Acute Infectious Diarrhea
Acute infectious diarrhea is caused by a large number of micro- organisms. Until the late 1960's, however, the etiology of most cases of diarrhea was obscure except for epidemics or localized outbreaks. In only 10 to 20% of sporadic cases could an enteric pathogen be recognized (Salmonella ,Shigella , or "enteropathogenic" serotypes of Escherichia coli. ). The rest were considered "nonspecific" or "viral" in etiology for lack of a more precise designation. This situation has now drastically changed, particularly with the recognition of rotavirus, enterotoxigenic E coli. , and several other recently implicated bacterial pathogens. Laboratories equipped to identify the entire range of known enteric pathogens can make an etiological diagnosis in 60 to 80% of cases. Obviously, there yet remains a smaller but significant "unknown" group of cases.
Acute infectious diarrhea may be caused by bacteria, viruses, and protozoa. Rotavirus is the major cause of severe diarrhea in children between the ages of 6 months and 2 years in both tropical and nontropical areas of the world.
The parvovirus-like agents are recognized primarily as a cause of localized epidemics of diarrhea in all ages, and newer viral etiologies are being described Entamoeba histolytica. and Giardia lamblia. are the only common protozoa that produce acute diarrheal illness; other larger parasites are not responsible for this syndrome.
In the United States, infectious diarrheas occur primarily as
i)sporadic cases with the peak incidence during the first 2 years of life,
ii)localized outbreaks in hospital nurseries (which have become very rare) or after ingestion of food or water contaminated with specific pathogens,
iii)a common illness among tourists visiting the developing countries, and
iv)small localized outbreaks of unusual organisms of great public health importance, such as that recently due to Vibrio cholerae. in Louisiana (3).
Since the etiological agent of infectious diarrhea may be suggested by several factors, including age, travel history, and recent ingestion of certain foods or water, input from the clinician is necessary for optimal laboratory evaluation. The laboratory cannot blindly search for all possible pathogens in the stool. The task of the clinical microbiology laboratory in the diagnosis of bacterial diarrhea is to isolate and identify from among the fecal microflora those bacteria most likely to be implicated as agents of diarrhea.
The feces of normal adults contain from 1011 to 1012 viable bacteria per g. The majority of these bacteria are found in the colon. In the small intestine, the fluid contents contain 102 to 103 bacteria per ml, and in the terminal ileum, the count usually does not exceed 107 bacteria per ml. The bacterial flora of the gastrointestinal tract is established within several weeks of birth and, except for intrusions by pathogens or antibiotics, remains relatively stable for life. The dominant bacteria in feces are non- sporeforming anaerobic bacilli, but gram-negative facultatively anaerobic bacteria, such as E. coli. and other members of the family Enterobacteriaceae are also normally present.
The acute diarrheal illnesses vary from mild to severe, and with a few exceptions, the clinical picture is usually not very helpful in suggesting an etiological agent. All have in common the passing of liquid stool in abnormally large volumes or with increased frequency. There are also asymptomatic carrier states described for all the enteric pathogens. Depending on the organism, this may follow recovery from clinical illness or be completely unassociated with it.
The clinical picture is dependent upon the pathogenesis and pathophysiology of the infection. Voluminous diarrhea, not containing blood or leukocytes and associated only with signs of dehydration and without significant fever, suggests small-bowel infection with a noninvasive, enterotoxigenic organism.
In contrast, frequent diarrheal stools of small quantities which contain pus cells or blood and are associated with abdominal pain, fever, and relatively mild dehydration suggest large-bowel infection with an invasive organism.
Immediate clinical management, fortunately, is not heavily dependent upon the etiological diagnosis. Replacement of fluid and electrolytes, either intravenously or orally, will correct the dehydration; antimicrobial therapy is indicated only when certain bacteria are isolated or highly suspected, such as in cholera or severe shigellosis. Etiological diagnosis, however, is important, both for clinical management, particularly in guiding the use of antimicrobial agents, and for the institution of public health control measures.
Pathogenesis and Pathophysiology
Current data indicate that bacteria cause diarrheal illness predominantly by either of two mechanisms:
i)the production of enterotoxins or
ii)invasion of intestinal mucosa.
These two processes are usually mutually exclusive, in that one mechanism or the other predominates.
In diarrheal diseases caused by Staphylococcus aureus. and Bacillus cereus. , a preformed enterotoxin, produced by the organism growing in food, is ingested. In all other cases, the enteric pathogen must be ingested, pass through the acid environment of the stomach, and then colonize or invade either the small or large bowel before the disease is manifested. The inoculum size causing disease varies tremendously and can only be estimated from experiments in volunteers. It is known that as few as 10 organisms of Shigella. may cause illness, whereas it takes 106 and 1010 organisms, respectively, of V. cholerae. and enterotoxigenic E. coli. to cause disease.
The acid barrier of the stomach is undoubtedly of major importance in all of these illnesses. Persons with a decreased stomach acid for any reason (i.e, through the taking of antacids or secondary to ulcer surgery) are at increased risk of developing infection with V. cholerae. , a highly acid-sensitive organism.
Once through the stomach, the organisms must colonize the intestine, i.e., attach to the mucosal surface and multiply there. Colonization of the small bowel byV. cholerae. and enterotoxigenic E coli. has been studied in some detail. Some enterotoxigenic strains of E coli. have specialized structural proteins (fimbria) which have been termed colonization factors and which aid in attachment to the small- bowel mucosa. Little is known about the colonization events of bacteria in the large bowel.
Once colonization has taken place, the organisms either elaborate enterotoxin(s) or invade the mucosa. The enterotoxins of V. cholerae and E coli. have been the best characterized. These enterotoxins bind to specific receptor sites on the mucosal cells and stimulate the hypersecretion of fluid and electrolytes through the stimulation of adenylate or guanylate cyclase.
The resulting outpouring of fluid from the small bowel overwhelms the absorptive capacity of the large bowel, and diarrhea results. The diarrheal fluid contains no inflammatory cells, very little protein, and mucus, which is discharged from the goblet cells of the small intestine. These enterotoxigenic organisms do not invade tissue, and biopsies from the secreting small bowel show an intact mucosa with only minimal inflammatory changes.
In contrast, the invasive organisms damage the mucosa, primarily of the large bowel, which results in the production of small amounts of diarrheal stool containing pus cells, blood, and large amounts of protein. No specific biochemical lesion has been identified to explain these events. Biopsy of the mucosa shows marked histological damage and inflammation, and bacteria can be identified in the lamina propria; occasionally, blood stream invasion may occur.
Some organisms, like Clostridium difficle , produce a cytotoxic enterotoxin which results in histological damage to the large bowel mucosa identical to that produced by an invasive organism like Shigella.
All of these illnesses are usually self-limited, lasting from 1 to 2 days in the case of enterotoxigenic E. coli. to 5 to 10 days in some cases of shigellosis Though it is postulated that the secretory immune response of the host is responsible for the termination of disease, this has yet to be conclusively demonstrated. It is clear that homologous immunity to reinfection develops after illness, that this protection is undoubtedly also mediated through the local immune response, and that it may not be reflected by serum antibody levels. Attempts at immunization against these diseases are also directed toward stimulation of the local intestinal immune response through the use of oral vaccines.
Specimen Collection and Transport
Specimens which may yield an etiological agent of infectious diarrhea include feces, duodenal contents, vomitus, and, with a few agents, bone marrow and blood. A screw-capped glass container has often been recommended for collection of fecal specimens; however, this is frequently not practical for many laboratories. Alternatively, feces may be passed directly into a clean, waxed cardboard container with a tight, leak-proof cover. Liquid stools should be collected in a glass or plastic container. Feces can also be collected from a sterile bedpan; however, the specimen is unsatisfactory if there is any residual soap, detergent, or disinfectant left in the pan or if there is contamination by urine. Those portions of the stool which contain pus, blood, or mucus should be submitted for examination.
If feces are not readily obtainable, a rectal swab may be submitted. The swab is passed beyond the anal sphincter, carefully rotated, and withdrawn. Swabbing lesions of the wall of the rectum or sigmoid colon during proctoscopy or sigmoidoscopy is preferred to a rectal swab blindly inserted through the anus.
Ideally, specimens are cultured as soon as possible since a significant number of bacteria, particularly Shigella. , will not survive the acidic pH change which occurs with a drop in temperature of the feces. If there is a delay of more than 2 to 3 h, a transport medium is necessary. In most instances, equal parts of 0.03 M phosphate buffer and glycerol (pH 7.0) may be used; however, this medium is not satisfactory for Vibrio parahaemolyticus. or Campylobacter fetus. subsp jejuni. , for which Cary-Blair transport medium is recommended. For convenience, the Cary-Blair formulation can be used for all stool specimens. One gram of feces should be emulsified in 10 ml of transport medium, to which phenol red is often added as a pH indicator. If the transport medium is yellow when it is received by the laboratory, the pH is too low and the specimen is unsatisfactory, particularly for isolation of Shigella. . A 1- or 2-g sample of feces is more than sufficient for culture.
Two to three separate specimens increase the probability of isolating a pathogen. If bacterial etiology is suspected and cultures are negative after several attempts, consultation between the microbiologist and clinician is recommended.
There are several instances when microscopic examination of the stool may be useful. A fleck of mucus or stool examined with Loeffler methylene blue stain for pus cells is helpful in indicating an invasive pathogen. Equal amounts of stool and the stain are mixed on a microscope slide with a wooden applicator stick. A cover slip is placed over the mixture, and, after 2 or 3 min, the slide is examined under a "high dry" objective for the presence of leukocytes. Since there are often false-positives and -negatives, the test's specificity is only good when the smear is strongly positive. In cases of staphyloccal enterocolitis, a Gram-stained smear may contain large numbers of clumps of gram- positive cocci. Otherwise, a Gram- stained smear may contain large numbers of clumps of gram-positive cocci. Otherwise, a Gram-stained smear of stool is not very useful.
It is generally agreed that all specimens should be cultured for Salmonella and Shigella. Many laboratories also perform cultures for C fetus. subsp jejuni. and Yersinia enterocolitica. Routine use of media for the isolation of other bacterial agents of diarrhea varies depending on the laboratory size, geographic location, and patient population. For example laboratories in areas where there is a high incidence of shellfish consumption may wish to screen for V. parahaemolyticus during the summer.
Specimens suscepted of containing either Salmonella or Shigella are inoculated onto a selective and a differential enteric agar medium (MacConkey or eosin methylene blue) and at least two selective agar media for enteric pathogens, preferably one of moderate selectivity and one of high selectivity. Moderately selective agar media include Hektoen enteric, deoxycholate, xylose-lysine-deoxycholate, and salmonella-shigella agars. Highly selective agar media include bismuth sulfite and brilliant green. Stool should also be inoculated into one of the enteric enrichment broths (gram-negative broth, Selenite-F broth, or tetrathionate broth). These broths are used primarily for enrichment of Salmonella , although Shigella. may grow in Gram-negative broth.
Enteric agar media and enrichment broths are incubated aerobically. After incubation, the broth is subcultured to the same types and number of agar media used for primary inoculation. The duration of incubation depends on the broth used. The inoculum size and incubation time suggested for each broth should be strictly followed. All plates should be incubated for a total of 48 h and examined at 24 and 48 h. Media for isolation of Salmonella. and Shigella. are incubated at 35 to 37oC. Other enteric pathogens may require lower or higher incubation temperatures.
Vibrio Cholerae O group 1
Vibrio Cholerae non O group 1
Vibrio cholera O group 1
V. cholerae. O group 1 is the only one of the enteric pathogens that has the potential to produce pandemic disease and is, thus, of unusual public health importance. It may, however, also produce sporadic cases in endemic areas. Although cholera is the most severe of all the diarrheal diseases, only a small percentage (approximately 1 to 10%) of persons infected with the organism develop the severe clinical syndrome. In its most severe form, the disease begins with vomiting and diarrhea, with the loss of large volumes of diarrheal stool. Within 12 h of the onset of illness, the patient can be in shock and near death from loss of fluid. If the disease is untreated, the mortality rate can be as high as 50 to 70%.
Characteristically in this syndrome, the stool is not at all fecal in character but rice-watery with a slightly fishy odour. In less severe forms of the disease, the illness may not be differentiated clinically from other mild to moderate forms of diarrhea.
In endemic areas, the disease occurs with greatest frequency in children, though cases may occur in all age groups. In areas newly infected with cholera, adult males usually have the highest attack rates. The disease is limited to humans; no animal reservoirs are known to exist. Contaminated water is felt to be the most common vehicle for transmission, although heavily contaminated food, particularly seafood, has also been shown to be important.
Culture attempts should be made when the patient has the typical cholera syndrome with rice-water stools or diarrheal disease after returning from a cholera endemic area and when there are public health reasons for suspecting cholera. In 1978 there were 11 cases of cholera in southwestern Louisiana.All of the cases had recently eaten crabs from coastal marsh areas. The organism was isolated from estuarine waters and sewage in these areas. Laboratories in the Gulf Coast area should, therefore, search for V. cholerae O group 1 in specimens from patients with severe watery diarrhea.
Although the stool from patients with cholera will have large number ofVcholerae. which will grow profusely in primary cultures, persons with less severe diarrhea or who are carriers have fewer organisms in feces, making enrichment necessary. Enrichment is best carried out by inoculating alkaline peptone water (1.5% peptone-0.4% NaCl, adjusted to pH 9.0 before autoclaving) heavily with stool, incubating it at 35oC for 6 to 12 h, and subculturing the alkaline peptone water to thiosulfate citrate bile salts (TCBS) agar.
TCBS agar is the optimal medium on which to recognize V. cholera. (and V parahaemolyticus. ). After overnight incubation at 35oC, the TCBS agar plates inoculated directly from stool and by subculture from the alkaline peptone water are examined.V. cholerae. colonies are yellow, flat, and large. There is little growth of indigenous fecal flora on this medium, although enterococci can form tiny, pin-point, yellow colonies, and Klebsiella-Enterobacter. can form opaque, heaped- up, dark-yellow, medium-sized colonies. These organisms can be distinguished easily, however, from V. cholerae. V. cholerae. non-O group 1 colonies may be similar in appearance to those of V. cholerae. O group 1, or may be green if sucrose is not fermented. Colonies of V. parahaemolyticus. also appear as green colonies (see below).
Typical colonies of V. cholerae. should be subcultured from TCBS agar to nutrient agar (or similar noninhibiting media) before performing a slide agglutination test. Group antisera, as well as anti-Inaba and anti-Ogawa sera, are tested. V. cholerae. O group 1 should agglutinate in the group sera, as well as in one of the specific antisera. (Agglutination can be done with colonies taken directly from TCBS agar, but the results may be misleading because of the granular nature of the colony, which will not homogenize well in saline.) Colonies taken from TCBS agar may yield a negative oxidate test.
Confirmatory tests include the "string test", in which the colony is emulsified in 0.5% sodium deoxycholate, and the tip of an inoculating needle is then lifted out of the emulsion; a mucus- like string connecting the tip of the needle and the emulsion will form. This reaction is specific for vibrios. In endemic areas, typical colonies which are oxidase positive, lysine and ornithine decarboxylase positive, and arginine dihydrolase negative, and which are inhibited by the vibrio-static agent 0/129 and agglutinated in 0 group 1 antisera, may be considered V cholerae However, the combination of a typical colony on TCBS agar and an appropriate slide agglutination reaction constitutes presumptive identification. In non-endemic areas, complete biochemical studies are necessary. Kits designed for testing gram-negative fermenters other than Enterobacteriaceae. may also be used.
Further studies, including bio- typing and testing for enterotoxin production, can be done at a reference laboratory, such as the Center for Disease Control.
Nearly all strains are susceptible to tetracycline, the drug of choice for treatment, so that antimicrobial susceptibility testing is not indicated, except in Tanzania and Bangladesh, where tetracycline- resistant strains have been isolated. If antimicrobial susceptibility testing is done, the standardized disk diffusion method may be used.
Acute- and convalescent-phase sera (obtained on day 10 to 14 of illness) may be tested for agglutinating or vibriocidal antibodies. A four-fold rise in titer is diagnostic of infection. Serological diagnostic procedures should be done in a reference laboratory.
V. cholera non-O group 1
Formerly called non-cholera vibrios or non-agglutinating vibrios, these organisms may produce a disease similar to cholera by the production of cholera-like protein entero- toxins. Much less is known of their human disease potential, since few laboratories have attempted to isolate them from stools in the United States. They exist in aquatic environments and have a wide distribution in the animal world. They are isolated on the same media asVcholera. and may form yellow or green colonies. They do not agglutinate in cholera antiserum. Strains of V. cholerae. non-O group 1 have biochemical reactions common to the genusVibrio. and can be classified according to serotype.
E coli. can cause diarrheal disease by both mechanisms discussed. Enterotoxigenic strains produce two types of enterotoxins, a heat-labile cholera-like enterotoxin and a small-molecular-weight, heat-stable enterotoxin. A strain can produce one or both of these enterotoxins. The genetic material which controls toxin production is plasmid-borne and can be transferred from one strain of E coli. to another. These strains may on occasion produce a clinical picture indistinguishable from cholera, although most of the time, the diarrhea is of mild to moderate severity.
Theoretically, an E. coli. strain (i.e. serotype) could carry the enterotoxin plasmid and thus be enterotoxigenic; however, for unknown reasons, only a relatively few serotypes seem to be consistently enterotoxigenic. These serotypes are quite different from the "enterophathogenic" serotypes described some 30 years ago.
E coli. may also be invasive and produce a disease indistinguishable, clinically from shigellosis. These strains also seem to be restricted to a certain few serotypes which are different from those already mentioned; they cannot be recognized with certainty except by demonstrating their invasive properties in the guinea pig eye conjunctiva model, a procedure beyond the scope of most clinical laboratories.
Some serotypes of E coli. were associated many years ago with outbreaks of diarrhea in nurseries and often with high fatality rates. Some of the original isolates also produced clinical illness when fed to volunteers.
Subsequently, on epidemiological grounds alone, approximately 16 serotypes came to be known as enteropathogenic (EPEC), despite the fact that apart from the outbreaks in nurseries, there was no good evidence that these strains caused diarrheal disease, particularly in sporadic cases. When enterotoxigenic E. coli strains were first recognized in the late 1960's, EPEC strains were also extensively studied for their ability to produce heat-labile and heat-stable enterotoxins and were found not to react in any of the animal model systems for detecting toxigenicity (15). Further studies in volunteers, however, have conclusively shown two recently isolated epidemic EPEC strains to be diarrheagenic, and the illness they produced was similar to an enterotoxin-mediated illness (24). This observation supported the earlier autopsy studies of infants dying with EPEC infection in which no histological evidence of invasion of intestinal mucosa was found.
At this time, therefore, there is evidence that epidemic strains of EPEC are probably enterotoxigenic due to as yet undescribed toxins which are qualitatively different from heat-labile and heat-stable enterotoxin. There is, however, little or no evidence to incriminate the EPEC strains as causes of sporadic cases of diarrhea. Since outbreaks of diarrheal disease in nurseries have largely disappeared in the United States, we no longer think of EPEC as an important cause of outbreaks. Whether they are at all important in causing sporadic cases of diarrhea remains to be proved.
Enterotoxigenic E.coli strains produce diarrheal disease predominantly in small children in the developing world and are rarely a cause of sporadic cases of infantile diarrhea in the United States. They are responsible for occasional outbreaks of diarrhea after ingestion of contaminated food and water, and they are the major cause of travellers' diarrhea in persons from the United States visiting the developing world (26). Outbreaks of illness due to invasive E. coli. strains largely occur after consumption of contaminated food products. These strains are probably also rare causes of sporadic illness.
Unfortunately, there are no serological tests that are useful in the retrospective diagnosis of E. coli. -mediated diarrheas. E coli. represents indigenous flora in asymptomatic persons. There are no morphological or biochemical differences between strains representing indigenous flora and those causing diarrhea; therefore, invasiveness or enterotoxigenicity must be demonstrated to identify diarrheagenic strains. Production of the heat-labile enterotoxin can be demonstrated either by an enzyme- linked immunosorbent assay, or in tissue culture (Y-1 adrenal or Chinese hamster ovary cells).
Infant mice are required to demonstrate production of the heat-stable enterotoxin. As already described, confirmation of invasiveness requires inoculation of the conjunctival sac of guinea pigs. Serotyping of E. coli is not widely available. For all of these reasons, it is unlikely that enterotoxigenic or invasive strains for E. coli will be identified in clinical laboratories in the near future.
There are certain instances, however, when disease due to E. coli should be suspected and strains should be isolated and sent to reference laboratories:
i)if an outbreak occurs, particularly in a nursery, and other enteric pathogens are not found, and
ii)when travellers have returned from the developing world with a diarrheal illness.
In either instance, approximately 10 colonies of E. coli should be subcultured for transport to a reference laboratory for tests on invasiveness, enterotoxigenicity, and serotyping. No enrichment procedures are necessary for isolating these strains. Enteric agar media used for routine primary cultures are adequate, and blood agar is not required.
Antimicrobial susceptibility tests are important only during an outbreak and can be performed with the standardized disk diffusion method. In sporadic cases of diarrhea, there seems to be no reason to try to isolate either entertoxigenic E. coli or EPEC, and routine sero- typing with EPEC antisera is discouraged.
S. aureus can produce acute diarrhea through either the production of enterotoxins or the invasion of the large intestinal mucosa. The first mechanism, which is by far the most common, has been well characterized.
Staphylococci growing in stored food products, usually meat or dairy products, produce heat-stable enterotoxins which when ingested mainly produce vomiting, although diarrhea may also occur. The onset of illness is usually within 2 to 6 h after eating the contaminated food, and the disease usually terminates within 12 to 18 h. This form of food poisoning is one of the most common, along with that produced byClostridium perfringens , and occurs in outbreaks when a number of persons ingest the same prepared foods. Individual cases are nearly impossible to recognize.
Staphylococci may possibly cause antibiotic-associated pseudomembranous enterocolitis. Although this entity was reported relatively frequently in the early 1950's, the role of C. difficile had not yet been recognized, and cultures forC. difficle. were not performed. Nonetheless, the possibility remains the antibiotic- resistant staphylococci overgrow the indigenous fecal flora which in inhibited during antimicrobial therapy, invade the mucosa, and produce marked inflammation and diarrhea. Further studies are required to elucidate the pathogenic role of S. aureus in antibiotic-associated colitis.
Attempts to isolate staphyloccci as the etiological agent in food poisoning should be made only when there is a recognized outbreak. In that case, the most important materials to culture are the suspected food article and any open lesions on the skin of any food handlers involved. Establishing the identity of isolates from food and food handlers requires phage typing.
Culture of the stool is not a sensitive method of making the diagnosis. The suspected food article should also be examined in a reference laboratory for the presence of preformed staphylococcal toxin. If staphylococcal food poisoning is suspected, the local health department would be notified so that the collection and culture of food samples can be performed by those specifically trained in these procedures.
V. parahaemolyticus. is a marine bacterium producing explosive, watery diarrhea, abdominal cramps, and sometimes nausea about 15 to 24 h after the ingestion of contaminated seafood. Vomiting occurs in about half the cases.
Symptoms may last for from several hours to over 10 days but usually subside without treatment in 3 days. Most isolates (96%) from patients are Kanagawa positive, i.e. they produce a hemolysin detectable on blood agar with a high salt content (Wagatsuma agar). Less than 1% of seafood isolates are Kanagawa positive. Although correlated with enteropathogenicity, the role of thehaemolysin and the pathogenesis of the infection have not been completely worked out. V. parahaemolyticus is a halophilic bacterium found in estuarine or coastal water. Its distribution is restricted by both water temperature and salinity in that growth is retarded at. 15`C and numbers of bacteria are greatest at a salinity of 15 to 20%.V. parahaemolyticus has been isolated from coastal waters throughout the world including the Pacific, Gulf, and Atlantic coastlines of the United States. Colwell has shown that during the winter the bacterium is found in sediment and in or on marine organisms found on the ocean bottom (9). When the water temperatures increases to 14+. 1`C,V. parahaemolyticus (and other Vibrio. species) are released from the sediment and attach to zooplankton. The bacteria utilize a slimy exudate coating the plankton, decompose the plankton, and are then released into the water. V. parahaemolyticus is found associated with many shellfish, including shrimp, clams, oysters, mussels, and crabs. Patients with V. parahaemolyticus gastroenteritis have a recent history of eating seafood. Although sporadic cases have been reported in the United States, the vast majority of cases are associated with common- source outbreaks. Since 1969, there have been 16 reported outbreaks, 2 occurring on cruise ships. In Japan, V. parahaemolyticus is the most common cause of food poisoning, with over 10,000 cases annually. Since V. parahaemolyticus infection is virtually always associated with seafood consumption, routine culture for the organism is not recommended. If the specimen (stool and vomitus) cannot be cultured within 8 h of collection, Cary-Blair transport medium should be used. Fresh specimens may be placed in alkaline peptone water (1% peptone-3% NaCl, pH 7.4) for 8 h. Specimens received in transport medium are plated directly. Although V. parahaemolyticus will grow on some enteric media and many strains will grow on mannitol salt agar, inoculation of TCBS agar is recommended since it is selective for Vibrio. spp. It is important to be aware that there is variation in brands and lots of TCBS agar in its ability both to inhibit fecal flora and to support the growth of vibrios (27). Frequently, prepared TCBS agar can be obtained from health department laboratories. After 24 h of incubation at 35`C,V. parahaemolyticus forms colonies 3 to 5 mm in diameter with green-blue centers. Preliminary screening tests of these colonies include Gram- stained smears, growth on triple sugar iron agar, motility, and oxidase reaction. The oxidase reaction of colonies from TCBS agar may be negative; therefore, the reaction should be confirmed by using growth from a nonselective medium. Oxidase-positive, curved, gram-negative, motile rods producing an alkaline slant and acid butt without gas or H2S production in triple sugar iron agar should be tested further. A complete biochemical description is found in theManual of Clinical Microbiology. Some media used for determination of biochemical characteristics of the species must be supplemented with NaCl. Minimal criteria for identification include the production of lysine decarboxylase, growth in 8% but not 0 or 10% NaCl, no agglutination in V. cholerae. O group 1 antiserum, no production of acetyl methyl carbinol, and no acid production from sucrose.
The majority of strains of V. parahaemolyticus produceB. - lactamase but are generally susceptible to tetracycline, chloramphenicol, streptomycin, kanamycin, polymyxin B, novobiocin, and oleandomycin (1). Tetracycline may be used for treatment of severe cases. Although susceptibility testing is not usually necessary, the standardized disk diffusion method (23) may be used if 2 to 3% NaCl is added to Mueller-Hinton agar. This modification does not appear to alter tetrcycline and chloramphenicol zone diameters of inhibition, but is does diminish gentamicin activity (21).
Two clinical syndromes are associated with B. cereus food poisoning: diarrhea or vomiting. A different enterotoxin is involved in each form. The diarrheal (more common) form has a longer incubation period (6 to 24 h) and is characterized by predominantly lower gastrointestinal tract symptoms, much as in C. perfringens food poisoning. The emetic form has a short incubation period (1 to 6 h) and is characterized by predominantly upper gastrointestinal tract symptoms, as in staphylococcal food poisoning. The initial clinical features, i.e. watery diarrhea without blood, abdominal cramps, and fever, suggest infection with an enterotoxigenic organism. Diagnosis can be confirmed by isolation of. 105 organisms per g from food.
Since B.cereus can be found in stools of healthy individuals, isolation of the organism from feces alone is not sufficient documentation of the cause of an outbreak, unless negative stools are obtained from a suitable control group (14,42). B. cereus is ubiquitous, aerobic, sporeforming, gram-positive bacillus. It has been recognized as an important cause of food- borne outbreaks of gastroenteritis in Europe but is uncommon in the United States.
The organism is found in uncooked rice and has been frequently isolated in short-incubation outbreaks after ingestion of fried rice. Heat-resistant spores survive boiling and germinate when boiled rice is left unrefrigerated; the vegetative forms multiply and produce toxin. Flash frying or brief rewarming before serving is often not sufficient to destroy the preformed, heat-stable toxin (42). Outbreaks of the diarrheal from are generally associated with consumption of contaminated meats and vegetables.
Laboratory diagnosis of possible B. cereus diarrhea is rarely attempted except when an outbreak occurs. Investigation and culture of suspect food and feces of affected individuals will often be performed by public health officials. Local health departments should be notified of a suspected outbreak. Since symptoms of B. cereus food poisoning are usually mild, self-limited, and mediated by enterotoxin, administration of antibiotics is not indicated.
C. perfringens type A accounts for approximately 15% of reported outbreaks of food-borne disease in the United States. Rarely, type F and some type C strains of C. perfringens cause a severe and sometimes fatal necrotizing enteritis.
The small bowel is the principal area involved in disease caused by type A strains. The incubation period ranges from 12 to 24 h. Diarrhea (probably the result of the production of enterotoxin by the bacteria) is the principal symptom. Fever and vomiting do not occur, and the illness is seldom serious enough to cause sequelae. The diagnosis is established by anaerobically culturing the incriminated food and growing large numbers of C. perfringens The same organism can usually be isolated from the feces of the patient.
Isolates from food and stool must be examined serologically to distinguish between the infecting organism and strains of C. perfringens normally present in stool. C. perfringens is an anaerobic, gram-positive rod present in human and animal feces and soil. Raw meat becomes contaminated with a small number of organisms from the soil or gastrointestinal tract at the time of slaughtering or processing. Inadequate cooking fails to kill all the organisms, and they multiply when the meat is allowed to sit at or slightly above room temperature; the food is then served without sufficient reheating. Large pieces of rolled or stuffed meat are commonly implicated. Quantitative C. perfringens colony counts should be performed on food samples. This is usually performed by health department laboratories. Antimicrobial therapy is usually not needed, because the illness is self-limited.
Antibiotic-associated P. enterocolitis
Pseudomembranous colitis, characterized by inflammatory plaques, has beenrecognized as a clinical entity in humans for years and has been linked to antimicrobial therapy since the 1950s. Only recently have experiments shown that this is an "ecological" disease. The majority of cases are due to the action of a cytotoxin produced by C. difficle . The disease is characterized by severe diarrhea, dehydration, fever, leucocytosis, abdominal pain, and gastro-intestinal bleeding. Fatality rates have ranged from 0 to 58%. Sigmoidoscopic changes may not always be present. When present, a characteristic membrane or patchy white material adherent to the mucosa can be seen.
In studies with conventional anaerobic techniques, C. difficle has been found consistently in the fecal flora of children but in that of only about 2% of healthy adults. It is possible that the organism is more commonly present but in numbers too small to be detected by conventional methods. Presumably when the indigenous flora is disturbed by antibiotic therapy, overgrowth of C. difficle may occur. Nearly all antibiotics have been implicated. Clinical laboratory procedures for the diagnosis of pseudomembranous colitis include examination of feces for the presence of C. difficle and for toxin. To date, all stools containing the toxin have been culture positive, but a small percentage of culture- positive stools do not contain toxin. Thus, isolation of the organism supports the diagnosis of C. difficle colitis, but demonstration of toxin in the stool is necessary to confirm the diagnosis. i)Culture. A variety of media have been formulated and evaluated for the isolation of C. difficle from feces, but two recent reports indicate that media containing cycloserine and cefoxitin are superior. These antibiotics are added to either an egg yolk- fructose agar base (500 ug of cycloserine per ml and 16 ug of cefoxitin per ml) or a brain heart infusion blood agar base (250 ug of cycloserine per ml and 10 ug of cefoxitin per ml). The latter is available commercially (Scott Laboratories, Fiskeville, R.I.). Enrichment procedures include incubation for 5 days in 0.2% paracresol enrichment broth before and after ethanol shock, followed by subculture. However, if the cycloserine-cefoxitin agar is used, the enrichment procedure does not seem necessary. Fresh or refrigerated feces are weighed, and 0.1 g is immediately added to 1.0 ml of phosphate-buffered saline (PBS). Serial 10-fold dilutions (10-1 to 10-7) are made, and 0.1 ml of each is plated on prereduced agar media. The plates may be incubated for 48 to 72 h in either an anaerobic chamber or jar. Ideally, the original dilution of feces should be made in an anaerobic chamber, but if a chamber is not available, the organism will tolerate some manipulation in an atmosphere of room air. Colonies on the egg yolk-fructose base are 4 to 8 mm in diameter, yellow, of ground-glass appearance, circular with a slightly filamentous edge, flat to low umbonate in profile, and lipase and lecithinase negative. Colonies on the blood agar base are 5 to 7 mm in diameter, flat, and filamentous, and have a greenish tinge. Characteristic colonies are identified by methods described in the Anaerobe Laboratory Manual. ii)Toxin assay. For toxin assay (8,33), the sample of feces is diluted 1:1 (wt/vol) in antibiotic-containing PBS (pH 7.2) and centrifuged for 30 min at 3,000 x g or for 10 min at 10,000 x g and then tested or stored at -70`C. The supernatant is filtered through a 0.22-um-pore size membrane filter and assayed in tissue culture for toxin. A number of tissue culture cell lines are acceptable, including WI-38 human diploid fibroblasts or HeLa cells. The filtrate is serially diluted (10-1 through 10-6) in PBS. If the titer of toxin is desired, 0.1 ml of undiluted filtrate and of each dilution is added to cell cultures containing 0.9 ml of maintenance medium, giving final dilutions of 1:20 to 1:20,000. Screening may be done by testing the undiluted filtrate and filtrate diluted 1:100. If the cells are grown in microtiter plates, then 0.01 ml of filtrate is added to each well. Inoculated cells are incubated overnight at 35`C. The toxin titer is the highest dilution showin a definite cytopathic effect, with rounded cells involving at leat 10% of the monolayer. To show the specificity of the reaction, one tube of tissue culture is inoculated with 0.1 ml of Clostridium sordellii antitoxin (Bureau of Biologics, Rockville, Md., or Virginia Polytechnic Institute, Blacksburg). Cytotoxicity due to C. difficle toxin is neutralized in the antitoxin tube. C. sordellii toxin is not responsible for pseudomembranous colitis, but the antitoxin does neutralize C. difficle toxin due to antigenic cross-reactivity. C. difficle antitoxin is not yet commercially available.
Testing C. difficle is not routinely recommended; however, since diarrhea that develops in hospitalized patients is more likely to be due to C. difficle and its toxin than to other enteric pathogens, laboratories may wish to screen these stools for C. difficle and its toxin, provided a history of antimicrobial therapy has been confirmed. Proctoscopic examination of stool for ova and parasites are also recommended.
Because vancomycin appears to be the drug of choice for treating pseudomembranous colitis, antimicrobial susceptibility testing is usually not indicated.
There are probably others yet to be described enterotoxigenic bacteria that are responsible for acute diarrheal disease. Until these become recognized through extensive research effort, however, it seems unwise for the clinical laboratory to be concerned with their detection in stools. Other organisms, such as Klebsiella, Enterobacter, Pseudomonas, and Pleisiomonas, have been associated with diarrhea; however, these associations frequently antedated the recognition of, and, therefore, specific attempts at detecting the presence of, other recently recognized enteric pathogens. Some laboratories do report a pure culture of Pseudomonas or Pleisiomonas, particularly if the patient is immunocompromised.
Aeromonas hydrophila has recently been isolated from a number of children with diarrhea and has been found to produce a cytotoxic enterotoxin. It is not yet possible to assign its importance in diarrheal disease, but Aeromonas can be cultured relatively easily. This organism grows on enteric agar media and forms low convex colonies, 1 to 2 mm in diameter, which are usually lactose negative. Lactose-positive colonies may resemble E. coli but can be distinguished from E. coli with an oxidase test (Aeromonas is positive). On TCBS agar Aeromonas forms small yellow colonies. It is differentiated from vibrios by decarboxylase reactions, salt tolerance, and agglutination in V. cholerae O group 1 antiserum.
Shigella
Salmonella
Enterocolitica
Camplytobacter fetus subsp. jejuni
Although toxin production has been demonstrated in Shigella dysenteriae 1, Shigella flexneri, and Shigella sonnei, the ability of the organism to penetrate and multiply in mucosa of the colon is its most important virulence factor. The diarrhea, which occurs after an incubation period of 1 to 3 days, is characterized by yellow-green stool containing mucus, inflammatory cells, and frequently blood. It is accompanied by fever, cramps, and abdominal pain. Vomiting is unusual. It is usually a self-limited infection, although in children dehydration can be a major complication. Shigellosis is, like cholera, associated with poor sanitation, and both food- and waterborne outbreaks occur. Person-to-person transmission is also common and is facilitated by the low infecting dose (10 to 200 organisms). In the United States, infection is most common in institutionalized populations, in the military, on Indian reservations, and in slum populations. About 85% of all Shigella isolates in the United States are S. sonnei. Other species may be expected in travellers returning from other countries. Inapparent infections and carriers also play a role in the spread of infection. Chronic carriage is rare.
Although Shigella will grow in gram-negative broth, the enrichment broths are designed for Salmonella. Shigella are more fragile than other Enterobacteriaceae, and attempts should be made to culture freshly passed feces. Several specimens may be necessary. Many microbiologists believe that xylose-lysine- doexycholate agar gives the highest yield, but as described earlier, several enteric agars should be utilized.
The use of antibiotics for the treatment of shigellosis is somewhat controversial, but they are definitely indicated when dysentery is present.
Antimicrobial susceptibility testing should be performed with all isolates, since ampicillin-resistant strains are common in some areas and multiply drug-resistant strains do occur. Alternative therapy for patients with ampicillin-resistant strains is trimethorpim-sulfamethoxazole or tetracycline.
Salmonellae penetrate the epithelium of the terminal ileum and migrate to the lamina propria. Thus, a methylene blue preparation of stool usually reveals only a few to a moderate number of inflammatory cells. Salmonellae also produce toxins which may be involved in the pathogenesis of the diarrhea, which infrequently may be watery. In humans, salmonellae may cause three pathological conditions:
i)a self-limiting gastroenteritis,
ii)enteric fever, and
iii)septicemia and metastatic infections.
The gastroenteritis is characterized by low-grade fever, diarrhea, and occasionally cramping abdominal pain beginning 1 to 3 days after ingestion of contaminated food. It is usually short-lived (1 to 4 days), but may be life-threatening in infants and geriatric patients due to fluid loss.
Salmonellae are very common in nature and have been isolated from most animal species, including poultry, rodents, livestock, domestic animals, arthropods, birds, and reptiles. Both sporadic cases and outbreaks are associated with contaminated foods (particularly powdered milk, meat products, cream cakes, and fresh and powdered eggs) and contact with infected animals (pet turtles, dogs, ducks, chicks, and livestock). Chronic carriers also act as an important reservoir.
Enrichment procedures for Salmonella should always be done. Enteric agar media, as previously described, are used for isolation. Bismuth sulfite agar is highly selective for Salmonella and useful for detecting lactose-fermenting strains. Procedures for isolation, screening, biochemical identification, and serological confirmation are described in the Manual of Clinical Microbiology. Even a small laboratory should be capable of determining the serogroup of the more common salmonellae (groups A,B,C,C1, D, and E). Complete serotype determination is important for public health reasons and is available at local or state health department laboratories.
The administration of antibiotics does not shorten the course of the diarrhea and may prolong the carrier state. Therefore, treatment is usually confined to infants, the very old, and those with systemic disease. Ampicillin is the drug of choice (chloramphenicol is the alternative), but susceptibility testing is necessary since resistant strains do occur.
The most common form of clinical infection with Y. enterocolitica in humans is acute gastroenteritis with abdominal pain and bloody or nonbloody diarrhea; fever may or may not be present. Other forms of illness include:
i)a syndrome of pseudoappendicitis, mesenteric lymphadenitis, or terminal ileitis,
ii)septicemia,
iii)meningitis, and
iv)urinary tract infection.
Symptoms include arthritis, erythema nodosum, and Reiter's syndrome. Asymptomatic cases do occur. The pathogenetic mechanisms of Y. enterocolitica are not fully understood.
There are three potential virulence factors, including the organism's ability to invade HeLa cells, to cause kerato- conjunctivitis in the guinea pig, and to produce a heat- stable enterotoxin (28). The portal of entry is probably the gastrointestinal tract. The organism has been isolated from a wide variety of animals (swine, dogs, rabbits, chinchillas, cows, sheep, horses, deer, cats, beavers, raccoons, various birds, and oysters) and has, therefore, the potential to contaminate food and water supplies. It has also been recovered from numerous rivers, reservoirs, lakes, and nonchlorinated well water. Y. enterocolitica probably can also be spread via person-to-person transmission. In Europe and Canada, patients with yersiniosis often give a history of pork ingestion or of direct contact with pigs; thus, the pig is thought to be an important reservoir of infection in these areas. In the United States, out- breaks have involved contaminated foodstuffs (chocolate milk) and interfamilial spread of the organisms. Each of these outbreaks resulted in several unnecessary appendectomies. The few studies regarding the incidence of Y. enterocolitica in the United States indicate that this organism may be a more common cause of bacterial diarrhea than previously suspected.
Specimens submitted for cultures for Y. enterocolitica may be refrigerated until cultured since the organism proliferates at 4 to 5`C and presumably outgrows other enteric flora. Microscopic examination of stool may reveal the presence of fecal leukocytes. Initial recovery requires inoculation of duplicate sets of MacConkey and salmonella-shigella agar, followed by incubation of one set each at 35 and 25`C for a minimum of 48 h. Agar surfaces should then be scanned with the aid of a stereoscopic microscope (magnification, x 13) using oblique illumination. Pinpoint colonies may then be selected for isolation as potential Y. enterocolitica. Y. enterocolitica can usually be isolated by direct culture of feces onto enteric media in acute cases of the disease. Recovery of the organism, particularly in adults, asymptomatic carriers, or convalescent patients, is enhanced by using cold enrichment techniques. However, it is not clear whether all the isolates obtained only by cold enrichment are clinically significant. Some are unequivocally pathogens, whereas others are of questionable significance. The use of cold enrichment needs further evaluation in more geographic areas before its routine use is accepted or rejected. For cold enrichment, a rectal swab or swab dipped in feces is placed into a test tube containing 5 ml of 0.067 M PBS (pH 7.6), which is held at 4 to 5`C for 3 weeks. Samples are cultured on MacConkey and salmonella-shigella agar after 7, 14, 21 days of enrichment. The sub- cultures are incubated at 25`C for 48 h because Y. enterocolitica will grow more readily at this temperature.
Recently, three selective agars, pectin agar, cellobiose- arginine-lysine agar, and Y medium, have been developed for isolation of Y. enterocolitica.
These media, although promising, have not yet been adequately evaluated in field trials. Several laboratories have successfully used cellobiose-arginine-lysine agar; it is commercially available as CAL medium (Remel-Regional Media Laboratories, Inc., Lenexa, Kans., and Scott, Laboratories, Fiskeville, R.I.). Y. enterocolitica is a relatively large (0.5 to 1.0 by 1 to 2 um), coccobacillary, ovoid- or rod-shaped, gram-negative rod. Colonies are 1 to 2 mm in diameter and are light pink to peach in colour on MacConkey agar after 24 h. Colonies on salmonella- shigella agar are similar in colour to those grown on MacConkey but are slightly smaller; after 48 h, colonies are smooth and colourless, resembling Shigella. Not all Y. enterocolitica will grow on salmonella-shigella agar. Triple sugar iron agar, urea agar, and two tubes (for incubation at 25 and 35`C, respectively) of motility medium are inoculated for presumptive identification of typical colonies, which should produce an acid butt and acid slant in triple sugar iron agar (due to fermentation of sucrose) with no gas or H2S production. Two biogroups are sucrose negative and will give an acid butt and alkaline slant reaction in triple sugar iron agar. Y. enterocolitica is urease positive and motile at 25`C but not at 35`C. Confirmatory biochemical reactions should be performed either by conventional methods (18) or by using commercially available kits.
Typical strains of Y. enterocolitica must be distinguished from two groups of rhamnose-positive strains which have been shown to be separate species on the basis of deoxyribonucleic acid hybridization studies; these strains, as well as sucrose-negative Y. enterocolitica, do not appear to be enteric pathogens.
The first species, tentatively designated Yersinia intermedia, is rhamnose, raffinose, melibiose, and sucrose positive; the second species, tentatively called Yersinia frederiksenii, is rhamnose and sucrose positive but raffinose and melibiose negative. Y. enterocolitica is sucrose positive and rhamnose, raffinose, and melibiose negative; the sucrose-negative biogroup (tentatively called Yersinia kristensenii) of Y. enterocolitica rhamnose, raffinose, melibiose, and sucrose negative.
There are a number of different schema fro biotyping Y. enterocolitica; these include those of Nile'hn, Wauters, and Knapp and Tal (5). Biotyping of isolates may be important. A recent study from Belgium indicates that most biotype 1 strains are not pathogenic, whereas reports from Canada, South Africa, and the United States (25, 34, 46) implicate biotype 1 strains in diarrheal disease.
There are presently 53 O (somatic) antigens and 29 H (flagellar) antigens of Y. enterocolitica. Groups 0:3 and 0:9 are the most common in Europe and Canada; 0:3 is the most common in Japan. A number of different 0 groups have been implicated in outbreaks in the United States, with 0:5 and 0:8 the most common. Serogrouping is beyond the scope of most clinical and public health laboratories. Similarly, detection of antibody response in the patient can be performed only where a complete collection of known serotype strains is available.
There also seems to be geographic variation in the incidence of Y. enterocolitica, although with the use of better selective media, this variation may disappear. We suggest that large laboratories screen for Y. enterocolitica in stool specimens for 6 to 12 months, including the summer and fall, to determine the necessity of routine screening in their area. Y. enterocolitica is usually susceptible in vitro to kanamycin, gentamicin, colistin, chloramphenicol, tetracycline, streptomycin, and trimetho-prim-sulfamethoxazole; variable susceptible to neomycin and ampicillin; and resistant to cephalothin. Infection is usually self- limited in uncomplicated gastroenteritis, and antimicrobial therapy is generally not indicated. Antimicrobial therapy may be beneficial in cases of chronic or fulminating illness; selection of agents to be used in therapy should be based on susceptibility testing of the isolate.
Camplytobacter fetus subsp. jejuni
Campylobacteriosis has been recognized as a disease of cattle, sheep, and birds for many years and as a human pathogen since 1947, but only in recent years has the association between C. fetus subsp. jejuni and infectious diarrhea been appreciated. The availability of special culture techniques and selective media has facilitated the isolation of this microorganism from human feces. The disease is characterized by rapid onset, abdominal pain, headache, fever, and watery diarrhea. Stools may contain blood, Enteric symptoms last from 1 day to 3 weeks, but in most patients, the total duration of illness is 1 week or less. Campylobacter infection is seen in all age groups. C. fetus subsp. jejuni may cause inflammatory bowel disease resembling ulcerative or pseudomembranous colitis.
Campylobacter enteritis is apparently spread by the fecal-oral route. A large waterborne epidemic has been reported (43). Poultry, raw milk, food, and infected children may also be a reservoir for the disease (4). Experience to date suggests that C. fetus subsp. jejuni may be isolated from the feces of patients with diarrhea at least as often as Salmonella and Shigella. Consequently, the clinical microbiology laboratory should culture stool specimens for C. fetus subsp. jejuni. Blaser et al. have suggested that enrichment in thioglycolate broth with 0.16% agar and vancomycin (10 mg/liter), trimethoprim (5 mg/liter), polymyxin B (2,500 IU/liter), amphotericin B (2 mg/liter), and cephalothin (15 mg/liter) increases the yield of C. fetus subsp. jejuni; however, a direct comparison between this procedure and direct inoculation of fecal material on basal medium containing sodium metabisulfite, sodium pyruvate, and ferrous sulfate (often referred to as FBP), as well as antimicrobial agents, has not been made. The FBP reagents reduce the levels of superoxide radicals and H2O2 in the medium and increase the aerotolerance of the organism. In laboratories able to prepare their own media, Brucella base containing 5% sheep blood and the following ingredients (per liter; modification of Skirrow's medium [37]) is suitable for primary inoculation; FeSO4 7H2O (0.25 g), sodium metabisulfite (0.25 g), sodium pyruvate (0.25 g), haemin (2 ml of a 5-mg/ml solution), vancomycin (10 mg), trimethoprim (5 mg), polymyxin B (1 mg), and cephalothin (5 mg). Cultures on this medium may remain exposed to room air for as long as 8 h without significant loss of organisms. Several agar media are commercially available. CS agar (Scott Laboratories, Inc., Fiskeville, R.I.; BBL Microbiology Systems, Cockeysville, Md.) is similar to the medium developed by Skirrow (37). The basal medium contains 10% horse blood, vancomycin (10 mg/liter), polymyxin B (2,500 IU/liter), and trimethoprim (5 mg/liter). Amphotericin B (2 mg/liter) may also be added to suppress yeast growth.
Campy BAP (Scott Laboratories, BBL Microbiology Systems, Remel- Regional Media Laboratories, Inc., GIBCO Diagnostics), formulated by Blaser et al. (4), contains 10% sheep blood in a Brucella agar base with amphotericin B and cephalothin (15 mg/liter) in addition to the above three antimicrobial agents. Published comparisons of the performance of these media are lacking. C. fetus subsp. jejuni is a strict microaerophile and grows best in an atmosphere containing no greater than 6% O2. It also requires 5 to 10% CO2 (38). Although this environment can be achieved with a GasPak hydrogen and CO2-generating envelope with a self-contained catalyst (CampyPak II, BBL Microbiology Systems) is now available for use in a GasPak jar specifically for the isolation of Campylobacter. If facilities for evacuating anaerobic jars are available, jars may be evacuated to 15 to 20 in. of Hg (ca. 50.8 to 67.7 kPa) twice and refilled each time with 10% CO2, a mixture of 10% H2 and 85% N2.
An alternative approach employs the Fortner principle (22), which is a technique used to reduce oxygen tension by coincubation of a Campylobacter-containing specimen with one of a group of facultatively anaerobic bacteria, such as Proteus, S. aureus, Pseudomonas aeruginosa, or Klebsiella pneumoniae. A simple method is to streak half the plate with the specimen and half with the facultatively anaerobic bacterium and seal the plate with autoclave tape. Growth is apparent on plates in 24 to 48 h.
Incubation of media at 42`C is recommended for optimal recovery of C. fetus subsp. jejuni. Subcultures should be made immediately since the organism may lose viability after 72 h on plating media.
Colonies of C. fetus subsp. jejuni may vary in size and may be pinpoint to spreading. Colonies are usually gray and nonhemolytic, but may be tan or slightly pink. The organism is oxidase positive and catalase positive and has a characteristic corkscrew-like darting motility when examined by phase-contrast or dark-field microscopy. Microscopically, it is a curved or spiral-shaped organism with a single polar flagellum at one or both ends. C. fetus subsp. jejuni is inert in carbohydrates and reduces nitrate to nitrite.
The ability to grow in Brucella broth (or fluid thioglycolate) at 42`C and not at 25`C distinguishes C. fetus subsp. jejuni from C. fetus subsp. intestinalis. C. fetus subsp. jejuni is susceptible to nalidixic acid (30 ug per disk), whereas C. fetus subsp. intestinalis is resistant.
Antimicrobial susceptibility testing of all isolates may not be necessary because the disease is self-limited and usually does not require treatment.
In persistent, recurrent, or severe cases, treatment with erythromycin is indicated. Susceptibility of the organism may be tested by either agar dilution or microdilution methods (23). Most isolates are resistant to the beta-lactam antibiotics but are susceptible to the amino-glycosides. Up to 8% of isolates may be resistant to erythromycin and tetracycline at safely achievable serum levels; however, concentration of these antimicrobial agents in the colon are substantially higher.
It is important that reports be an accurate reflection of the work that was performed. If the specimen was examined only for Salmonella or Shigella, for example, then the report should read "No Salmonella or Shigella isolaged". A report stating "No enteric pathogens isolated" is vague and may mislead the clinician. It is necessary to be specific in listing those organisms for which culture attempts were made and to inform physicians that if they suspect infection with other enteric pathogens, they should consult the microbiologist regarding further work-up.