Travel Medicine Advisor Traveler's Diarrhea: Prevention and Self-Management Charlcs D. Ericsson, MD Medical Director for the University Center for Travel Medicine, Herman Hospital, University of Texas, Houston, the author has written extensively on prevention and treatment of traveler's diarrhea. Philip C. Johnson, MD Assistant Director for the University Center for Travel Medicine, Herman Hospital, University of Texas, Houston, the author has written on treatment of traveler's diarrhea and Norwalk infection. For each week that travelers from developed countries reside in developing nations, their risk of travelcr's diarrhea is 20-25%. In studies lasting several weeks, diarrhea rates of 40-60% have been reported. Although mortality is not described with this disease, individuals may experience a considerable degree of morbidity. On average, untreated traveler's diarrhea lasts three to four days, and an affected person will pass an average of 17 unformed stools. In addition, many individuals experience abdominal cramps or nausea, and some are distressed with vomiting or fever. Definitions A normal person will pass less than 100 ml of stool per 24 hours in the form of one large or several small evacuations. A person with diarrhea often passes more than 200 ml of stool water per day in one or more bowel movements, which frequently are unformed. One or more unformed evacuations totaling less than 100 ml in 24 hours generally is defined as loose stools rather than diarrhea. By the same token, passage of formed stools with a total volume greater than 100 ml is not diarrhea. Although some patients with dysentery pass frequent but small-volume, bloody, mucoid stools, dysentery is still considered a form of diarrhea. All of these definitions are somewhat academic. Because physicians must instruct patients when to start treatment of traveler's diarrhea, more clinically useful definitions of diarrhea have been employed in treatment and prevention studies. In studies, traveler's diarrhea is often defined at clinical presentation as passage of at least three unformed stools in a 24-hour period and at least one symptom of enteric disease (abdominal pain, nausea, vomiting, or fever). Passage of fewer than three unformed stools per day is considered mild illness. Although similar enteropathogens cause both diarrhea and mild illness, the natural history of untreated mild illness is only two to three days. Both mild illness and traveler's diarrhea respond to antimicrobial agents and symptomatic treatment. We instruct travelers to begin symptomatic treatment promptly when they pass two or more unformed stools and have at least one symptom of enteric disease. Travelers should also be instructed to observe their stools for the presence of blood and mucus; this finding implies dysentery, especially when the traveler also experiences temperature elevation. This presentation is classic (but not specific) for organisms such as Shigella, Salmonella, Campylobacter, or enteroinvasive Escherichia coli (EIEC). Risk Factors Studies of the epidemiology of traveler's diarrhea have shown that travelers from developed countries visiting developing nations are at highest risk. Only a small percentage of persons from developing nations who visit developed countries acquire a form of traveler's diarrhea. In early studies, fewer than 10% of Mexicans traveling to the United States developed diarrhea. This diarrhea was short-lived and was believed to be due to changes in diet rather than to enteric infections. Travel rs to rural areas and tropical climates may be at higher risk for traveler's diarrhea compared with persons traveling to cooler climates and more urban settings. The major source of infection appears to be food; cultures of food obtained in some developing countries have the same level of bacterial contamination as that of stool. Epidemiologic studies have shown an increased risk of traveler's diarrhea in persons who eat fruit and vegetables that cannot be peeled and those who eat food from street vendors. Buffet-style meals served on cruise ships and in hotels may involve the highest risk. The general rule that heat kills bacteria is of prime importance to travelers considering what foods to eat. Water may be a source of traveler's diarrhea caused by rotavirus, Norwalk virus, or parasites; it is a less likely source of bacterial enteropathogens. Although ice can be a source of contaminated water, a recent study showed that freezing decreases levels of bacteria in water. Alcoholic beverages made with contaminated ice have lower levels of bacteria over time, compared with nonalcoholic cold drinks. However, alcoholic drinks made with ice should still be considered risky. Heating water to the boiling point for one minute effectively removes most bacterial, viral, and parasitic pathogens; boiling for five minutes is generally recommended to guarantee killing of all pathogens. Drinking tap water from the hot water faucet is not a substitute for boiling. A number of filters and resin ion-exchange cups designed to remove enteropathogens from water havc been studied and can be recommended for travelcrs on backpacking trips. However, attention must be paid to the limitations of such devices; although they adequately control the risk of giardiasis, most fail to remove viruscs from water. Etiologic Agents The etiologic agents responsible for traveler's diarrhea in studies conducted in Mexico are shown in Table 1. Studies from different parts of the world have shown a varied percentage of the same organisms. Both a temporal and a seasonal variation in these organisms may occur from country to country and even within the same country. The most common cause of traveler's diarrhea, regardless of the country visited, is enterotoxigenic E. coli (ETEC). ETEC produces either a heat-stable or a heat-labile enterotoxin, or both. These toxins bind to cells in the small intestine and activate cyclic AMP or cyclic GMP, which causes an outpouring of fluid and electrolytes from the enterocyte. The heat-labile enterotoxin of E. coli is similar to cholera toxin, and the pathogenesis of watery diarrhea caused by Vibrio cholerae and ETEC is the same. Classically, stools are watery, with multiple large-volume evacuations. Although cells in the small intestine are affected, fluid and electrolytes can still be absorbed in the small intestine. This absorption is facilitated by solutions containing glucose as well as electrolytes. Oral rehydration therapy has been life-saving in treating children with diarrhea in developing countries; however, its usefulness in travelers is not as well-studied. Table 1. Common Enteropathogens Causing Acute Diarrhea in Travelers to Mexico________________ Organism Percent Bacterial Enterotox igenic E. coli 40 Shigella spp 15 Salmonella spp 7 Enteroadherent E. coli 7 Entcroinvasivc k. coli <5 Campylobacter jejuni <5 Viral Rotavirus 5-10 Norwalk virus 5-10 Parasitic Giardia lamblia 2 Entamoeba histolytica 1 Cryptosporid ium <1 Unknown <15 ______________________________________________________________________________ Shigella, Salmonella, Carnpylobacter, and E. coli can produce a diffuse colitis. As few as lOO viable organisms are sufficient to cause shigellosis, but the pathogenesis of shigellosis is not completely understood. There are four species of Shigella: S. sonnei, S. boydii, S.flexneri, and S. dysenteriae. Shiga toxin is in part responsible for diarrhea produced by S. flexneri. Because the organism is infrequently isolated from the blood, it is thought to be noninvasive. Studies have shown that Shigella penetrates the epithelium of the intestinal mucosa, but it does not extend as far as the muscularis mucosa or the lamina propria. Shigellosis is typically a biphasic disease, starting with frequent passage of high-volume stools and followed by passage of lower-volume, bloody, mucoid stools as the colon becomes involved. Symptoms of fever and tenesmus, and blood and mucus in the stool suggest shigellosis. Salmonella Several thousand strains of Salmonella can produce gastroenteritis. Salmonella typhi, the cause of typhoid fever, is an infrequent cause of traveler's diarrhea (occurring in fewer than one in lO,OOO travelers). Other strains of Saltnonella can produce an enteric fever picture or can cause an enteritis. Typically, nontyphoidal salmonellosis has a broad range of clinical presentations, ranging from an invasive picture indistinguishable from shigellosis to a secretory presentation that mimics ETEC disease. Campylobacter Campylobacter jejuni is frequently associated with dairy products and poultry. Although it is an Invasive pathogen, C. jejuni is infrequently cultured from blood. Diffuse colitis may occur, and the chronic form of Campylobacter illness may last several weeks. Enteroinvasive E. coli Enteroinvasive E. coli causes up to 5% of all traveler's diarrhea. This E. coli is similar to Shigella in its ability to produce a Shiga toxin and cause conjunctivitis in guinea pigs. Certain serotypes of E. coli (0157:H7 and 027:H11) have been associated with outbreaks of hemorrhagic colitis that typically is not accompanied by fever. Like enteroinvasive E. coli, hemorrhagic E. coli strains produce Shiga toxin. Some diarrheagenic E. coli are recognized by their adherence properties in assay systems. The relationship of adherence to virulence is uncertain, but such enteroadherent E. coli (EAEC) appear to cause up to one third of the cases of traveler's diarrhea in which no etiologic agent is found. Clostridium Clostridium difficile is an anaerobic bacterium that may produce diarrhea in travelers who have taken antibiotics for prophylaxis of diarrhea or for treatment of some other infection. The hallmark of C. difficile diarrhea is pseudomembranous enterocolitis, although other presentations have been noted. Diarrhea due to C. difficile may present several weeks to months after a traveler returns home. Detection of C. di~icile toxin in stool is the preferred diagnostic test, because the organism can be grown anaerobically from stools of asymptomatic persons. Clostridium perfringens is a common cause of food poisoning and ought to cause a proportion of traveler's diarrhea. However, the organism is not well-studied in travelers. Because illness typically lasts fewer than 24-36 hours, C. perfringens disease may not have been enrolled or sought in many studies. Other Agents Other causes of traveler's diarrhea include viruses and parasites. Norwalk virus has been associated with up to 10% of cases of traveler's diarrhea. In one study of Panamanian travelers visiting Mexico, Norwalk virus accounted for 15% of cases of diarrhea. Other small, round gastroenteritis viruses may also cause traveler's diarrhea. For example, rotavirus has been isolated in 5-10% of traveler's diarrhea cases. However, because viruses often are found in association with bacterial enteropathogens, their exact role in causing disease is not always clear. In certain areas of the world, parasites have been a major cause of traveler's diarrhea. Both Giardia larnblia and Cryptosporidiurn have been associated with disease in travelers to Leningrad. However, G. Iarnblia, Entarnoeba histolytica, and Cryptosporidiurn account for only 2-4% of diarrhea in travelers to Mexico. The long incubation period of these parasites makes them more likely causes of diarrhea in travelers who have returned to a developed country; as such, they may be underreported as causes of traveler's diarrhea. Most cases of traveler's diarrhea in which no etiologic agent is found respond to treatment with an antimicrobial agent. Bacterial infections appear to account for the majority of cases of diarrhea of unknown cause. Chemoprophylaxis of Traveler's Diarrhea Tables 2 and 3 show the various agents that are currently used in short-term prophylaxis of traveler's diarrhea. Bismuth subsalicylate (BSS) tablets and liquid effectively prevent traveler's diarrhea in a dose-responsive fashion. The effective dosage for liquid BSS is one 8-ounce bottle per day in divided doses. BSS tablets are reported to be as effective as the liquid at half the dosage. Virtually everyone taking BSS will experience darkening of stools due to formation of insoluble bismuth salts in the gut; black coating may form on the tongue. The level of salicylate found in these preparations is a potential problem because BSS dissociates in the gut, and nearly all the salicylate is available for absorption. Travelers must not take salicylate-containingproducts (such as aspirin or aspirin-containing medications) if they use BSS prophylaxis. In one study, tinnitus was experienced as frequently by subjects taking placebo as by those taking BSS. Because BSS can interfere with the bioavailability of doxycycline, travelers requiring doxycycline for malaria prophylaxis should not take BSS. The benefit of BSS prophylaxis can be maximized if travelers are also careful about where they eat. Antimicrobials (Table 3) are the most effective agents in preventing traveler's diarrhea. Of currently recommended antimicrobials, doxycycline was the first to be studied extensively; reduced rates of traveler's diarrhea were commonly seen. Side effects of doxycycline include photosensitivity, gastrointestinal (GI) distress, and vaginal candidiasis. The emergence of resistance of many enteropathogens to doxycycline has raised questions about its usefulness. Trimethoprim alone and trimethoprim-sulfamethaxozole (TMP-SMX) have been used for prophylaxis of traveler's diarrhea; both have been highly effective. Rashes occur in approximately 4% of persons taking one double-strength tablet of TMP-SMX per day; GI distress and vaginal infections are also reported. Although the possibility that a patient may develop Stevens-Johnson syndrome with TMP-SMX use is small (incidence < 1/50,000), the potential consequences of this syndrome are drastic. Quinolone agents such as norfloxacin and ciprofloxacin can be used for prevention of traveler's diarrhea. No emergence of resistance to norfloxacin was noted during a two-week study. Resistance to the quinolone derivatives is believed to be chromosomal rather than plasmid-mediated, which theoretically makes them more desirable as prophylactic agents. Table 2. Bismuth Subsalicylate (BSS) in Prevention of Traveler's Diarrhea_____________________ Total daily BSS dose (mg) Dosing regimen Percent reduction in diarrhea 1050 2 tablets BID 35 1050 1 tablet QID 40 2100 4 tablets BID 41 2100 2 tablets QID 65 4200 2 ounces QID 62 ______________________________________________________________________________ Table 3. Antimicrobial Agents in Prevention of Traveler's Diarrhea___________________________ Agent Daily dose Percent reduction Comments in diarrhea Doxycycline 100 mg 58-83 Resistance high in some areas of the world. Trimethoprim 100 mg 52 Resistance increasing. Trimethoprim/ 160/800 mg 73-95 Resistance increasing. sulfamethoxazole Norfloxacin 400 mg 89 Very low resistance. Expensive. ______________________________________________________________________________ In 1985, the NIH Consensus Development Panel did not recommend antimicrobials for prevention of traveler's diarrhea. This decision was based primarily on 1) concern that prophylaxis of traveler's diarrhea may cause emergence of resistant organisms in developing countries, and 2) the number of side effects associated with antimicrobials. Other agents, such as lactobacillus preparations and activated charcoal, have been studied for prevention of traveler's diarrhea; they have not been found to be effective and are not recommended. Despite concerns of the NIH Consensus Development Panel, a cost-effectiveness argument can be made for antimicrobial prophylaxis of traveler's diarrhea in persons at high risk or those with the most to lose from illness. Persons planning once-in-a-lifetime vacations or important business trips may be candidates for chemoprophylaxis. Travelers with inflammatory bowel disease, diabetes mellitus, renal insufficiency, or AIDS might also be considered for prophylaxis. Our approach is to explain the risks and benefits; although we are willing to prescribe antimicrobials, we more often recommend prophylaxis with BSS tablets and prescribe antimicrobials for early treatment of traveler's diarrhea when (and if) it occurs. Treatment Replacement of fluid and electrolytes is the hallmark of therapy for diarrhea. Home-produced oral rehydration solutions can be useful in adults, but their use in children is discouraged because of the possibility of inadequate measurement of salts. Commercial oral rehydration solutions usually are not necessary for adults with traveler's diarrhea because significant dehydration is uncommon; nevertheless, such solutions are becoming readily available in pharmacies around the world. Flavored mineral waters (which supply glucose and a hypotonic salt solution) are also readily available in most countries and often suffice for rehydration in a typical case of traveler's diarrhea. We advise patients to replace stool losses with enough oral liquid to maintain adequate urination. We also advise travelers to adjust their diets during illness, although scientific justification for this advice is generally lacking. We recommend avoiding miL~ or milk products during the active phase of diarrhea; ingestion of these products may result in additional stool losses due to lactase deficiency and lead to an osmotic diarrhea. Simple carbohydrates and liquids are recommended for the first day or so after illness, followed by addition of protein and finally fats to the diet. Nonspecific, symptomatic treatment of traveler's diarrhea is described in Table 4. BSS is effective, perhaps because of its antimicrobial effects and dissociation products. Loperamide is more effective than BSS in reducing the number of unformed stools and abdominal cramps. Patients should be cautioned not to use antimotility agents such as loperamide to treat dysentery. Results of studies in volunteers and animal models have shown that loperamide treatment may make shigellosis and C. difficile disease worse. In one study, treatmenl of nondysenteric shigellosis with loperamide did not appear to prolong symptoms; however, in anotherstudy, diphenoxylate prolonged the course of diarrhea in patients presenting with bloody diarrhea. Although agents containing kaolin-pectin increase the bulk of a stool, they do not decrease the number of stools passed. Activated charcoal and anticholinergics are not effective in treating traveler's diarrhea. Table 4. Nonspecific Treatment of Traveler's Diarrhea____________________________________ Agent Dosage Benefit Comments Kaolin/pectin 4 oz after each More formed stool. Poor palatability. loose stool. No change in diarrhea Safe. frequency. Bismuth l oz each 1/2 h Relief of diarrhea Black tongue and stools. subsalicylate (BSS) for 8 doses. comparcd to placebo. Salicylate absorption. Loperamide 4 mg STAT, then Better relief of diarrhea Avoid use in dysentery. 2 mg after each compared to BSS. loose stool. (Do not exceed 16 mg/day.) ______________________________________________________________________________ Antimicrobial agents have proved very effective for treatment of traveler's diarrhea. Recommended agents are shown in Table 5. A three-day regimen is usually sufficient; at least one study of traveler's diarrhea showed that one large dose of an antimicrobial is as effective as three-day treatment. TMP-SMX, norfloxacin, and ciprofloxacin are currently the agents of choice. Doxycycline is not recommended. Although TMP-SMX is cheap and effective, resistance to this agent is increasing in developing countries, and its position as a front-line recommendation may change in the next few years. Furazolidonc, which is effective against bacterial agents and G. Iamblia, is a back-up recommendation; it is only modestly effective, and an Antabuse-like reaction may occur when alcoholic beverages are consumed. Appropriate treatment of C. jejuni is still controversial; this organism is sensitive to the quinolones but resistant to TMP-SMX. In one st udy, treatment with erythromycin improved symptoms of disease. In regions such as Asia, where Campylobacter disease is prevalent, quinolones may be the antimicrobial drugs of choice, despite their high expense. Treatment of non typhoidal salmonellosis with antimicrobials remains controversial. Past studies have documented prolongation of the carrier state after use of antibiotics (especially ampicillin). Whether this prolongation occurs as commonly with the quinolone derivatives as it did with ampicillin awaits further study. Combination therapy using an antibiotic plus loperamide (Table 5) has recently been studied and is our current recommendation to travelers. This approach provides fast relief of symptoms (afforded by the rapidly absorbed loperamide) and elimination of the offending organism. Combination therapy is also highly effective for patients with blood in their stools or fever. Table 5. Antimicrobial Agents in the Treatment of Traveler's Diarrhea________________________ Agent Dosage* Average duration of Comment diarrhea (h)* € Trimethoprim/ 160/800 mg 34 Not effective against sulfamethoxazole BID x 3 days. Camplylobacter. 320/1600 mg STAT. 28 Not studied in combination with loperamide. € Ciprolloxacin 500 mg BID x 5 days. 29 Single large dose or 500 mg BID x 3 days probably effective but unstudied. € Furazolidone 100 mg QID x 5 days. 57 Antabuse-like reaction Effective against Giardia (not the agent of choice). € Trimethoprim/ 160/800 mg 1 Avoid loperamide in sulfamethoxazole BID x 3 days. dysentery. plus Loperamide 4 mg STAT, then 2 mg after each stool. (Do not exceed 16 mglday.) * In published studies. Practical Patient Preparation Our approach to travelers is to educate them about the various risk factors associated with diarrhea. We stress that early treatment often can control symptoms of traveler's diarrhea within 24 hours, and we openly discuss the risk of prophylaxis versus early treatment. A common recommendation is BSS tablets for prevention and, if diarrhea develops, early treatment with an antimicrobial agent. Suggested components of a travel kit include a three-day supply of TMP-SMX (or a quinolone agent), capsular loperamide (or liquid loperamide, which can now be bought over the counter), BSS tablets, and oral rehydration packets (if the traveler will be visiting areas remote from pharmacies). Immunity One important remaining question is whether natural immunity develops as a traveler continues to be exposed to enteric pathogens in developing countries. Epidemiologic studies have shown that traveler's diarrhea occurs early in the course of travel, and that the incidence decreases over time. These findings imply that immunity to at least some common etiologic agents eventually develops. In a recent study, travelers appeared to lose this immunity a few months after returning to a developed country. Host immunity to agents such as ETEC, Shigella, and Salmonella is still not fully understood. In ETEC disease, antibody responses develop to enterotoxins, colonization factors, and lipopolysaccharide; however, the exact protective response (if any) remains to be revealed. Presumably, immunologic protection for all four species of Shigella must be developed. Immunologic protection against salmonellosis may never be practical because of the large number of different strains. No active vaccine has been developed to protect against the common causes of traveler's diarrhea. One promising development is that bovine immunoglobulins, taken before and after an experimental challenge with ETEC, appear to protect volunteers against clinical disease. Efforts are currently under~vay to immunize cattle against various pathogens causing traveler's diarrhea and then to use the bovine colostral immunoglobulins as passive protection against traveler's diarrhea. References Bandres JC, et al: Heat susceptibility of bacterial enteropathogens. Implications for the prevention of traveler's diarrhea. Arch Intern Med 1988;148:2261-2263. DuPont HL: Bismuth subsalicylate in the treatment and prevention of diarrheal disease. Drug Intell Clin Pharm 1987;21 :687-693. Ericsson CD, et al: Treatment of traveler's diarrhea with sulfamethoxazole and trimethoprim and loperamide. JAMA 1989;263:257-261. Ericsson CD, et al: Ciprofloxacin or trimethoprimsulfamethoxazole as initial therapy for traveler's diarrhea. Ann Intern Med 1987;106:216-220. Ericsson CD, et al: The role of location of food consumption in the prevention of traveler's diarrhea in Mexico. Gastroenterology 1987;79:812-816.