Treatment of infectious diarrhea: the
role of antimicrobial therapy
Michael Bennish, MD
Departments of Pediatrics & Medicine, New England Medical Center and Tufts University School of Medicine, Boston,
Infectious diarrhea remains one of the most common maladies of both adults and children in industrialized countries,
and is among the leading causes of death in developing countries, especially among children. Although the industrialized
world experienced a dramatic decline in mortality and morbidity from diarrhea following the introduction of effective
sanitary measures in the first half of this century, the incidence of diarrhea remains high in the United States.
Two significant reasons for this continued high incidence are the vast expansion of day-care facilities for pre-school
children, and the "industrialization" and the "internationalization" of the food supply in
this country. High rates of diarrheal disease in day-care centers reflect the inherent difficulties in controlling
fecal contamination in crowded settings, particularly with children still in diapers. It reflects as well the under-staffing
and under-funding of many day-care centers, with the resultant inability to effectively implement measures limiting
cross-infections. Children who acquire enteric infections in a day-care center often serve as the source of infections
for other members of their families, and thus help sustain mini-outbreaks of diarrheal disease in the larger community.
The "industrialization" of the food supply has had important consequences for the spread of enteric infections
in the United States. Examples include the risk of salmonellosis from eating undercooked commercially produced
eggs, the risk of Campylobacter
infection from processed chickens, and the risk of infection with enterohemorrhagic Escherichia
coli (which causes bloody diarrhea and can lead to the development
of the hemolytic-uremic syndrome) from undercooked hamburgers and other foodstuffs. Importation of foods such as
lettuce and cantaloupes from developing countries has led to a number of food-borne outbreaks of enteric infections.
Together with changes in habits, the recent emergence of previously unrecognized entericpathogens has contributed
to the continued high prevalence of diarrhea in the United States. Enterohemorrhagic E.
coli is one such newly recognized pathogen, and Cryptosporidium parvum,
which until 20 years ago had only rarely been known to cause diarrhea, has now become an increasingly important
pathogen. It received its most recent notoriety as the cause of a massive outbreak of diarrhea in Milwaukee, in
which an estimated 400,000 persons were infected via the municipal water supply.
In developing countries, where the majority of the world's population lives, diarrhea incidence rates remain high
due to inadequate sanitation and inadequate knowledge of personal hygiene. Although other infectious diseases in
developing countries have been effectively controlled with vaccines (polio and tetanus being two examples), effective
vaccines for the prevention of infectious diarrhea remain elusive.
In spite of the infectious nature of most episodes of diarrhea, antimicrobial therapy has a much more limited role
to play in its treatment than in most other infectious syndromes. Five issues can be identified in discussing diarrheal
diseases: the division of diarrhea into clinical syndromes that allow for a logical approach to therapy; the indications
for antimicrobial therapy of infectious diarrhea; the effect of increasing antimicrobial resistance on the choice
of treatment; non-antimicrobial therapy for infectious diarrhea; and simple measures that can be taken to prevent
Diarrhea can be divided into two clinical syndromes - watery (i.e.
diarrhea that is watery in character) or dysenteric (characterized
by bloody-mucoid stools of relatively small volume, the passage
of which is often accompanied by tenesmus). Although sometimes
indistinct, the watery vs. dysenteric difference is a useful initial
step in the assessment and management of the patient with diarrhea,
and important differences between the two can be noted (Table
1). Although commonly used, classification according to infecting
pathogen is less useful in the clinical approach to the patient
with diarrhea, because there are no pathognomonic findings associated
with any of the enteric infections.
Watery diarrhea is usually the result of infection in the small bowel, while dysentery affects the colon. Fluid
fluxes are much greater in the small bowel than in the colon, and thus infection of the colon produces a lower
volume of diarrhea than that incurred with infection of the small intestine. An important difference between dysentery
and watery diarrhea is in the pathogenesis of the infectious process, a difference that has implications for antimicrobial
therapy. In small bowel diarrheas (e.g. cholera), infection remains confined to the intestinal lumen (or at most
to the superficial epithelial layer) and disease is a result of alterations in cellular function caused by production
of an exotoxin that attaches to receptors on the surface of the cell. Cell death does not occur, and for the most
part the cells remain morphologically normal. Treatment with an antimicrobial agent does not reverse the effects
of toxin that is already bound, but will only eradicate organisms present in the gut lumen. The major complication
from watery diarrhea is the dehydration that results from the loss of water in the stool, a derangement that can
be corrected with the replacement of these fluids by mouth with oral rehydration salts, or by the use of intravenous
fluids in severe cases.
In contrast, dysentery is the result of invasion and destruction of epithelial cells by the infecting enteropathogen.
This process can extend to the lamina propria and results in colonic ulcerations. In the case of infection with
or Yersinia enterocolitica,
invasion of gut lymphoid tissues occurs. Infections with these "invasive" diarrheal pathogens can lead
to a number of different systemic complications, including bacteremia and infection at sites distant from the gut.
Treatment of dysentery or other invasive diarrheas with an antimicrobial agent can eradicate the infecting organism,
halt the process of inflammation and intestinal cell death, and prevent systemic complications from occurring.
Table 2 lists diarrheal
pathogens by the type of diarrhea they usually cause. There can
be considerable overlap in the symptoms that are caused by certain
infecting pathogens (Shigella
for instance, can cause watery diarrhea initially) and the pathogenesis
and clinical manifestations caused by these infections do not
always fit neatly within the two categories outlined in Table
1. Nonetheless, for purposes of broad categorization, such a classification
therapy of diarrhea
The majority of enteric infections causing diarrhea do not require
treatment with an antimicrobial agent. For some infections, such
as rotavirus and Cryptosporidium
effective antimicrobial therapy is not available. In infections
the utility of antimicrobial therapy is unproven. In other infections
(e.g. non-typhoid Salmonella
infections), treatment with an antimicrobial actually lengthens
the duration of excretion of the organism in the stool without
affecting the duration of clinical illness.
It is impossible to distinguish on clinical grounds the subset
of patients which requires antimicrobial therapy. Therefore routine
empiric antimicrobial therapy of diarrheal illness is contraindicated
for both public health reasons (unnecessary costs incurred and
the risk of selecting out antimicrobial resistant strains of enteric
pathogens) and for clinical reasons (the patient is exposed to
the potential adverse effects of antimicrobial therapy without
evidence of benefit). Routine empiric antimicrobial therapy is
indicated during outbreaks of known pathogens which require treatment
(e.g. dysentery caused by S.
type 1 in refugee camps and in locales where certain infections
are hyperendemic, such as dysentery and shigellosis in south Asia).
But these situations are only rarely encountered in clinical practice
in Western countries. Thus, empiric antimicrobial therapy is not
In the normal host, most episodes of diarrheal illness are self-limiting,
and therefore supportive therapy (as discussed below) is indicated.
This is especially true for watery diarrhea, where the probability
of identifying the causative pathogen is very low. In dysentery,
which occurs less commonly, the probability of finding a pathogen
requiring treatment is higher, and diagnostic efforts should be
focused on these patients.
The assessment of the patient with diarrhea should include four
basic components: a history focused on identifying any epidemiological
risk factors for infection (such as eating raw seafood, having
been at a picnic or gathering at which others were sick, or ingestion
of an antimicrobial agent putting the patient at risk of pseudomembranous
colitis from Clostridium
infection); a physical examination focusing on assessment of dehydration
or the less likely presence of systemic complications; a gross
and microscopic examination of stool, which will help establish
whether the diarrhea is invasive (gross blood, erythrocytes and
leukocytes on microscopic examination) and whether a parasitic
infection is present; and a bacterial culture of a stool specimen.
In most clinical laboratories only a limited number of agents
are routinely identified when a culture of stool is requested;
these usually include Shigella,
Salmonella, Campylobacter jejuni,
In addition most clinical laboratories now are capable of screening
for enterohemorrhagic E.
if requested by the physician. If a patient is suspected of having
infection secondary to antibiotic consumption, most clinical laboratories
can test for the presence of C.
Those infections for which antimicrobial therapy is routinely
indicated are listed in Table
3. As can be seen, with the exception of Vibrio
(which is rarely encountered in industrialized countries) and
most of the infections for which antimicrobial therapy is routinely
indicated, cause infection of the colon and dysentery. Knowledge
of local antimicrobial resistance patterns is important in selecting
therapy, but the drug susceptibilities for an isolate will generally
not be known until 72 hours after a sample is submitted. Therefore,
antimicrobial therapy is usually initiated before susceptibility
results are known.
Because of low yield, there is limited utility in submitting routine
stool samples for culture from all patients with diarrhea. At
the New England Medical Center, as in other clinical laboratories,
a pathogen will be identified in less than 5% of all samples,
the majority of which do not require antimicrobial therapy. In
this era of managed care, such relatively profligate use of the
clinical laboratory is unlikely to be sustainable (a stool culture
at the New England Medical Center costs $46,more if susceptibility
tests are required). Stool cultures should be obtained from patients
with dysentery, prolonged diarrhea, immune deficiencies (e.g.
patients with HIV infection), and those with a special predilection
for developing complications from diarrheal illness, such as patients
of antimicrobial resistance on the treatment of enteric infections
Table 4 lists the
antibiotics of choice for use in those enteric infections requiring
them. Because of the dramatic increase in antimicrobial resistance,
this list of recommendations is radically different from a list
of 10 years ago. Formerly, trimethoprim-sulfamethoxazole, ampicillin
or nalidixicacid would have been the drugs of choice for the treatment
of shigellosis; now, in most of the world,resistance to these
agents is so common as to preclude their routine use. Likewise,
for cholera, thedrug of choice was tetracycline, and for Salmonella
either chloramphenicol, trimethoprim-sulfamethoxazole or ampicillin.
Now resistance to all these agents is common in Asia and Africa.
The development of widespread antimicrobial resistance among enteric
pathogens in developing countries has important implications for
treatment of diarrhea in the United States and other industrialized
countries. Among United States citizens, infection with pathogens
such as Shigella
type 1, Salmonella
is usually acquired when traveling abroad, or by eating imported
foodstuffs. Perhaps more importantly, the genetic elements encoding
resistance to these agents can be transferred to other organisms,
especially to normal gut flora which can be readily disseminated
between humans. The reasons for this rapid increase in antimicrobial
resistance are not known with certainty, but extensive and inappropriate
use of antimicrobials, with the resultant "ecologic pressure"
(natural selection) against susceptible strains undoubtedly plays
a major role. In the United States the dire implications of antimicrobial
resistance are typically associated with enterococci and Streptococcus
but the problem of antimicrobial resistance among enteric pathogens
is no less critical. For some strains of Shigella,
for instance, the only effective remaining oral antimicrobial
agents are the newer fluoroquinolones. Should resistance to these
agents develop, there would be no readily available alternative
for outpatient treatment of these potentially lethal infections.
therapy of diarrhea
There are two important non-antimicrobial therapies of diarrhea;
one is the replacement of the water and salts that are lost in
the diarrheal stool, and the other is the use of agents that inhibit
gut motility and perhaps decrease gut fluid secretion. Fluid replacement,
especially with oral rehydration salts, is a cornerstone of child
health programs in developing countries supported by UNICEF and
the World Health Organization. Packets of oral rehydration salts
to be reconstituted with water can be found at almost any pharmacy
in developing countries. Oral rehydration remains vastly underutilized
in the United States, however, perhaps because a major commercial
producer appeared only recently, and marketing targeted only the
infant population. The use of these salts is effective in restoring
fluid balance in both children and adults (I can give personal
testimony to the latter), and is preferred over household fluids
(tea, soft drinks, fruit juices) because it provides salt and
water in an appropriate balance. It also avoids the need for physicians
to give an "antimicrobial placebo" in response to patient
pressure. Packets of oral rehydration salts can be obtained from
CERA Products Inc., P.O. Box 801, Columbia, MD. 21044, telephone
(410) 997-2334. Salts already reconstituted in water are available
in many pharmacies and grocery stores under a variety of tradenames.
Antiperistaltic agents, of which loperamide is the best known,
can decrease the frequency of diarrheal stools. Their use should
be limited to watery diarrheas. With dysentery, the major objective
should be establishing the etiology of the infection and implementing
effective therapy rather than masking the symptoms of the infection.
Prevention of a case of diarrhea precludes having to face the
dilemma of how to treat a patient infected with a multiply resistant
enteric pathogen. Prevention of diarrhea among travelers is discussed
in detail in another paper in this series, but prevention of enteric
infections within developed countries follows similar principles.
Namely, this constitutes attention to personal hygiene and the
thorough cooking of all foods that are likely to transmit enteric
pathogens. Although in the United States, unlike most developing
countries, the safety of the food supply is in part a governmental
responsibility, it is clear that, even given current guidelines,
contamination of the food supply does occur. Important food vehicles
for transmission of enteric infections, include eggs (Salmonella),chicken
and hamburger (enterohemorrhagic E.
Eating these foods undercooked may provide gustatory pleasure
(where would the French kitchen be without its sauces using partially
cooked eggs, or a bar or cafe that doesn't offer a rare hamburger)
but ingesting undercooked foods clearly carries with it a small,
but important and perhaps increasing, risk of enteric infection.
- Bennish ML, Levy SB. Antimicrobial resistance of enteric
pathogens. In: Blaser, M.J. P.D. Smith,J.I. Ravdin, H.B. Greenberg, R.L. Guerrant, eds. Infections of the gastrointestinal
tract. New York,New York: Raven Press. 1995;1499-523.
- Bennish ML, Salam AM, Hossain MA, et al. Antimicrobial
resistance among Shigella isolates from Bangladesh, 1983-1990; increasing frequency of strains multiply resistant
to ampicillin, trimethoprim-sulfamethoxazole, and nalidixic acid. Clin Infect Dis 1992;14:1055-60.
- Bennish ML, Salam MA. Rethinking options for the treatment
of shigellosis. J AntimicrobChemother 1992;30:243-7.Salam MA, Seas C, Khan WA,
- Bennish ML. Treatment of shigellosis, IV: cefixime
ineffective inshigellosis in adults. Ann Intern Med 1995;123:505-8.
- Taylor DN, Blaser MJ, Echeverria P, et al. Erythromycin-resistant
Campylobacter infections inThailand. Antimicrob Agents Chemother 1987;31:438-442.
- Glass, RI, Lew JF, Gangarosa RE, et al. Estimates of
morbidity and mortality rates for diarrheal diseases in American children. J Pediat 1991;118:27-33.
- Segreti J, Gootz TD, Goodman, LJ, et al. High-level
quinolone resistance in clinical isolates of Campylobacter jejuni. J Infect Dis 1992;165-667-670.
- Rodrigue DC, Cameron DN, Puhr ND, et al. Comparison
of plasmid profiles, phage types, and antimicrobial resistance patterns of Salmonella enteritidis isolates in the
United States. J ClinMicrobiol 1992;30:854-857.
- Ryan CA, Nickels MK, Hargrett-Bean NT, et al. Massive
outbreak of antimicrobial-resistantsalmonellosis traced to pasteurized milk. JAMA 1987;258:3269-3274.
- Centers for Disease Control and Prevention. The management
of acute diarrhea in children: oralrehydration, maintenance, and nutritional therapy. MMWR 1992;41(No. RR-16).
- Church DL, Cadrain G, Kabani A, et al. Practice guidelines
for ordering stool cultures in a pediatricpopulation. Am J Clin Pathol 1995;103:149-153.
- MacKenzie WR, Hoxie NJ, Proctor ME, et al. A massive
outbreak in Milwaukee of cryptosporidiuminfection transmitted through the public water supply. New Eng J Med 1994;331:161-7.
- St. Louis ME, Morse DL, Potter ME, et al. The emergence
of grade A eggs as a major source of Salmonella enteritidis infections: new implications for the control of salmonellosis.