You are on page 1of 18

Shigella is a Very, Very Nasty Bacteria

By Bill Marler on October 20, 2015


Introduction to Shigella

Shigella is a species of enteric bacteria that causes disease in humans and other primates. [16, 20] The
disease caused by the ingestion of Shigella bacteria is referred to as shigellosis, which is most typically
associated with diarrhea and other gastrointestinal symptoms. [11, 16] “Shigella infection is the third
most common cause of bacterial gastroenteritis in the United States, after Campylobacter infection
and Salmonella infection and ahead of E. coliO157 infection.” [19]

The global burden of shigellosis has been estimated at 165 million cases per year, of which 163 million
are in developing countries. [23] More than one million deaths occur in the developing world yearly due
to Shigella infection. [23, 29] By one estimate, Shigella infections are responsible for 300,000 illnesses
and 600 deaths per year in the United States. [25] By another estimate, each year 450,000 Americans
are infected with Shigella, causing 6,200 hospitalizations and 70 deaths. [27]

In general, Shigella is one of the most communicable and severe forms of the bacterial-induced
diarrheas. [18] No group of individuals is immune to shigellosis, but certain individuals are at increased
risk. [16] Small children acquire Shigella at the highest rate, and [24, 28] persons infected with HIV
experience shigellosis much more commonly than other individuals. [4]
Shigella is easily spread person-to-person because of its relatively tiny (compared to other bacteria)
infectious dose. [16, 23] Infection can occur after ingestion of fewer than 100 bacteria. [1, 16, 17]
Another reason Shigella so easily cause infection is because the bacteria thrive in the human intestine
and are commonly spread both by person-to-person contact and through the contamination of food.
[11, 22, 32]

The Discovery and Naming of Shigella

The several types of Shigella bacteria have been named after the lead workers who discovered each
one. [11, 16, 20] The first bacterium to be discovered, Shigella dysentariae,was named after Kiyoshi
Shiga, a Japanese scientist who discovered it in 1896 while investigating a large epidemic of dysentery in
Japan. [22, 37] The bacterium was also referred to more generally as the dysentery bacillus (the term
“bacillus” referring to a genus of Gram-positive, rod-shaped bacteria of which Shigella is a member). [37]

In a summary published annually, the CDC provides an overview of the classification of various types
(species) of Shigella bacteria, as follows:

There are 4 major subgroups of Shigella, designated A, B, C and D, and 44 recognized serotypes.
Subgroups A, B, C and D have historically been treated as species: subgroup A for Shigella dysenteriae;
subgroup B for Shigella flexneri; subgroup C for Shigella boydii and subgroup D for Shigella sonnei. These
subgroups and serotypes are differentiated from one another by their biochemical traits (ability to
ferment D-mannitol) and antigenic properties. The most recently recognized serotype belongs to
subgroup C (S. boydii). [12]

S. sonnei, also known as Group D Shigella, accounts for over two-thirds of shigellosis in the United
States. Shigella flexneri, or group B Shigella, accounts for almost all the rest. [11, 19] More specifically,
according to one recent study, “From 1989 to 2002, S. flexneri accounted for 18.4% of Shigella isolates
submitted to CDC. [4] From 1973 to 1999, only 49 S. flexneri-associated outbreaks of foodborne disease
were reported.” [32] In contrast, in developing countries, S. flexneri is the most predominant cause of
shigellosis, but S. dysinteriae type 1 is still a frequent cause of epidemic throughout the developing
world. [1, 16, 23, 37]

The Incidence of Shigella Infection

The number of shigellosis cases reported annually to the Centers for Disease Control and Prevention
(CDC) has varied over the past several years, from more than 17,000 during 1978–2003, to an all-time
low of 14,000 in 2004, to almost 20,000 in 2007. [11, 19] But a majority of cases go undiagnosed or
unreported. [12, 16, 37] In one study done in Oregon, 430 confirmed Shigella cases from July 1995
through June 1998 were examined. [30] Among the several findings about those most likely to fall ill was
the following:

Of 430 isolates, 410 were identified to the species level: Shigella sonnei accounted for 55% of isolates,
and Shigella flexneri, for 40%. The overall annual incidence of shigellosis was 4.4 cases per 100,000
population. Children aged [less than] 5 years (annual incidence, 19.6 cases per 100,000 population) and
Hispanics (annual incidence, 28.4 cases per 100,000 population) were at highest risk. [30]

The CDC estimates that 450,000 total cases of shigellosis occur in the U.S. every year. [4, 11, 21]
Shigellosis is also characterized by seasonality, with the largest percentage of reported cases occurring
between July and October, and the smallest proportion occurring in January, February, and March. [19]
Sporadic (or non-outbreak) infections account for the majority of cases and, in general, the exact means
by which persons are infected (risk factors) are not yet well documented or understood. [21, 36]

Shigella is an especially common cause of disease among young children, in large part because it is
difficult to control the spread of the bacteria in daycare settings. [16, 28] The symptoms of shigellosis
vary so widely that children shedding Shigella in their stool may exhibit no symptoms of infection. A
person infected with Shigella can be asymptomatic (show no symptoms of illness), suffer from moderate
to severe diarrhea, or suffer complications up to and including death. [11, 17, 26]

More on Incidence Rates and How Shigella is Transmitted

As previously noted, Shigella species are transmitted by the fecal-oral route, and most infections are
transmitted from person to person, reflecting the low infectious dose. [19] As also noted, as few as
ten Shigella bacteria can result in clinical infection. [17] Where persons who are infected may be
present, the risk of transmission and infection increases with poor hand hygiene, ingestion of
contaminated food or water, inadequate sanitation and toileting, overcrowding, and sexual contact. [4,
14, 21, 24, 28] Shigella bacteria are present in the stools of infected persons while they are sick and for
up to a week or two afterwards. [11, 16, 17] It is estimated that up 80% of all infection is the result of
person-to-person transmission. [17]

Because of its quite common person-to-person spread, shigellosis has long been associated with
outbreaks in daycare centers, nursing homes, institutional settings (like prison), and cruise ships. [11, 14,
17, 23, 24] Explaining the significance of daycare centers as a source of Shigella infection, one well-
respected study explains as follows:

High shigellosis rates in children are attributable to several factors. Young children are unable to practice
good personal hygiene and have not yet acquired immunity to S. sonnei. The infectious dose is as low as
10–200 organisms, and person-to-person transmission is highly effective. Day-care centers play an
important role in the person-to-person spread of shigellosis and its subsequent dissemination in
communities. Inadequate hand washing, diapering practices, and fecal contamination of water-play
areas, such as kiddie pools, have been associated with S. sonnei transmission in day-care centers. [19]

Several studies have demonstrated an increased frequency of shigellosis cases in young adult men
residing in urban settings who have little, if any, exposure to these traditionally recognized risk groups.
[4, 36] Although some of these studies indicated that sex between men can be a risk-factor, most of
these studies occurred before the HIV epidemic. [4, 23]

Shigella infections also may be acquired from eating contaminated food. A study published in 2010
estimated more than one-third of U.S. shigellosis cases annually might be caused by the consumption of
contaminated food. [21] In the United States, incidence of foodborne illness is documented through
FoodNet, a reporting system used by public health agencies that captures foodborne illness in over 13%
of the population. [8, 9] Of the 10 pathogens tracked by FoodNet, Salmonella, Campylobacter,
and Shigella are responsible for most cases of foodborne illness. [27] An estimated 20% of the total
number of cases of shigellosis involve food as the vehicle of transmission. [27]

In one oft-cited study summarizing food-related illness and death in the United States, the following
synopsis is set forth at the end, summarizing Shigella.
Reported cases: Outbreak-related cases based on reports to CDC, 1983-1992. Passive surveillance
estimate based on average number of cases reported annually to CDC, 1992-1997. Active surveillance
estimate based on extrapolation of average 1996-1997 FoodNet rate to the 1997 U.S. population.

Total cases: Because Shigella frequently causes bloody diarrhea, total cases assumed to be 20 times the
number of reported cases, based on similarity to E. coli O157:H7.

Hospitalization rate: Based on hospitalization rate for culture-confirmed cases reported to FoodNet,

Case-fatality rate: Average case-fatality rate among cases reported to FoodNet, 1996-1997 (23,24).

Percent foodborne: Assumed to be 20%. Although most cases are due to person-to-person transmission
(60),foodborne outbreaks are responsible for a substantial number of cases [27]

According to the CDC, Shigella is the third most common pathogen transmitted through food. In
FoodNet surveillance areas in 2008, the rate of Shigella was 6.6 per 100,000 population, exceeded only
by Salmonella (15.2/100,000) and Campylobacter (12.7/100,000). [10] During 2006, public health
officials reported a total of 1,270 foodborne-related outbreaks from 48 states in the U.S. [9]
Although Shigella was responsible for only 10 (1%) of those outbreaks, 183 confirmed cases of
shigellosis were nonetheless reported. [9] This reporting rate contrasts with an average of 659 cases
annually in the previous five years, making it potentially an aberration or outlier.

Shigella is also responsible for a substantial portion of foodborne outbreaks on cruise ships. [16, 34] In a
review of cruise ship outbreaks worldwide over several years, 16% of outbreaks were attributed
to Shigella, affecting over 2,000 passengers. [34] Sanitation violations related to food handling and
communicable disease have decreased substantially, however, over the past 15 years. [14]

Symptoms of Shigella Infection

Most people who are infected with Shigella develop diarrhea, fever, and stomach cramps after being
exposed to the bacteria. [11, 16, 26] Symptoms may start 12 to 96 hours after exposure, usually within 1
to 3 days. [1, 16] Diarrhea may range from mild to severe, and it usually contains mucus. [16] When
more severe, the diarrhea is bloody 25% to 50% of the time. [1, 16, 22] Rectal spasms, which are
technically referred to as “tenesmus,” are common. [16]

Shigellosis usually resolves in 5 to 7 days. [1, 11, 26] A severe infection with high fever may be
associated with seizures in children less than two years old. [16, 19] Some persons who are infected may
have no symptoms at all, but may still pass the Shigella bacteria to others. [11, 17]

Persons with shigellosis in the U.S. do not often require hospitalization, although the hospitalization rate
has been estimated to be in excess of 50,000 per year. [27] Predictably, the hospitalization rate tends to
be highest among older individuals. [9, 10, 16] Those who are immune-compromised, like persons
infected with HIV, are also more likely to face hospitalization because of the risk of complications. [4]

The relationship between HIV infection and the subsequent risk for shigellosis has yet to be conclusively
evaluated, although it is known that “HIV-associated immunodeficiency leads to more severe clinical
manifestations of Shigella infection.” [23] Moreover, persons infected with HIV “may develop persistent
or recurrent intestinal Shigella infections, even in the presence of adequate antimicrobial therapy. They
also face an increased risk of Shigella bacteraemia, which can be recurrent, severe or even fatal.” [23]

What are the serious and long-term risks of Shigella infection?

Persons with diarrhea caused by S. sonnei in particular usually recover completely, although it may be
several months before their bowel habits are entirely normal. [1, 11, 26] About 2% of persons who are
infected with S. flexneri later develop pains in their joints, irritation of the eyes, and painful urination—
something typically diagnosed as Reiter’s Syndrome. [1, 6]

Reiter’s syndrome is more generally referred to as reactive arthritis, a complication that accompanies
other kinds of bacterial infections as well. [27, 37] This complication occurs because the immune system,
intending to fight Shigella, attacks the body instead. [6, 31] Reactive arthritis is most common in
persons with the HLA-B27 gene. [31] (About 80% of people with reactive arthritis have the HLA-B27
gene. Only 6% of people who do not have the syndrome have the HLA-B27 gene.) Reactive arthritis can
last for months or years and may be difficult to treat. [6]

Once someone has suffered a Shigella infection, a certain level of immunity develops, meaning that the
person is not likely to get infected with that specific type again for at least several years. [16] This
temporary immunity does not, however, protect against other types of Shigella. [16, 29]

Shigella bacteria multiply in the human intestinal tract and invade the cells, which results in much tissue
destruction. [29] Many strains produce a toxin called Shiga toxin, which is very potent and destructive.
[16, 22] Shiga toxin is very similar to the verotoxin of E. coliO157:H7. Complications of shigellosis include
severe dehydration, seizures in small children, rectal bleeding, and invasion of the blood stream by the
bacteria (bacteremia or sepsis). [1, 11, 16, 26] In some cases, the bacteria that cause shigellosis may also
cause inflammation of the lining of the rectum (proctitis) or rectal prolapse. [26]

In rare cases (but more common in S. dysenteriae infection), there can also be a deadly complication
called “toxic megacolon.” [1, 26] This rare complication occurs when the colon becomes paralyzed,
preventing bowel movements or passing gas. [16, 26] Signs and symptoms include abdominal pain and
swelling, fever, weakness, and disorientation. [26] Untreated, the colon may rupture and cause
peritonitis, a life-threatening condition requiring emergency surgery. [26]

The other relatively rare complication that can occur with a Shigella infection is the development of
hemolytic uremic syndrome. This rare complication is more commonly caused by E. coli O157:H7, and it
can lead to a low red blood cell count (hemolytic anemia), low platelet count (thrombocytopenia), and
acute kidney failure. [26, 37] It is more common to develop HUS after being infected with S. dysenteriae.

Diagnosis and Treatment of Shigella Infections

Because the symptoms of a Shigella infection are consistent with a fairly large number of potential
illnesses, including most foodborne infections, a diagnosis must be confirmed by a laboratory test. [5,
11, 26] First a stool sample must be obtained from the potentially infected person, and then the sample
is placed on a medium to encourage the growth of bacteria. If and when there is growth, the bacteria
are identified, usually by looking at the growth under a microscope. [20, 26] The laboratory can also do
special tests to tell which species of Shigella the person has, and which antibiotics would be best to treat
the infection. [16, 22, 30] Antibiotic-sensitivity tests are important because Shigella is often resistant to
multiple antibiotics. [16, 30]

More advanced testing and surveillance methods, such as plasmid profiling and chromosomal
fingerprinting, can also be used. [11, 20, 29] So-called “genetic fingerprinting” of the bacterial isolate,
using pulsed-field gel electrophoresis (PFGE) is a molecular technique that can help to
characterize Shigella isolates, whether obtained from human or food samples. [11, 27] Taken together,
all of these tests can assist public health officials in determining whether cases (confirmed infections)
are isolated or associated with common-source outbreaks. [19, 20, 27]

Efforts to identify outbreaks of foodborne illness—whether caused by Shigella or other pathogens—are

important to preventing the secondary spread of infection, especially with bacteria as highly
communicable as Shigella. [1, 11, 21] One major advance in these efforts was the creation of FoodNet,
an active surveillance system for foodborne illness. As described by the CDC,

FoodNet workers regularly contact more than 300 laboratories for confirmed cases of foodborne
infections in 10 states encompassing a population of more than 44 million persons. In addition to
monitoring the number of Shigella infections, investigators monitor laboratory techniques for isolation
of bacteria, perform studies of ill persons to determine exposures associated with illness, and administer
questionnaires to people living in FoodNet sites to better understand trends in the eating habits of
Americans. [11]

Although shigellosis is usually a self-limited illness, antibiotics can shorten the course, and in the most
serious cases, might be life-saving. [1, 16, 22] Historically, the antibiotics commonly used for treatment
of bacterial infections, like those caused by Shigella, are ampicillin, trimethoprim/sulfamethoxazole
(TMP-SMZ, also known as Bactrim or Septra), or ceftriaxone (Rocephin). [1, 11, 26] Ciprofloxacin is also
used commonly to treat adults who are infected. [11, 26, 30].

Unfortunately, Shigella bacteria have become resistant to one or more of these antibiotics. [16, 30] This
means some antibiotics might not be effective for treatment, and that using (or overusing) antibiotics to
treat shigellosis can sometimes make the bacteria more resistant. [30] As noted in one recent study,

Of 369 isolates tested, 59% were resistant to TMP-SMZ, 63% were resistant to ampicillin, 1% were
resistant to cefixime, and 0.3% were resistant to nalidixic acid; none of the isolates were resistant to
ciprofloxacin. Thirteen percent of the isolates had multidrug resistance to ampicillin, chloramphenicol,
streptomycin, sulfisoxazole, and tetracycline. Infections due to multidrug-resistant shigellae are endemic
in Oregon. [30]

This study therefore suggests that “[n]either ampicillin nor TMP-SMZ should be considered appropriate
empirical therapy for shigellosis any longer; when antibiotics are indicated, a quinolone or cefixime
should be used.” [30]

More on Antimicrobial Resistance in Bacteria

Antimicrobial resistance in bacteria is an emerging and increasing threat to human health. [2, 3]
Physicians are increasingly aware that antimicrobial resistance is increasing in foodborne pathogens and
that, as a result, patients who are prescribed antibiotics are at increased risk for acquiring antimicrobial-
resistant foodborne infections. [3, 29] Indeed, “increased frequency of treatment failures for acute
illness and increased severity of infection may be manifested by prolonged duration of illness, increased
frequency of bloodstream infections, increased hospitalization or increased mortality.” [3]

The use of antimicrobial agents in the feed of food animals is estimated by the FDA to be over 100
million pounds per year. [3] It is estimated that 36% to 70% of all antibiotics produced in the United
States are used in food animal feed or in prophylactic treatment to prevent animal disease. [2, 3] In
2002, the Minnesota Medical Association published an article by David Wallinga, M.D., M.P.H. who

According to the [Union of Concerned Scientists], 70 percent of all the antimicrobials used in the United
States for all purposes—or about 24.6 million pounds annually—are fed to poultry, swine, and beef
cattle for nontherapeutic purposes, in the absence of disease. Over half are “medically important”
antimicrobials; identical or so closely related to human medicines that resistance to the animal drug can
confer resistance to the similar human drug. Penicillin, tetracycline, macrolides, streptogramins, and
sulfonamides are prominent examples. [33]

Based on recent research, it is now recommended that doctors avoid the use of commonly prescribed
antibiotics, like tetracycline and ampicillin, in favor of drugs for which Shigella has not shown resistance,
such as ciprofloxacin. [29] European countries have reduced the use of antibiotics in animal feed and
have seen a corresponding reduction in antibiotic-resistant illnesses in humans. [3]

The Economic Impact of Shigella Infections

The USDA Economic Research Service (ERS) published its first comprehensive cost estimates for sixteen
foodborne bacterial pathogens in 1989. [32] Five years later, it was estimated that the medical costs and
productivity losses that Shigella infections caused each year ran from $907 million to over $1 billion,
based on an estimate of 2.1 million cases and between 120-360 deaths. [13] The average length of a
related hospital stay was 4.6 days, with the cost (based on a 1990 average cost per day of $687) was
$16,888. [13]

Using a different kind of economic analysis, this same 1996 study estimated that the annual cost
of Shigella infections was $63 million, while the average cost of each confirmed and treated infection
was $390; however, these estimates are based on significantly lower (and outdated) incidence and
death rates. [13] Most recent estimates are all much higher. For example, a study published in 2010
estimated the cost per case (in 2009 dollars) for a treated Shigella infection to be $7,092, with an
estimate of 96,686 cases and 1,227 deaths per year, and a total cost to U.S. residents of $686 million.

Real Life Impacts of Shigella Infection

Because the illnesses caused by the ingestion of Shigella bacteria range from mild to severe, the real life
impacts of Shigella infection vary from person to person.

While anyone can become ill with Shigella infection, very young children, the elderly, and persons with
compromised immune systems are most likely to develop severe illness.

 About 2% of persons who are infected with one type of Shigella, Shigella flexneri, later develop
pains in their joints, irritation of the eyes, and painful urination. This is called post-infectious
arthritis, or reactive arthritis. The arthritis can last for months or years, and can lead to chronic
arthritis. Post-infectious arthritis is caused by a reaction to Shigella infection that happens only
in people who are genetically predisposed to it. [11]

 An unknown percentage of patients with Shigella infections develop digestive disorders,

including irritable bowel syndrome.

Although most patients with Shigella infections recover within a few months, some continue to
experience complications for years.

How to Prevent Shigella Infection

According to the World Health Organization, “Despite the continuing challenge posed by Shigella, there
is room for optimism as advances in biotechnology have enabled the development of a new generation
of candidate vaccines that shows great promise for the prevention of Shigella disease.” [23] But such a
vaccine has yet to be perfected. Thus, in the meantime, preventing infection is the best approach, and
that means implementing proper sanitation measures. [1, 14] Indeed, as noted in one authoritative text
summarizing the research:

A safe water supply is important for the control of shigellosis and is probably the single most important
factor in areas with substandard sanitation facilities. Chlorination is another factor important in
decreasing the incidence of all enteric bacterial infections. Of critical importance to the establishment of
a safe water supply is the general level of sanitation in the area and the establishment of an effective
sewage disposal system. [16]

As previously noted, it takes but a few—far less than 100—Shigella bacteria to cause infection. [17]
Moreover, a person can be infectious even if there are no symptoms, either because he remained
asymptomatic (never exhibited symptoms of shigellosis), or because the person continued to shed the
bacteria in his stool for a week or two after recovering. [11, 16, 17]

The spread of Shigella from an infected person to other persons can be avoided by frequent and careful
hand-washing with soap and hot water. [1, 14, 15] Hand-washing among children should be frequent
and supervised by an adult in daycare centers and homes with children who have not been fully toilet
trained. [24, 28]

If a child in diapers has shigellosis, everyone who changes the child’s diapers should be sure the diapers
are disposed of properly in a closed-lid garbage can, and should wash his or her hands and the child’s
hands carefully with soap and warm water immediately after the diaper has been changed. [15, 39]
After use, the diaper changing area should be wiped down with a disinfectant such as diluted household
bleach, Lysol, or bactericidal wipes. [15, 24, 39] When possible, young children with a Shigella infection
who are still in diapers should not be in contact with uninfected children. [1, 11, 15]

Basic food safety precautions and disinfection of drinking water should prevent Shigellabacteria from
contaminating food and water. [11, 15, 39] Nonetheless, it should go without saying that people with
shigellosis should not prepare food or drinks for others until they have been confirmed (by a stool
culture) to no longer be shedding Shigella bacteria in their stool. [1, 15] At swimming beaches, there
should be bathrooms and handwashing stations near the swimming area to help keep the water from
becoming contaminated. [14, 29] Daycare centers should not provide water play areas. [24]
Simple precautions taken while traveling to the developing world can prevent shigellosis. [1, 39] Drink
only treated or boiled water, and eat only cooked hot foods or fruits you peel yourself. [1, 11] The same
precautions prevent other types of traveler’s diarrhea. [39]

Shigella Infections in Household Contacts of Pediatric Shigellosis Patients in Rural Bangladesh

On This Page

 Methods

 Results

 Discussion

 Cite This Article


 Table 1

 Table 2

 Table 3

 Table 4

 Table 5


 PDF[440 KB - 8 pgs]

 RIS[TXT - 2 KB]

Christine Marie George , Shahnawaz Ahmed, Kaisar A. Talukder, Ishrat J. Azmi, Jamie Perin, R.
Bradley Sack, David A Sack, O. Colin Stine, Lauren Oldja, Mohammad Shahnaij, Subhra Chakraborty,
Tahmina Parvin, Sazzadul Islam Bhuyian, Edward Bouwer, Xiaotong Zhang, Trisheeta N. Hasan,
Sharmin J. Luna, Fatema Akter, and Abu S.G. Faruque

Author affiliations: Johns Hopkins University, Baltimore, Maryland, USA (C.M. George, J. Perin, R.B. Sack,
D.A. Sack, L. Oldja, S. Chakraborty, E. Bouwer, X. Zhang); icddb,b, Bangladesh (S. Ahmed, K.A. Talukder,
I.J. Azmi, M. Shahnaij, T. Parvin, S.I. Bhuyian, T.N. Hasan, S.J. Luna, F. Akter, A.S.G. Faruque); University
of Maryland School of Medicine, Baltimore (O.C. Stine)

Cite This Article


To examine rates of Shigella infections in household contacts of pediatric shigellosis patients, we

followed contacts and controls prospectively for 1 week after the index patient obtained care.
Household contacts of patients were 44 times more likely to develop a Shigella infection than were
control contacts (odds ratio 44.7, 95% CI 5.5–361.6); 29 (94%) household contacts of shigellosis patients
were infected with the same species and serotype as the index patient’s. Pulsed-field gel electrophoresis
showed that 14 (88%) of 16 with infected contacts had strains that were indistinguishable from or
closely related to the index patient’s strain. Latrine area fly counts were higher in patient households
compared with control households, and 2 patient household water samples were positive for Shigella.
We show high susceptibility of household contacts of shigellosis patients to Shigella infections and found
environmental risk factors to be targeted in future interventions.

In South Asia and Africa, an estimated 88.5 million diarrhea episodes are attributed to Shigella infections
annually (1). Shigellosis occurs most often in children <5 years of age (2,3). Two recent multicountry
studies found that Bangladesh has the highest rates of shigellosis (4,5). In the recent Global Enteric
Multicenter Study (GEMS) conducted at a study site in Mirzapur, Bangladesh, Shigella was the third
leading cause of moderate-to-severe diarrhea in children 12–23 months of age and the second leading
cause of moderate-to-severe diarrhea in children 24–59 months of age (5). In addition, hospital-based
surveillance of Shigella in Bangladesh found that patients from rural health facilities have higher rates
of Shigella isolates than patients from urban facilities (e.g., 3% in urban Dhaka vs. 12% in rural Mirzapur
during 2011) (6). Previous studies have identified risk factors for developing shigellosis, such as age (7–
9), high fly counts (10–12), contaminated food (13,14), and recent overnight travel (8). Furthermore, a
recent study in rural Bangladesh found that 10% of tube wells sampled had detectable Shigella (15).

Household studies have indicated that family members of shigellosis patients are at much higher risk for
developing a Shigella infection than the general population (13–19 cases/100 shigellosis patient
households vs. 7 cases/100 control households) (7,8). However, little research has been done to identify
clinical and environmental transmission routes for Shigella infection in this susceptible population.

Shigella includes 4 species and numerous serotypes: S. flexneri (17 serotypes), S. dysenteriae (16
serotypes), S. boydii (20 serotypes), and S. sonnei (1 serotype) (16). A study in Wisconsin found that
isolates from family members of index shigellosis patients were always the same serotype as the index
patient’s (7). In contrast, studies in rural (8) and urban (9) Bangladesh found that 75% and 72%,
respectively, of infected household contacts of shigellosis patients excreted serotypes different from the
index patient’s serotype. These studies suggest that Shigella infections in Bangladesh are attributable to
both secondary transmission and external infecting sources. To examine the rate of Shigella infection
within households of shigellosis patients and to investigate risk factors for infection, we prospectively
observed a cohort of household contacts of pediatric shigellosis patients and community controls in
rural Mirzapur, Bangladesh.


This study was conducted in Mirzapur, a subdistrict of Bangladesh’s Tangail district, at a field site of the
icddr,b. Mirzapur is the Bangladesh site of the GEMS Demographic Surveillance System. We received
ethical approval for this study from the icddr,b ethical review committee and an exemption from the
Institutional Review Board at the Johns Hopkins Bloomberg School of Public Health (Baltimore, MD,
USA). Written informed consent was obtained from all study participants or their guardians.

A cohort of household contacts of index shigellosis patients and matched community controls was
followed prospectively for 7 days after each index patient visited the Kumudini Women’s Medical
College and Hospital in Mirzapur for health care. Our sample size was determined by the number of
pediatric shigellosis patients that could be recruited during October 2013–July 2014. Patients with
suspected pediatric shigellosis were identified as children <5 years of age with dysentery, which was
defined as having >1 stool containing blood during the previous 24 hours, as reported by a guardian or
observed by research personnel. Community controls were matched to index shigellosis patients on the
basis of age (within 3 months) and village location and were randomly selected by using the GEMS
Demographic Surveillance System. Stool samples were collected from patients with suspected shigellosis
and from community controls at time of enrollment in the study and were immediately sent in a cooler
to the icddr,b Enteric and Food Microbiology Laboratory in Dhaka, Bangladesh, for bacterial culture
analysis to detect Shigella.

Households of suspected shigellosis patients whose samples were found to be negative for Shigella by
culture were excluded from the study. All index patients enrolled in the study received ciprofloxacin as
part of their standard course of care at the Kumudini Hospital.

After enrollment of pediatric shigellosis patients, we recruited household contacts of shigellosis patients
and of community controls. Household contacts were defined as persons sharing the same cooking pot
as the index shigellosis patient or control during the previous 3 days. To be eligible for the study,
household contacts had to report that they would be present in the household of the patient or control
for the next 7 days and be present for most household visits. Household contacts were followed
prospectively for clinical and environmental surveillance by conducting household visits at days 1, 3, 5,
and 7 after the initial visit of the index shigellosis patient Kumudini Hospital.

For clinical surveillance, household contacts were asked to report whether, during the previous 48
hours, they had diarrhea (>3 loose stools during a 24-hour period), dysentery (blood in stool observed
by caregiver or research personnel), or vomiting. A stool sample was collected from enrolled household
contacts at every visit. In addition, at each household visit, a questionnaire was administered to enrolled
contacts to collect information on previously identified risk factors for Shigella infection. For
environmental surveillance, on days 1 and 5, a water sample was collected directly from the household’s
primary drinking water source, and a second sample was collected from drinking water stored in the
home for immediate consumption. The water samples were tested for Shigella spp. by bacterial culture
and PCR. Trained field assistants also conducted fly counts by using a Scudder grill at all surveillance
visits over a period of 30 minutes at the kitchen and latrine area of each household, according to
previously published methods (12). Weekly fly counts for each household were the total number of flies
observed during surveillance visits.

All stool and water samples collected were analyzed at the icddr,b Enteric and Food Microbiology
Laboratory in Dhaka. The laboratory received no information identifying whether samples were from
patient or control households. For isolation of Shigella, stool and water samples were cultured on
MacConkey and Shigella-Salmonella agar media, and Shigella was isolated and serotyped by using
standard microbiologic and biochemical methods described previously (4). In brief, water samples of
1,000 mL were filtered through 0.22 μm pore-size filters. The filter paper was enriched in 25 mL gram-
negative broth at 37°C overnight and then analyzed by culture. Template DNA was prepared from the
enriched broth and tested for the ipaH gene, according to previously published methods (16).

To determine genetic relatedness of Shigella strains isolated within households, pulsed-field gel
electrophoresis (PFGE) was performed on all Shigella–positive water and stool samples, according to the
PulseNet protocol (17). Agarose–embedded genomic DNA of Shigella strains was digested by using XbaI,
and fragments were separated by using a CHEF-DR II apparatus (Bio-Rad, Hercules, CA, USA) (18).
Genetic relatedness was determined on the basis of previously published methods (19). To compare
strains within a single household, 4 categories of genetic relatedness were used: a) “indistinguishable”
(all fragments matched); b) “closely related” (1–3 fragments differed); c) “possibly related” (4–6
fragments differed); and d) “unrelated” (>7 fragments differed).

A Shigella-infected person was defined as a person with a stool sample positive for Shigella spp. by
culture. Various tests were used for household-level variables: a McNemar test for paired binary
variables; a Friedman test for clustered categorical variables with >2 levels; and a Wilcoxon signed-rank
test for paired continuous variables (Tables 1, 2). For individual-level variables, a Fisher exact test was
used for categorical variables and a Wilcoxon rank-sum test for continuous variables (Tables 1, 2).

Logistic regression was used to estimate the odds of developing a Shigella infection. Generalized
estimating equations were used to account for clustering within households and in patient–control pairs
and to estimate an odds ratio (OR) and approximate 95% CIs. Clusters in this analysis are the 27 patient–
control pairs. A bivariate analysis was performed in which index patient or control child status in the
household was used as the single predictor, and a binary outcome was used to determine whether
household members developed a Shigella infection. All analyses were performed by using SAS, version
9.3 (SAS Institute, Inc., Cary, NC, USA).


During October 2013–July 2014, a total of 27 shigellosis patients with 83 household contacts and 27
community controls with 77 household contacts were followed prospectively. Of the initial 61 suspected
shigellosis patients who were screened, 29 were excluded because cultures were negative for Shigella,
and 5 were excluded because the caregiver was too busy or uninterested in the study. Of the 33
community controls screened, 6 were excluded because they did not pass stool on visit 1. Of 88
shigellosis–patient household contacts and 81 control household contacts who were screened for
eligibility, 5 (6%) and 4 (5%), respectively, were excluded from the study because they did not pass stool
on visit 1.

Median age for household contacts was 27 years for patient households and 30 years for control
households (Table 1). Of the 83 household contacts in patient households, 48 (58%) were women,
compared with 47 (61%) of the 77 contacts in control households. Patient and control households did
not differ significantly by age, sex, number enrolled, or number of total contacts (Table 1). Among
household contacts of patients and control children, 52 (33%) were mothers, 31 (19%) fathers, 23 (14%)
brothers, 22 (14%) sisters, and 32 (20%) other relatives. Contact relationship to the patient or control
child did not differ significantly (p = 0.88).

During the surveillance period, patient contacts reported spending more time outside the home than
did control contacts (median time outside home during previous 48 hours, 2 hours vs. 1 hour; p = 0.09);
eating more food outside the home (61% vs. 47%; p = 0.08); and consuming more uncooked vegetables
and fruits (27% vs. 16%; p = 0.12). These differences were not statistically significant (Table 1). The rate
for acquiring stool samples was 97% for enrolled household contacts and did not differ significantly for
patient and control households (p = 0.15).

We found no significant differences between patient and control households in caregiver’s education
level, type of latrine or floors, or presence of soap at the household latrine area, a proxy measure for
handwashing with soap (Table 2). All households relied on tube wells as their primary drinking water
source. The latrine areas of patient households had significantly higher weekly fly counts compared with
those of control households (p = 0.001), but weekly kitchen fly counts did not differ significantly (p =
0.47). In patient households, 33% had concrete floors compared with 15% of control households; this
difference was not statistically significant (p = 0.13). Household water samples from 2 (7%) patient
households were positive by PCR for the ipaH gene of Shigella and were culture positive for non–type
1 S. dysenterae and S. sonnei during the surveillance period, compared with 1 PCR-positive (for
the ipaH gene) household water sample and 2 PCR-positive samples from household water sources in
control households.

Of the 27 patient households, 16 (59%) had >1 Shigella-infected contact during the 7-day surveillance
period, compared with 1 (4%) control household, in which an asymptomatic Shigella infection was
detected on day 1 of clinical surveillance (Table 3). In a bivariate model that used patient household as
the predictor, the odds of developing a Shigella infection were 44 times higher for patient contacts than
for control contacts (OR 44.7, 95% CI 5.5–361.6). The 16 patient households had 31 Shigella-infected
contacts, compared with 1 Shigella-infected contact in 1 control household (Table 4). Four (15%) patient
households had 6 contacts with symptomatic Shigella infection (i.e., having diarrhea, vomiting, or blood
in stool during the previous 48 hours). Shigellainfections for 13 (42%) of 31 Shigella-infected patient
contacts (asymptomatic and symptomatic) were first detected on day 1 of clinical surveillance; 6 (19%)
were detected on day 3, 5 (16%) on day 5, and 7 (23%) on day 7.

In patient households, a Shigella infection developed in 18 (51%) of 35 male contacts and in 13 (27%) of
48 female contacts during the surveillance period (p = 0.02). Five of the 6
symptomatic Shigella infections were in men, and 4 of the infections were first detected on day 1 or 3 of
surveillance. Difference in day of initial detection by sex was not significant (p = 0.53), but male contacts
spent significantly more time outside their homes during the surveillance period (p<0.0001) and
reported drinking significantly more water outside their homes (p<0.0001) than did female contacts.

During the surveillance period, 4 patient contacts reported using antibacterial drugs; 3 of those had a
symptomatic Shigella infection, 1 of whom was hospitalized for symptoms. This person tested positive
on visit 1. To estimate duration of shedding for patient household contacts, we observed the time
during which shedding occurred for 6 household contacts with a Shigella infection first detected on visit
2. Of these 6 contacts, 5 (83%) had detectable shedding for <2 days: 3 for 1 day, 2 for 2 days, and 1 for 3

Among the 31 patient household contacts in whom Shigella infection developed, 29 (94%) were infected
with the same species and serotype as the index patient in their household (Table 5). Twenty (65%) of
the 31 patient household contacts with detectable Shigella in stool by culture had S. flexneri (2 S.
flexneri 1b, 3 S. flexneri 1c, 12 S. flexneri 2a, and 3 S.flexneri 3a); 9 (29%) had S. sonnei; and 2 (6%) had S.
bodyii (1 S. boydii 7 and 1 S. boydii of an unknown serotype) (16–18). In the 16 patient households with
infected household contacts, 7 (44%) contacts had strains indistinguishable from those of the index
patient by PFGE analysis; 7 (44%) had closely related strains; and 2 (12%) had unrelated strains (Table 5).


We found that the odds of developing a Shigella infection were >44 times higher for contacts of
pediatric shigellosis patients than for control contacts (OR 44.7, 95% CI 5.5–361.6). We also observed
that 94% of patient household contacts had the same species and serotype as the index patient.
Consistent with this finding, PFGE analysis found that 88% of households with infected household
contacts had strains that were indistinguishable from or closely related to the index patient’s strain. In
contrast, 2 previous studies in Bangladesh found that only one quarter of household contacts had the
same species and serotype as the index patient in their home (8,9). Our finding suggests that a single
infectious pathogen is being spread in study households; however, whether the infections are caused by
a shared environmental source or secondary transmission from an infected household member is

We observed that most (59%) patient households had >1 household contact with a Shigella infection
over the 7-day surveillance period, and 25% of these households had contacts who had symptomatic
infections. In comparison, 1 (4%) control household contact had a Shigella infection, which was
asymptomatic. The proportion of shigellosis patient households with Shigella-infected contacts in our
study is higher than has been previously reported. In a study in Peru, 34% of households of index
shigellosis patients during the 1-week surveillance period had infected contacts (20). Another study
conducted in rural Bangladesh in 1974 found that 19% of members of household compounds (i.e.,
multiple households living together) developed Shigellainfection over a 10-day period after
identification of the index shigellosis patient, compared with 7% of members of control household
compounds (21). The reason our study found a higher rate of infection in household contacts of
shigellosis patients is unknown but may reflect differences in environmental risk factors or population

Despite no significant difference in toilet type, we found significantly higher fly counts in the latrine
areas of patient households than in control households during the surveillance period (Table 2). This
finding likely suggests that patient households had latrines with poorer sanitary conditions than did
control households. Similarly, a recent study conducted in the same site found a significant association
between seasonal fly densities and peaks in pediatric shigellosis patients (12). Furthermore, these
significant associations are consistent with the growing body of literature that implicates houseflies as
vectors of shigellosis. The 2 flies thought to be responsible for transmission of Shigella are Musca
domestica, because of its mobility, and M. sorbens, because it commonly breeds in human feces.
Previous studies in Bangladesh, Myanmar, Egypt, and the United States have detected Shigella in flies by
culture (11). An intervention study conducted on 2 military field bases in Israel found that baited fly
traps reduced fly counts by 64% and rates of shigellosis by 85% compared with fly counts and shigellosis
rates on the control base (10). These studies suggest that fly control could be a promising intervention
to reduce Shigella transmission in the study population. Future studies should more closely evaluate the
potential of flies to be a vector of Shigella in our study population by culturing flies and conducting fly
species identification.

Our study also found 2 (7%) patient households with stored water samples positive for Shigella. In 1
household, detectable S. sonnei by culture was found in stool from the patient and 2 household contacts
and in stored water. This finding suggests potential secondary contamination of stored water by a
household member, particularly because none of the corresponding source water samples had
detectable Shigella by culture. In another household, S. dysenteriae was found in stored water but was
not detected in any household members. A potential explanation for this finding is that
the Shigella came from the household tube well. A previous study in rural Bangladesh
identified Shigella by PCR in 10% of tube wells; that study also found that 40% of these tube wells were
contaminated with rotavirus, 10% with Vibrio spp., and 8% with pathogenic Escherichia coli (15). The
mostly likely source of Shigella in tube well water is fecal contamination from latrines, which are
commonly located near tube wells used for drinking in rural Bangladesh. Further research is needed to
evaluate whether groundwater is a major environmental transmission route for shigellosis and other
enteric infections in this population.

We observed that male patient contacts were twice as likely as female contacts to
develop Shigella infection during the surveillance period (51% vs. 27%); all but 1 symptomatic infections
in contacts were in men. The reason for this higher rate of infection among male household contacts is
unknown, but a possible explanation is that men may introduce the infection into the home. Future
studies should investigate the role of sex in susceptibility and transmission of Shigella infection.

Among patient households, 41% had >1 contact who developed an initial Shigella infection after day 1 of
surveillance. This finding suggests a potential opportunity to intervene in Shigella transmission in
households with shigellosis patients. An intervention study that promoted handwashing with soap in
Dhaka reduced the secondary infection rate for Shigella by 69% in the 10-day period after identification
of the index patient compared with a control group (22). Future studies should evaluate whether this
intervention would be effective in rural settings such as Mirzapur.

This study has several limitations. First, our small sample size limited our ability to detect significant
differences in environmental risk factors for Shigellainfection at the household level and to detect
differences in behavioral risk factors at the individual level. Second, our analysis focused on pediatric
index shigellosis patients, so our findings are not necessarily generalizable to older index patients. Third,
we did not collect longitudinal stool samples from index patients and therefore cannot determine how
long their shedding may have continued through the 1-week surveillance period. However, because all
index patients received antibacterial drugs, we suspect that the shedding was minimal and that the
source of Shigella infection in these households during the surveillance period was more likely from
household members already infected or from a shared environmental source that we did not measure.
Fourth, we used bacterial culture to detect Shigella in the stool samples. This method limited our
analysis to infections with sufficiently high bacteria quantity in stool to be detected by culture. Future
studies should use bacterial culture and quantitative PCR on collected stool samples. Finally, future
studies should obtain >1 isolate from stool samples collected from each household contact to determine
whether a person can shed multiple PFGE genotypes.

A main strength of our study was the environmental surveillance of household water sources, stored
household water, and fly counts in study households. Second, we included households of both
shigellosis patients and controls; this approach enabled us to examine rates of Shigella infection in
patient households, compared with control households, and to investigate household-level risk factors
for shigellosis. Third, we followed up with study households at 4 specific times during a 1-week period to
obtain detailed information on potential behavioral and environmental risk factors for Shigella infection.
Fourth, we used PFGE to conduct genetic characterization of strains within households.

In rural Bangladesh, household contacts of shigellosis patients are highly susceptible to Shigella infection
during the week after the index patient visits a health facility for care. Our findings suggest that each
shigellosis patient household represents the spread of a single infectious pathogen. Therefore,
interventions for household-level risk factors, such as fly control, water treatment, and hygiene
practices, could potentially reduce Shigella transmission in this high-risk population.
Dr. George is an assistant professor of International Health at the Johns Hopkins Bloomberg School of
Public Health. Her research focuses on identifying environmental transmission routes for enteric
infections and developing approaches for interventions in identified transmission routes of infection.
She has conducted 3 cluster randomized controlled trials of safe drinking water interventions in rural
and urban Bangladesh.


We thank all study participants and staff involved in this project. We are grateful to Mrs. Sherrilyn Fisher
and Dr. Paul G. Auwaerter for their support of our study.

This work was supported by a grant from the Sherrilyn and Ken Fisher Center for Environmental
Infectious Diseases at the Johns Hopkins Department of Medicine.


1. Lamberti LM, Bourgeois AL, Fischer Walker CL, Black RE, Sack D. Estimating diarrheal illness and
deaths attributable to Shigellae and enterotoxigenic Escherichia coli among older children,
adolescents, and adults in South Asia and Africa. PLoS Negl Trop Dis. 2014;8:e2705. DOIPubMed

2. Kotloff KL, Winickoff JP, Ivanoff B, Clemens JD, Swerdlow DL, Sansonetti PJ, Global burden
of Shigella infections: implications for vaccine development and implementation of control
strategies. Bull World Health Organ. 1999;77:651–66.PubMed

3. Bardhan P, Faruque AS, Naheed A, Sack DA. Decreasing shigellosis-related deaths

without Shigella spp.–specific interventions, Asia. Emerg Infect Dis. 2010;16:1718–
23. DOIPubMed

4. von Seidlein L, Kim DR, Ali M, Lee H, Wang X, Thiem VD, A multicentre study
of Shigella diarrhoea in six Asian countries: disease burden, clinical manifestations, and
microbiology. PLoS Med. 2006;3:e353. DOIPubMed

5. Kotloff KL, Nataro JP, Blackwelder WC, Nasrin D, Farag TH, Panchalingam S, Burden and
aetiology of diarrhoeal disease in infants and young children in developing countries (the Global
Enteric Multicenter Study, GEMS): a prospective, patient–control study. Lancet. 2013;382:209–
22. DOIPubMed

6. Das SK, Ahmed S, Ferdous F, Farzana FD, Chisti MJ, Leung DT, Changing emergence
of Shigella sero-groups in Bangladesh: observation from four different diarrheal disease
hospitals. PLoS ONE. 2013;8:e62029. DOIPubMed

7. Wilson R, Feldman RA, Davis J, LaVenture M. Family illness associated with Shigella infection:
the interrelationship of age of the index patient and the age of household members in
acquisition of illness. J Infect Dis. 1981;143:130–2. DOIPubMed

8. Boyce JM, Hughes JM, Alim AR, Khan M, Aziz KM, Wells JG, Patterns of Shigella infection in
families in rural Bangladesh. Am J Trop Med Hyg. 1982;31:1015–20.PubMed
9. Khan AI, Talukder KA, Huq S, Mondal D, Malek MA, Dutta DK, Detection of intra-familial
transmission of shigella infection using conventional serotyping and pulsed-field gel
electrophoresis. Epidemiol Infect. 2006;134:605–11. DOIPubMed

10. Cohen D, Green M, Block C, Slepon R, Ambar R, Wasserman SS, Reduction of transmission of
shigellosis by control of houseflies (Musca domestica).Lancet. 1991;337:993–7. DOIPubMed

11. Levine OS, Levine MM. Houseflies (Musca domestica) as mechanical vectors of shigellosis. Rev
Infect Dis. 1991;13:688–96. DOIPubMed

12. Farag TH, Faruque AS, Wu Y, Das SK, Hossain A, Ahmed S, Housefly population density correlates
with shigellosis among children in Mirzapur, Bangladesh: a time series analysis. PLoS Negl Trop
Dis. 2013;7:e2280. DOIPubMed

13. Martin DL, Gustafson TL, Pelosi JW, Suarez L, Pierce GV. Contaminated produce—a common
source for two outbreaks of Shigella gastroenteritis.Am J Epidemiol. 1986;124:299–305.PubMed

14. Kapperud G, Rørvik LM, Hasseltvedt V, Høiby EA, Iversen BG, Staveland K, Outbreak of Shigella
sonnei infection traced to imported iceberg lettuce.J Clin Microbiol. 1995;33;609–14.PubMed

15. Ferguson AS, Layton AC, Mailloux BJ, Culligan PJ, Williams DE, Smartt AE, Comparison of fecal
indicators with pathogenic bacteria and rotavirus in groundwater. Sci Total
Environ. 2012;431:314–22. DOIPubMed

16. Talukder KA, Azmi IJ. Population genetics and molecular epidemiology of Shigella species. In:
Faruque SM, editor. Foodborne and waterborne bacterial pathogens epidemiology, evolution
and molecular biology. Norfolk (UK): Caister Academic Press; 2012. p. 63–76.

17. Ribot EM, Fair MA, Gautom R, Cameron DN, Hunter SB, Swaminathan B, Standardization of
pulsed-field gel electrophoresis protocols for the subtyping of Escherichia
coli O157:H7, Salmonella, and Shigella for PulseNet. Foodborne Pathog Dis. 2006;3:59–
67. DOIPubMed

18. Talukder KA, Islam MA, Dutta DK, Hassan F, Safa A, Nair GB, Phenotypic and genotypic
characterization of serologically atypical strains of Shigella flexneri type 4 isolated in Dhaka,
Bangladesh. J Clin Microbiol. 2002;40:2490–7. DOIPubMed

19. Tenover FC, Arbeit RD, Goering RV, Mickelsen PA, Murray BE, Persing DH, Interpreting
chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for
bacterial strain typing. J Clin Microbiol. 1995;33:2233.PubMed

20. Fernandez-Prada CM, Venkatesan MM, Franco AA, Lanata CF, Sack RB, Hartman AB, Molecular
epidemiology of Shigella flexneri in a diarrhoea-endemic area of Lima, Peru. Epidemiol
Infect. 2004;132:303–16. DOIPubMed

21. Hughes JM, Boyce JM, Levine RJ, Khan M, Aziz KM, Huq MI, Epidemiology of eltor cholera in
rural Bangladesh: importance of surface water in transmission. Bull World Health
Organ. 1982;60:395–404.PubMed
22. Khan MU. Interruption of shigellosis by hand washing. Trans R Soc Trop Med Hyg. 1982;76:164–
8. DOIPubMed


 Table 1. Characteristics of household contacts of pediatric shigellosis patients and of community

controls, rural Bangladesh

 Table 2. Demographic and environmental characteristics of households of pediatric shigellosis

patients and of community controls, rural Bangladesh

 Table 3. Household infection characteristics of pediatric shigellosis patients and of community

controls, rural Bangladesh

 Table 4. Characteristics of household contacts with Shigella infections for pediatric shigellosis
patients and community controls, rural Bangladesh

 Table 5. Patient households with Shigella-infected household contacts by Shigella species and
serotype and visit number at which infection was detected, rural Bangladesh

Cite This Article

DOI: 10.3201/eid2111.150333

Table of Contents – Volume 21, Number 11—November 2015