infect-hatchette_Layout 1 02/02/11 10:33 AM Page 339

CMAJ Practice
Primer

Infectious diarrhea: when to test and when to treat

Todd F. Hatchette MD, Dana Farina MD

Competing interests: None

A
cute gastrointestinal illness is common. mation acquired from thorough history-taking
declared.
An estimated 1.3 episodes per person and physical examination.
occur each year in Canada, which trans- Clinical features that help direct management This article has been peer
reviewed.
lates to more than 40 million incidents at an esti- include characteristics of the stool (e.g., bloody,
mated cost of $3.7 billion annually.1 Acute diar- watery), associated symptoms (e.g., fever, tenes- Correspondence to:
rhea is the second most common reason travellers mus) and, in some instances, the intensity of the Dr. Todd F. Hatchette,
todd.hatchette@cdha
returning from developing countries seek medical vomiting. A detailed description of the clinical .nshealth.ca
attention.2 In most instances, infectious diarrhea features associated with specific pathogens can
CMAJ 2011. DOI:10.1503
is self-limiting and treatment does not depend on be found in recent reviews.4–6 /cmaj.091495
identification of the responsible pathogen. The Additional information that may help in
challenge for front-line clinicians is to recognize deciding whether testing is necessary and what
who requires testing and treatment. testing is required includes underlying comor-
The diagnostic yield of stool cultures is rela- bidities, travel history, ingestion of unfiltered
tively low, estimated to range from 1.5% to water, recent use of antibiotics, recent contact
5.6%, which for example translates to a cost of with a sick individual, dietary history focusing
US$952 to US$1200 per positive sample in the on foods that are at high risk for transmitting
United States.3 Therefore, selective testing has diarrheal illness (raw or undercooked meat,
been used to improve the cost-effectiveness of shellfish, eggs or milk), employment (daycare
testing. When deciding whether testing is war- worker, food handler, health care worker) and
ranted, clinicians need to consider two questions: residence in a closed facility (e.g., long-term care
Will identification of the pathogen influence facility).3–5
patient management, such as treatment or infec- Evidence of volume depletion is important to
tion-control measures? Is identification of the assess the severity of the illness. Reduced skin
causative agent important from a public health turgor and dry mucous membranes may reflect
perspective? This primer is a summary of infor- mild volume depletion. Tachycardia, an orthosta-
mation obtained from recent reviews, clinical tri- tic drop in blood pressure, hypotension, changes
als and guidelines. in mental status, listlessness in children and
weight loss in infants are all suggestive of severe
dehydration.3 Examination of the abdomen may
What factors are important on reveal focal or diffuse tenderness. Rebound ten-
history and examination? derness suggestive of peritoneal inflammation
should warrant further assessment to rule out
Ingestion of many different bacteria, viruses, severe colitis or perforation.
parasites and bacterial toxins can all lead to acute
presentation of vomiting and diarrhea. Infectious Key points
diarrhea can be classified in several ways. One
• Acute infectious diarrhea is a common, usually self-limiting condition
simple approach is to categorize patients into that is underreported to public health authorities.
two groups: those who present with bloody diar- • Testing should be considered when patients present with diarrhea of
rhea and those who present with nonbloody diar- more than one day’s duration associated with bloody stools, fever,
rhea (Table 1).3–5 symptoms of sepsis or evidence of dehydration, recent antibiotic use or
The initial assessment of anyone presenting underlying immunocompromised state, or when identification of the
causative agent is important from a public health perspective.
with infectious diarrhea should focus on severity,
need for hydration and historical clues as to the • Although antibiotics can reduce the duration of symptoms in
nonbloody travellers’ diarrhea, use of antibiotics to treat this otherwise
cause in the patient. The decision to test for a self-limited illness may not justify the risk of antimicrobial resistance in
specific pathogen, notify public health or start the community.
empirical treatment can be influenced by infor-

© 2011 Canadian Medical Association or its licensors CMAJ, February 22, 2011, 183(3) 339
infect-hatchette_Layout 1 02/02/11 10:33 AM Page 340

Practice

Who requires testing? What treatment should be

offered?
Because most instances of acute diarrhea are
self-limiting and because the diagnostic yield of Treatment of infectious diarrhea can be divided
stool cultures is relatively low, testing everyone into supportive treatment and pathogen-directed
who presents with infectious diarrhea is not treatment.
necessary.4 However, guidelines from the Infec-
tious Disease Society of America suggest that Supportive treatment
stool cultures are important in certain settings The mainstay of treatment is oral hydration. The
because the identification of a pathogen can re- World Health Organization (WHO) has laid out
duce unnecessary tests or treatments, allow for clear guidelines regarding the composition of
appropriate treatment where there is docu- effective oral rehydration solutions.7 Commercial
mented benefit and prevent inappropriate treat- solutions are available for treating dehydration.
ment that could be harmful. 3 In addition, When commercially available preparations are
diagnostic testing is a critical part of ongoing unavailable, homemade oral rehydration solu-
surveillance, is essential to identify outbreaks, tions can be prepared by mixing one teaspoon
can provide public health officials with poten- (5 mL) of salt and eight teaspoons (40 mL) of
tial clues as to origin and can be useful when sugar in 1 L of water.8 Alternatively, a combina-
implementing public health measures, if appro- tion of water, fruit juices and salted crackers or
priate (e.g., a Shigella infection in a person em- soup may help replace fluids and electrolytes in
ployed as a food handler).3 patients with mild dehydration.4,9 In contrast,
In general, testing should be considered most sweat-replacement sports drinks do not
when patients present with diarrhea of more meet the WHO standards, because they contain
than one day’s duration that is associated with too many carbohydrates and do not have enough
bloody stools, fever, symptoms of sepsis or evi- replacement electrolytes.10,11
dence of dehydration, recent antibiotic use or Although many suggest a diet that contains
underlying immunocompromised state. Testing easily digested foods such as bananas, rice,
should also be considered when the diarrhea has applesauce and toast (the BRAT diet), data
important public health implications (e.g., the from a pilot study show that dietary restriction
patient is employed as a food handler or daycare does not enhance recovery.12 Over-the-counter
worker, or the investigation of clusters of a sus- medications such as bismuth subsalicylate
pected outbreak is underway).3 Figure 1 outlines (525 mg four times daily) and loperamide
an approach to deciding who to test and what to (4 mg initially, followed by 2 mg after each
test for. loose stool; maximum 16 mg/d) are available
to treat acute diarrhea; 13 however, an open-
label comparison between these two medica-
Table 1: Possible causes of acute infectious diarrhea3–5 tions published in 1990 found that loperamide
was more effective.14 Loperamide is generally
Agent Nonbloody diarrhea Bloody diarrhea
not recommended for use in patients with diar-
Bacterium • Enterotoxigenic Escherichia • Aeromonas spp rhea associated with fever or bloody stools and
coli (traveller’s diarrhea) • Campylobacter spp should not be used in children less than three
• Vibrio parahaemolyticus • E. coli producing Shiga-like years old.13,15 Bismuth subsalicylate is not rou-
• Shigella spp toxin (e.g., E. coli O157:H7 tinely recommended for use in children, be-
• Salmonella spp and other strains)
cause the salicylate component could predis-
• Yersinia spp • Shigella spp
pose them to Reye syndrome.
• Salmonella spp
• Yersinia spp
Pathogen-directed treatment
Virus • Norovirus
There are many considerations in the decision to
• Rotavirus
• Adenovirus
use antibiotics. Infectious diarrhea is a classic
• Astrovirus
example of weighing the risks to the patient ver-
sus the benefits of the treatment.
Parasite • Giardia lambia • Entamoeba histolytica
Because results of stool culture can take 48
• Cryptosporidum
• Isospora or Cyclospora spp
hours, the treating physician needs to decide
whether empirical treatment is required before
Toxin • Clostridium difficile
confirming the diagnosis. When clinicians are
• Staphylococcus aureus
• Bacillus cereus
considering the use of empirical therapy, it is
• Clostridium perfringens
helpful to categorize patients into the following
groups: travellers with nonbloody diarrhea, non-

340 CMAJ, February 22, 2011, 183(3)

infect-hatchette_Layout 1 02/02/11 10:33 AM Page 341

Practice

travellers with nonbloody diarrhea and people the combined data from six trials showed that
with bloody diarrhea (Figure 2).3,5,13 diarrhea in patients with travel-associated, non-
The most common causes of community- bloody diarrhea resolved one to two days earlier
acquired, infectious diarrhea are viral and short- on average with antibiotic use; however, the
lived. Conservative management with hydra- incidence of side effects was greater among
tion, with or without loperamide, is the most these patients than among those who were not
appropriate approach in most instances. Al - given antibiotics.16 Despite these data, in an era
though guidelines suggest that antibiotic treat- of increasing rates of antimicrobial resistance,
ment can be considered for patients with fever the small improvement in recovery from an oth-
and moderate or severe disease,3 the benefits erwise self-limited illness may not justify the
need to be weighed against the potential disad- risk of antimicrobial resistance in the com -
vantages of antibiotic use. A meta-analysis of munity. However, if symptoms are severe or

Patient presents
with diarrhea

Suspect outbreak: notify

public health authorities

Nonbloody diarrhea Bloody diarrhea

Is the patient one of Order stool culture for:

the following? • Salmonella
• Resident in a closed • Shigella
facility (e.g., long- • Campylobacter
term care facility) • Escherichia coli O157
• Daycare worker • Yersinia
• Food handler Consider EIA for Shiga-
• Health care worker like toxin

No Yes History of
antibiotic use in
Assess severity: does Order stool last few weeks?
the patient have any culture for:
of the following? • Salmonella
• Dehydration • Shigella Order assay
• Fever • Campylobacter for C. difficile
• Underlying comorbid • E. coli O157 toxin
illness • Yersinia

Yes
No

Symptomatic Nausea History of History of Diarrhea > 7 days,

treatment; prominent ingesting antibiotic use or associated with
no need for symptom raw shellfish in last 3 history of travel or
testing months community outbreak

Add test for: Add culture Add assay for Add stool
• Norovirus for Vibrio C. difficile examination
• Rotavirus toxin for ova and
• Adenovirus parasites

Figure 1: Algorithm for deciding when testing is required in patients who present with infectious diarrhea.3–5
Note: C. difficile = Clostridium difficile, EIA = enzyme immunoassay.

CMAJ, February 22, 2011, 183(3) 341

infect-hatchette_Layout 1 02/02/11 10:33 AM Page 342

Practice

worsen, empirical therapy can be considered In patients with a positive culture result, the
(Figure 2). decision to treat depends on the pathogen, the
Empirical treatment should be avoided in length of time from symptom onset, and comor-
patients presenting with bloody stools, because bidities. Documented bacteremia requires antibi-
potential causes include Escherichia coli O157 otic treatment. In Shigella and Campylobacter
or other strains producing Shiga-like toxin. infections, antibiotic treatment is associated with
Although the data analyzed in a recent sys- reduced severity and bacterial shedding.4 How-
temic review of antibiotic use and risk of ever, in Campylobacter infections, a study pub-
hemolytic uremic syndrome in patients with a lished in 1982 showed that antibiotics must be
toxin-producing strain of E. coli O157 were given within four days after symptom onset,21
conflicting, most clinicians recommend against and a recent meta-analysis showed that antibiotic
antibiotic treatment.9,17,18 (Box 1 outlines the treatment shortened the duration of symptoms by
clinical features suggestive of infection with only 1.3 days.22
E. coli O157.19,20) In addition, Shigella infec- Although treatment of Salmonella carries
tions are rarely life-threatening; thus, waiting with it an increased risk of chronic carriage, bac-
for the results of the stool culture in a patient teremia can occur in 2%–4% of patients.4 There-
with bloody stools is reasonable before starting fore, treatment should be considered in individu-
therapy.20 als with a positive culture result and who are at

Patient presents with diarrhea

Nonbloody diarrhea, Nonbloody diarrhea, Bloody

travel history no travel history diarrhea

Consider loperamide in Assess severity, Assess severity,

adults (4 mg, then 2 mg identify risk factors identify risk factors
after each loose stool; for severe disease for severe disease
maximum 16 mg/d).
If decision is made to use
antibiotics:*
• Adults: ciprofloxacin
Mild Antibiotic use in Mild symptoms Severe
750 mg orally once; if
disease last 3 mo or risk factors symptoms,
symptoms persist after
for E. coli O157 unlikely to be
24 h, 500 mg orally
(Box 1) E. coli O157
twice daily for 3 d.
No Yes
Consider azithromycin
1000 mg once† or
500 mg/d for 3 d for No Fever with • Mild symptoms: No
travel in Southeast Asia treatment moderate metronidazole treatment
• Children: azithromycin or severe 500 mg orally
10 mg/kg once daily disease 3 times daily for 10 d
for 3 d • Severe symptoms:
vancomycin 125 mg
orally 4 times daily
for 10 d
• Reassess after
C. difficile test
results are known

• Adults: ciprofloxacin 500 mg orally twice daily

for 3 d
• Children: azithromycin 10 mg/kg once daily for 3 d
• Reassess once pathogen identified

Figure 2: Algorithm for deciding whether empirical treatment of infectious diarrhea is required.3,5,12 *For nonbloody diarrhea with a his-
tory of travel, antibiotic use only reduces symptoms by 1–2 days and is associated with side effects; such treatment may not be justi-
fied in an era of antimicrobial resistance. †1000 mg of azithromycin can cause nausea. Note: C. difficile = Clostridium difficile, E. coli =
Escherichia coli.

342 CMAJ, February 22, 2011, 183(3)

infect-hatchette_Layout 1 02/02/11 10:34 AM Page 343

Practice

risk of serious infection or complications of bac- When should public health

teremia (Box 2).3,5
Patients with a positive test result for Clostrid-
officials be notified?
ium difficile cytotoxicity should receive treatment
with either metronidazole, if the infection is mild, Acute gastrointestinal illness is underreported to
or oral vancomycin, if the infection is severe.23 As public health officials. Although laboratories are
mentioned previously, antibiotic treatment in obligated to report isolates of many of the patho-
patients with documented infection with E. coli gens, the true incidence of disease is underrepre-
O157:H7 or another strain producing Shiga-like sented. Recent data from the National Studies on
toxin is not recommended because of the potential Acute Gastrointestinal Illness suggest that, for
increased risk of hemolytic uremic syndrome.19 every patient with verotoxigenic E. coli, Salmo-
nella or Campylobacter infection detected by the
How can transmission be national surveillance system in Canada, up to 49
people with such infections are missed in the
prevented? community.1 In addition, the identification of
clusters of parasitic infections can be important
Although infectious diarrhea is often associ- indicators of food- or water-borne outbreaks,
ated with food-borne illness, many pathogens such as the Cryptosporidium outbreak in North
such as norovirus or Shigella can be readily Battleford, Saskatchewan, in 2001.25
transmitted person to person. Patients who Because primary care physicians are often the
have acute gastrointestinal illness should be first individuals to recognize clusters of cases, it
reminded of the importance of hand hygiene is particularly important for them to report cases
as a way to prevent transmission to others. involving patients who would be at risk of trans-
People at high risk of transmitting disease mitting infection or who may be indicators of
(e.g., those employed as food handlers, day- potential outbreaks. This would include employ-
care workers or health care workers) should ees who are food handlers, day care workers,
stay off work until symptoms resolve. How- health care workers and residents in closed facil-
ever, even in healthy people, many pathogens, ities. The reporting requirements may differ from
including norovirus, can shed for weeks fol- province to province; clinicians should check
lowing infection, which reinforces the need for with local public health officials regarding re-
attention to hand hygiene.24 porting obligations.
Patients with underlying comorbidities, such
as immunosuppression or chronic liver disease, When should follow-up testing
are at increased risk of severe infections and
should use safe food-preparation practices and be done?
avoid eating high-risk foods (e.g., raw shellfish
and raw or undercooked meat). Contaminated Although follow-up testing is not routinely rec-
swimming pools are well recognized as the ommended, there may be situations where con-
source of infectious diarrhea in some outbreaks; firmation is needed to show that a patient is no
therefore, patients with a Cryptosporidium infec- longer shedding the pathogen. Follow-up cul-
tion should avoid swimming while symptomatic tures may be recommended for food handlers or
and for two weeks after the diarrhea stops, health care workers before they are allowed to
because of the potential for persistent shedding
of oocytes.6
Box 2: Risk factors where treatment of
diarrhea associated with nontyphoidal
Box 1: Clinical features suggestive of
Salmonella infection is indicated3,5
infection with Escherichia coli O15719,20
• Age < 6 months
• Bloody diarrhea preceded by one to three
days of nonbloody diarrhea • Age > 65 years

• Five stools in 24 hours • Immunosuppression

• Abdominal tenderness • Corticosteroid use

• Worsening pain on defecation • Inflammatory bowel disease

• No fever on presentation • Prosthetic joint or vascular material

• No granulocytic reaction seen in differential • Hemoglobinopathy (e.g., sickle cell disease)

count of leukocytes • Hemodialysis

CMAJ, February 22, 2011, 183(3) 343

infect-hatchette_Layout 1 02/02/11 10:34 AM Page 344

Practice
14. DuPont HL, Flores Sanchez J, Ericsson CD, et al. Comparative
return to work.1 Recommendations regarding efficacy of loperamide hydrochloride and bismuth subsalicylate
which pathogens and patients should prompt fol- in the management of acute diarrhea. Am J Med 1990;88:15S-9S.
low-up testing can vary among provinces. As 15. Li ST, Grossman DC, Cummings P. Loperamide therapy for
acute diarrhea in children: systematic review and meta-analysis.
such, clinicians should consult their local public PLoS Med 2007;4:e98.
health officials for further guidance. 16. De Bruyn G, Hahn S, Borwick A. Antibiotic treatment for trav-
ellers’ diarrhea. Cochrane Database Syst Rev 2000;3:CD002242.
17. Panos GZ, Betsi GI, Falagas ME. Systematic review: Are antibi-
References otics detrimental or beneficial for the treatment of patients with
Escherichia coli O157:H7 infection? Aliment Pharmacol Ther
1. Thomas MK, Majowicz SE, Pollari F, et al. Burden of acute gas-
2006;24:731-42.
trointestinal illness in Canada, 1999–2007: interim summary of
18. Bavaro MF. Escherichia coli O157: What every internist and
NSAGI activities. Can Commun Dis Rep 2008;34:8-15.
gastroenterologist should know. Curr Gastroenterol Rep 2009;
2. Freedman DO, Weld LH, Kozarsky PE, et al. GeoSentinel Sur-
11:301-6.
veillance Network. Spectrum of disease and relation to place of
19. Tarr PI, Gordon CA, Chandler WL. Shiga-toxin-producing
exposure among ill returned travelers. N Engl J Med 2006;354:
Escherichia coli and haemolytic uraemic syndrome. Lancet
119-30.
2005;365:1073-86.
3. Guerrant RL, Van Gilder T, Steiner TS, et al. Infectious Dis-
20. Holtz LR, Neill MA, Tarr PI. Acute bloody diarrhea: a medical
eases Society of America. Practice guidelines for the manage-
emergency for patients of all ages. Gastroenterology 2009;136:
ment of infectious diarrhea. Clin Infect Dis 2001;32:331-51.
1887-98.
4. Thielman NM, Guerrant RL. Acute infectious diarrhea. N Engl J
21. Anders BJ, Lauer BA, Paisley JW, et al. Double-blind placebo
Med 2004;350:38-47.
controlled trial of erythromycin for treatment of Campylobacter
5. DuPont HL. Clinical practice. Bacterial diarrhea. N Engl J Med
enteritis. Lancet 1982;1:131-2.
2009;361:1560-9.
22. Ternhag A, Asikainen T, Giesecke J, et al. A meta-analysis on
6. Davies AP, Chalmers RM. Cryptosporidiosis. BMJ 2009;339:
the effects of antibiotic treatment on duration of symptoms
b4168.
caused by infection with Campylobacter species. Clin Infect Dis
7. Oral rehydration salts: production of the new ORS. Geneva
2007;44:696-700.
(Switzerland): World Health Organization; 2006. Available:
23. Gerding DN, Muto CA, Owens RC Jr. Treatment of Clostridium
http://whqlibdoc.who.int/hq/2006/WHO_FCH_CAH_06.1.pdf
difficile infection. Clin Infect Dis 2008;46:S32-42.
(accessed 2010 Dec. 15).
24. Atmar RL, Opekun AR, Gilger MA, et al. Norwalk virus shed-
8. Rehydration Project. Oral rehydration solutions made at home.
ding after experimental human infection. Emerg Infect Dis 2008;
The Mother and Child Health and Education Trust; 2010. Avail-
14:1553-7.
able: http://rehydrate.org/solutions/homemade.htm (accessed
25. Stirling R, Aramini J, Ellis A, et al. Waterborne cryptosporidio-
2009 Nov. 26).
sis outbreak, North Battleford, Saskatchewan, Spring 2001. Can
9. Al-Abri SS, Beeching NJ, Nye FJ. Traveller’s diarrhoea. Lancet
Commun Dis Rep 2001;27:185-92.
Infect Dis 2005;5:349-60.
10. Atia AN, Buchman AL. Oral rehydration solutions in non-
cholera diarrhea: a review. Am J Gastroenterol 2009;104:2596- Affiliations: From the Division of Microbiology, Depart-
604. ment of Pathology and Laboratory Medicine (Hatchette), and
11. Harris C, Wilkinson F, Mazza D, et al. Health for Kids Guide-
the Division of Gastroenterology, Department of Medicine
line Development Group. Evidence based guideline for the man-
agement of diarrhoea with or without vomiting in children. Aust (Farina), Capital District Health Authority, Halifax, NS; and
Fam Physician 2008;37:22-9. the Departments of Pathology (Hatchette) and Medicine
12. Huang DB, Awasthi M, Le BM, et al. The role of diet in the (Farina), Dalhousie University, Halifax, NS
treatment of travelers’ diarrhea: a pilot study. Clin Infect Dis
2004;39:468-71. Contributors: Both authors contributed equally to the writ-
13. Hill DR, Ryan ET. Management of traveler’s diarrhoea. BMJ ing and revision of the manuscript, and both approved the
2008;337:a1746. final version submitted for publication.

344 CMAJ, February 22, 2011, 183(3)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.