodium (and water) transport and provides nutrients for colonic absorptive cells. By the time feces reach the left colon, most of the available carbohydrate has been fermented so that sodium absorption becomes dependent on exhangers and on sodium - channels.

Dietary protein also enhances sodium and water absorption by mechanisms similar to those described for carbohydrate. Amino acids and sodium are absorbed by coupled transport, and short-chain fatty acids derived from amino acids in the right colon enhance sodium (and water) absorption.

Colonic bacteria do not salvage appreciable amounts of unabsorbed long-chain fatty acids (LCFA). In fact, the double bonds of dietary LCFA may be hydroxylated so that the excreted LCFA bears little resemblance to the dietary LCFA,

and they may become more potent inhibitors of colonic absorption.

A final consideration is the mouth-to-anus transit time which can be derived by measuring the transit of the head of the meal (HOMTT), or of the whole meal (WMTT). WMTT involves ingesting a number of radio-opaque, or isotopically-labeled pellets whose average mouth-to-anus transit time is calculated. WMTT is 48-72 hours in normal subjects [Cummings, 1976], [Metcalf, 1987]. Pellets have the longest residence in the colon, and fecal weights are inversely proportional to the time of colonic residence [Vassallo, 1992]. Head of meal transit time (HOMTT) is measured with a poorly-absorbed colored substance, and it is the time between ingestion and the first appearance of the color in the stools. HOMTT averaged 36 hours after 14 healthy subjects ingested carmine red with an English breakfast [Read, 1980]; diarrhea ensued when the HOMTT was experimentally reduced to less than 12 hours [Read, 1980].

The punch-line: the overall balance for the absorption of sodium is 99%; of starch, 99%; of protein, 95%; and of LCFA, 95%, and these remarkable efficiencies depend on adequate lumenal digestion, absorptive cell surface, and transit time.

Mechanistically, absorption may be impaired by poorly absorbed, osmotically active solutes in the intestinal lumen, by alteration in absorptive cell function, by increases in crypt cell secretion, and by too rapid transit of intestinal contents. Most often, absorption is impaired by mechanisms acting in concert. For example, excessive volume of intestinal contents can speed intestinal transit; cytokines from mural inflammatory cells can enhance cryptal secretion, and can influence the enteric nervous system to speed transit; bile salts, and long-chain fatty acids, malabsorbed in the small intestine, can block water and electrolyte absorption in the colon.

The colon employs several mechanisms to ensure it delivers to the rectosigmoid a formed stool, probably the most important factor in fecal continence. The colon has reserve capacity by which it can absorb 2-3 extra liters of water and electrolyte delivered from the small intestine in a day [Debongnie, 1978].

Colonic bacteria ferment soluble carbohydrate and protein, which escaped small intestinal absorption, into absorbable gases and short-chain fatty acids. Otherwise, these unfermented, unabsorbable solutes would be osmotically active in colonic contents, and would cause diarrhea.

3.0 Acute Diarrhea

Acute diarrhea may be defined clinically as any sudden and significant increase in the frequency or decrease in the consistency of the stool lasting less than four weeks. Acute enteric illness is preferable to gastroenteritis, since the underlying pathophysiology may not involve any inflammation [Guerrant, 1985]. Most patients presenting with acute diarrhea will improve spontaneously without any specific tests or treatment. These patients must be readily distinguished from those at risk who require urgent intervention. It is, therefore, incumbent on the primary care provider to carefully select both diagnostic tests and therapeutic modalities. This is one area of clinical medicine where careful history and physical are still of paramount importance.

3.1 Background

3.1.1 Incidence

The overall incidence of acute diarrhea in adults is difficult to determine. Most episodes are self-limited and probably do not involve the health care provider. Epidemiologists have examined the impact of acute diarrheal disease on the health care system by looking at the number of adult patients hospitalized with the diagnosis of diarrhea using the International Classification of Diseases, Ninth revision (ICD-9) codes for both infectious and presumed non-infectious diarrhea. One study, in the southeast United States found that, of a total 4.06 million hospitalizations in 1985, 2.1% listed diarrhea among the top three-discharge diagnoses [Gangarosa, 1992]. Diarrhea was strongly associated with 9% of all hospitalizations of children and 1.5% of hospitalizations of adults.

Certain circumstances are associated with an especially high incidence of acute diarrheal disease. Operation Desert Shield involved the deployment of 200,000 US military personnel in Saudi Arabia. The rate of diarrheal disease in this group soared to as much as 100 cases per 1000 soldiers per week in some units [Hyams, 1991]. 57% of surveyed troops reported at least one episode of diarrhea and 20% reported to be temporarily disabled due to diarrhea. Enterotoxigenic E. coli, Shigella and Norwalk virus appeared to be the most common etiological agents.

3.1.2 Morbidity and Mortality

Diarrhea is a major cause of childhood disease in the developing world. Global mortality estimates from diarrhea and its complications range from 1.5 to 5.1 million deaths per year for children under the age of five [Bern, 1992]. Acute diarrheal illness is also associated with adult mortality in the United States. In fact, age over 70 years is the most important risk factor for death related to diarrhea.

The National Center for Health Statistics data indicates more than 25,000 persons died of diarrhea in the US between 1979 and 1987 [Lew, 1991]. Seventy-eight per cent of these individuals were 55 years of age or older. Being white, female and residing in a nursing home were independent risk factors for death among the elderly.

3.1.3 Foodborne Disease

Foodborne-disease is a very important cause of acute diarrheal illness. It is difficult to reliably estimate the true incidence, because most episodes are not reported. The Centers for Disease Control maintains a surveillance program to track the occurrence of foodborne-disease outbreaks (FBDOs), defined as the occurrence of two or more cases of a similar illness resulting from the ingestion of a common food [Bean, 1996].

Between 1988 and 1992 a total of 2,423 FBDOs were reported. The etiology was determined in only 41% of all outbreaks. Many of those without a defined cause may have been viral because appropriate testing may not have been available. The CDC also tracks the vehicle of transmission for FBDO’s where the most common single sources are chicken, eggs, and salads made of poultry, fish or eggs. Recent outbreaks and the increasing reliance on commercially prepared foods with nationwide distributions have contributed to increased concern with food safety [Hedberg, 1994].

3.1.4 Costs

The economic impact is considerable as well. The Food and Drug Administration estimated that in 1985 acute intestinal infection causing vomiting, diarrhea, or both, resulted in more than $1 billion in medical costs and more than $20 billion in lost productivity [Garthright, 1988].

3.2 Differential Diagnosis

The differential diagnosis of acute diarrhea in adults is listed in (Table 1). The differential diagnosis can be narrowed based on the predominant symptoms at presentation. The three major symptom complexes are as follows:

Acute diarrhea occurring in travelers and persons with AIDS warrants certain special considerations and will be discussed later.

3.2.1 Acute Watery Diarrhea with Nausea and Vomiting

Acute watery diarrhea associated with nausea and vomiting is typically called gastroenteritis. Most cases are infectious, due to either bacteria or viruses. Staphylococcus aureus

Staphylococcus aureus infection is characterized by the sudden onset of abdominal pain, nausea, vomiting, and diarrhea and in a minority of cases fever (Table 2). Symptoms are due to ingestion of preformed toxin which is usually in foods which are cooked and then improperly stored at room temperature [Holmberg, 1984]. Antimicrobial therapy is not recommended for S. aureus associated diarrhea because symptoms are due to preformed toxin and are self-limited. Viral Gastroenteritis

Acute diarrhea, vomiting and fever are the hallmarks of viral gastroenteritis, which accounts for 30-40% of all acute diarrheal illness in the US (Table 2). Symptoms result from reduced absorptive surface in the small intestine due to viral mediated lysis of enterocytes. Most viral gastroenteritis is caused by one of five viruses. Rotavirus accounts for 30-60% of all severe watery diarrhea in infants and young children and is a major cause of morbidity and mortality worldwide. Rotavirus, enteric Adenovirus, Calcivirus and Astrovirus are uncommon causes of diarrhea in adults in the United States. Norwalk virus, however, is much more common in adults and older children. It is more common during the winter and may be found in contaminated drinking or swimming water, poorly cooked shellfish, and contaminated foods [Kohn, 1995]. Norwalk virus has been responsible for nearly 40% of all outbreaks of acute vomiting and diarrhea in schools, camps, hospitals, nursing homes and on cruises [Blacklow, 1991]. The virus is spread via the fecal-oral route, though airborne transmission has also been suggested. Norwalk virus infection causes vomiting, cramps headache and diarrhea beginning 1-2 days after exposure and usually lasting 1-2 days. Norwalk viral infection is associated with marked delayed gastric emptying and small intestinal mucosal damage causing both carbohydrate and fat malabsorption. Symptoms may persist for 2 weeks. Specific diagnostic tests remain research tools at the present time. Symptoms are self-limited and there is no effective or recommended anti-viral therapy. Non-Infectious Causes of Acute Diarrhea and Vomiting

There are also non-infectious causes of watery diarrhea and vomiting. For example, acute heavy metal poisoning due to ingestion of copper, zinc, iron or cadmium may cause nausea, vomiting, cramps and diarrhea occurring 5 to 60 minutes after ingestion [Bishai, 1993].

3.2.2 NonInflammatory Diarrhea

The hallmark of noninflammatory diarrhea or watery nonbloody diarrhea, is large volume stools. Upper gastrointestinal symptoms including nausea, vomiting and cramps may occur but significant abdominal pain and fever are usually absent. Acute noninflammatory diarrhea is usually associated with small bowel infections. Leukocytes are typically absent from the stool. Bacterial enteric pathogens, which cause acute nonbloody diarrhea in adults, include Vibrio cholerae, enterotoxigenic E. coli (ETEC), Clostridium perfringens and Bacillus cereus. Commonly implicated protozoa include Giardia lamblia, Cryptosporidium parvum, Cyclospora and Dientamoeba fragilis (Table 3). Frequent noninfectious causes of acute nonbloody watery diarrhea include medications, poorly absorbed carbohydrates (lactose, sorbitol, xylitol), caffeine, and alcohol. Vibrio cholera

Vibrio cholera is a gram-negative bacteria found in marine or brackish waters. Seafood, especially crabs, oysters and shrimp, is a natural reservoir. Cholera toxin binds to mucosal cells causing a secretory diarrhea mediated by cyclic AMP. Symptoms may range from asymptomatic infection to profuse watery diarrhea. Individuals with impaired gastric acid secretion are at increased risk. The syndrome of "cholera gravis" is characterized by "rice water" stool with volumes over 1L/hour leading to rapid dehydration and death. Fortunately, cases of mild to moderate disease are 4 - 5 times more common [Morris, 1985]. Rehydration is the mainstay of treatment although antibiotics may shorten the duration of symptoms and excretion of bacteria [Farthing, 1996]. Vibrio parahaemolyticus

Vibrio parahaemolyticus causes hemolysis associated with acute self-limited and sometimes bloody diarrhea, vomiting and fever. Vibrio vulnificus may cause diarrhea but is far more commonly reported as a cause of primary septicemia in individuals with underlying liver disease [Klontz, 1988]. Other noncholera vibrios have also been reported causes of food poisonings and traveler’s diarrhea. Enterotoxigenic E. coli (ETEC)

Enterotoxigenic E. coli (ETEC) is the leading bacterial cause of travelers diarrhea. It can also be a cause of domestic outbreaks. ETEC infection is transmitted by ingestion of contaminated food or water. Disease results from adherence and colonization of the small intestine and the subsequent elaboration of two distinct enterotoxins (heat-labile, LT and heat-stable, ST). ETEC is responsible for the classical syndrome of traveler’s diarrhea including watery diarrhea, nausea, abdominal cramps and low grade fever [Levine, 1987]. Definitive diagnosis remains largely confined to research laboratories. It requires identification of a specific toxin by EIA (enzyme immunoassay) or DNA probe of the toxin gene. Empiric therapy, in the appropriate setting, is very common but often unwarranted. Clostridium perfringens

Clostridium perfringens is a spore forming, enterotoxin producing, gram positive bacillus. It is a frequent cause of food poisoning in institutions serving cooked but improperly stored meats [Lund, 1990]. It has also been reported as a cause of sporadic and nosocomial diarrhea. Symptoms are often mild but may include watery diarrhea and abdominal pain lasting 12-24 hours. The diagnosis depends on detection of enterotoxin in the stool. Antimicrobial therapy is generally not recommended. Bacillus cereus

Bacillus cereus is an aerobic gram positive rod which can produce two different toxins responsible for two distinct clinical syndromes. The emetic syndrome associated with ingestion of preformed toxin causes sudden onset of nausea, vomiting and abdominal pain. Improperly stored rice and pasta are the most commonly reported sources. B. cereus emetic toxin has also been associated with fulminant liver failure and death [Mahler, 1997]. The diarrheal syndrome, in contrast, is due to elaboration of toxin within the lumen of the small intestine. This toxin causes profuse watery diarrhea and abdominal cramps lasting approximately 24 hours. Antimicrobial therapy is not recommended for B. cereus infections. Giardia lamblia

Giardia lamblia is the most common intestinal protozoal pathogen in the industrial world causing 7% of all cases of acute diarrhea [Farthing, 1993]. The trophozoites localize in the small intestine and cause diarrhea. Contaminated food, water or direct person to person contact may transmit Giardia. Fecal-oral spread is prevalent in day care centers and residential institutions [Pickering, 1984]. Most infections are clinically inapparent though symptoms may include watery diarrhea, nausea, anorexia and bloating. Chronic infection may present with symptoms that mimic irritable bowel syndrome. The sensitivity of the microscopic examination of stool for ova and parasites is unacceptably low. A commercially available ELISA for the detection of G. lamblia associated antigens has sensitivity and specificity of >90% and has emerged as the test of choice [Addiss, 1991]. Treatment with metronidazole 250mg qid for 7 days is generally recommended [DuPont, 1997]. Cryptosproridium parvum

Cryptosproridium parvum, a coccidian protozoa became recognized as a worldwide pathogen after its description in patients with AIDS. Improved diagnostic techniques have lead to a marked increase in the frequency of diagnosis in both healthy and immunocomprimised individuals. Groups at increased risk include day care age children [Tangermann, 1991], travelers [Jokipii, 1985] and the elderly [Bannister, 1989]. In 1993 a large outbreak of acute watery diarrhea was traced to Cryptosporidium transmitted through the water supply due to inadequate filtration [MacKenzie, 1994]. Since then four foodborne outbreaks have been reported associated with either fresh pressed apple cider or chicken salad. The potential for outbreaks is significant, because the oocysts are fastidious and only a small inoculum (100 organisms) is needed to produce disease. Cryptosporidiosis has been associated with relentless and sometimes fatal diarrhea in AIDS. Healthy persons infected with cryptosporidium develop self-limited watery diarrhea, cramps, and nausea. Associated symptoms may include fatigue, low-grade fever, myalgias and headaches. The incubation period (median of one week) is similar to Giardia but the duration of illness is considerably shorter. Diagnosis is based on the finding of oocysts on acid-fast stains of concentrated stool. Commercially available enzyme immunoassays (EIA) to detect Cryptosporidium specific antigen promise improved sensitivity and specificity [Dagan, 1995]. No antimicrobial has proven efficacious though, paromomycin has been used successfully in persistent cases and in immunocomprimised hosts [Bissuel, 1994]. Cyclospora

Cyclospora, another coccidian protozoa, was first positively identified as a human enteric pathogen in 1993 [Ortega, 1993]. Cyclospora is similar to Cryptosporidium in its morphological appearance, staining characteristics and effect on small bowel mucosa. Cyclospora is distinguished from Cryptosporidium by its larger size. Both cause self limited diarrhea in healthy individuals and intractable diarrhea in persons with AIDS [Dagan, 1995]. The first reported outbreak of diarrheal illness due to Cyclospora occurred among housestaff physicians living in a hospital dormitory who were exposed to contaminated tap water [Huang, 1995]. Symptoms included explosive watery diarrhea, cramps, and myalgias and less frequently nausea, vomiting and low grade fevers. Cycles of remission and relapse often last several weeks. Treatment with trimethoprim-sulphamethoxazole for 7 days has been shown to be effective in a double-blinded randomized controlled trial [Hoge, 1995]. Dientamoeba fragilis

Dientamoeba fragilis, another protozoa previously thought to be harmless is now recognized as a true diarrheal pathogen. Symptoms typically include diarrhea and abdominal pain. Microscopic examination of 3 fresh stools preserved immediately with polyvinyl alcohol fixative and stained appropriately has a sensitivity of 70-85% [Butler, 1996]. D. fragilis infection is highly associated with and probably transmitted by Enterobius vermicularis or pinworm. Positive identification of D. fragilis should prompt one to look for and treat E. vermicularis. Diiodohydroxyquin, is often recommended though metronidazole and tetracycline are also effective [Butler, 1996]. Nonpathogenic Organisms

Normally the stool contains many nonpathogenic organisms. Some have suggested that Candida species when present in high concentration may cause watery diarrhea. Several reports have presented cases wherein individuals with diarrhea and no other identifiable enteric pathogen are cured after eradication of Candida with antifungal medications [Gupta, 1990]. Blastocystis hominis similarly has often been presented as a cause of diarrhea when large numbers are present in the stool. To date there is insufficient evidence to definitively implicate either of these organisms as causes of diarrhea [Shlim, 1995].

3.2.3 Acute Inflammatory Diarrhea

Acute inflammatory diarrhea typically presents as frequent bowel movements often accompanied by blood, left lower quadrant cramping, abdominal pain and urgency. Fecal leukocytes are often present. The typical pathogens associated with inflammatory diarrhea include Salmonella, Shigella, Campylobacter jejuni, enterohemorrhagic E. coli 0157:H7 (EHEC), enteroinvasive E. coli (EIEC) and Clostridium difficile. Less common causes include Yersinia, E. histolytica, Aeromonas species and Plesiomonas shigelloides (Table 4). Salmonella

Salmonella species are gram-negative rod shaped bacteria. S. typhi is the principal cause of enteric or typhoid fever causing fever, delirium and abdominal pain. Typhoidal salmonella is not a particularly important cause of diarrhea. The nontyphoidal Salmonella species, however include some of the most important causes of foodborne illness in the U.S. [Chalker, 1988]. They may account for more than 25% of all reported cases of FBDO and 50% of all FBDO deaths in which an organism can be identified [Bean, 1996] (Table 5). Approximately 40,000 cases of nontyphoidal Salmonella are reported to the CDC annually, probably representing only a small fraction of the actual incidence. Most transmissions have been traced to raw or partially cooked eggs, poultry, beef, pork or milk. Documented sources have also included contaminated medical equipment, marijuana, and pets. Salmonella infections cause diarrhea principally by direct invasion of both small and large intestinal mucosa, however an enterotoxin, which causes secretory diarrhea, has also been described. Salmonella enteritis usually presents with watery diarrhea and abdominal pain. Grossly bloody diarrhea is surprisingly uncommon. Constitutional symptoms including headache, fever, chills and malaise may occur [Baird-Parker, 1990]. Salmonella enteritis is usually self-limited with resolution of diarrhea within one week. While fecal leukocytes are common, the diagnosis is based upon isolation of Salmonella from the stool. Antibiotic therapy has not been proven to shorten the course of the acute illness and may prolong excretion and thereby transmission of the organism [Neill, 1991]. Antibiotic treatment is therefore not recommended for healthy persons with mild symptoms. Antimicrobial therapy is reserved for patients who are bacteremic or at increased risk of the consequences of bacteremia including extremes of age and those with sickle cell disease, AIDS, cancer, prosthetic valves or other serious underlying disease [DuPont, 1997]. Shigella

Shigellae include the four species; S. dysenteraie, S. flexneri, S. boydii and S. sonnei. Shigellae are gram negative bacilli found in contaminated food and water. Fecal-oral and person to person transmission is common. High rates of infection are associated with poor sanitation and overcrowding. Day-care facilities are a frequent location of outbreaks in the US with S. sonnei as the predominate species. A very low inoculum is required to transmit infection and an attack rate of up to 20% of household contacts has been reported. Shigellosis is the classic example of bacillary dysentery characterized by fever, abdominal cramps and watery diarrhea rapidly progressing to bloody mucoid stool, fecal urgency and tenesmus. Shigella causes disease by invasion and destruction of colonic mucosa. Its cytotoxic activity is likely related to elaboration of an enterotoxin. S. dysenteriae serotype 1 is the most severe and has been associated with hemolytic-uremic syndrome. Shigellosis is typically self-limited but may be associated with severe complications particularly in infants and those with altered immune status. Microscopic stool examination will typically reveal sheets of polymorphonuclear leukocytes. Sigmoidoscopy may reveal nonspecific signs of colitis with increased rectal erythema, friability, and sometimes ulceration. Definitive diagnosis depends on stool culture. Though caution should be exercised, recent evidence suggests antimotility agents such as loperamide may be used safely in patients with dysentery [Murphy, 1993]. Antibiotic therapy is always indicated for confirmed Shigellosis in order to shorten the duration of illness and the likelihood of transmission. Trimethoprim/sulfamethoxazole is adequate if the infection is acquired in the US but because of high levels of resistance, a quinolone (Table 14) is necessary for persons acquiring the disease elsewhere [DuPont, 1997], [Tauxe, 1990]. Campylobacter

Campylobacter are mobile gram-negative rods. Campylobacter jejunii is a common cause of diarrhea in the United States [Blaser, 1983]. Most cases occur following ingestion of incompletely cooked chicken or cross contamination of other foods because of improper handling of raw chicken. Person to person transmission and outbreaks in day-care centers are uncommon. The mechanism of disease is believed to be direct tissue invasion of the small and large intestine causing a nonspecific enterocolitis. Several enterotoxins have been described but their pathogenic significance is unknown. The most common clinical manifestations include fever, abdominal pain and diarrhea, which may be loose, watery, or bloody. The illness generally lasts up to a week. Complications are uncommon but relapses can occur [Skirrow, 1990]. Fecal leukocytes are common and sigmoidoscopy will typically reveal changes of acute colitis. The diagnosis is made on stool culture. Antibiotic treatment is recommended to reduce the duration of symptoms. Erythromycin is recommended though azithromycin may be effective as well [DuPont, 1997], [Kuschner, 1995]. Significant quinolone resistance has been reported. Escherichia coli

There are six major groups of diarrheagenic Escherichia coli. Enteropathogenic E. coli (EPEC) causes watery diarrhea in infants and children. Enterohemorrhagic (EHEC) and enteroinvasive (EIEC) strains both produce bloody diarrhea (see below). enterotoxigenic E. coli (ETEC) is the major cause of traveler’s diarrhea (see below). Less common enteroaggregative E. coli (EaggEC) and enteroadherent E. coli (EAEC) have been reported as causes of travelers diarrhea and AIDS associated diarrhea. Escherichia coli O157:H7 (EHEC)

Escherichia coli O157:H7 was first recognized as a cause of diarrhea in humans in 1982 [Riley, 1983]. It is now well known as one of the most common bacterial causes of diarrhea. Numerous outbreaks have occurred with potentially devastating complications, especially in young children. The term enterhemorrhagic E. coli (EHEC) is used to describe E. coli serotype O157:H7 because of its propensity to produce hemorrhagic colitis . The pathogenesis of E. coli O157:H7 is due to the production of a Shiga-like toxin, which causes a vasculitis responsible for the colitis and the hemolytic-uremic syndrome (HUS). In one study E. coli O157:H7 was the most commonly isolated organism from stool specimens that were visibly bloody [Slutsker, 1997]. Outbreaks have occurred related to poorly cooked meat in fast-food restaurants, unpasteurized apple cider, swimming in public pools and lakes, day care centers and consumption of raw vegetables including potatoes and alfalfa sprouts [Su, 1995]. Person to person transmission during outbreaks is common particularly in day care centers and nursing homes. The clinical manifestations of E. coli O157:H7 typically include the sudden onset of abdominal cramps followed by watery diarrhea, which becomes bloody. Diarrhea usually lasts 2 to 14 days. The clinical presentation, endoscopic findings and histology of mucosal biopsies are identical to that of ischemic colitis. Nausea, vomiting and fever are common. HUS characterized by microangiopathic hemolytic anemia, thrombocytopenia and renal failure complicates 2 - 7 % of E. coli O157:H7 cases [Griffin, 1991]. Progression to HUS and thrombocytopenic thrombotic purpura occurs predominantly in young children. Only very recently, have most clinical laboratories begun to identify E.coli O157:H7 routinely on stool cultures. Antimicrobials have not been shown to be effective and are not recommended. Enteroinvasive E. Coli (EIEC)

Enteroinvasive E. coli (EIEC) is distinguished by the presence of a plasmid containing genes that code for Shiga-like toxin which permits invasion, proliferation and ultimately destruction of epithelial cells. EIEC is contracted by ingestion of contaminated food or water. Though it has been responsible for outbreaks in both adults and children, it is much less common than either ETEC or EHEC. The clinical presentation closely resembles that of Shigellosis with fever, abdominal cramps, tenesmus and bloody diarrhea lasting 5 to 7 days. Most clinical laboratories do not test for EIEC which can be identified by culture and specialized serogrouping, or PCR. Treatment recommendations are identical to that proposed for Shigellosis. No controlled trials have examined the role of anti-microbial therapy though use of trimethoprim and sulfamethoxazole has been suggested [DuPont, 1997]. Clostridium difficile

Clostridium difficile associated diarrhea accounts for only 15-20% of all antibiotic associated diarrhea [Kelly, 1994]. In fact, most antibiotic associated diarrhea is mild, self limited and not associated with C. difficile [Fekety, 1997]. C. difficile is a spore forming obligate anaerobic bacillus found ubiquitously in soil, water and health care institutions. C. difficile rarely invades intestinal mucosa. Instead, it causes intestinal damage via the elaboration of 2 toxins which adhere to the mucosal surface. Most strains of C. difficile elaborate both toxins A and B. The two toxins vary in potency, however both act to cause disease. Approximately 8% of all patients admitted to the hospital are asymptomatic carriers of C. difficile [McFarland, 1990]. C. difficile associated diarrhea occurs sporadically and in outbreaks. Almost every antibiotic has been associated with C. difficile diarrhea. The most frequently involved are ampicillin, other penicillin derivatives, cephalosporins, and clindamycin. Even metronidazole and vancomycin have been implicated as causes of C. difficile diarrhea. C. difficile associated diarrhea may not manifest itself until 6-8 weeks after the discontinuation of antibiotics. It has also been associated with chemotherapeutic agents and less frequently with underlying leukemia and lymphoma without prior exposure to antibiotics or chemotherapy. Patients typically present with cramps and watery diarrhea. Occult bleeding may be noted, but significant hematochezia is rare [Kelly, 1994]. Colitis more commonly involves the left colon but isolated right sided colitis may occur and can cause marked right sided abdominal pain and tenderness with little diarrhea. The term "C. difficile associated diarrhea" may be used to describe the entire spectrum of disease from a purely secretory diarrhea to colitis with or without pseudomembranes and even fulminant colitis. Complications may include dehydration, electrolyte disturbances, reactive arthritis, toxic megacolon and colonic perforation. Clinical signs and symptoms can include fever, abdominal tenderness, and profound leukocytosis. Flexible sigmoidoscopy may reveal a nonspecific colitis. Pseudomembranes, when present, appear as yellowish plaques which may become confluent. C. difficile Diagnosis

The gold standard for the diagnosis is the stool cytotoxin test with sensitivities of 94-100% and specificities of 99% [Doern, 1992]. This is a tissue culture assay wherein toxin B in stool causes cell lysis. Several commercially available simple, rapid and inexpensive test have become available for the detection of C. difficile in the stool. Enzyme immunoassays (EIA) detect toxins A and/or B in the stool. Sensitivities reported range from 60% -95%, depending on the individual lab [Jacobs, 1996]. Most studies have compared EIA to clinical or endoscopic diagnosis rather than to stool cytotoxin tests. EIA tests may fail to diagnose on the average 10% (range 5-33%) of clinically apparent C. difficile diarrhea [Fekety, 1997], [Jacobs, 1996]. The latex agglutination test detects glutamate dehydrogenase produced by C. difficile. It does not detect toxin A or B and therefore cannot distinguish between toxigenic and nontoxigenic strains of C. difficile [DiPersio, 1991]. Stool culture for C. difficile is not recommended because the frequency of nontoxigenic C. difficile may approach 25%. Difficulties arise because tests are reported simply as positive or negative and clinical decisions are commonly made without knowledge of the assay used or the sensitivities and specificities in the particular laboratory performing the test.

False negative C. difficile tests may occur with any of the available methods. Recently published practice guidelines recommend additional stool specimens be submitted for the same or other testing methods if the initial test is negative and the clinical suspicion is strong. This approach may have clinical merit but there are no available studies evaluating its utility. Overall, stool testing for C. difficile in hospitalized patients with acute diarrhea has a much higher yield (20%) than either stool culture for enteric pathogens or examination for ova and parasites (1-3%).

There is abundant clinical experience regarding the utility of sigmoidoscopy in the diagnosis of C. difficile associated diarrhea. Flexible sigmoidoscopy in a patient without enema preparation may reveal the characteristic appearance of pseudomembranous colitis and allow a rapid diagnosis when sensitive stool testing is either delayed or unavailable. The utility of sigmoidoscopy compared to stool testing has not been subjected to scientific study, however.

The diagnosis of C. difficile associated diarrhea should be suspected in anyone who was recently hospitalized or received antibiotics. Data regarding the epidemiology of community acquired C. difficile associated diarrhea is limited. A recent retrospective cohort study by the Harvard Community Health Plan found an overall incidence of 7.7 cases/100,000 person years [Hirschhorn, 1994]. The overall risk was <1 in 10,000 antibiotic prescriptions and 82% of cases were treated as outpatients. Risk factors included increasing age, and exposure to more than one antibiotic. Several well designed studies have demonstrated C. difficile associated diarrhea is a nosocomial infection. A prospective comprehensive study of 487 patients admitted to a general medicine ward found 29/428 (7%) of patients had positive C. difficile cultures on admission. 21% acquired C. difficile while in the hospital and of those 31% developed antibiotic associated diarrhea [McFarland, 1990]. Diarrhea, fever, leukocytosis and fecal leukocytes were often absent even in the most severe cases. Risk factors for the acquisition of C. difficile associated diarrhea by hospitalized patients included increasing age, severity of underlying illness, use of cephalosporins, penicillins, enemas, gastrointestinal stimulants and stool softeners.

Recently published practice guidelines advise testing for C. difficile in anyone who has received antibiotics within the prior two months and/or whose diarrhea began 72 hours or more after hospitalization [Fekety, 1997] (Table 6). The use of a two month time limit for prior antibiotic exposure appears to be reasonable based on clinical experience, however there is insufficient evidence to support this as a strict cutoff. The recommendation that C. difficile be considered only in those with diarrhea beginning 72 hours after hospitalization does not appear to be supported by available studies. C. difficile Treatment

Diarrhea may resolve obviating the need for specific treatment if antibiotics can be discontinued. Similarly, mild C. difficile associated diarrhea may resolve spontaneously and does not need to be treated. Moderate to severe diarrhea should be treated with oral metronidazole (250mg qid for 10 days). Oral vancomycin (125 mg qid for 10 days) is more expensive and should be reserved for pregnant women, children under age ten and patients who are critically ill due to pseudomembranous colitis [Fekety, 1997]. It may also be prescribed for patients who do not respond, cannot tolerate or are infected with organisms that are resistant to metronidazole.

Clinical relapse will occur in 10 to 20% of patients who have initially responded to therapy. Relapse will usually occur within three weeks after completion of therapy. Relapses should be confirmed with repeat stool C. difficile toxin assay. Asymptomatic patients should not be retested. Positive toxin tests in patients without symptoms do not need to be treated [Kelly, 1994]. If possible, antibiotics should be discontinued and avoided for at least two months. Relapses are usually due to incomplete eradication or reinfection rather than the development of resistance. Metronidazole should also be the treatment of choice for relapses. Though there are few randomized placebo controlled trials, treatment approaches for patients with multiple relapses include the following: 1- 2 months of either oral metronidazole or oral vancomycin every other day with a gradual taper; oral vancomycin plus rifampin; oral yogurt, lactobacillus preparations, cholestyramine or Saccharomyces boulardii [Fekety, 1997].

Prevention of C. difficile associated diarrhea depends primarily upon proper handwashing by hospital personnel and avoidance of unnecessary and potentially high risk antibiotics [Pear, 1994]. Saccharomyces boulardii, a nonpathogenic yeast was evaluated in a double blind-blind placebo controlled randomized trial for the prevention of antibiotic associated diarrhea [McFarland, 1995]. Patients receiving new ß-lactam antibiotic prescriptions were randomized to placebo or S. boulardii (500 mg po bid) beginning within 72 hours of antibiotic initiation and continuing until 3 days after the antibiotic was discontinued. Significantly fewer patients taking S. boulardii developed antibiotic associated diarrhea (RR = 0.29 95% CI = 0.08,0.98). S. boulardii does not appear to have any side effects. Amebiasis

Amebiasis is one of the most common parasitic diseases worldwide. Infection is acquired by ingestion of protozoal cysts of Entamoeba histolytica. The trophozoite produced in the small bowel invades the epithelium of the colon causing ulceration.

Risk factors in the United States include sexual promiscuity and colonic irrigation. Amebiasis is unlikely to be contracted during short-term foreign travel though high rates of infection exist in the Indian subcontinent, southern and western Africa, the Far East and South and Central America. The clinical presentation of acute amebiasis includes abdominal pain, tenesmus, and frequent loose to watery stools with blood and mucus. Rarely fulminant colitis may intervene with rapid onset of severe bloody diarrhea, fever and abdominal tenderness progressing to colonic perforation and death. The diagnosis is complicated by the recent discovery of 2 distinct but morphologically identical species, E. dispar and E. histolytica. E. dispar is noninvasive, only associated with an asymptomatic carrier state and 10 times more common worldwide. It is nonpathogenic even in patients with AIDS [Ravdin, 1995]. The pathogenic and nonpathogenic species cannot be distinguished under the microscope unless one identifies E. histolytica trophozoites, which have engulfed erythrocytes.

Virtually all patients with acute amebic colitis have fecal occult blood tests, but fecal leukocytes are rare due to the phagocytic activity of the parasite [Gonzalez-Ruiz, 1994]. Serum anti-amebic antibodies are usually present but cannot distinguish between past and present infection. A negative antibody test, however is very helpful is eliminating acute amebic colitis from the differential diagnosis in a patient with acute bloody diarrhea [Ravdin, 1990]. Colonoscopy with biopsy of the ulcer edge and histological demonstration of ameba remains the definitive test for the diagnosis of amebic colitis. Recommended treatment of invasive colitis is metronidazole 750 mg po tid for 10 days. Less severe colitis may be treated with tetracycline 250 mg po tid for 10 days or erythromycin 500 mg po qid for 10 days. All treatments should be followed by diiodohydroxyquin 650mg po tid for 20 days to eliminate all residual cysts [Ravdin, 1995]. Aeromonas and Plesiomonas

Aeromonas hydrophilia, Aeromonas sobria and Plesiomonas shigelloides are gram negative facultative anaerobic bacteria, which are likely but not universally, accepted causes of acute diarrhea. The most common source of Aeromonas is untreated drinking water and patients typically present within one week of exposure with watery diarrhea, abdominal cramps and sometimes vomiting [Ravdin, 1990], [Holmberg, 1986]. Blood in the stool associated with the sigmoidoscopic appearance of non-specific colitis has been described [Deutsch, 1997]. Plesiomonas infection is associated with eating raw shellfish and causes an illness characterized by severe abdominal cramps and diarrhea, which can be bloody. Antimicrobial therapy is generally not needed, although Aeromonas and Plesiomonas are both sensitive to quinolones. Yersinia enterocolitica

Yersinia enterocolitica is a gram negative bacteria that is an uncommon cause of acute diarrhea in adults in the United States. It is an invasive organism and though enterotoxins have been described their role is not yet well established. Transmission of the disease may occur via a fecal-oral route from contaminated food and water. The most frequent source in the US in recent years has been consumption of raw pork intestines (chitterlings). Children under five will typically present with inflammatory diarrhea [Lee, 1991]. Older children may develop mesenteric adenitis and inflammation of the terminal ileum clinically mimicking acute appendicitis. Adults rarely develop clinically apparent Yersinosis which may include self-limited diarrhea followed by arthritis and erythema nodosum. Antimicrobial therapy has not been shown to be effective.

3.3 Traveler's Diarrhea

The ease with which people can move around the world has dramatically increased the frequency of travelerís diarrhea, now affecting up to one third of individuals who travel to a developing area. Travelerís diarrhea has been defined as the passage of at least three unformed stools in a 24-hour period during travel or during the first seven to ten days after returning home. Associated symptoms may include nausea, vomiting, abdominal pain, fecal urgency, tenesmus, and bloody or mucoid stools. Individuals at highest risk include young children, adults ages 15-29 years, and those with high gastric pH (achlorhydria, postgastrectomy, and proton-pump inhibitor use). The spectrum of infectious agents varies from country to country, but overall, the most common pathogens in order of decreasing frequency include enterotoxigenic Escherichia coli (ETEC), enteroaggregrative E coli (EAEC), Shigella species, Campylobacter jejuni, Rotavirus, Aeromonas species, Plesiomonas shigelloides, Salmonella species, noncholera Vibrios, and Noravirus [Jiang, 2002]. Giardia, Cryptosporidium, and Microsporidia are less frequent causes of travelerís diarrhea but may cause prolonged symptoms. Cyclospora may also be an important cause of prolonged traveler's diarrhea particularly for those traveling to Nepal. However, in ~40% of travelers with severe diarrhea, no cause of infection is found even with sophisticated microbial testing [Jiang, 2002]. Severity of travelerís diarrhea is mostly independent of the organism recovered, with the possible exception of Campylobacter, which leads to more systemic symptoms and a more prolonged course than other common pathogens [Sanders, 2002]. Most travelers fear contaminated water as the source of disease, but contaminated food may be a much more common vehicle of transmission for both bacteria and viruses.

3.3.1 Prophylaxis

Education of travelers in the adage "Öboil it, peel it, cook it or forget itÖ" may be one of our best first-line defenses. Prophylactic use of fluoroquinolones in low doses (250 mg ciprofloxacin, or 400 mg norfloxacin per day is 90% effective in preventing diarrhea when traveling in areas where the resident bacteria have not become resistant to these drugs (Southeast Asia, particularly Thailand, for example). [Rendi-Wagner, 2002]. Bismuth subsalicylate has been demonstrated to have 65% effectiveness when taken prophylactically, in a dose of 2 tablets 4 times daily [Graham, 1983], [DuPont, 1987]. Prophylaxis with antibiotics may be effective, but is not recommended for healthy travelers because of the risk of side effects and the increasing problem of drug resistance. Exceptions are usually made for those with serious underlying medical conditions and occasionally for those travelers unable or unwilling to risk a brief illness.

3.3.2 Treatment of Traveler's Diarrhea

The evaluation and approach to the treatment of travelerís diarrhea should be similar to that employed in other causes of acute diarrhea, as discussed below. Most persons have mild symptoms, which are self-limited and require only attention to oral hydration. Nonantibiotic therapies including bismuth subsalicylate and loperamide may reduce stool frequency, liquidity and the duration of illness (Table 7). In general, oral rehydration plus bismuth subsalicylate or loperamide are adequate therapy for mild to moderate diarrhea (less than four stool per day). Antibiotics should generally be reserved for persons with travelerís diarrhea who have moderate to severe symptoms. Double blind randomized studies have demonstrated the efficacy of several antibiotic regimens in treating acute travelerís diarrhea: single doses of either levofloxacin 500 mg or azithromycin 1000 mg, or twice-daily dosing of rifaximin 200 mg or ciprofloxacin 500 mg, for three days, appear to be roughly equivalent [DuPont, 2001], [Adachi, 2003]. In countries where bacteria are likely to be resistant to fluoroquinolones, azithromycin or rifaximin should be used for empiric treatment. Over half of enteric bacterial isolates from patients with travelerís diarrhea are resistant to Trimethoprim -sulfamethoxazole, which has limited utility for treatment of travelerís diarrhea.

Download a PowerPoint Presentation to learn more about Traveler's Diarrhea.

3.4 Diarrhea in Patients with AIDS

Diarrhea is one of the most common manifestations of acquired immunodeficiency syndrome (AIDS). AIDS associated diarrhea is typically chronic, however it is also usually infectious and therefore inclusion in this discussion of acute diarrhea is warranted. An exhaustive review of the causes and treatment of diarrhea in AIDS is however, beyond the scope of this review and appears elsewhere [Lew, 1997]. Diarrhea associated with HIV infection differs most notably by the fact that multiple etiologic agents may be present simultaneously [Antony, 1988]. Though some infectious agents remain without proven effective treatment, evaluation is justified by studies which demonstrate the benefit of specific therapy [Smith, 1992], [Smith, 1988]. Bacterial agents including Salmonella species, Shigella flexneri and Campylobacter jejunii may cause chronic watery diarrhea with blood and mucus. These infections occur more frequently in persons infected with HIV and may often be chronic, recurrent and associated with bacteremia. Various strains of E. coli have also been identified as important cause of diarrhea in AIDS. Mycobacterium avium complex causes diarrhea usually in association with disseminated disease. Some parasites including Cryptosporidia, Microsporidia and Isospora belli may cause profuse watery diarrhea and weight loss in persons with AIDS. Giardia lamblia and E histolytica do not appear to be more frequent or severe in HIV infected patients [Albrecht, 1995]. Blastocystis hominis is probably not an enteric pathogen, even in severely immunocompromised hosts. Adenovirus may also cause watery diarrhea with increased stool frequency in association with HIV infection. Cytomegalovirus [See Figure 7A ] can cause either an enteritis or more commonly a colitis with bloody diarrhea, fever, and weight loss. Herpes simplex virus may cause perianal disease and proctitis [ Figure 7B ]

but is much less frequently a cause of significant diarrhea [Smith, 1988]. Indications for specific antibiotic therapy are listed in (Table 8). A stepwise approach to the diagnostic evaluation of nonbloody diarrhea in persons with AIDS has been suggested [Smith, 1993]. An initial approach which includes only a stool culture appears to be the most cost effective when subjected to medical decision analysis [Johanson, 1990]. When necessary, further workup should include stool culture, C. difficile toxin, and microscopic examination for ova and parasites using saline, iodine, trichrome and acid-fast preparations. A complete evaluation including colonoscopy and esophagogastroduodenoscopy with multiple biopsies should be reserved for those patients in whom a specific agent cannot be identified, and significant symptoms persist despite nonspecific therapy.

3.5 Diagnostic Evaluation

3.5.1 History

The initial evaluation of the adult patient presenting with acute diarrhea should focus on the setting, nature and severity of the acute diarrheal illness as well as the patient's age and overall health status. Early intervention is warranted for individuals who are elderly (>70 years), debilitated or immunocompromised. Even previously healthy adults with acute diarrhea should see their health care provider early if they experience dehydration, fever, gastrointestinal bleeding, abdominal pain or neurologic symptoms.

Recently published practice guidelines recommend prompt medical evaluation in those patients who present with dehydration, bloody stools, profuse watery diarrhea, fever greater than 38.5° C (101.3° F), severe abdominal pain, >6 unformed stools during a 24 hr period, or duration of illness of >48 hours [DuPont, 1997]. These guidelines are derived from available scientific evidence but in this case they are largely based upon consensus. The impact of >6 unformed stools or duration of illness >48 hours on either the likelihood of a positive stool culture or the ultimate clinical outcome of the patient has not been studied.

The patient should be asked if they have noted any symptoms of dehydration including lightheadedness, dizziness, dry mouth, excessive thirst, decreased urine output, increased heart rate or changes in mental status. Initial history should also elicit any associated nausea, vomiting, fever, shaking chills or signs of significant upper or lower gastrointestinal bleeding (coffee ground emesis, hematemesis, melena, hematochezia), joint pain or a new skin rash. If this initial assessment does not indicate anything requiring immediate intervention, the evaluation can proceed to focus on the character of the illness, the setting in which it began, and other factors which may point to a specific diagnosis.

In order to assess the severity the diarrhea one should ask specifically about the frequency, consistency and volume of the stool. Patients should be asked to quantify the number of bowel movements in a 24-hour period. It may be helpful to stratify the severity of diarrhea. Mild diarrhea be defined as less than three unformed bowel movements in a 24 hour period. Moderate diarrhea is three or four per day and severe diarrhea is more than four unformed bowel movements per day. In the absence of a specific diagnosis, the type of empiric treatment (bismuth subsalicylate, loperamide or antimicrobials) may be predicated on this type of stratification scheme [Gorbach, 1997]. A report of >6 stools is a sign of severity and should prompt one to pursue further medical evaluation. Loose consistency correlates directly with quantity of stool water. This is one area where the patient’s description may be difficult to interpret. Verbal descriptors such as "solid" and "liquid" appear to have reproducible meaning while the meaning of "loose" and "semiformed" are variable [Mertz, 1995]. Some directed questions may help the patient describe their current stool appearance.

Diarrhea Questionnaire

A. Stool Form Assessment

1.) Have you been having diarrhea in the past 7 days?


2.) In the past 7 days have your stools typically been:

Semi-formed (very soft but retains some form)
Loose ( no form, breaks apart)
Liquid (mushy like applesauce or watery)

3.) How often have your bowel movements looked like each of the following pictures in the past 7 days?

All of the time
Most of the time
Some of the time
None of the time

[A pictorial representation may be very helpful.]

Nonspecific antidiarrheal therapy may reduce the looseness of the stool without causing measurable decreases in stool water [Wenzl, 1995]. The risk of dehydration may not change in the patient using nonspecific antidiarrheal therapy even though the stool may appear more formed. In a very general sense, it is useful to know if the bowel movements are small or large in volume. The severity of diarrhea strongly correlates with the presence of fecal incontinence and nocturnal bowel movements. Fecal incontinence in the setting of acute diarrhea is extremely common. Patients may hesitate to volunteer this as a symptom unless they are asked specifically.

The pattern of the diarrhea and the relationship to meals or medication times may be helpful. A history of intermittent diarrhea with some completely normal stools may suggest lactose intolerance or some other dietary or medication intolerance. Patients should always be asked about the presence of blood or mucus. Tenesmus, the sensation of a recurrent sometimes painful and often ineffectual need to defecate may indicate the presence of proctitis. Associated symptoms of nausea, vomiting, abdominal pain, fever, chills, anorexia, and loss of weight should be elicited.

The setting in which the diarrhea began is crucial to the history in adults with acute diarrhea. The health care provider should ask patients about recent travel, exposure to untreated water, recent antibiotics, and contacts with day-care age children, nursing home residents or other individuals with diarrhea. Dietary history should include recent ingestion of raw or poorly cooked foods (eggs, meat, shellfish, fruits and vegetables) or foods that may have been improperly handled or stored (restaurants, buffets or picnics). Patients should be asked specifically about ingestion of dairy products and sugarless food or candy containing sorbitol or xylitol. The use of public swimming pools and contact with domestic or farm animals may also be important. Patients should be asked about any changes in diet, alcohol or medications (prescribed, over-the-counter, complimentary, herbal or botanical) which were initiated before the onset of diarrhea. A number of new medications including recent diet medications that hamper fat absorption may cause diarrhea in addition to a large number of medications that have been in use for many years. (Table 9) Sexual history may be relevant particularly if there has been oral-genital or oral-anal contact. Past medical history should focus on any prior history of diarrheal illness, significant underlying medical problems (e.g. AIDS, diabetes, cirrhosis, sickle cell disease, cancer, endocrine or autoimmune disease), prior radiotherapy and immunological status. It is very important to ask about any risk factors for AIDS since diarrhea may be the initial presenting manifestation of HIV infection. Questions about past surgical history should specifically include cholecystectomy, intestinal resection and surgery for peptic ulcer disease. A very complete list of medications is crucial including all prescription, ophthalmologic, nonprescription and herbal medications.

3.5.2 Physical Examination

The initial physical examination in the adult with acute diarrhea should focus primarily on the assessment for any signs of significant dehydration. Hypotension, resting tachycardia and orthostatic changes should be sought by measuring heart rate and blood pressure in both the supine and standing positions after 3 minutes have elapsed to allow for equilibration. Changes of 20 points or more in the pulse or blood pressure indicate significant volume depletion. Other indications of dehydration include dry mucus membranes, poor skin turgor and if severe, changes in mental status. Physical examination should also include a careful abdominal and rectal examination to look for signs of obstipation, obstruction, peritonitis, or anorectal disease.

3.5.3 Laboratory Evaluation

An otherwise healthy adult with mild acute diarrhea of less than 3 days duration without any obvious cause may be treated symptomatically. Individuals with significant underlying medical disease or moderate to severe diarrheal illness of greater than 48 hours should undergo further diagnostic evaluation. This group would include individuals with bloody diarrhea, significant dehydration, fever or abdominal pain. Further evaluation is also indicated for individuals with recent travel, antibiotic use, day-care exposure or nursing home workers, food handlers, and those present during a community outbreak or following exposure to high risk foods. Stool Examination

The initial evaluation of the stool in the setting of acute diarrhea may include visual examination for evidence of gross blood, testing for occult blood, and microscopic examination of the stool for red and white blood cells. Visual indicators of gastrointestinal hemorrhage include bright red blood, maroon or black tarry stools. Dark stools, which are not black and tarry, do not indicate bleeding. Though hematochezia is nonspecific, one study done in Bangladesh demonstrated visible blood was found significantly more often in those infected with Shigella or Entamoeba histolytica than with other pathogens [Stoll, 1983]. In the United States, E. coli O157:H7 is the most commonly isolated organism from stool specimens that are visibly bloody.

Testing of the stool for occult blood is problematic. A study of US citizens studying in Mexico and presenting with acute diarrhea demonstrated the finding of a negative fecal occult blood test was a reliable indicator of the lack of invasive bacterial infection [McNeely, 1996]. The percentage of false positive exams, however is likely increased in the setting of anal trauma associated with acute diarrhea and there is insufficient positive predictive value. Newer tests for fecal occult blood using latex agglutination of fecal hemoglobin may improve the specificity but are currently too complicated and expensive for general use [Beltinger, 1997].

Direct examination of the stool for fecal leukocytes has been used for decades as an indicator of intestinal inflammation [Harris, 1972]. The methodology is simple enough that someone with microscopy skills and a very basic laboratory can perform the test reliably. Use of a fresh stool sample in a cup rather than one obtained with a rectal swab is recommended. An aliquot of fresh stool on the tip of a wooden stick is mixed on a glass slide with normal saline until it is barely transparent. A drop of methylene blue or Grams stain may facilitate identification of fecal leukocytes. The finding of >10 fecal leukocytes per high power field is considered positive. One US study retrospectively examined the utility of fecal leukocytes as an indicator of a positive stool culture. The sensitivity and specificity were 40% and 78%, respectively. The positive predictive value was only 20%. Several other studies have also shown fecal leukocytes are a very poor predictor of enteric infection. Fecal leukocytes are also present in noninfectious inflammatory disorders of the colon including inflammatory bowel disease, ischemia and radiation proctitis.

The measurement of fecal lactoferrin by latex agglutination has been proposed as a more sensitive test for the detection of fecal leukocytes [Guerrant, 1992]. Stool Hemoccult, fecal leukocyte and fecal lactoferrin all have poor positive predictive value.

Routine stool culture for enteric pathogens should identify Salmonella, Shigella, Campylobacter, E coli O157:H7, and Yersinia. Some labs have added Aeromonas and Plesiomonas.

Clinical predictors of a positive culture include diarrhea of greater than 24 hours duration, fever, and either blood in the stool or abdominal pain with vomiting [Koplan, 1980]. A multicenter review of nearly 60,000 stools submitted for culture revealed an overall positivity rate of 6.4%. Multiple specimens are frequently submitted from the same patient, however, 96.9% of positives were found on the first specimen and 99.0% after the second [Valenstein, 1996]. No more than 2 specimens are necessary to exclude an enteric infection.

Stool cultures are frequently sent to evaluate hospitalized patients who develop diarrhea. At least 5 retrospective and 1 prospective study have evaluated the likelihood of a positive stool culture in relationship to the number of days in the hospital [Siegel, 1990], [Rohner, 1997], [Yannelli, 1988]. While the overall positivity rate was approximately 6%, the yield of stool culture in those developing diarrhea after the 3rd hospital day was only 0-1% in all studies [Valenstein, 1996], [Rohner, 1997]. Similarly, specimens submitted for ova and parasites (O&P) had an overall yield of 2.5%. This decreased to 0.7%, if submitted from patients hospitalized for more than 4 days [Valenstein, 1996]. Guidelines have been proposed to limit the use of stool cultures to patients in the hospital <3 days unless they are immunocompromised. Outcome studies using this type of guideline have reduced the number of specimens submitted by 37% [Chitkara, 1996].

The examination for ova and parasites (O&P) is recommended for individuals with persistent diarrhea who are at risk because of foreign travel, high-risk sex, immunocomprimised status, exposure to unfiltered water, day care aged children or a community outbreak [DuPont, 1997]. The evaluation is time consuming and the yield may vary considerably with the available level of expertise. Typically 3 specimens are collected on successive days. In those patients in whom a parasite was ultimately identified by stool examination, 97.6% of specimens were positive after 2 specimens were examined. The yield increased to 99.8% after three specimens were examined. Stool examination for ova and parasites identifies an organism in only 2.5% of specimens submitted. Specimens from immunocompetent persons are seldom positive if submitted after the fourth hospital day [Valenstein, 1996], [Mohr, 1992].

In contrast, testing for Clostridium difficile toxin in appropriate patients has a high frequency of positivity in both outpatients and inpatients regardless of length of stay [Barbut, 1996]. C. difficile toxin was detected more frequently from inpatients than any other bacterial pathogen even though it was requested on only 50% of the specimens submitted for culture [Valenstein, 1996]. Testing for C. difficile toxin is appropriate in individuals who are receiving or have received antibiotics within the previous 2 months.

The evaluation of an adult with watery non-bloody diarrhea and exposure to day care aged children or unfiltered water should focus on giardiasis. The sensitivity of the microscopic examination for Giardia lamblia, however, is only 50-70%. In lieu of stool for O&P, one should submit a single stool specimen for Giardia specific antigen. This commercially available ELISA has sensitivity and specificity of 96% and 100%, respectively [Rosoff, 1989].

Routine stool studies for outpatients with mild to moderate diarrhea may not be necessary and can in the aggregate greatly add to the cost of medical care. (Table 10) Stool culture for enteric pathogens is indicated if the patient has any of the alarm signals outlined in (Table 11).

3.5.4 Endoscopy

Endoscopy is generally unnecessary in the evaluation of adults with acute diarrhea. Flexible sigmoidoscopy may be recommended in the evaluation of patients with persistent diarrhea and one of the following:

The endoscopic and histologic appearance of infectious colitis

is generally not specific but biopsies may be helpful, particularly in distinguishing acute from chronic colitis.

Upper endoscopy with biopsies may be helpful in selected patients with persistent watery nonbloody diarrhea and negative routine evaluations [DuPont, 1997].

3.6 Treatment

Excellent treatment guidelines have been published recently by the American College of Gastroenterology and the British Society for the Study of Infection [Farthing, 1996], [DuPont, 1997]. The initial approach to the treatment of acute diarrhea is predicated upon a careful evaluation of the patient. Attention must be paid to the patient’s age, underlying medical condition, severity of current illness, and degree of volume depletion. The determination of volume status is critical in those who are elderly or debilitated. The clinical indications for medical evaluation in adults with acute diarrhea are summarized in (Table 11).

3.6.1 Oral rehydration

Most adults presenting with acute diarrhea will have only mild sodium and water depletion and will respond appropriately to oral hydration. Clear liquids such as dilute fruit juice, carbonated beverages, and sports drinks may suffice for mild self-limited diarrhea. These drinks however, should not be recommended for patients with moderate to severe diarrhea because they have an inappropriate ratio of sodium to carbohydrate [Avery, 1990]. The physiological principle explaining the coupled transport of sodium and glucose underlies most current recommendations. The ideal solution would contain sodium 60-90 mEq/L, potassium 20mEq/L, citrate 30mEq/L and glucose 20g/L. Rehydration formulas which incorporate these principles include the WHO-UNICEF oral rehydration salt packets and commercially available products including Infalyte® , Lytren®, Pedialyte® and Resol®. The constituents of several often utilized formulations are presented in (Table 12). Homemade solutions may be used effectively if clean water is available and care is taken to avoid errors in mixing (Table 13). Continued intake of food is now considered an important part of oral therapy for diarrhea and should not be discouraged [Avery, 1990]. Boiled starches and cereals (potatoes, noodles, rice, wheat, and oats), crackers and bananas are ideal. Milk is often avoided, however the development of clinically important lactose intolerance in the setting of acute diarrhea is rare [DuPont, 1997].

3.6.2 Antidiarrheal Therapy

Nonspecific antidiarrheals are very commonly used in mild to moderate acute diarrhea. These may include bismuth subsalicylate, diphenoxylate HCl with atropine sulfate (Lomotil®), loperamide (Imodium®), attapulgite (KaoPectate®) and cholestyramine (Questran®). Anticholinergics are generally ineffective and are not recommended because of side effects.

Bismuth subsalicylate (Pepto-Bismol®) may have both anti-secretory and anti-microbial activity. Salicylate absorption can occur but short term use in adults appears to be safe [Gorbach, 1990]. In 1977 DuPont et al. published a randomized double blind placebo controlled study of bismuth subsalicylate suspension in the treatment of acute diarrhea in US students who had recently arrived in Mexico [DuPont, 1977]. Bismuth subsalicylate suspension was associated with a significant decrease in the number of unformed stools and subjective complaints of diarrhea, nausea and abdominal cramps. There was no significant difference in water content or total weight of the stools between the two groups. A subsequent study of healthy volunteers who ingested enterotoxigenic E. coli (ETEC) demonstrated bismuth subsalicylate tablets taken four time a day caused a significant reduction in the incidence of acute diarrhea [Graham, 1983].

ETEC was rarely recovered in the stools of subjects given bismuth subsalicylate. A second study on US students in Mexico demonstrated bismuth subsalicylate (2 tabs four times a day in doses identical to Pepto-Bismol®) given within 48hours of arrival was associated with a significant reduction in the incidence of traveler’s diarrhea [DuPont, 1987]. This regimen accounted for a protection rate of 65%. Side effects reported included only darkening of the tongue and stool. There was no statistically significant increase in the development of tinnitis. Bismuth subsalicylate may also be effective in reducing symptoms in patients with viral gastroenteritis [Steinhoff, 1980].

Loperamide is a modified opiate that does not significantly penetrate the central nervous system. It is available over-the-counter and exerts its antidiarrheal effect by reducing intestinal motility. It has no demonstrated antisecretory activity. Loperamide (Imodium A-D®) has been compared to bismuth subsalicylate (Pepto Bismol®) and attapulgite (Kaopectate®) in two open labeled field trials for the treatment of acute watery nonbloody diarrhea. Loperamide was more effective and worked faster than bismuth subsalicylate. 75% of patients achieved relief from symptoms of diarrhea within only one day of therapy [Johnson, 1986], [DuPont, 1990].

Diphenoxylate-HCl with atropine (Lomotil®) is another opiate-like antimotility agent. Although data is limited, diphenoxylate-HCl with atropine may be contraindicated in patients with bloody diarrhea. In one study, diphenoxylate-HCl with atropine prolonged fever and illness in patients with bloody diarrhea due to Shigella [DuPont, 1973]. A subsequent study did not confirm this finding however [Murphy, 1993].

Attapulgite (Kaopectate®) is a naturally occurring purified hydrated aluminum magnesium silicate which is not absorbed systemically. It was tested in a double blind placebo controlled study of adults with mild to moderate acute diarrhea not due to bacteria or protozoa. Attapulgite (600mg) two tabs after each bowel movement for a maximum of 72 hours was associated with a significant reduction in the frequency of diarrhea [Zaid, 1996]. Stool volume was not evaluated. These adsorbents may add form to the stool without affecting overall fluid losses and should therefore be used only in patients with mild diarrhea.

Cholestyramine is a nonabsorbable resin used as a cholesterol lowering agent with the side effect of constipation. Limited studies have demonstrated cholestyramine is effective as a nonspecific treatment for acute diarrhea [McCloy, 1971]. A specific role may exist for treatment of relapsing pseudomembranous colitis secondary to Clostridium difficile.

3.6.3 Antibiotic Therapy

Empiric antibiotic therapy may be indicated in selected adults with acute diarrhea. Pending culture results, empiric antibiotic therapy may be considered in patients with fever >38.5° C (101° F), dysentery or bloody diarrhea (excluding amebiasis), and moderate to severe traveler’s diarrhea. The rationale for empiric antibiotic therapy in moderate to severe traveler’s diarrhea, is that most of these patients harbor a bacterial pathogen. Few studies have evaluated the benefit of empiric antimicrobial therapy for acute diarrhea in adults who are not recent travelers. One such study evaluated 202 patients treated with either ciprofloxacin, trimethoprim/sulfamethoxazole or placebo for 5 days in a randomized double blind fashion [Goodman, 1990]. Ciprofloxacin, but not trimethoprim/sulfamethoxazole shortened the duration of illness and had a higher cure rate than placebo. 82% of patients who received ciprofloxacin had improved or were cured by the third day. Despite a lack of strong scientific evidence, empiric antibiotic therapy is also commonly used for patients over age 60 with moderate to severe diarrhea or those with significant underlying medical conditions [Ericsson, 1987], [Wistrom, 1992].

When indicated, the drug of choice for antimicrobial therapy of diarrhea of unknown cause is a quinolone such as norfloxacin 400mg, ciprofloxacin 500mg or ofloxacin 300mg bid for 3-5 days [Farthing, 1996], [DuPont, 1997]. Unfortunately, as feared, reports of quinolone resistance have begun to appear.

The decision to use antimicrobial therapy, after stool studies reveal a causative agent, will depend on whether or not symptoms have persisted. The benefits of antibiotics are most well demonstrated for Shigella, C. difficile and Giardiasis. The drugs of choice for specific antimicrobial agents are listed in (Table 14).

4.0 Chronic Diarrhea

4.1 Introduction

Chronic diarrhea implies an increased frequency of passing looser stools for more than a month.

Healthy, young Americans eating controlled diets containing 20g of dietary fiber have stool weights of about 100 g daily [Saunders, 1988]. The stools are 70% water by weight, but the water entrapped in fibrous residue or in bacteria comprises 30-75% of the wet weight of stools [Stephen, 1980].

Healthy humans can have heavier stools if their diet contains food stuffs which enhance fecal bacteria, or contribute to fecal fiber. Vegetarians may have stool weights normally in excess of 200g. On the other hand, patients may have abnormally loose stools whose daily weight is less than 200g. The consistency of stools is determined by the water content and by the ability of fecal insoluble solids to bind the fecal water [Wenzl, 1995].

The punch-line: a decreased consistency of feces should be the major characteristic in the definition of diarrhea rather than an arbitrarily defined excessive daily fecal weight (exceeding 200g per day, for example), or increased frequency of defecation.

Diarrhea is said to be chronic after one month because most infectious causes of acute diarrhea resolve within this period of time [Donowitz, 1995], except in patients who are immunosuppressed.

We present an approach to the differential diagnosis of chronic diarrhea based on groups of diseases, although pathophysiologic mechanisms can overlap considerably. However, clinicians should be able to use this scheme to develop a hypothesis and to test the hypothesis in a logical manner.

4.2 The Diarrhea History

Physicians begin to generate hypotheses as soon as a patient is greeted, and as the presenting complaint is heard. The ensuing questioning will depend on the initial hypotheses. The essence of this disease-oriented approach to chronic diarrhea is the emphasis on eliminating expeditiously diseases causing hematochezia and diseases with malabsorption, so that the difficult group of predominantly watery diarrheas can be considered without distraction.

4.2.1 Immunocompetence

It is essential to establish that the patient with chronic diarrhea is immunocompetent. Otherwise, an infectious etiology for the diarrhea would be of paramount concern. Questions about blood transfusions, intravenous drug use, occupational or recreational exposure to HIV, and immunosuppressive medical therapy must be asked.

4.2.2 Onset

An abrupt onset may connect the symptom of chronic diarrhea to the cause such as cyclospora (mountaineering in Nepal) or as lactose intolerance (following viral enteritis).

4.2.3 Frequency of Defecation Diarrhea After Meals

Diarrhea after meals suggests a heightened gastro-colic reflex. Patients with irritable (idiopathic) bowel syndrome often voice this complaint. Awakening from Sleep

Diarrhea which awakens a patient from sleep is an alarm signal for organic disease. Diarrhea rarely awakens the patient with functional (idiopathic) bowel disease. Daily Diarrhea

Daily diarrhea suggests organic disease. Patients with irritable (idiopathic) bowel syndrome often have good days interspersed with bad ones.

4.2.4 Volume of Diarrhea Teaspoons or Tablespoons?

Is the patient passing teaspoons or tablespoons (small volume) of stools? Affirmation suggests anorectal dysfunction (incontinence), or proctitis (especially if tenesmus or hematochezia is present). Cupfuls?

Is the patient passing cupfuls of stool (large volume)? Patients who pass more than 1 L of stool daily have small bowel mucosal disease, or a combination of small and large bowel mucosal disease, or a hypersecretory state; these inferences are based on the absorptive capacity of the normal colon [Debongnie, 1978].

4.2.5 Relationship to Eating Worsened by Eating?

Is the diarrhea worsened by eating, and, if so, is the diarrhea worsened by eating fatty foods? An affirmation suggests steatorrhea. Lessened by Fasting?

Is the diarrhea lessened by fasting? An affirmation suggests malabsorption, but patients with irritable (idiopathic) bowel syndrome often give a positive response. Unaffected by Fasting?

Is the diarrhea unaffected by fasting? A positive response suggests an exudative enteropathy (inflammatory bowel disease), or a hypersecretory state.

4.2.6 Fecal Characteristics Blood in Stools?
Has the patient seen blood in stools? (Inflammatory bowel disease) Stool Consistency?

What is the consistency of the stools? The patient should be challenged to describe the stools as semi-formed, or as mushy (like applesauce), or as loose (like thin soup). A formal diarrhea questionnaire is offered by Mertz et al [Mertz, 1995]. Steatorrheic stools are sometimes mushy rather than loose [Bo-Linn, 1984], [Hofmann, 1985]. Color and Odor?

The color and the odor of stools are only informative if these characteristics have changed dramatically. Excessive Flatus?

Floating stools and excessive flatus suggest carbohydrate malabsorption. Stools float because of their content of gas, not of fat [Levitt, 1972]. Oily Droplets?

Oily droplets in the toilet water indicate steatorrhea, and especially malabsorption of dietary triglyceride. Recognizable Food Stuffs?

Has the patient recognized food stuffs in the stool, and of additional importance, is there an estimate of elapsed time between ingestion and expulsion? This head of meal transit time can be very helpful. A HOMTT less than 12 hours has been associated with experimentally induced diarrhea [Read, 1980]. Food stuffs such as corn kernels, or whole beets, which color the stools red, can be used as markers.

4.2.7 Previous Abdominal Surgery or Irradiation Intestinal Resection

Gastric resection used to be a major condition associated with malabsorption. Nowadays, intestinal resection is much more common. It is important to obtain past records; the amount of right colon resected can exceed in importance the amount of ileum resected in the pathogenesis of the post-operative diarrhea [Cummings, 1973].

4.2.8 History of Medication Use

Virtually any medicine can impair the 99% efficiency of normal salt and water absorption. Major culprits include antibiotics, antiarrhythmics, antimetabolites, and alcohol (Table 9).

4.2.9 Characteristics of the Patient Does the Patient Appear Well?

If the patient appears well, has a normal screening physical exam, and has no alarm signals, irritable (idiopathic) bowel syndrome can be pursued further. Family History

A family history of inflammatory bowel disease, celiac sprue, or cancer can strengthen suspicions about the etiology of chronic diarrhea. Risk factors for immunosuppression?

4.2.10 The Diarrhea Diary

A patient can be instructed to keep a diary for a week or more which lists the times of defecation, some fecal characteristics, relationship to meals, etc (Table 15). Such a questionnaire could be sent to a patient prior to the office visit.

4.3 Simple Stool Studies

These simple tests are used to accept or reject initial hypotheses about the cause(s) of chronic diarrhea. Surprisingly, collecting a stool specimen is often a neglected essential in the investigation of the patient.

4.3.1 Head of Meal Transit Time

Head of a meal transit time (HOMTT) can be assessed by asking a patient to measure the time elapsed between swallowing half a cup of corn kernels, or eating four whole canned beets. A HOMTT averaged 22 hours (median, 20 hours; range 9-45 hours) in 24 presumably healthy medical students who observed a red discoloration of their stools after eating four whole canned beets with a meal [Saunders, Unpublished Observations]. The clinician is interested only in abnormally rapid transit times in the work-up of diarrhea. Diarrhea can be induced experimentally when the HOMTT becomes less than 12 hours [Read, 1980]. Furthermore, steatorrhea (up to 14 g of fat per day) can be induced in normal subjects by osmotic laxatives [Fine, 1992] which would be expected to hasten transit of chyme through the small intestine. So having an idea of intestinal transit time contributes to understanding the etiology of diarrhea, and the interpretation of measurements of fecal fat.

4.3.2 The Sudan Test for Fecal Fat

This test can be performed on a spot specimen, or on a timed collection of stool. A representative sample of stool is placed on a glass slide, and it is acidified with glacial acetic acid to protonate long-chain fatty acids so that they are converted from insoluble salts to fatty crystals. After adding an ethanolic solution of Sudan III and applying a cover slip, the slide is gently heated to melt the fatty acids into oily droplets [Drummery, 1961]. Up to 100 tiny (<4 microns) droplets may be seen in normal stool at a magnification of 400x. If the number and size of the orange fatty droplets is increased, a second fecal slurry in water should be prepared. This specimen is not acidified; it is examined directly after adding the Sudan III [Drummery, 1961].

The Sudan Test is very sensitive for the detection of fatty acid (Part One of the test) and of triglyceride (Part Two of the test) [Khouri, 1989]. Therefore, if Part Two (and Part One) of the Sudan Test is positive, the clinician should suspect maldigestion of dietary triglyceride (pancreatic insufficiency, small bowel resection). A negative Part Two of the Sudan Test does not exclude pancreatic insufficiency. Mineral oil, and the unabsorbable fat substitute, sucrose polyester (Olestra ) could cause false positive Sudan Tests (Parts One and Two).

The bottom line: If a patient is ingesting fat (> 80 g per day), the Sudan Test is excellent for proving clinical suspicion of fat malabsorption [Drummery, 1961], and it has the added advantage of being able to suggest maldigestion of dietary triglyceride.

4.3.3 The pH of Fecal Water

The pH of fecal water can be estimated with color-fast indicator strips (for example, ColorpHast, EM Science, 480 Democrat Road, Gibbstown, NJ 08027) which can be dipped into a fresh specimen of liquid stool and washed briefly under tap water. Fecal pH of <5.5 is diagnostic of carbohydrate malabsorption [Eherer, 1992], because it is caused by the fecal excretion of excessive amounts of short-chain fatty acids from colonic fermentation of carbohydrate.

4.3.4 Fat Balance

This test should be designed to obtain a timed collection of feces while the digestive and absorptive functions of the small intestine are being challenged, that is, when the etiological hypothesis is predominantly fatty diarrhea.

A critical factor in a meaningful result from a fat balance is knowing the patient’s fat intake. Consider if the patient has a coefficient of fat absorption of 80% (normal, 95%) and yet is eating only 40 g of fat a day: a potential additional excretion of 8 g of fat in 24 hours can be underwhelming. The patient should have dietary counseling to ensure that > 80 g of fat are ingested daily during the test. A Big Mac® with large fries is nearly 50 g of fat. Some patients think that a prescription of two Big Macs plus fries per day is inspired; other patients plead cruel and unusual punishment.

A fat balance can be combined with the HOMTT test, and with collections of stool samples for searching for parasites (Table 16). Patients are instructed in how to place feces in precalibrated vials containing a fixative-preservative which allows storage without obscuring subsequent microscopic search for eggs, cysts, and trophozoites (Ecofix, Meridian Diagnostics Inc., Cincinatti, Ohio 45244). Another portion of stool can be placed in a plastic jar so that giardia antigen can be sought by immunoassay. One adequate fecal specimen is sufficient for the detection of Giardia antigen. Three stools should be examined if E. histolytica trophozoites are being sought.

The weight of a 24 hour stool can be helpful. Patients with irritable (idiopathic) bowel syndrome rarely have daily stool weights over 500 g; low stool weights may indicate that a patient’s true trouble is fecal incontinence. On the other hand, patients with hormonally-driven, hypersecretory diarrhea often have daily stool weights over 1000 g.

If the fecal fat has been quantified, and the weight of the 24 hour stool is known, a concentration of more than 8 g of fat per 100 g of stool suggests pancreatic insufficiency [Bo-Linn, 1984], although, as a test, fecal fat concentration is neither very sensitive or specific [Hofmann, 1985].

4.3.5 Measuring Fecal Electrolytes

Fecal water has about the same osmolality as plasma, 290 mosmols / kg. K+ is the predominant cation, and short-chain fatty acids the predominant anion (~ 120mm) in normal fecal water. Fermentation continues in stored stools; consequently fecal osmolality rises as short chain fatty acids (SCFA) continue to be generated. Therefore it is essential to use a freshly passed specimen of stool, and to obtain the supernatant solution as quickly as possible if fecal osmolality is to be measured. Alternatively, a fecal osmolality of 290 mosmol / kg can be assumed [Eherer, 1992]. The osmotic gap is estimated by subtracting the sum of the concentrations of K+ plus Na+ (multiplied by a factor of 2 to allow for the accompanying anions) from 290. In experimentally induced hypersecretory diarrhea, the osmotic gap shrinks to less than 50 mosmols / kg [Eherer, 1992] as the electrolyte composition of the watery stools comes to resemble that of plasma. On the other hand, osmotic gaps are widened when unmeasured solutes such as lactose, or as Mg++ are present in fecal water.

One perquisite of measuring the osmolality of fecal water is that factitial diarrhea can be discovered if the fecal osmolality is less than 290 mosmols / kg; water or dilute urine must have been added to the fecal specimen [Eherer, 1992], [Phillips, 1995].

4.3.6 Diarrheal Response to Fasting

Patients with hormonally-driven, hypersecretory diarrhea usually have daily stool volumes of over 700 ml [Donowitz, 1995] so they may need to be hospitalized for correction of hypovolemia and hypokalemia. Typically, hypersecretory diarrhea continues unabatedly even if the patient is fasting. The hospitalized patient can be "fed" intravenously during a controlled fast. A modified fast can be accomplished in selected out-patients by proscribing food for 24 hours while prescribing sufficient Na+ and water to prevent hypovolemia should the diarrhea continue. Drinks such as Gatorade® contain insufficient Na+. Solutions such as Pedialyte (Na+, 45 meg / L; K+, 20 meg / L; glucose, 25 g / L) can be drunk in quantities sufficient to keep urinary output above one L per day while the patient keeps a diarrhea diary.

4.4 Types of Diarrhea

4.4.1 Sugary Diarrhea

New experimental data, and clinical experience allows definition of this common cause of chronic diarrhea. Patient Profiles

A non-caucasian patient is advised to drink two glasses of whole milk a day to provide needed calcium for strong bones.

A patient chews six packs of sugarless gum a day in order to satisfy her sweet tooth without gaining weight.

A patient’s diarrhea begins with the ripening of the cherries and her indulgence of 1-2 lbs of cherries a day.

All of these patients complained of excessive flatus, of a feeling of abdominal bloating, and of passing 3-4 mushy stools a day which floated in the toilet water. Their diarrhea resolved when the offending sugars (lactose; sorbitol; fruit sugars such as fructose, mannitol, sorbitol, raffinose, etc.) were eliminated. They had no alarm signals of other illness. Pathophysiology

Colonic bacteria ferment poorly absorbed osmotically active carbohydrate into absorbable gases (hydrogen, carbon dioxide, methane), and short-chain fatty acids. Diarrhea results when the fermentative capacity of colonic bacteria for soluble carbohydrate and the absorptive capacity of colonocytes for SCFA are overwhelmed [Saunders, 1981], [Rao, 1988], [Hammer, 1989].

Excessive flatus is a consequence of colonic fermentation. Even small amounts of carbohydrate such as 5 g of lactulose daily are flatogenic [Levitt, 1996]. Fermentation gases contribute the most to the volume of normal flatus [Tomlin, 1991].

Malabsorption of carbohydrate may contribute importantly to the volume of diarrhea in patients with combined small and large bowel resections [Hammer, 1990]. Many medicinal elixirs contain sorbitol which is often the cause of diarrhea in tube-fed patients [Edes, 1990]. Diagnosis

Malabsorption of carbohydrate should be suspected when diarrhea is accompanied by excessive flatus.

Finding a pH of < 5.5 in the fecal water of a freshly-passed stool would be corroborative evidence. Removal of the offending sugar should eliminate the diarrhea.

A tolerance test with 50 g of lactose, glucose-galactose is rarely necessary. Twenty-five g of glucose plus 25 g of galactose might be used as a control. The lactase-deficient individual will have abdominal cramps, flatulence, and diarrhea after the lactose challenge, but not after the component monosaccharides.

4.4.2 Predominantly Fatty Diarrhea

Steatorrhea should be suspected from questioning the patient, be proven by simple stool studies, and then the precise etiologic defect in the absorptive process should be defined. Profile of a Patient

A 60 year old housewife presented with diarrhea of three months’ duration. She was passing 2-3 mushy stools per day whose frequency and volume were worsened by eating, especially foods rich in fat. Diarrhea awakened her from sleep; she had lost 10 pounds in weight despite even though she forced herself to eat extra quantities of food. Physical exam revealed only a wan lady with no abdominal scars.

Investigations elsewhere included normal hematocrit, thyroid screen, stool ova and parasites. Upper endoscopy and colonoscopy with biopsies were normal. Courses of loperamide and metronidazole were unhelpful.

That this lady was malabsorbing should have been suspected from the history of weight loss with, presumably, an adequate dietary intake (hyperthyroidism, and diabetes were excluded), and the worsening of her diarrhea with fatty foods. Even though she was able to eat only one hamburger and fries each day, she presented us with 700 g of mushy stools. The Sudan Stain was positive, and, significantly, Part Two of the stain revealed triglyceride droplets. An abdominal CT scan was performed because she did not have risk factors for pancreatic disease such as alcoholism, or familial predisposition. The final diagnosis was cancer blocking her main pancreatic duct. Pathophysiology

Steatorrhea should be an initial clinical suspicion. When steatorrhea is proven, the definitive cause can be discovered by remembering the way stations along the route of fat digestion and absorption. Stomach
After pyloroplasty and vagotomy, the stomach empties abnormally rapidly [Carvajal, 1994]. Intestinal transit may be hastened so that fat absorption is less efficient. Small Bowel Lumen

Long-chain fatty acids in the duodenum elicit an outpouring of pancreatic bicarbonate and enzymes, and of bile. The intraluminal pH is crucial; in patients with excessive gastric acid secretion (as in Zollinger-Ellison syndrome), pancreatic enzymes are denatured, bile salts are precipitated, and absorptive cells are injured so that steatorrhea results [Shimoda, 1968].

Bile salts, by forming mixed micelles, improve the efficiency of LCFA absorption. This efficiency is impaired when the concentration of luminal bile salts falls [Porter, 1971]. as in ileal resection, or in cholestasis.

Deficiency of pancreatic enzymes will impair hydrolysis of triglycerides which have no solubility in luminal water even in the presence of bile salts. Some hydrolysis of triglyceride occurs in the stomach with gastric lipase [Aoubala, 1993], so that patients with severe pancreatic insufficiency do absorb 30-70% of their dietary LCFA [Shimoda, 1974]. Small Bowel Mucosa

Mucosal surface may be lost by resection or by disease such as Crohn’s disease.

Absorptive cells may be injured by antigens such as gluten in celiac sprue, or by abnormal bile salts and bacterial enzymes in small intestinal stasis with bacterial overgrowth [Ament, 1972].

Duodenal and jejunal mucosa is abnormal when the lamina propria is infiltrated with macrophages which may be laden with Tropheryma whippelii, or with Mycobacterium avium intracellulare. Small Intestinal Transit Time

It is difficult to delineate the contribution of hastened motility to the pathogenesis of steatorrhea. We know that experimentally-induced diarrhea can cause mild steatorrhea [Fine, 1992] so it is reasonable to assume that hastened transit will further impair digestion and absorption whose efficiencies are already decreased by an underlying disease. Summary

In small intestinal diseases, it is not only the excessive amounts of fat which contribute to the weight of the stools. Malabsorbed LCFA [Ammon, 1973], and bile salts [McJunkin, 1981], if solubilized in fecal water, can block absorption of salt and water by colonocytes. Diagnosis

The importance of gastric, or of intestinal resection to the etiology of steatorrhea is evidenced by surgical scars. The clinical challenge is to distinguish between small bowel luminal defects and mucosal diseases. Pancreatic Insufficiency

Pancreatic insufficiency is the chief cause of small bowel luminal defects if the intestinal tract is intact; a history of cystic fibrosis or of recurrent attacks of abdominal pain associated with ethanolic excesses would be pertinent. Proving pancreatic insufficiency is more difficult. A plain abdominal film might reveal pancreatic calcification, but the absence of calcification is unhelpful; not all patients with calcifications have steatorrhea [Lankisch, 1986].

In our patient, triglyceride in her stool incriminated the pancreas, and the CT scan revealed an abnormal head of the pancreas. Fecal triglyceride, however, is an insensitive test for pancreatic insufficiency [Khouri, 1989]. The Secretin Test involves placing a tube in the proximal duodenum to aspirate pancreatic juice after an intravenous injection of secretin intravenously; gastric contents must be prevented from entering the duodenum. Recent modifications [Heij, 1986] of the Secretin Test (continuous infusion of secretion and CCK-octapeptide) make the test even more laborious but they allow a sensitivity of 83% and a specificity of 89% for detecting exocrine pancreatic insufficiency.

The Bentiromide Test has a sensitivity of about 80% in severe chronic pancreatitis when compared to the Secretin Test [Niederau, 1985]. Bentiromide is benzoyl-tyrosyl-para-amino-benzoic acid which, after ingestion, is hydrolyzed by pancreatic chymotrypsin; the released p-amino benzoic acid is absorbed, and excreted in urine where it can be measured. False positive tests have been reported in renal insufficiency, and in small intestinal disease [Niederau, 1985]. Pancreatic insufficiency is not reliably detected by the bentiromide test until pancreatic chymotrypsin is less than 5% of normal.

Measuring fecal chymotrypsin has been used as an indirect test of pancreatic insufficiency especially in cystic fibrosis [Niederau, 1985]. Like the Bentiromide Test, it is insensitive in mild to moderate disease.

A therapeutic trial of gastric acid inhibition with supplements of pancreatic enzymes makes good sense although it has not been rigorously tested. The patient keeps a diarrhea diary while eating a constant diet, and omitting anti-diarrheal medicines. After a baseline period of 3 days, a proton-pump inhibitor (for example, omeprazole 20 mg twice daily) is added for another 3 days. Then for a final 3 days, pancreatic enzymes (for example, Cotazyme capsules of 8000 lipase units) 4 capsules with meals and 2 capsules with snacks are added to the PPI. An advantage of this trial is that a definite improvement in the PPI period directs the clinician to consider Zollinger-Ellison syndrome. Mucosal Diseases

Among mucosal diseases, celiac sprue would be suggested by a childhood history of diarrhea, presence of diabetes, of iron-deficiency anemia that fails to respond to oral iron, and oral ulcers.

If our patient did not have the clue of triglyceride in her stools, she might have been screened for celiac sprue by searching for IgA endomysial antibodies. The sensitivity and specificty of this immunofluoresent test approaches 100% in specialized laboratories, but there are concerns about the reliability of commercial assays [Grodzinsky, 1994]. Definitive diagnosis of celiac sprue depends on small bowel biopsy, and a clinical response to a gluten-free diet.

A trial of a gluten-free diet has no place in the investigation of patients with suspected celiac sprue unless the endomysial antibody test is postitive, or the small bowel biopsy is compatible with the diagnosis. Some patients with gastrointestinal symptoms such as flatulence and abdominal distension due to functional (idiopathic) bowel disease can feel better when glutenous (cereals and the attendent poorly absorbed carbohydrates) are avoided.

4.4.3 Bloody Diarrhea

Exudative diarrhea is usually obvious because of bloody stools. Sometimes, however, Crohn’s disease is overlooked as a cause of chronic diarrhea if the stools are not grossly bloody. Profile of a Patient

For the past year an eighteen year old college student had episodes of diarrhea (3-4 mushy to watery stools daily) which lasted 1-4 weeks. He was thought to be lactose-deficient, and, indeed, 50 g of a lactose drink provoked abdominal cramps, flatus, and diarrhea. The diarrhea was improved, but not eliminated by avoiding lactose. The discovery of occult blood (Hemoccult® ) in his stools led to flexible sigmoidoscopy. The rectosigmoid mucosa was macroscopically normal, but biopsy specimens contained granuloma typical of Crohn’s disease. The patient continues to have occasional episodes of diarrhea which are helped by loperamide. Pathophysiology

Inflammation can cause disruption of the mucosal surface so that blood or plasma leaks into the lumen and increases the volume of the fecal stream. Anemia, and hypoalbuminemia can result. Inflammatory cytokines, and hastened colonic transit impair colonic absorption of salt and water.

Malabsorption could also be a factor in the diarrhea of Crohn’s disease. Involvement of ileal mucosa might impair the absorption of bile salts so that the concentration of bile salts in fecal water becomes sufficient to block absorption of Na+ and water by colonocytes [Hofmann, 1972]. If hepatic synthesis of bile acids can compensate for a minor disruption of the enterohepatic circulation, the diarrhea can be mainly watery, rather than fatty [Hofmann, 1972]. Diagnosis

Chronic bloody diarrhea would prompt investigation of idiopathic inflamatory bowel disease, or of parasitic disease (E. histolytica, Schistosomiasis). Less obvious is the patient who has chronic inflammation of the colonic mucosa without gross blood. A family history of idiopathic inflammatory bowel disease can be helpful, as well as a past history of hematochezia. Crampy abdominal pain, and right lower quadrant tenderness or mass suggest Crohn’s disease.

A history of recent therapy with an antibiotic might incriminate C. difficile and its exotoxins which should be sought in the patient’s stool.

The presence of fecal occult blood (or excessive leukocytes), unexplained by C. difficle or ameba, necessitates at least a flexible sigmoidoscopy with mucosal biopsy in a young patient, and a colonoscopy in the older patient. The colonoscopist should try to obtain biopsies of the terminal ileum which can be informative even if the colon is normal [Geboes, 1998].

4.4.4 Predominantly Watery Diarrhea Diarrhea Due to Endocrinopathies

This group deserves pride of place, not because of its prevalence, but because thyroid disease, and adrenal disease are eminently treatable. Thyroid Disease

Hyperthyroidism is associated with chronic diarrhea with, and without steatorrhea, and with a hypersecretory state [Donowitz, 1995]. Hypoadrenocorticalism

Gastrointestinal symptoms may be overshadowed by profound fatigue, hypotension, and hyperpigmention. The clinical chemistry laboratory reports peripheral blood eosinophilia, and hyponatremia with hyperkalemia. Diabetes mellitus

Typically, diarrhea begins after 8 years of diabetes; peripheral neuropathy is usually present. Diarrhea is often nocturnal. Profile of a Patient

A 36 year old manager was referred for an opinion about a diarrheal syndrome of 7 months’ duration. Defecation usually occurred urgently during supper; formed stools would be followed progressively by increasingly liquid motions. About half the time, he was awakened during the night. Glucosuria was discovered 7 months ago, and his diabetes was being treated with glyburide (Micronase® ).

A 24 hour stool specimen weighed 760 g, and it contained 5 g of fat. Beets colored his stools within 6 hours. He began to complain of impotence, and testing of his autonomic nervous system revealed an abnormal RR interval. Hemochromatosis was dismissed by a normal hepatic iron index. We assumed that his diabetes had been subclinical for many years. Pathophysiology

Clinicians, puzzling over the cause of diarrhea in diabetes mellitus must consider malabsorption (celiac sprue, pancreatic insufficiency), and watery diarrhea due to impaired adrenergic function in the myenteric plexus associated with autonomic (and, usually, peripheral) neuropathy. Our patient’s diarrhea responded to loperamide which would improve salt and water absorption by slowing intestinal transit. The Irritable (Idiopathic) Bowel Syndrome

The Irritable (idiopthic) Bowel Syndrome is a subset of the functional bowel disorders; it is associated with disturbed defecation, and with symptoms of abdominal pain and bloating. Patient Profile

A 45 year old engineer had 20 years of intermittent diarrhea which was worse when he had to travel to administrative meetings. Diarrhea may lessen to one stool a day when he was on vacation. He was dissatisfied with medical reassurances that his troubles were "due to stress." General physical exams were normal as were routine laboratory screenings (hemogram, chemistry battery); negative fecal occult blood and fat.

A diarrhea diary, kept meticulously over one month, revealed that about half the days were notable for obstipation. Colonoscopy was performed finally because our patient wanted to insure that "nothing was missed"; it was normal. His bowel function is now also normal since he relocated to a less stressful job. Pathophysiology

Whilst there is general agreement that the physiology of defecation is disturbed in patients with irritable (idiopathic) bowel syndrome (IBS), how one can have diarrhea alternating with constipation remains puzzling. Some evidence suggests that patients with IBS have exaggerated intestinal motility responses when they are challenged by environmental, or psychological stimuli [Drossman, 1994]. Whole gut transit time is faster in IBS patients with diarrhea [Cann, 1983] The transit of radiolabelled pellets through the ascending, and transverse colons is accelerated in IBS patients with diarrhea, and the faster the transit, the higher is the fecal weight [Vassallo, 1992].

Patients with IBS may have exaggerated intestinal responses to sugars such as fructose, lactose, and to the more complex carbohydrates in fruits and legumes [Camilleri, 1992]. Diagnosis

The recommended approach to patients with suspected IBS involves a tentative diagnosis based on the history, judicious testing to exclude organic disease, and careful follow-up to insure that the patient’s symptoms are improving.

IBS should be suspected when symptom criteria [Camilleri, 1992], [Manning, 1978] are present (Table 17). Many patients with IBS remember symptoms beginning in their early teens, and the intermittency of their intestinal complaints. Anxiety, or depression are frequent accompaniments.

A complete examination should include physical exam, hemogram, and stool for occult blood, ova and parasites. A collection of stool for 24 hours (Table 16) should be weighed (usually <500 g in IBS), and should be searched for fat (Sudan Stain), and for ova and parasites.

Our patient had diarrhea for many years without exhibiting any alarm signals. His diarrhea diary over one month was remarkable for meticulous detail, the intermittency of his diarrhea (he had as many good days as bad ones), the association of his diarrhea with travel, and the realization that he was never awakened from sleep by diarrhea. Loperamide was helpful during his episodes of diarrhea, which was greatly improved when he changed to a less stressful job. Diarrhea Due to Medications

All patients presenting with chronic (or acute) diarrhea must be questioned thoroughly about their use of medicines. It is sometimes amazing that medication-induced diarrhea can become chronic without the association being suspected by the patient or the clinician. Profile of a Patient

A 34 year old executive was referred for flexible sigmoidoscopy to pursue a diagnosis for a diarrheal syndrome which had not resolved or been explained for 3 weeks. This healthy-appearing woman said that she was passing about a cupful of mushy stools three times a day; she was awakened from sleep once or twice by the need to defecate. During the pre-procedure chat, we uncovered that an anti-inflammatory drug (Clinoril® ) had been prescribed three weeks ago because she sprained her ankle playing badminton. Stopping the Clinoril was associated with her return to normal bowel function within 45 hours. Pathophysiology

The extent of the association between medications and diarrhea is emphasized by the Physician’s Desk Reference Drug Interactions and Side Effects Index, which lists over 600 medications that can cause diarrhea (Table 9). Mechanisms are diverse and often poorly understood, but they often involve osmotically active agents (such as Mg++, citrate, sulfate); drugs which interfere with salt and water absorption; agents which alter intestinal motility; and drugs which stimulate intestinal secretion (phenolphthalein as in Ex-Lax®; dioctyl sodium sulfosuccinate as in Colace®; bisacodyl as in Dulcolax®).

Ethanol can block salt and water absorption in the human small intestine [Mekhjian, 1977].

Patients treated with prolonged courses of antibiotics risk developing diarrhea because the antibiotic perturbs the colonic microflora: Antibiotic Use C. difficile

C. difficile can colonize the colon, and its exotoxins can cause mucosal inflammation. In the absence of mucosal ulceration, the diarrhea is watery. C. difficile toxins are usually present in the stools. Diarrhea usually subsides when the offending antibiotic is withheld [Mitty, 1994], but may require treatment with oral metronidazole, or vancomycin. Suppression of Fermentation

Colonic bacteria which ferment carbohydrate are suppressed [Clausen, 1991] so that osmotically active carbohydrate obligates salt and water in the colonic lumen. The wetter stools usually become normal when the antibiotic is discontinued. Diagnosis

Usually the diagnosis of medication-induced diarrhea is suggested by the history, and it is proven by withdrawing the offending substance.

Uncommonly (except at tertiary care hospitals), a patient may be abusing laxatives, a puzzle whose solution may require careful sleuthing, and analyses of stool, and of urine. Infectious Causes

These invaders can often be suspected from the history, and their presence can be confirmed by examination of the stools. Common Pathogens Giardia

Contact with children in day-care; intimate sexual contact; drinking pristine water in the countryside; foreign travel [Shandera, 1990] should prompt a search for Giardia lamblia in the stools by immunodetection of the Giardia antigen in fresh, or in formalin-preserved stools [Thompson, 1993]. C. difficile

Use of antibiotics currently, or in the recent past, should prompt a search for the exotoxins of C. difficile in a fresh stool specimen. Colonic dilatation with bloody diarrhea is more specific for the diagnosis of overgrowth of this organism. E. histolytica

Foreign travel, or a history of oral-anal intimacies can incriminate E. histotytica whose trophozoites can be found in fresh smears of rectal mucus, or in preserved specimens of freshly-passed stool. IgM antibodies to E. histolytica can be sought in the patient’s serum. History of Less Common Agents

Less common agents can also be suspected from the patient’s history: A. hydrophilia and Cyclospora

Drinking untreated water (Aeromonas hydrophilia, [Holmberg, 1986]), or contaminated water and fruit (Cyclospora cayetanensis [Huang, 1995]), and travel to Nepal (Cyclospora [Hoge, 1993]) have been associated with outbreaks. Spore-Forming Protozoa

Cryptosporidium parvum, Isospora belli, Cyclospora cayetanensis, Enterocytozoon bieneusi, and Septata intestinalis are intestinal spore-forming protozoa that invade intestinal mucosal cells. Cellular injury and inflammatory cytokines impair salt and water absoprtion, and, probably, enhance secretion. Immunodeficient patients are predisposed to prolonged infection with these protozoa. In immunocompetent individuals, diarrhea with Cyclospora can persist after acute infection [Goodgame, 1996]. Diagnosis

Examination of the stools is the most important diagnostic test. The laboratory should be asked specifically to screen for spores or oocysts; multiple samples increases the diagnostic yield [Goodgame, 1996]. Subtle Inflammation

The rubric, microscopic colitis, probably includes other descriptions (lymphocytic colitis, collagenous colitis) of macroscopically normal colonic mucosa whose biopsies are abnormal. Profile of a Patient

A 69 year old lady presented with a history of diarrhea of 3 years duration. Stools were negative for fat and for parasites. Colonoscopy was said to be within normal limits. Many therapeutical trials had been performed; supplemental pancreatic enzymes were thought to lessen the diarrhea.

She was studied according to the protocol outlined in (Table 16). Beets colored her stools within 4 hours. The 24 hour specimen weighed 440 g; it looked like brown, mucoid pudding which tested negative for occult blood (by Hemoccult®), and for fat (by Sudan Stain). Colonoscopy was repeated because colonic disease seemed the most likely cause of her diarrhea. Eleven biopsies were taken throughout the length of a macroscopically normal colon. A spotty mucosal inflammation was discovered (mainly with eosinophils in the lamina propria and with lymphocytes in the surface epithelium); marked thickening of the subepithelial collagen plate was found in many areas. Based on the diagnosis of microscopic colitis, an anti-inflammatory agent (5-amino salicyclic acid) was prescribed; her diarrhea gradually resolved. Pathophysiology

The cause of microscopic colitis is unknown. Speculations include impending inflamatory bowel disease (ulcerative colitis, or Crohn’s colitis), or the sequel of an infectious colitis. It seems certain that the abnormal mucosa is a factor in the pathogenesis of diarrhea, because colons of patients with microscopic colitis malabsorb salt and water infused under steady state conditions [Bo-Linn, 1985]. Diagnosis

Microscopic colitis should be considered in patients when malabsorption and medication-induced diarrhea have been excluded. Irritable (idiopathic) bowel syndrome would not be expected to present de novo in the middle-aged, or elderly. Colonoscopy with biopsy is probably more informative than flexible sigmoidoscopy because the right colonic mucosa can be more involved than that in the left colon [Janda, 1991]. Multiple (>12) mucosal biopsies should be obtained. Perhaps Subtle Colitis is a better rubric than Microscopic Colitis, because all colitides have microscopic abnormalities. Hypersecretory (Hormonally-Induced) Diarrhea

This category of predominantly watery diarrhea is uncommon, and physicians often order searches for vasoactive intestinal polypeptide (VIP), without first ascertaining if the diarrhea is voluminous (it usually exceeds 1000 ml a day), if the diarrhea improves with fasting (it should not), or if the osmotic gap in a fresh fecal supernate is less than 50 mosmols / kg. Laxative abuse is much more commonly encountered than VIPomas [Read, 1980].

A history of facial flushing may direct a search for carcinoid tumor, or for pheochromocytoma.

4.5 Diarrhea of Unknown Cause

For this phrase to have any credence, it must be carefully defined, much as is Fever of Unknown Origin in Infectious Disease. By definition, then, patients with DUO have an increased frequency of passing stools of decreased consistency for more than 2 months. Thorough investigation of the groups of diseases outlined in Sections 4.4 is unrewarded. In particular, there is no evidence for immunosuppression. Finally, the patient’s diarrhea is difficult to control with customary non-specific measures. A virtue of an appellation of DUO is that there are a small group of diseases that can be revisited so that selected study can establish a diagnosis [Schiller, 1991].

4.5.1 Fecal Incontinence

Incontinence is probably paramount if an elderly, parous lady admits to involuntary passage of small amounts of stool. Corroborative evidence might be obtained in a direct physical exam. An anal "wink" (cutaneo-anal reflex) indicates that neural pathways are intact. Visual and digital exam can assess the tissue mass of the pudenal body, and the strength of the external anal sphincter. A descent of the pelvic floor of more than 1 - 2 inches during straining (in the left lateral position) denotes muscle weakness. The toilet test (whereby the patient strains on a toilet while the pelvic floor is viewed with a hand mirror) can highlight abnormal descent of the pelvic floor, hemorrhoids, and rectal prolapse. A 24 hour stool might weigh 100 - 200 g. Such a patient should be referred to a specialized clinic for electrophysiological and manometric tests; biofeedback training can help many patients.

4.5.2 Overlooked Malabsorption

A fat balance may give a spuriously high coefficient of absorption because the stool collection was incomplete, or because the patient was eating insufficient amounts of fat. Furthermore, the underlying disease may have progressed since the initial testing so that steatorrhea is now readily detectable.

Repeating the inquiry outlined in (Table 16) involves no risk or major financial penalty. Profile of a Patient

A 47 year old lawyer’s diarrhea began one year ago at a time when his daughter contracted diarrhea in a daycare center. He did not respond to two courses of metronidazole based on a suspicion of giardiasis. Random stool specimens were said to be negative for blood, fat, and leukocytes. Sigmoidoscopy with biopsy was unremarkable. The family’s pediatrician suggested a diet of bananas, applesauce, and rice on which the diarrhea seemed to improve.

Six months after his initial work-up, he consented to the protocol in (Table 16). The 24 hour stool weighed 1070 g, containing 21 g of fat. The Sudan Stain revealed sudanophilic droplets after acidification only.

The hypothesis of celiac sprue was unsupported by the duodenal mucosal biopsies which contained normal villi; however, patchy areas of inflammation and mucosal erosions were seen. Additionally, a lesser curve gastric ulcer was encountered by the endoscopist. The suspicion of Zollinger-Ellison symdrome was supported by a serum gastrin of 860 pg / ml (normal 46 — 142 pg / ml), and a dramatic improvement in his diarrhea with a proton pump inhibitor. Subsequent CT-directed biopsy of a hepatic mass revealed neuroendocrine tumor.

4.5.3 Induced and Factitious Diarrheas

As emphasized in, many medications can cause diarrhea. Laxative abuse can be especially vexatious; it should be suspected in patients who seem overly concerned about maintaining a thin body. Phenolphthalein can be readily detected by alkalinizing fecal supernate, or urine with a drop of 1 N NaOH. The clinical chemistry laboratory can search for bisacodyl, Mg++, sulfate, phosphate, and anthraquinones. Profile of a Patient

A 40 year old licensed practical nurse was referred with a diarrheal syndrome of 3 years’ duration. Multitudinous investigations had been unrewarded, and her diarrhea continued even when she fasted. Therapy with prednisone 20 mg a day ameliorated her diarrhea, but she was now confined to a wheelchair because of osteopenic fractures.

While fasting, and on intravenous feeding, her 24 hour stool weight was 770 g. The supernate of this liquid stool had an osmolality of 316 mosmols / kg; [Na+], 31 meq / L; [K+], 53 meq / L. The osmotic gap of 148 was found to consist of Mg++. Bottles of milk of magnesia were uncovered in her dresser drawer and in her suitcase; she said that she was using milk of magnesia to regulate her bowels. Profile of a Patient

A 36 year old roofer complained of diarrhea since a laparotomy for abdominal pain 4 years previously. A Meckel’s diverticulum and a small carcinoid were resected. Extensive work-ups were unrewarded; 24 hour stools of 300 - 400 g contained no excess fat, blood, or leukocytes. Because he was applying for disability (as a roofer with diarrhea!) he was admitted to hospital for a 48 hour fast. He said that beets passed through his gut within 10 minutes. A random stool specimen had a [Na+] of 28 meg / L, [K+] of 41 meg / L, and an osmolality of 168 mosmols / kg.

Our patient had factitious diarrhea. A whole gut transit time of 10 minutes might be compatible with an oro-anal fistula, but a fecal osmolality of 168 mosmols / kg can be achieved only by watering the stool specimen!

4.5.4 Malabsorption of Bile Salts

This penultimate category of DUO contains some inconsistencies. Malabsorption of bile salts can occur in a wide variety of intestinal troubles such as post-vagotomy, post-cholecystectomy, post-abdominal irradiation, diabetes, and chronic idiopathic diarrhea [Schiller, 1991]. However, not all of these patients are improved when cholestyramine, an anion-binding (bile salt binding) exchange resin is ingested therapeutically. This inconsistency can be explained by our ignorance of the concentration of bile salts in fecal water which must be high enough to block salt and water absorption by colonocytes. In other words, bile acid malabsorption as measured with radiolabeled bile salts does not predict the response to cholestyramine.

If a therapeutic trial of cholestyramine is undertaken, sufficient resin must be ingested (for example, cholestyramine as colestipol is supplied in packets of 5 g; 2 packets with each of three meals, and at bedtime). The dosage of this unpalatable therapy should be reduced promptly to the lowest effective amount. A confounding factor of a therapeutic trial with cholestyramine is that the resin non-specifically may improve the consistency of the stools just as psyllium does [Wenzl, 1995].

4.5.5 Occult Infections and Inflammations

The boundaries encompassed by chronic idiopathic diarrhea will shrink as knowledge grows for "all experience is an arch where through gleams that untraveled world, whose margins fade for ever and forever when — (we) — move" [Tennyson, Poems, 1842]. At present, the clinician must pursue an epidemiological history, and an alliance with the gatroenterologist and mucosal histopathologist.

The sudden onset of diarrhea which becomes chronic suggests an infectious etiology because this syndrome occurs in foreign travelers, and in local folk who ingest untreated water, or raw milk. Brainerd diarrhea [Osterholm, 1986] is a rubric which is used to embrace such an illness persisting for many months without an infectious or non-infectious etiology being found. A relationship between acute and chronic disease can be found by colonoscopy and histopathology: focal areas of inflammation are present in a patchy manner mainly in the right colon. In the one patient studied, these abnormalities were still present two years later [Janda, 1991]. We can speculate that an unknown infectious agent altered the balance between the mucosal immune system, the intestinal neuroendocrine system, and the inflammatory cascade.

The punch line: after a careful review and sagacious repetition of simple stool studies, the gastroenterologist and histopathologist should combine forces to study systematically the proximal small mucosa, and especially, the entire colonic mucosa. Distal ileal biopsies should be obtained for good measure. If the mucosal biopsies are normal, distraught patients with DUO may be assured that their chronic diarrhea will eventually resolve [Afzalpurkar, 1992].

5.0 When a Gastroenterologist Should Be Consulted

5.1 Bloody Diarrhea For More Than Two Weeks — Stool Culture is Negative