Contagious Acute Gastrointestinal Infections
new england journal medicine The of review article medical progress
Contagious Acute Gastrointestinal Infections Daniel M. Musher, M.D., and Benjamin L. Musher, M.D.
n our ever-shrinking world, widespread media coverage of in- From the Medical Service, Infectious Dis- ease Section, Michael E. DeBakey Veterans fections, ranging from the severe acute respiratory syndrome (also known as
Affairs Medical Center, and the Depart- SARS) and influenza in Asia to acute gastroenteritis on cruise ships and outbreaks ments of Medicine and Molecular Virology
i
and Microbiology, Baylor College of Medi- in day-care centers in the United States, has raised public interest in contagious diseas- cine — both in Houston (D.M.M.); and es to new heights. Our purpose in this article is to examine contagion (from the Latin, the Department of Medicine, University of
tangere
, to touch) — direct human-to-human spread — of acute gastrointestinal illness, Pennsylvania School of Medicine, Philadel- phia (B.L.M.). Address reprint requests to defined as a syndrome of vomiting, diarrhea, or both, that begins abruptly in otherwise
Dr. Daniel Musher at the Infectious Disease healthy persons and is most often self-limited. Section, Veterans Affairs Medical Center,
1 Unlike agents that cause contagious respiratory infections, which are largely or ex- Houston, TX 77030, or at daniel.musher@ med.va.gov. clusively indigenous to humans, agents that cause acute gastrointestinal illness (Table 1) may spread from person to person or may be acquired from a common food or envi-
N Engl J Med 2004;351:2417-27. ronmental source, often water; they may also result from exposure to animals. Food or Copyright © 2004 Massachusetts Medical Society. water may serve as a primary source of contagion or may, in turn, have been contami- nated by contact with an infected person or animal. Thus, the epidemiology of acute gastrointestinal illness is complex.
Different ways of gathering, analyzing, and presenting data have generated very dif- ferent estimates of the frequency of acute gastrointestinal illness, leading to seemingly contradictory results. Estimates based on extrapolation from isolation of known diar- rheal pathogens and the numbers of stool samples submitted for study suggest that
2 there might be 38 million cases of acute gastrointestinal illness each year. In contrast, a carefully conducted questionnaire survey asking about acute, self-limited illness characterized by vomiting, diarrhea, or both found that about 1.05 cases occur per per-
2,3 son per year in the United States. When this number was reduced by 25 percent on the basis of estimates that a respiratory infection is the responsible agent in about one quarter of persons with symptoms of acute gastrointestinal illness, the resulting 0.79 case per person per year translated to 211 million cases of acute gastrointestinal illness nationally in 1997, the year for which data were available.
Earlier data from the United States and questionnaire-based studies in the Nether-
3 lands and the United Kingdom yielded similar results. On the basis of reports to pub- lic health authorities and an exchange of information between the Centers for Disease
2,3 Control and Prevention and a network of participating laboratories (FoodNet), there are thought to be about 76 million cases per year of foodborne infection. If this number
3 and an additional 13 million cases of waterborne illness are subtracted, there may well be 122 million cases of acute gastrointestinal illness each year in the United States for which human-to-human transmission is responsible. As noted above, a varying pro- portion of foodborne and waterborne outbreaks are also ultimately attributable to hu- man contamination. n engl j med 351;23 www.nejm.org december 2, 2004
2417 new england journal medicine The of
b a c t e r i a l c a u s e s
Table 1. Agents That Commonly Cause Acute Gastrointestinal Illness.* salmonella Bacteria
Because many principles of contagion with respect
Salmonella
to enteric organisms were elucidated in studies of
Shigella
typhoid fever, it seems appropriate to begin our dis-
Campylobacter Escherichia coli O157:H7
cussion of causes of acute gastrointestinal illness
Clostridium difficile
with Salmonella typhosa ( S. enterica serotype typhi). Al-
Viruses
though physicians do not always associate this or-
Caliciviruses (Norwalk-like and related viruses) Rotavirus†
ganism with a typical syndrome of acute gastroin-
Adenovirus types 40 and 41
4 testinal illness, some studies suggest that diarrhea
Astrovirus
5,6 predominates in the majority of cases. S. typhi is
Protozoa Giardia
highly adapted to humans. Infection is virtually al-
Cryptosporidium
ways acquired by transmission from one person to
Entamoeba histolytica
another; an inviolable rule of epidemiology is that the occurrence of a case of typhoid fever implies an
- These organisms commonly cause acute gastrointesti- nal infection in otherwise healthy children and adults in
epidemiologic link to another person who either is
developed countries. The frequency of infection is simi-
actively infected or is chronically carrying the organ-
lar among such countries — for example, the United
ism and shedding it in feces. When cases result from States, United Kingdom, France, and Argentina.
† Symptomatic disease usually occurs only in infants or
food ingestion, individual food handlers, such as 7 very young children. the infamous cook known as Typhoid Mary, are usually found to be responsible. An infection from drinking contaminated water can also usually be traced to one or more infected persons whose ex- The early implications of the watchwords “fingers,
8-10 creta have entered the water supply. food, and flies,” and the frequent spread from pa-
The current rarity of typhoid fever in the United tients to nurses and physicians in the era before an-
8 States reflects good hygiene, lack of crowding, and tibiotics, are consistent, at least in some instances high public health standards for home and indus- of natural infection, with low-inoculum contagion, trial sewage. During the late 1990s, a breakdown of under the assumption that large numbers of organ- the public health infrastructure in the former Sovi- isms would not be transmitted in these situations. et Union led to a cessation of chlorination, the pi- Infections with most other types of salmonella, rating of water lines with the use of substandard except for S. paratyphi, derive from environmental pipe fittings, and the crossing of these fittings by sources, principally poultry and livestock. Despite sewage lines, which culminated in an outbreak of the frequency with which these organisms cause
11 10,000 cases of typhoid fever. acute gastrointestinal illness, there are remarkably
The likelihood of direct contagion depends on few documented examples of person-to-person 17-19 the number of organisms in feces or contaminated spread. An outbreak in a day-care facility was foods, their ability to survive, replicate, or both, and associated with an uncertain number of secondary
19 the infectivity of the species and the specific strain. cases, and long-term surveillance of 54 perma-
6
9 Chronic carriers of S. typhi have 10 to 10 colony- nent carriers of nontyphoidal salmonella identified
12
13
20 forming units (CFU) per gram or more in their 10 instances of transmitted infection. On the ba-
3 feces. In experimental studies, ingestion of 10 CFU sis of epidemiologic studies, the infective dose of of the Quailes strain of S. typhi was not infectious in nontyphoidal salmonella is thought to be small,
21,22 volunteers, whereas nearly 50 percent of volunteers not exceeding 100 CFU. The paucity of docu-
5
7 were infected by ingesting 10 or 10 CFU, and 96 mented instances of contagion may reflect the dif-
8 percent were infected by ingesting 10 or more ficulty of distinguishing person-to-person spread
14,15 CFU. The results of these experimental studies from that due to a common food source, rather than indicate that a large inoculum is infective. However, the true absence of human transmission. infection in the real world will depend on the infec- tivity of the strain studied. In nature, such strains are shigella almost certainly heterogeneous, as has been shown Like S. typhi , shigella has no reservoir in nature and
16
1 for other enteric and for respiratory pathogens. spreads from person to person (usually child to n engl j med 351;23 www.nejm.org december 2 , 2004
2418
24 The high level of contagious-
n engl j med 351;23 www.nejm.org december 2, 2004
47 Although the disease in some persons results from the proliferation of an endogenous strain, infection is clearly contagious; in hospitals, both human vectors and environmen- tal contamination are implicated in the spread.
E. coli –related diarrhea
linked to hamburger consumption in an epidemic that occurred during 1992 and 1993, 48 infections (about 10 percent) were secondary.
42 Person-to- person transmission occurs in day-care centers,
43 among families,
44 and in mental institutions
45 ; an attack rate — the rate of appearance of symptoms in exposed persons — of around 20 percent has been reported. It is worth noting that this form of acute gastrointestinal infection is associated with a substantial incidence of the hemolytic–uremic syn- drome, affecting up to 13 percent of young chil- dren with the infection.
46
clostridium difficile Clostridium difficile is a major cause of nosocomial
colitis, generally occurring after antibiotic-induced alterations of bowel flora.
C. difficile to 50 percent
48 In day-care settings, an infection in one child may be followed by the spread of
medical progress 2419
of the classmates, in nearly all of whom diarrhea then develops
49 ; contagion is greatly facilitated by the ingestion of antibiotics. Caregivers may acquire
C. difficile colitis while caring for patients who have
this disease. We treated an elderly woman for acute
C. difficile colitis; she had been caring for her hus-
band during his bout of
C. difficile colitis, and she had
not been taking antibiotics. Her stools contained * These factors are discussed in Pickering et al. 30 Table 2. Factors That May Contribute to Contagion
of Acute Gastrointestinal Illnesses within Day-Care Centers.* Presence of one or more cases of acute gastrointestinal
illness
Lack of gloving or handwashing during or after diaper changes or helping at toilet Lack of policy to isolate or send home children with acute gastrointestinal illness Common area for diaper changing Larger groups of children Carpeted flooring Shared toys and classroom objectsTransmission of Escherichia coli O157:H7 occurs pri- marily through the consumption of contaminated meat, but secondary infection does occur, and a small bacterial inoculum may lead to clinical illness. For example, of 501 cases of
escherichia coli o157:h7
41 has been shown to be contagious among homosexual men.
40 and shigella ,
child) after direct contact or the ingestion of con- taminated food. Shigellosis is highly contagious; as few as 200 CFU can cause infection,
23 and the role of a small inoculum is supported by early observa- tions, which emphasized spread by casual contact and insect vectors.
ness of shigellosis may be inferred from the large number of secondary cases that follow a document- ed outbreak; persons who have varying degrees of contact with infected patients are likely, themselves, to become infected.
25 A very young child is the usu- al source.
26 Not surprisingly, shigella readily spreads within families,
24 in custodial institutions,
27 and within and among children’s day-care centers.
28,29 Day- care centers provide remarkable natural settings in which contagion in acute gastrointestinal illness can be studied
28 (Table 2). In these settings, shigellosis
(Tables 1 and 2) may affect from one third to two thirds of children,
30 with severe diarrhea increasing the likelihood of contagion, reflecting high fecal counts of bacteria and increased chances of soil- ing.
26 At least one additional case of shigellosis is recognized in the families of about 25 percent of infected children.
28 The current widespread use of medications that reduce gastric acidity (which nor- mally eradicates salmonella and shigella) probably increases the risk of spread
14 to parents of infected children or to adults who work in day-care centers.
campylobacter
The epidemiology of infection due to campylobac- ter, perhaps now the most common bacterial cause of acute gastrointestinal illness,
31,32 is similar to that of nontyphoidal salmonella. Most infections are traced to poultry, meat, dairy products, or con- taminated water.
33 Although fewer than 1000 CFU
may cause infection,
16 massive foodborne out- breaks are not often recognized, in part because this organism does not replicate in food
34 and in part because ingestion even of large numbers of organ- isms may cause symptoms in only a small propor- tion of subjects.
16 Contagion within the home has
been described occasionally, 35-37 and in one house- hold, an infant was infected with the same strain that caused diarrhea in a newly acquired puppy.
38 As
with nontyphoidal salmonella, the paucity of de- scriptions of human-to-human spread may reflect the difficulty of studying organisms that are present in so many food sources. Infection by campylobac- ter ,
39 as well as by S. typhi n engl j med 351;23 www.nejm.org december 2 , 2004
The new england journal of medicine
51 In contrast, in underdeveloped na-
Within families, acute gastrointestinal illnesses are spread chiefly by young children, whose hygiene is not as consistently good as that of adults and who are dependent on, and therefore in intimate contact with, their parents and caregivers.
69 In another example, 43 members of a football team contracted acute gastrointestinal illness from a sin- gle food source. Eleven members of an opposing v i r a l c a u s e s
68 Contagion by Norwalk-like viruses has been docu- mented in other circumstances as well. When Brit- ish soldiers with acute gastrointestinal illness were airlifted out of a combat zone, two flight medics and one hospital staff member subsequently became ill; fecal samples from both the patients and the medi- cal personnel contained Norwalk-like viruses.
67 traced to a food handler in a college dining hall, about 20 percent of all cases occurred after the dining hall was closed and were therefore thought to reflect secondary person-to- person spread (Fig. 1). In a large community out- break in Sweden, secondary cases appeared in one third of the households in which a case occurred.
62 indicate person-to-person, rath- er than foodborne, transmission. For example, in a hyperacute outbreak
66 Well-documented sec- ondary outbreaks
Norwalk-like virus, the importance of person-to- person transmission has been recognized since the initial identification of this organism in an outbreak that affected one third of family members and 50 percent of school contacts.
64,65 Although consumption of contaminated food or water causes large outbreaks of infection with
62,63 Attack rates have been as high as 41 percent, reflecting the propensity of infection with Norwalk-like viruses to cause emesis and vo- luminous stools, the large number of organisms in stools and vomitus, and the low inoculum (fewer than 100 viral particles) required to produce infec- tion. The extent of spread in such closed environ- ments may involve nearly 100 percent of exposed persons, since experimental ingestion of infectious material causes symptoms in only 50 to 80 percent of subjects.
2,61 As already noted, if the same percentages are applied to cases of acute gastrointestinal illness identified by questionnaire, the incidence of calicivirus-induced infection is far greater; there may be a total of 74 million cases each year in the United States. Outbreaks have been reported in nursing homes and on military bases and school campuses, but Norwalk-like viruses on cruise ships have made national headlines in the past few years.
Calciviruses, of which Norwalk-like viruses are the prototype, cause more than 90 percent of outbreaks of acute gastrointestinal illness in the United States and account for about 23 million cases of diarrheal disease per year, according to the pathogen-associ- ated method of calculation.
caliciviruses
50 20 percent of persons have sympto- matic infection after exposure to a family member with acute gastrointestinal illness. The likelihood of secondary infection increases from 10 percent when symptoms are mild to 30 percent if severe vomiting and diarrhea are present, reflecting increased vol- umes of infective excreta that presumably contain higher concentrations of infective particles.
Dingle et al.,
60 Features of contagion by these agents are summa- rized in Table 3.
2420
O157:H7 each in 2 percent or less of fecal samples
C. difficile toxin, and she responded to treatment with metronidazole (unpublished data).
At least since the end of the Second World War, in developed countries, viruses have been thought to cause the vast majority of cases of acute gastroin- testinal illness, whether sporadic or part of an out- break. In the 1950s, a definitive family study by Dingle et al.
50 found no isolates of salmonella or shigella in 77 cases of acute gastrointestinal illness; these investigators concluded that most cases were due to viruses, although, at the time, they were un- able to isolate them. At that time, techniques were not available to identify campylobacter or
E. coli
O157:H7. The relative infrequency of bacterial acute gastrointestinal illness in developed coun- tries was confirmed by prospective studies that iden- tified salmonella, shigella, campylobacter, and
E. coli
32 ; these numbers have steadily declined in the past several years.
57-59 have also been implicated. Caliciviruses and astrovi- ruses are more prevalent among outpatients, where- as rotavirus is a common cause of hospitalization.
tions, one of the aforementioned bacteria, vibrio, enteropathogenic
E. coli, protozoa, or intestinal par-
asites cause the majority of cases of acute gastroin- testinal illness.
In the United States, the United Kingdom, north- ern Europe, and Japan, caliciviruses such as the Norwalk and Sapporo viruses are the most common cause of sporadic acute gastrointestinal illness in patients of all age groups except infants and tod- dlers, in whom rotaviruses predominate.
52-54 Ade- novirus types 40 and 41
55,56 and astroviruses
50 As shown by
7 days Very low Downloaded from www.nejm.org on November 16, 2006 . Copyright © 2004 Massachusetts Medical Society. All rights reserved.
RT-PCR 2 days (1–3) 3 days (1 day to weeks) Very high Rotavirus Very low Human contact 2 days Fever, vomiting, diarrhea (in infants) EIA, latex agglutination 4 days
Microscopical exam- ination of feces 10–12 days (3–60)
(1–2 cysts) Tap water, human contact 7 days (1–14 days) Diarrhea, abdominal pain, headache, fever
Microscopical exam- ination of feces 1–8 wk 3 wk–6 mo High Cryptosporidium Very low
Tap water, human contact 9 days (1–2 wk) Abdominal discomfort, diarrhea
(1–7) 5 days (3–11 days) Low Giardia Low (≤10 2 organisms)
40 and 41 Unknown Human contact (feces, possibly vomitus) 2–3 days Diarrhea, vomiting, fever EIA (not commer- cially available) 2–4 days
2–5 days (1–14) 4 days (1 day to weeks) High Adenovirus types
(3–9) 4 days (2–7 days) Very high Astrovirus Unknown Human contact 1–2 days Diarrhea EIA (not commer- cially available)
1 day (1–2 days) Diarrhea, vomiting, fever
n en gl j m ed 351;23 www.nejm.org december 2, 2004
(1–7) 50% negative after 3 wk Very low Calicivirus Low Human contact (feces, vomitus), prepared food
Days to weeks Very high Campylobacter Low Poultry, milk, tap water 3 days (1–7 days) Diarrhea, fever Fecal culture 3 days
3 days (1–7 days) Diarrhea, fever Fecal culture 3 days (2–6)
(8–24 hr) Diarrhea, fever Fecal culture 2–4 days 5 wk, rarely lifetime Very low Shigella Low (≤10 2 CFU) Human contact, prepared food, contaminated water
Salmonella (nontyphoidal) Low (10 2 –10 3 CFU) Poultry, eggs, meat 24 hr
(10 5 CFU)‡ Human contact, prepared food, contaminated water 5–14 days Fever, abdominal pain, diarrhea Blood culture, fecal culture 3–4 wk 2–6 wk, rarely lifetime High
Table 3. Relevant Features of Selected Acute Gastrointestinal Infections.* Agent Quantity of Inoculum to Cause Disease Usual Mode of Transmission Incubation Period Usual Symptoms Diagnostic Methods Duration of Symptoms Duration of Shedding Probability of Human- to-Human Spread† Salmonella typhi High
† This column reflects the authors’ assessment of the likelihood of human-to-human spread, based on all the available sources of information as presented in the text. ‡ Experimental studies show a high inoculum, but some clinical observations suggest a low inoculum.
medical progress 2421
- This table is subject to the limitations of the medical literature (for some organisms, clinical studies are more detailed, whereas for other organisms the documentation may not exist, al- though the clinical syndromes may be very similar). In the entries in the columns “Incubation Period,” “Duration of Symptoms,” and “Duration of Shedding,” the numbers in parentheses indicate the range. “Shedding” is the time during which the infectious agent can be recovered from feces after the end of illness. CFU denotes colony-forming units, RT-PCR reverse-tran- scriptase–polymerase-chain-reaction assay, and EIA enzyme immunoassay.
stances have been cited in other locales, such as
Questionnaire A
72
73
74 cruise ships, hospitals, and restaurants, which
120
suggests that a small inoculum can spread disease
100
by aerosol. Caliciviruses persist in an infective form
75
80
in the environment and are resistant to deactiva-
76 tion by ordinary cleansing agents, although they
60
are inactivated by exposure to household bleach di-
40 No. of Cases
77 luted 1:10. This explains why, once they are in the
20
environment, for example in a day-care setting or a cruise ship, they are so difficult to eradicate.
1
2
3
4
5
6
7
8
9
10
11
12 Day of Outbreak
rotaviruses
Rotaviruses are a prominent cause of severe diar-
B Infirmary rheal disease in children under the age of two years.
160
Infection is highly contagious, indicating that a
140
very small inoculum is infectious, since the feces of
120
infected children usually contain no more than 100
100
CFU per gram. When a rotavirus is introduced into a family, about 50 percent of exposed children and
80
15 to 30 percent of exposed adults become infected,
60 No. of Cases
although some proportion of infected children and
40
28,78-80 most infected adults remain asymptomatic.
20 Most adults who are infected become so within the
family, whereas most infections in very young chil-
1
2
3
4
5
6
7
8
9
10
11
12
dren are acquired outside the family — for exam-
Day of Outbreak
28,81 ple, in day-care settings. Like caliciviruses, ro-
82 taviruses survive well on environmental surfaces
Figure 1. Contagion (Primary and Secondary Infection) in a Foodborne Out-
83 and are difficult to inactivate, although diluted break of Infection.
84 household bleach seems to be effective. The con-
An outbreak of a presumed calicivirus infection was traced to a single food 67
gruence of the small size of the inoculum required
handler who prepared salads in a college dining hall. Panel A shows the
85,86
number of cases of acute gastrointestinal infection that developed each day, for infection, the survival of the pathogen in
as reported on a questionnaire by persons who ate or worked in the diningthe environment, and its resistance to most com-
hall. Panel B shows the numbers of persons who presented to the college in-
mon cleansing agents renders this virus very diffi-
firmary each day with symptoms of acute gastrointestinal disease. The dining
cult to control in closed populations; the same is
hall was closed at the end of day 4. Patients who presented to the college in- true of the Norwalk-like viruses. firmary on days 5 through 7 were presumed to have acquired the infection in the dining hall. The long “tail” on the right side in Panel B is thought to reflect transmission of infection from persons initially infected by ingestion of con- adenovirus types 40 and 41 taminated food (primary cases) to other students who did not ingest the con- 67 Enteric adenoviruses, types 40 and 41, which have taminated food (secondary cases). The graphs are adapted from Kilgore et al.
been identified only recently by application of novel techniques, are found in the feces of about 3 per- cent of all young children with acute gastrointesti-
55,87,88 team (17 percent) later had acute gastrointestinal nal illness. These viruses are readily trans- illness due to a Norwalk-like virus with an identical mitted from child to child, with disease developing
70 genogroup. in about half of infected children; most infected
89-91 Whereas bacteria causing diarrheal disease are adults remain asymptomatic. In a prospective,
91 presumably shed exclusively in feces, caliciviruses five-year investigation, adenovirus type 40 or 41
67 are detected in vomitus and feces, and contact was found in all 10 outbreaks in which other or- with either source may result in infection. Airborne ganisms had not been identified; 38 percent of all transmission may have caused an outbreak in a ger- fecal samples studied were positive. Nevertheless, iatric facility in which 9 of 14 employees who con- one family study suggested that this organism is
55 tracted acute gastrointestinal illness had no direct much less contagious than rotavirus. In one pro-
71 contact with the feces of residents. Similar in- longed outbreak of acute gastrointestinal illness in n engl j med 351;23 www.nejm.org december 2 , 2004
2422
57 This ap-
e. histolytica
When adults are infected, the risk for secondary in- fection in families is less than 5 percent 108
; in part, this low rate of risk is consistent with the better hy- giene of adults, as compared with children, and, in part, it may be due to other, uncertain causes. Food handlers may also spread this organism.
109
giardia
Outbreaks of infection with giardia in child-care settings are associated with overall attack rates (in- cluding clinical and subclinical cases) of 17 to 47 percent among attendees and 10 to 35 percent among adult workers.
110 When a young child be- comes infected, there is a 5 to 25 percent chance that one or more family members will contract the disease.
28,110 Severe giardiasis occurs most com- monly in young children and women of childbear- ing age,
111 probably reflecting host susceptibility together with the effect of the size of the inoculum.
Giardia also spreads among participants in swim- ming classes 112 and among homosexual men.
113
Outbreaks of E. histolytica infection in schools are generally traced to contaminated water sources. Person-to-person spread has, however, been docu- mented in homes, schools, and day-care centers, as well as among homosexual men.
Cryptosporidium may infect 40 percent of house- hold members who have contact with young chil- dren with diarrhea, but fewer than 10 percent of household members whose contact is with asymp- tomatic carriers become infected
30,114,115 Never- theless, somewhat surprisingly, documented spread within families is unusual.
116,117 In nearly all instances, transmission of acute gastro- intestinal illness is due to organisms that are present transiently on the hands.
118 The distinction between transient and resident flora is important in under- standing apparent discrepancies relating to trans- mission of acute gastrointestinal illness. Washing the hands for 30 seconds with soap or detergent and water may not substantially reduce the total num- ber of bacteria that are present on relatively clean hands
119 ; in contrast, handwashing reduces by about 95 percent the numbers of bacteria or viruses that are applied to the hands experimentally
120,121 or that are acquired exogenously under natural con- ditions
122 ; and handwashing clearly reduces the spread of acute gastrointestinal illness in day-care and family settings.
123-125 The explanation is that exogenously acquired organisms or transient flora
(the ones that are likely to transmit infection) are removable by washing, whereas resident flora (the ones that are normally present) are not.
Whereas the antibacterial substances in house- hold soaps do not prevent acute gastrointestinal ill- ness,
126 additional field studies with alcohol-based p r o t o z o a l c a u s e s p r e v e n t i o n
107 — again illus- trating the importance of diarrhea in contagion.
n engl j med 351;23 www.nejm.org december 2, 2004
medical progress 2423
92,99,100 hospital wards have been involved.
persons hospitalized for long periods, rotavirus and adenovirus type 40 or 41 were isolated in nearly equal proportions.
92
astrovirus
Astroviruses, which are perhaps less well studied than the viruses already described, cause outbreaks of acute gastrointestinal illness — generally, but not always,
93 by person-to-person spread. Day-care
94 and kindergarten
57 attendees, military recruits,
95 and mothers and children in maternal-care facili- ties
96 have been implicated, and pediatric
97,98 and geriatric
During outbreaks in day-care centers, 50 to 90 per- cent of children and up to 25 percent of adults may have disease
106 Nevertheless, person- to-person spread of cryptosporidium
57,94,96 ; secondary cases occur in the families of one third of affected children.
parently high rate of contagion belies results show- ing disease in only a very small proportion of hu- man volunteers after experimental ingestion of astrovirus
101 ; the lower rate in the study is perhaps attributable to differences between naturally ac- quired strains and those used experimentally.
Cryptosporidium, Giardia lamblia, and Entamoeba his-
tolytica cause acute diarrheal disease, with transmis-
sion via a small inoculum (fewer than 100 organ- isms).
102-104 Once regarded as waterborne,
105 these organisms are now known to spread through day- care centers by way of the fecal–oral route with a substantial likelihood of secondary infection among family members, especially women of childbear- ing age.
cryptosporidium
Because it can be difficult to eradicate cryptospo- ridium from drinking water, large outbreaks of infection have occurred.
107 is well doc- umented in homes, schools, and day-care centers. n engl j med 351;23 www.nejm.org december 2, 2004
The new england journal of medicine
mon respiratory infections? N Engl J Med 2003;348:1256-66.
6. Mathieu JJ, Henning KJ, Bell E, Frieden
Rahman H, Stoll BJ. Diarrhea associated with typhoid fever. J Infect Dis 1985;151: 1138-43.
5. Roy SK, Speelman P, Butler T, Nath S,
Braunweld E, Fauci AS, Kasper DL, Jameson JL, eds. Harrison’s principles of internal med- icine. 15th ed. New York: McGraw-Hill, 2001: 970-5.
4. Lesser CF, Miller SI. Salmonellosis. In:
Burden of self-reported acute diarrheal ill- ness in FoodNet surveillance areas, 1998- 1999. Clin Infect Dis 2004;38:Suppl 3:S219- S226.
3. Imhoff B, Morse D, Shiferaw B, et al.
related illness and death in the United States. Emerg Infect Dis 1999;5:607-25.
2. Mead PS, Slutsker L, Dietz V, et al. Food-
1. Musher DM. How contagious are com-
2424
r e f e r e n c e s
We are indebted to Marsha Sullivan and the staff of the Medical Library at the Michael E. DeBakey Veterans Affairs Medical Center, Houston, for their help. s u m m a r y a n d c o n c l u s i o n s
Dr. Daniel Musher reports having received Merit Review Funding from the Department of Veterans Affairs, grant support from Ro- mark Laboratories, and consulting fees from Aventis.
Person-to-person transmission is best prevented by the practice of excellent personal hygiene both by infected persons and by those exposed to them. Fecal–oral transmission is the usual route of spread of acute gastrointestinal illness, but caliciviruses and probably adenoviruses are present in vomitus, so kissing or sharing utensils should also be avoid- ed. Dilution by handwashing reduces the inoculum of causative organisms, greatly diminishing the risk of contagion. There is no apparent benefit from the antibacterial agents in soaps, although the regular use of alcohol-based gels will probably reduce trans- mission. The use of diluted household bleach on environmental surfaces may be necessary to inter- rupt transmission of viral or protozoal agents.
127 Within families, young children are the usual source for contagion because of their exposure to other children, their imperfect personal hygiene, and their dependence on adults. Severely affect- ed persons are more contagious because they dis- charge greater volumes of infective material that contain large numbers of infectious particles. The likelihood of contagion varies with the concentra- tion of organisms in excreta, the capacity of the or- ganisms to survive and replicate in food or persist in the environment, and the number required to in- fect. Spread of acute gastrointestinal illness is com- mon and problematic in all closed environments such as day-care centers, schools, and cruise ships.
The likelihood of contagion depends on the age and self-reliance of an infected person, the nature of the social interaction within the potentially in- volved group, the intensity of the symptoms, the concentration of organisms in the potentially in- fective material, the likelihood that the organism will survive direct transmission or survive in the en- vironment, and other, less well understood factors. The immune status of the host undoubtedly plays a role in determining whether symptomatic disease or subclinical infection results, but the nature of such immune factors is poorly understood.
As a general matter, the failure to identify a common source for most sporadic, presumably vi- ral, acute gastrointestinal illnesses does not exclude the possible link to an unrecognized foodborne out- break. The essential point remains, however, that — even though the visibility of an outbreak tends to focus attention on foodborne infection — the great majority of cases are sporadic and spread from per- son to person. Although free-living protozoa, such as cryptosporidia or giardia, are widespread in na- ture, contagion is also well documented.
Acute gastrointestinal illness is exceedingly com- mon; viruses, bacteria, and protozoa are the princi- pal recognized causes. Some causative organisms, such as calicivirus, rotavirus, astrovirus, adenovirus types 40 and 41, S. typhi, and shigella, are indige- nous to humans; person-to-person spread follows direct contact or human contamination of food or water. In contrast, nontyphoidal salmonella, cam- pylobacter, and pathogenic E. coli are prevalent in meat, poultry, and dairy foods; human-to-human spread is documented infrequently relative to the total number of cases of infection with these bacte- rial agents. This lower rate of documentation may reflect the difficulty, in an individual case, of deter- mining whether some common food source is re- sponsible or in distinguishing an environmental source from a human source.
gels may be warranted in day-care centers and other sites where the risk of person-to-person transmis- sion is particularly high. As noted above, washing environmental surfaces with solutions containing diluted household bleach (1:10) greatly reduces the counts of bacteria and viruses that are implicated in acute gastrointestinal illness, but this type of appli- cation is not always practicable.
TR. Typhoid fever in New York City, 1980 through 1990. Arch Intern Med 1994;154: 1713-8. n engl j med 351;23 www.nejm.org december 2, 2004
medical progress 2425
Point source outbreaks of Campylobacter jejuni infection — are they more common than we think and what might cause them? Epidemiol Infect 2003;130:367-75.
Sexual transmission of typhoid fever: a mul- tistate outbreak among men who have sex with men. Clin Infect Dis 2003;37:141-4.
40. Reller ME, Olsen SJ, Kressel AB, et al.
erythromycin- and ciprofloxacin-resistant Campylobacter jejuni subsp. jejuni from 1999 to 2001 in men who have sex with men, Quebec, Canada. Clin Infect Dis 2003; 37:131-6.
39. Gaudreau C, Michaud S. Cluster of
Neonatal sepsis by Campylobacter jejuni: genetically proven transmission from a household puppy. Clin Infect Dis 2001;32: E97-E99.
38. Wolfs TF, Duim B, Geelen SP, et al.
Huisman J. Epidemiological investigations on Campylobacter jejuni in households with a primary infection. J Hyg (Lond) 1984;93: 325-32.
37. Oosterom J, den Uyl CH, Banffer JR,
36. Gillespie IA, O’Brien SJ, Adak GK, et al.
who have sex with men — San Francisco, California, 2000–2001. MMWR Morb Mor- tal Wkly Rep 2001;50:922-6.
landson AL. Outbreaks of Campylobacter enteritis in two extended families: evidence for person-to-person transmission. J Pediatr 1981;98:254-7.
35. Blaser MJ, Waldman RJ, Barrett T, Er-
34. Skirrow MB. Campylobacter. Lancet 1990;336:921-3.
al. Factors associated with increased and decreased risk of Campylobacter infection: a prospective case-control study in Norway. Am J Epidemiol 2003;158:234-42.
33. Kapperud G, Espeland G, Wahl E, et
JG, Hutwagner L, Griffin PM. Escherichia coli O157:H7 diarrhea in the United States: clinical and epidemiologic features. Ann In- tern Med 1997;126:505-13.
32. Slutsker L, Ries AA, Greene KD, Wells
lard RA, Allen JR. Campylobacter enteritis in the United States: a multicenter study. Ann Intern Med 1983;98:360-5.
31. Blaser MJ, Wells JG, Feldman RA, Pol-
41. Shigella sonnei outbreak among men
42. Bell BP, Goldoft M, Griffin PM, et al.
30. Pickering LK, Bartlett AV, Woodward
Peterson LR, Gerding DN. Recurrences of Clostridium difficile diarrhea not caused by the original infecting organism. J Infect Dis 1989;159:340-3.
53. Hedlund KO, Rubilar-Abreu E, Svens-
demiology of enteric caliciviruses from humans: a reassessment using new diag- nostics. J Infect Dis 2000;181:Suppl 2:S254- S261.
52. Glass RI, Noel J, Ando T, et al. The epi-
RV. Surveillance for sporadic foodborne dis- ease in the 21st century: the FoodNet per- spective. Clin Infect Dis 2004;38:Suppl 3: S115-S120.
51. Allos BM, Moore MR, Griffin PM, Tauxe
ness in the home: a study of 25,000 illnesses in a group of Cleveland families. Cleveland: Press of Western Reserve University, 1964.
50. Dingle JH, Badger GF, Jordan WS Jr. Ill-
breaks of diarrhea associated with Clostridi- um difficile and its toxin in day-care centers: evidence of person-to-person spread. J Pedi- atr 1983;102:376-82.
49. Kim K, DuPont HL, Pickering LK. Out-
48. Johnson S, Adelmann A, Clabots CR,
A multistate outbreak of Escherichia coli O157:H7-associated bloody diarrhea and hemolytic uremic syndrome from hamburg- ers: the Washington experience. JAMA 1994;272:1349-53.
47. Bartlett JG. Antibiotic-associated diar- rhea. N Engl J Med 2002;346:334-9.
of hemolytic uremic syndrome after sporad- ic Escherichia coli O157:H7 infection: re- sults of a Canadian collaborative study. J Pe- diatr 1998;132:777-82.
46. Rowe PC, Orrbine E, Lior H, et al. Risk
Hemolytic-uremic syndrome during an out- break of Escherichia coli O157:H7 infec- tions in institutions for mentally retarded persons: clinical and epidemiologic obser- vations. J Pediatr 1990;116:544-51.
45. Pavia AT, Nichols CR, Green DP, et al.
Wells JG, Gunn RA, Blake PA. Hemolytic uremic syndrome and diarrhea associated with Escherichia coli O157:H7 in a day care center. J Pediatr 1986;109:287-91.
44. Spika JS, Parsons JE, Nordenberg D,
Transmission of Escherichia coli O157:H7 infection in Minnesota child day-care facili- ties. JAMA 1993;269:883-8.
43. Belongia EA, Osterholm MT, Soler JT, Ammend DA, Braun JE, MacDonald KL.
WE. Acute infectious diarrhea among chil- dren in day care: epidemiology and control. Rev Infect Dis 1986;8:539-47.
Drummond WK, Pavia AT. Characterization of endemic strains of Shigella sonnei by use of plasmid DNA analysis and pulsed-field gel electrophoresis to detect patterns of transmission. J Infect Dis 1997;175:864-70.
7. Leavitt JW. Typhoid Mary: captive to the
TE, DuPont HL, Dawkins AT, Snyder MJ. Ty- phoid fever: pathogenesis and immunolog- ic control. 2. N Engl J Med 1970;283:739- 46.
17. Ager EA, Top FH Sr. Salmonellosis. In:
Hughes TP, Blaser MJ. Experimental Cam- pylobacter jejuni infection in humans. J In- fect Dis 1988;157:472-9.
16. Black RE, Levine MM, Clements ML,
30.
Bradley DJ. Infecting dose and severity of ty- phoid: analysis of volunteer data and exami- nation of the influence of the definition of illness used. Epidemiol Infect 1995;115:23-
15. Glynn JR, Hornick RB, Levine MM,
immunologic control. N Engl J Med 1970; 283:686-91.
14. Idem. Typhoid fever: pathogenesis and
13. Hornick RB, Greisman SE, Woodward
18. Buchwald DS, Blaser MJ. A review of hu-
berg HS, Wood WBJ. Microbiology. New York: Harper & Row, 1967.
12. Davis BD, Dulbecco R, Eisen HN, Gins-
A massive epidemic of multidrug-resistant typhoid fever in Tajikistan associated with consumption of municipal water. J Infect Dis 1999;179:1416-22.