Available via license: CC BY 4.0
Content may be subject to copyright.
SUPPLEMENT ARTICLE
S196 • CID 2020:71 (Suppl 3) • Saha etal
Clinical Infectious Diseases
Correspondence: S. K. Saha, Child Health Research Foundation, House # 23/2, SEL Huq
Skypark, 5th Floor, Dhaka 1207, Bangladesh (samir@chrfbd.org).
Clinical Infectious Diseases® 2020;71(S3):S196–204
© The Author(s) 2020. Published by Oxford University Press for the Infectious Diseases Society
of America.This is an Open Access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted
reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
DOI: 10.1093/cid/ciaa1356
Hospitalization of Pediatric Enteric Fever Cases, Dhaka,
Bangladesh, 2017–2019: Incidence and RiskFactors
Shampa Saha,1 K.M.Ishtiaque Sayeed,1 Senjuti Saha,1,2 MdShafiqul Islam,1 Afiour Rahaman,1 Maksuda Islam,1 Hafizur Rahman,1 Raktim Das,1
MdMahmudul Hasan,1 MohammadJamal Uddin,1 ArifMohammad Tanmoy,1 A.S.M.Nawshad Uddin Ahmed,1,3 StephenP. Luby,4 JasonR. Andrews,4
Denise O. Garrett,5 and SamirK. Saha1,3
1Child Health Research Foundation, Dhaka, Bangladesh, 2Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, USA, 3Bangladesh Institute of Child Health, Dhaka Shishu
Hospital, Sher-E-Bangla Nagar, Dhaka, Bangladesh, 4Division of Infectious Diseases and Geographic Medicine, Stanford University School of Medicine, Stanford, California, USA, and 5Applied
Epidemiology, Sabin Vaccine Institute, Washington, DC, USA
Background. Enteric fever causes substantial morbidity and mortality in low- and middle-income countries. Here, we analyzed
Surveillance for Enteric Fever in Asia Project (SEAP) data to estimate the burden of enteric fever hospitalization among children
aged <15years and identify risk factors for hospitalization in Bangladesh.
Methods. SEAP used hospital surveillance paired with a community-based health-care utilization assessment. In SEAP hospital
surveillance, blood was obtained for culture from children aged <15years with ≥3days of fever. In the hospital catchment area, a
health-care utilization survey (HCUS) was conducted to estimate the proportion of febrile children hospitalized at the study hos-
pitals. We analyzed hospital surveillance and HCUS data to estimate the health care–adjusted incidence of enteric fever hospitaliza-
tion, and conducted univariable and multivariable logistic regressions.
Results. From July 2017 through June 2019, 2243 laboratory-conrmed enteric fever cases were detected in 2 study hospitals; 673
(30%) were hospitalized. e health care–adjusted incidence of enteric fever hospitalization among children <15years old was 303/100000
children/year (95% condence interval [CI], 293–313). Salmonella Typhi contributed most to the enteric fever hospitalization incidence
(277/100000 children/year; 95% CI, 267–287). e incidence was highest among children aged 2 to <5years (552/100000 children/year;
95% CI, 522–583), followed by those aged <2years (316/100000 children/year; 95% CI, 288–344). Factors independently associated with
enteric fever hospitalization included fever duration, diarrhea, vomiting, abdominal pain, and leukocytopenia.
Conclusions. We estimated a high burden of hospitalization due to enteric fever among children aged <5years in Bangladesh.
e introduction of a typhoid conjugate vaccine would protect children from typhoid and avert typhoid hospitalizations.
Keywords. enteric fever; typhoid; hospitalization; incidence; risk factors.
Typhoid and paratyphoid, collectively called enteric fever, are
caused by the bacteria Salmonella enterica serovars Typhi and
Paratyphi A, B, or C. Enteric fever remains a major cause of
morbidity and mortality in low- and middle-income coun-
tries [1, 2]. In 2017, S. Typhi caused approximately 11 million
illnesses and 117000 deaths, whereas S. Paratyphi caused over
3 million illnesses and 19000 deaths globally [3, 4]. The burden
of enteric fever was greatest in South Asia and among children
aged <15years [3].
A heat-killed, phenol-preserved, whole-cell typhoid vaccine
for the prevention of typhoid fever has been available since
1896. e ecacy of this vaccine was established in 1960 [5–7].
Live attenuated Ty21a and Vi typhoid polysaccharide vaccines
are eective against typhoid fever, but these vaccines are not re-
commended for children aged <2years and have a short du-
ration of protection [5]. A new-generation typhoid conjugate
vaccine (TCV) containing Vi polysaccharide conjugated to a
tetanus-toxoid protein carrier showed 55% protective ecacy
in a human challenge model [8]. Considering its improved im-
munologic properties, suitability for use in infants and young
children, and expected longer duration of protection, the
World Health Organization (WHO) Strategic Advisory Group
of Experts recommended the use of TCV over other available
vaccines against typhoid in October 2017 [9]. In January 2018,
the WHO prequalied the rst TCV for use in countries with a
high burden of typhoid fever [10]. e Phase III clinical trial of
the WHO-prequalied TCV in Nepal from 2017–2019 showed
that a single dose of TCV was 82% eective in reducing S. Typhi
bacteremia in children 9months to 16years of age [11].
For the introduction of a new vaccine into a country’s national
immunization program, the disease burden is a critical consid-
eration [12]. As the case fatality rate of typhoid fever is less than
1%, hospitalization may serve as a proxy of severe disease [13]. In
resource-poor countries, there is a erce competition for hospital
Enteric Fever Hospitalization Incidence • CID 2020:71 (Suppl 3) • S197
beds. For example, in Bangladesh there are only 3 beds per 10000
people [14]. e lack of hospital beds oen means that patients
with serious illnesses—for example, pneumonia, meningitis, birth
asphyxia, and preterm birth–related complications—are not ad-
mitted [15]. e prevention of hospitalization of enteric fever cases
would free up these needed hospital beds, and would ultimately
reduce the burden on the health system. erefore, hospitalization
can be a consideration for decision-making for the introduction of
TCV in the national immunization program.
In Bangladesh, several studies have documented a high
burden of enteric fever at hospitals and in the community
[16–20]. However, data on the incidence of enteric fever hos-
pitalization are scarce. Hospitalizations were reported in 2
population-based studies that were conducted in urban slums
and were limited by representation of the urban slum popula-
tion. ese studies captured disease at an early stage by active
surveillance in the community, hence interrupting the natural
course of disease that might have led to hospitalization [19, 20].
e Surveillance for Enteric Fever in Asia Project (SEAP), a
multi-country, multi-site prospective study, used a hybrid sur-
veillance approach with facility-based surveillance paired with
a community-based health-care utilization survey (HCUS) to
assess the burdens of enteric fever in Bangladesh, Pakistan,
and Nepal. e community-based HCUS included all socioec-
onomic levels and did not interrupt the natural course of di-
sease, as it did not involve active, population-based surveillance
[21]. Here, we analyzed SEAP data to estimate the burden of
enteric fever hospitalization among children aged <15years in
an urban population and to identify risk factors for hospitaliza-
tion in Bangladesh.
METHODS
Hospital Surveillance
Study Design, Sites, and Procedures
SEAP surveillance was conducted at 2 pediatric hospitals in
Dhaka, Bangladesh: Dhaka Shishu (children) Hospital and
Dr MR Khan Shishu Hospital. With 660 beds, Dhaka Shishu
Hospital is the largest pediatric hospital in Bangladesh that
provides primary to tertiary levels of care to patients aged
<18years. With 250 beds, Dr MR Khan Shishu Hospital is the
second largest pediatric hospital in the country and provides
patients aged <15years with primary and secondary levels of
care. Acatchment area of these 2 hospitals that had been pre-
viously identified was used for the HCUS (Figure1) [21, 22].
We enrolled children from the inpatient department (IPD),
outpatient department (OPD), and the hospital laboratory. All
children living in the hospital catchment area and presenting at
the OPD with ≥3days fever were eligible to participate in the
study. e inpatient eligibility criteria included clinical suspi-
cion of enteric fever by clinicians or a blood culture positive for
S. Typhi or S. Paratyphi A.Eligible cases were enrolled in SEAP
IPDs and OPDs if they had a blood culture in the study hospital
and gave consent to participate in the study. Additional lab-
conrmed cases that were identied at the hospital laboratories
and missed at the IPDs and OPDs were enrolled retrospectively
in the study as hospital lab–enrolledcases.
Figure 1. Maps showing population density in Bangladesh by (A) districts and location of Dhaka district, (B) location of the study catchment area in Dhaka district, and (C)
location of the study hospitals in the study catchment area.
S198 • CID 2020:71 (Suppl 3) • Saha etal
Detailed demographic, clinical, and laboratory information
of enrolled cases was collected electronically.
Information regarding symptoms of enteric fever—for ex-
ample, fever start date, diarrhea, vomiting, abdominal pain,
cough, constipation, headache, seizure, bloody stool, and dif-
culty breathing—were collected directly from the patient or
caregiver by a study physician at the time of enrollment of the
patient into SEAP. Diarrhea was dened as the passage of 3 or
more loose or liquid stools per day or more frequent passage
than is normal for the individual. Clinical signs were assessed
and recorded by the study physician, and laboratory informa-
tion was collected from hospital medical records.
Detection of S.Typhi and S.Paratyphi in Blood Samples
A sample of 3 milliliters of blood was obtained under aseptic
conditions for culture from all children enrolled in this study.
Blood culture was performed utilizing the BACTEC (Becton
Dickinson and Company) automated culture system. The
bottle was incubated at 37o C for a maximum of 5days. After
incubation, positive samples were subcultured on sheep blood,
chocolate, and MacConkey agar plates. Standard biochemical
tests and agglutination with Salmonella serovar–specific anti-
sera (Ramel, Thermo Fisher Scientific) were used to confirm
Salmonella Typhi/Paratyphi isolates [16].
Health-care UtilizationSurvey
The hospital catchment area for the HCUS included 22 out
of 90 administrative wards of Dhaka city corporation and
covered approximately 51 square kilometers. The catchment
area was divided into 2524 clusters by an overlaying geo-
graphic information system (GIS grids), and then 100 clus-
ters were randomly chosen for conducting HCUS. In SEAP, 2
HCUSs were conducted; the first during April–July 2017 and
the second during September–December 2018, during which
a new set of clusters were selected. During both HCUSs, each
household included in the selected clusters was approached
for interview, with no replacement of households in cases
of absence or refusal. We used standardized questionnaires
to collect health care–seeking behavior of children aged
<18years who had symptoms consistent with enteric fever—
which included fever for >3days in the last 8 weeks for out-
patients and hospitalization due to febrile illness in the last
1 year for inpatient cases—and estimated the proportion
of individuals with a febrile illness compatible with enteric
fever who sought care at the health facilities participating in
SEAP hospital surveillance [21].
Data Analysis
We analyzed data generated from July 2017 to June 2019
through SEAP hospital surveillance and HCUS. We included
children aged <15years in the analysis. We used medians ± in-
terquartile ranges (IQRs) for summarizing continuous variables
and frequencies with percentages for summarizing categorical
variables. Comparative statistics included independent samples
t-tests, chi-square tests, and Fisher’s exact tests, as appropriate.
We estimated the number of children aged <15 years living
within the catchment area by multiplying the number of chil-
dren identified in the HCUS by the inverse of the fraction of the
area sampled(4.3%).
We estimated the crude incidence rate of enteric fever hos-
pitalization in the hospital catchment area by dividing the
number of culture-positive typhoid and paratyphoid cases
hospitalized by the population, multiplied by time. We ad-
justed the crude incidence rate of enteric fever hospitaliza-
tion for 3 factors: the proportion who sought care at study
hospitals (facility coverage), the proportion captured by
health facilities (facility capture), and blood-culture sen-
sitivity. Facility coverage was estimated by the household
survey and facility capture was estimated at study hospitals.
In order to adjust for blood-culture sensitivity, we used the
61% sensitivity of blood culture that was reported by a recent
systematic literature review [21]. We estimated the adjusted
incidence rate of enteric fever hospitalization by dividing
the crude incidence of enteric fever hospitalization by the
facility coverage, multiplied by facility capture, multiplied by
blood-culture sensitivity.
We calculated 95% condence interval (CIs) for incidence
rates using a normal approximation to Poisson distribution
method. For the risk-factor analysis, we included children aged
<15years who were enrolled in SEAP hospital surveillance, re-
gardless of residence. Risk factors associated with hospitaliza-
tion were analyzed using univariable and multivariable logistic
regressions. We conducted 2-tailed statistical tests at an alpha
level of 0.05.
Ethical Consideration
This study was approved by the Ethical Review Committee of the
Bangladesh Institute of Child Health, Dhaka. Informed written
consent for participation in the study was obtained from the
guardian of each child. In addition, assent was obtained from
the children aged ≥11years.
RESULTS
Hospital Surveillance
A total of 11354 children were enrolled and had a blood cul-
ture performed at IPDs and OPDs of Dhaka Shishu Hospital
and Dr MR Khan Shishu Hospital (Figure 2). Among these,
1629 lab-confirmed enteric fever cases were detected: 1447
(89%) S. Typhi and 182 (11%) S. Paratyphi. Additionally, 614
enteric fever cases were identified at the hospital lab who were
not enrolled at an IPD or OPD. All 275 S. Paratyphi that were
isolated during the reporting period were S. Paratyphi A.Of the
2243 lab-confirmed typhoid and paratyphoid cases, 673 (30%)
Enteric Fever Hospitalization Incidence • CID 2020:71 (Suppl 3) • S199
were hospitalized, and 91% (n = 612) of the hospitalizations
were due to S. Typhi. Enteric fever occurred throughout the
year, with slight decreases in the numbers of cases and hospi-
talizations during winter (December–February; Figure3). The
majority (n = 523; 78%) of the hospitalized typhoid and par-
atyphoid cases came from the hospital catchment area where
HCUS had been conducted.
Of 673 hospitalized, lab-conrmed S. Typhi and S. Paratyphi
cases, 361 (54%) were aged <5years (Table1). All had fever at
the time of hospitalization, with a median duration of 6days
(IQR, 4–8). Additional predominant clinical and labora-
tory ndings at the time of hospitalization included vomiting
(262/672, 39%), diarrhea (204/671, 30%), abdominal pain
(165/672, 25%), cough (99/672, 15%), hepatomegaly (27/54,
50%), splenomegaly (11/54, 20%), leukocytopenia (25/323,
8%), and thrombocytopenia (25/323, 8%). Nearly all (667/673,
99%) isolates identied from the hospitalized children were
ciprooxacin-resistant, whereas the rate of multidrug resist-
ance, dened as resistance to chloramphenicol, ampicillin, and
trimethoprim-sulfamethoxazole, was 17% (115/673). e me-
dian duration of stay in the hospital was 7days (IQR, 5–9). No
deaths were reported among the hospitalized cases. Detailed
characteristics of hospitalized and nonhospitalized children
are shown in Table1.
Figure 3. Distribution of total number of enteric fever cases (n = 2243) and hospitalizations (n = 673) by month at Dhaka Shishu Hospital and Dr MR Khan Shishu Hospital,
Dhaka, Bangladesh, July 2017–June 2019.
8742 blood cultures
1126 (13%) Enteric fever
989 (88%) S. Typhi
137 (12%) S. Paratyphi
106 of 1126 (9.4%)
S.Typhi and S. Paratyphi
were hospitalized
106 (100%) came from
hospital catchment area
2612 blood cultures
503 (19%) Enteric fever
458 (91%) S. Typhi
45 (8.9%) S. Paratyphi
503 of 503 (100%)
S.Typhi and S. Paratyphi
were hospitalized
368 (80%) came from
hospital catchment area
614 Enteric fever
521 (85%) S. Typhi
93 (15%) S. Paratyphi
64 of 614 (10%) S. Typhi
and S. Paratyphi were
hospitalized
49 (77%) came from
hospital catchment area
Enrollment locations at hospitals (n=11968)
OPDIPD Hospital lab
614 blood cultures
Figure 2. Enteric fever cases aged <15years enrolled (n = 11968) by recruitment location, with Salmonella Typhi and S. Paratyphi isolated, and hospitalized at Dhaka
Shishu Hospital and Dr MR Khan Shishu Hospital, Dhaka, Bangladesh, July 2017–June 2019. Abbreviations: IPD, inpatient department; OPD, outpatient department; S.,
Salmonella.
S200 • CID 2020:71 (Suppl 3) • Saha etal
Although young children were more likely to be hospital-
ized, age was not identied as a risk factor for hospitalization
in the multivariable logistic regression. Factors independently
associated with hospitalization included fever duration, diar-
rhea, vomiting, abdominal pain, and leukocytopenia (Table2).
No signicant dierences in resistance to chloramphenicol,
ampicillin, and trimethoprim-sulfamethoxazole were found
between hospitalized and nonhospitalizedcases.
Community Health-care Utilization
In the hospital catchment area, the estimated number of chil-
dren aged <15years in 2017 and 2018 was 1148076. During
Table 1. Demographic, Clinical, and Laboratory Characteristics
Characteristics
Hospitalized Nonhospitalized
Total P valuen % n %
Demographic characteristics
Age
<2years 93 14% 176 11 % 269 .01 9a
2 to <5years 268 40% 568 36% 836
5 to <15years 312 46% 826 53% 11 4 1
6months to <15years 672 100 % 1568 70% 2243
Median age (IQR) in months 48 (24–84) … 60 (36–84) … 60 (36–84) <.001a
Sex
Male 363 54% 866 55% 1229 .594
Resident of catchment area 523 78% 1438 91% 1961 <.001a
Clinical and laboratory findings
Median duration of fever (IQR) in day at the time of
enrollment
7 (5–9) … 5 (4–8) … 6 (4–8) .009a
Vomiting 262 39% 2 74 17% 536 <.001a
Diarrhea 204 30% 158 10 % 362 <.001a
Abdominal pain 165 25% 2 17 14% 382 <.001a
Cough 99 15% 319 20% 418 .006a
Constipation 23 3.4% 60 3.8% 83 .615
Headache 18 2.7% 74 4.7% 92 .032a
Rash 10 1.5% 5 .3% 15 .006a
Jaundice 4 .6% 2 .1% 6 .055
Seizure 3 .4% 1 .1% 4 .055
Bloody stool 2 .3% 0 0% 2 .037a
Difficult breathing 2 .3% 4 .3% 6 .376
Hepatomegaly on ultrasonogram 28 52% 5 71% 33 .305
Splenomegaly on ultrasonogram 12 22% 1 14% 13 .645
Leukocytopenia, total WBC < 4000/mm325 7.7% 10 1.9% 35 <.001a
Thrombocytopenia, platelet < 150000µl25 7.7% 13 2.5% 38 <.001a
Organism
S. Typhi 612 91% 1356 86% 1968 .003a
S. Paratyphi 61 9.1% 214 14% 275
Severity
Days not able to conduct usual activities due to illness 3 (0–8) … 3 (0–12) … … .902
Hours of bed rest during worst day of illness 3 (2–6) … 2 (1–5) … … .003a
Drug resistance
Ampicillin 202 30% 373 24% 575 .002a
Chloramphenicol 126 19% 248 16% 3 74 .088
Cotrimoxazole 118 18% 237 15 % 355 .206
Ciprofloxacin 667 99% 1534 97% 2201 .042a
Azithromycin 9 1.3% 40 2.5% 49 .072
Cefixime 1 0% 3 0% 4 .827
Ceftazidime 0 0% 0 0% 0
MDR 115 17 % 222 14% 337 .073
Data are of hospitalized (n = 673) and nonhospitalized (n = 1570) S.Typhi and S.Paratyphi cases isolated at Dhaka Shishu Hospital and Dr MR Khan Shishu Hospital, July 2017–June 2019.
Abbreviations: IQR, interquartile range; MDR, multidrug resistance; S., Salmonella; WBC, white blood cell.
aStatistically significant.
Enteric Fever Hospitalization Incidence • CID 2020:71 (Suppl 3) • S201
2 rounds of HCUS, data on hospitalization due to febrile ill-
ness in the last 1year were collected from 36142 children aged
<15years living in the selected clusters (Table3). Atotal of 335
(0.9%) hospitalizations due to febrile illness were identified; 177
(52.8%) were hospitalized at a study hospital.
Incidence of Hospitalization Due to EntericFever
The overall crude incidence of hospitalization due to enteric
fever among children aged <15years in the hospital catchment
area in urban Dhaka during 2017–2018 was 46 per 100000 chil-
dren per year (95% CI, 42–49; Table4). The estimated incidence
of enteric fever hospitalization after adjustment for facility cov-
erage, facility capture, and blood-culture sensitivity was 303 per
100000 children per year (95% CI, 293–313). Salmonella Typhi
was the main contributing factor for hospitalization. There was
no difference in the hospitalization rates between males and
females. The hospitalization rate was highest among children
aged 2 to <5years, followed by those aged <2years.
DISCUSSION
Our data showed, over the 2-year study period, that 30% of the
lab-confirmed pediatric enteric fever cases visiting the study
hospitals required hospitalization. The incidence of enteric
fever hospitalization was 303 per 100000 children per year;
S. Typhi was the main contributing factor for hospitalization.
The hospitalization rate was highest among children aged 2 to
<5years, followed by those aged <2years. The duration of fever
and the presence of diarrhea, vomiting, abdominal pain, and
leukocytopenia were identified as risk factors for enteric fever
hospitalization.
Similar to our ndings of 30% hospitalization, a study con-
ducted from 2004 through 2016 in a similar setting reported that
32% of the typhoid cases and 21% of the paratyphoid pediatric
cases required hospitalization [16]. Previous data showed that
3.6% of all admissions at large pediatric hospitals in Bangladesh
were either laboratory-conrmed or clinically diagnosed en-
teric fever cases [23]. is could be because Bangladesh has a
high burden of enteric fever, with incidence rates ranging from
2.7–18.7 per 1000 children [19, 20, 24].
In concordance with previous studies evaluating the clinical
prole of enteric fever, we found that the predominant signs,
symptoms, and laboratory ndings of hospitalized enteric fever
cases include fever, vomiting, diarrhea, abdominal pain, cough,
hepatomegaly, and splenomegaly [23, 25–27]. But in contrast
to the systematic review conducted by Azmatullah et al. [27]
Table 3. Health-care Utilization Survey–Identified Hospitalizations of Children <15 Years Old for Febrile Illness at Dhaka Shishu Hospital and Dr MR Khan
Shishu Hospital, Dhaka, Bangladesh, 2017–2018
HCUS 2017 HCUS 2018 Total
n%n%n%
<15year old children interviewed for hospitalization due
to febrile illness in last 1year
20603 … 15539 … 36142 …
Hospitalized anywhere 141 .7% 194 1.2% 335 .9%
Hospitalized at Dhaka Shishu Hospital 22 15.6% 49 25.3% 71 21.2%
Hospitalized at Dr MR Khan Shishu Hospital 53 37.6% 53 27.3% 10 6 31.6%
Hospitalized at other places 66 46.8% 92 47.4% 158 47.2%
Abbreviation: HCUS, health-care utilization survey.
Table 2. Risk Factors for Hospitalization of Children With Laboratory-confirmed Typhoid and Paratyphoid at Dhaka Shishu Hospital and Dr MR Khan
Shishu Hospital, Dhaka, Bangladesh, July 2017–June 2019
Characteristics
Univariable Multivariable
Odds Ratio (95% CI) P value Odds Ratio (95% CI) P value
Agea2 to <5years .89 (.67–1.19) .44 1.02 (.64–1.63) .938
5 to <15years .71 (.54–.95) .02 .82 (.52–1.32) .419
Fever duration 1.01(1.00–1.03) .025 1.07 (1.03–1.11) .0 01b
Diarrhea 3.89 (3.09–4.92) <.001 3.34 (2.26–4.95) <.001b
Vomiting 3.42 (2.53–4.61) <.0 01 2.03 (1.44–2.86) <.001b
Abdominal pain 2.03 (1.62–2.54) <.001 1.62 (1.11–2.37) .013 b
Headache .56 (.33–.94) .028 .63 (.28–1.38) .244
Leukocytopenia 4.22 (2–8.91) <.001 3.34 (1.47–7.59) .0 04b
Thrombocytopenia 3.22 (1.63–6.41) .001 2.32 (1.10–4.87) .027
Abbreviations: CI, confidence interval.
aReference: <2years.
bStatistically significant.
S202 • CID 2020:71 (Suppl 3) • Saha etal
Table 4. Crude and Adjusted Incidence Rate of Hospitalization of Laboratory-confirmed S. Typhi and S. Paratyphi Among Children Aged <15 by Organism, Sex, Age Group and Recruitment Location, at the
Catchment Area of Dhaka Shishu Hospital and Dr MR Khan Shishu Hospital, Dhaka, Bangladesh, July 2017–June 2019
Characteristics
Number of
hospitalized
cases Population
Crude incidence rate
per 100000/year
(95% CI)
Febrile cases
hospitalized at
study hospitals
Health-care seeking
adjusted incidence
per 100000/year
(95% CI)
Facility captured
febrile cases at
study hospitals
Facility capture
adjusted incidence
per 100000/year (95% CI)
Blood culture
sensitivity
Blood culture sensitivity
adjusted incidence
per 100000/year
(95% CI)
Overall enteric
fever
523 1148076 46 (42–49) 53% 86 (81–92) 47% 185 (177–193) 61% 303 (293–313)
Organism
S. Typhi 478 1148076 42 (38–45) 53% 79 (74–84) 47% 169 (161–177) 61% 277 (267–287)
S. Paratyphi A 45 1148076 3.9 (2.8–5.1) 53% 7.4 (5.8–9) 4% 16 (14–18) 61% 26 (23–29)
Sex
Male 276 584730 47 (42–53) 54% 88 (80–95) 47% 185 (174–196) 61% 303 (289–317)
Female 247 563346 44 (38–49) 51% 86 (78–93) 46% 188 (176–199) 61% 308 (293–322)
Age
<2years 70 155915 45 (34–55) 58% 78 (64–92) 40% 193 (171–215) 61% 316 (288–344)
2 to <5years 209 225453 93 (80–105) 57% 163 (146–179) 48% 337 (313–361) 61% 552 (522–583)
5 to <15years 244 766708 32 (28–36) 46% 69 (63–75) 50% 138 (129–146) 61% 225 (215–236)
6months to
<15years
522 1110388 47 (43–51) 51% 91 (86–97) 47% 195 (187–203) 61% 319 (309–330)
Enrollment location
OPD 106 1148076 9.2 (7.5–11) 53% 17 (15–20) 44% 40 (36–43) 61% 65 (60–70)
IPD 368 1148076 32 (29–35) 53% 61 (56–65) 84% 72 (67–77) 61% 118 (112–125)
Hospital lab 49 1148076 4.3 (3.1–5.5) 53% 8.1 (6.4–10) 10 0 % 8.1 (6.4–10) 61% 13 (11–15)
Abbreviations: CI, confidence interval; IPD, inpatient department; OPD, outpatient department; S., Salmonella.
Enteric Fever Hospitalization Incidence • CID 2020:71 (Suppl 3) • S203
and a study conducted in Fiji [26], we observed lower propor-
tions of children with signs and symptoms of complications,
including gastrointestinal bleeding (6–8% in prior studies vs
0.3% in our study) and jaundice (2.8% in prior studies vs 0.6%
in our study), and no mortality (1–6% in prior studies vs 0%
in our study) due to enteric fever among children, which is
possibly due to the widespread over-the-counter availability of
antibiotics in Bangladesh. Relatively less severe disease in our
population could also possibly be explained by a hypothesis
that frequent subclinical exposure to S. Typhi and S. Paratyphi
from contaminated water and food may provide a level of im-
munity against the disease to the people living in the endemic
area. However, our data demonstrated that only fever duration,
diarrhea, vomiting, abdominal pain, and leukocytopenia were
associated with enteric fever hospitalization among children
aged <15years. Although there are reports on coinfections with
dengue, malaria, and so forth, data on the role of coinfection
with other pathogens on hospitalization of enteric fever cases is
scarce [28, 29]. In the blood cultures of the enteric fever cases
enrolled in SEAP, we did not nd any other bacterial patho-
gens. However, based on the clinical criteria and advice of clin-
icians, only 74 cases were tested for dengue. Of them, 2 (2.7%)
were positive for either an NS1 antigen test or anti-dengue
immunoglobin M.As the number of dengue-positive cases was
too small, we did not include coinfection with dengue in the
risk factor analysis.
We reported that the adjusted incidence of hospitalization
due to typhoid and paratyphoid was 303 per 100000 children
per year in children aged <15years. is rate is higher than the
rates observed by previous studies conducted in Bangladesh,
Kenya, and Spain [19, 20, 25, 30]. In Bangladesh, Brooks etal.
[20] gathered 12407 person-years worth of data for 10months,
but found no hospitalizations, complications, or deaths among
blood culture–conrmed typhoid fever patients. Naheed
etal. [19] observed only 4 hospitalizations among blood cul-
ture–conrmed typhoid fever patients over 1year and 19 710
person-years worth of data. is translates to 20 enteric fever
hospitalizations per 100000 population per year. ese studies
utilized active, population-based surveillance that captured
cases at an early stage of the disease and, thus, prevented hos-
pitalization by treating them early. Our study utilized a hybrid
method that did not interfere in the natural course of the di-
sease and included a combination of passive surveillance at the
study hospitals and HCUS in the hospital catchment area for
adjustment in the care seeking seen at the hospitals.
Our study demonstrated the substantial burden of enteric
fever in children aged <15years, with the greatest burden in
children aged <5years, which is consistent with other studies
[19, 20, 24, 30]. We estimated 277 hospitalizations due to S.
Typhi per 100 000 children per year, and a median of 7days
of hospital stay. is translates to 3180 hospitalizations due to
typhoid, with an estimated 22261 hospital bed-days in a year
in the hospital catchment area. is has important policy im-
plications, as a TCV with 80% ecacy and 90% vaccination
coverage would be able to avert 16028 hospital bed-days per
year in the hospital catchment area. e resulting availability of
hospital beds at secondary- and tertiary-level hospitals would
mean children with other serious illnesses, at risk of disability
and mortality, would have the opportunity to receive treatment
at the hospital [15]. us, TCV would reduce the burden on the
overall healthsystem.
Our study included all hospital departments: OPDs, IPDs,
and hospital laboratories. Although most of the cases contrib-
uting to the numerator came from the IPDs, the inclusion of
OPDs and hospital labs into the study increased the capture of
cases by 40%. is nding illustrates the importance of a com-
prehensive surveillance system for the estimation of the disease
burden [31].
is study had some limitations. First, this surveillance was
conducted in a city where we know the water is contaminated
with S. Typhi [32]. erefore, the catchment area of the study
hospitals may be a high-incidence area for enteric fever. e
burden of the disease in other communities is likely to be dif-
ferent. erefore, we would not extrapolate this incidence rate
to a national estimate. However, our study includes a represen-
tative urban population of all socioeconomic levels. Second,
we used a 61% sensitivity rate to adjust the incidence of enteric
fever hospitalization for all subgroups, as information on sensi-
tivity rates by subgroups was not available. Sensitivity rates may
vary by subgroups, and thus aect our estimates for dierent
subgroups.
CONCLUSIONS
We estimated that 303 per 100000 children aged <15years with
enteric fever require hospitalization per year. Children aged
<5years are affected most. The introduction of TCV in the na-
tional immunization program would protect children from ty-
phoid and avert related admissions at pediatric hospitals, thus
allowing children with other severe diseases to be treated. Thus,
TCV would not just prevent morbidity and mortality due to ty-
phoid fever, it would also save more lives and prevent disability
from other severe diseases, and therefore reduce the burden on
the health system and improve child health overall.
Notes
Acknowledgments. e authors thank all physicians, laboratory per-
sonnel, and research assistants working at Dhaka Shishu Hospital and
Dr MR Khan Shishu Hospital, especially Dr Dipalok Mukharjee, Dr
Shamsun Nahar, Dr Shammin Azmery, Dr Humaira Susmita, Dr Shumaiya
Ferdaus, and Mr Anik Sarker, who played active roles in implementation
of the Surveillance for Enteric Fever in Asia Project hospital surveillance;
Ms Salma Akhter for coordination of the health-care utilization survey
(HCUS); all research assistants who collected quality HCUS data; Ms Ruma
Datta for administration support; and Mr. Md. Shariful Islam and Mr Argho
Sarker for data management support.
S204 • CID 2020:71 (Suppl 3) • Saha etal
Financial support. is work was supported by the Bill and Melinda
Gates Foundation (Investment ID OPP1113007).
Supplement sponsorship. is supplement is sponsored by the Sabin
Vaccine Institute and made possible by a grant from the Bill & Melinda
Gates Foundation.
Potential conicts of interest. e authors: No reported conicts of
interest. All authors have submitted the ICMJE Form for Disclosure of
Potential Conicts of Interest.
References
1. MogasaleV, MaskeryB, OchiaiRL, etal. Burden of typhoid fever in low-income
and middle-income countries: a systematic, literature-based update with risk-
factor adjustment. Lancet Glob Heal 2014; 2:e570–80.
2. AntillónM, WarrenJL, CrawfordFW, etal. The burden of typhoid fever in low-
and middle-income countries: a meta-regression approach. PLOS Negl Trop Dis
2017; 11:e0005376.
3. StanawayJD, ReinerRC, BlackerBF, etal. The global burden of typhoid and para-
typhoid fevers: a systematic analysis for the Global Burden of Disease Study 2017.
Lancet Infect Dis 2019; 19:369–81.
4. WierzbaTF, SandersJW. The global burden of enteric fevers in the age of typhoid-
conjugate vaccines. Lancet Infect Dis 2019; 19:340–1.
5. FraserA, GoldbergE, AcostaCJ, PaulM, LeiboviciL. Vaccines for preventing ty-
phoid fever. In: FraserA, ed. Cochrane database of systematic reviews. Chichester,
United Kingdom: John Wiley & Sons, Ltd, 2007. Available at: http://doi.wiley.
com/10.1002/14651858.CD001261.pub2. Accessed 22 February 2020.
6. Yugoslav Typhoid Commission. A controlled field trial of the effectiveness of
phenol and alcohol typhoid vaccines: final report. Bull World Health Organ 1962;
26:357–69.
7. World Health Organization. State of the art of vaccine research and development.
2006. Available at: https://apps.who.int/iris/bitstream/handle/10665/69348/
WHO_IVB_06.01_eng.pdf;jsessionid=910C9043CA1D327048D2C391905A32B
F?sequence=1. Accessed 22 May 2020.
8. FeaseyNA, LevineMM. Typhoid vaccine development with a human challenge
model. Lancet 2017; 390:2419–21.
9. World Health Organization. Weekly epidemiological record. Typhoid vaccines:
WHO position paper—March 2018. 2018. Available at: https://apps.who.int/iris/
bitstream/handle/10665/272272/WER9313.pdf?ua=1. Accessed 12 November
2019.
10. World Health Organization. Typbar TCV® from Bharat Biotech, world’s first ty-
phoid conjugate vaccine prequalified by WHO. 2018. Available at: https://www.
who.int/medicines/news/2017/Bharat-Biotech-TypbarTCV-WHO-PQ-Press-
Release-Global-Final.pdf?ua=1. Accessed 12 November 2019.
11. ShakyaM, Colin-JonesR, Theiss-NylandK, et al. Phase 3 efficacy analysis of a
typhoid conjugate vaccine trial in Nepal. N Engl J Med 2019; 381:2209–18.
12. World Health Organization. Principles and considerations for adding a vaccine
to a national immunization programme. Available at: https://apps.who.int/iris/
bitstream/handle/10665/111548/9789241506892_eng.pdf;jsessionid=F69C31CC
406B69A1BA2647EA67DC3E34?sequence=1. Accessed 15 November 2019.
13. YuAT, Amin N, Rahman MW, Gurley ES, RahmanKM, LubySP. Case-fatality
ratio of blood culture–confirmed typhoid fever in Dhaka, Bangladesh. J Infect Dis
2018; 218:S222–6.
14. World Health Organization. World health statistics 2009. 2010. Available at:
https://www.who.int/whosis/whostat/2009/en/. Accessed 23 February 2020.
15. SahaS, SantoshamM, Hussain M, Black RE, Saha SK. Rotavirus vaccine will
improve child survival by more than just preventing diarrhea: evidence from
Bangladesh. Am J Trop Med Hyg 2018; 98:360–3.
16. SahaS, IslamMS, SajibMSI, etal. Epidemiology of typhoid and paratyphoid: im-
plications for vaccine policy. Clin Infect Dis 2019; 68:S117–23.
17. SahaSK, BaquiAH, HanifM, etal. Typhoid fever in Bangladesh: implications for
vaccination policy. Pediatr Infect Dis J 2001; 20:521–4.
18. Saha S, SahaS, Das RC, et al. Enteric fever and related contextual factors in
Bangladesh. Am J Trop Med Hyg 2018; 99:20–5.
19. NaheedA, RamPK, BrooksWA, etal. Burden of typhoid and paratyphoid fever in
a densely populated urban community, Dhaka, Bangladesh. Int J Infect Dis 2010;
14:e93–9.
20. BrooksWA, HossainA, G oswamiD, etal. Bacteremic typhoid fever in children in
an urban slum, Bangladesh. Emerg Infect Dis 2005; 11:326–9.
21. AndrewsJR, BarkumeC, YuAT, etal. Integrating facility-based surveillance with
healthcare utilization surveys to estimate enteric fever incidence: methods and
challenges. J Infect Dis 2018; 218:S268–76.
22. LubySP, HalderAK, SahaSK, etal. A low-cost approach to measure the burden of
vaccine preventable diseases in urban areas. Vaccine 2010; 28:4903–12.
23. SahaS, UddinMJ, IslamM, DasRC, GarrettDO, SahaSK. Enteric fever cases in
the two largest pediatric hospitals of Bangladesh: 2013–2014. J Infect Dis 2018;
218:S195–200.
24. DewanAM, CornerR, HashizumeM, OngeeET. Typhoid fever and its associa-
tion with environmental factors in the Dhaka metropolitan area of Bangladesh: a
spatial and time-series approach. PLOS Negl Trop Dis 2013; 7:e1998.
25. BreimanRF, CosmasL, NjugunaH, etal. Population-based incidence of typhoid
fever in an urban informal settlement and a rural area in Kenya: implications for
typhoid vaccine use in Africa. PLOS One 2012; 7:e29119.
26. Getahun SA, Parry CM, Crump JA, et al. A retrospective study of patients
with blood culture–confirmed typhoid fever in Fiji during 2014–2015: epide-
miology, clinical features, treatment and outcome. Trans R Soc Trop Med Hyg
2019; 113:764–70. Available at: http://www.ncbi.nlm.nih.gov/pubmed/31638153.
Accessed 15 November 2019.
27. AzmatullahA, QamarFN, ThaverD, ZaidiAK, BhuttaZA. Systematic review of
the global epidemiology, clinical and laboratory profile of enteric fever. J Glob
Health 2015; 5:020407. doi:10.7189/jogh.05.020407
28. SharmaY, AryaV, JainS, KumarM, Deka L, Mathur A. Dengue and typhoid
co-infection- study from a government hospital in North Delhi. J Clin Diagn Res
2014; 8:DC09–11.
29. Birhanie M, Tessema B, Ferede G, Endris M, Enawgaw B. Malaria, typhoid
fever, and their coinfection among febrile patients at a rural health center in
Northwest Ethiopia: a cross-sectional study. Adv Med 2014; 2014:531074.
doi:10.1155/2014/531074
30. GilR, ÁlvarezJL, GómezC, ÁlvaroA, GilÁ. Epidemiology of typhoid and para-
typhoid fever hospitalizations in Spain (1997–2005). Hum Vaccin 2009; 5:420–4.
31. SahaS, IslamM, Sa haS, etal. Designing comprehensive public health surveillance
for enteric fever in endemic countries: importance of including different health-
care facilities. J Infect Dis 2018; 218:S227–31.
32. Saha S, Tanmoy AM, Andrews JR, et al. Evaluating PCR-based detection of
Salmonella Typhi and Paratyphi Ain the environment as an enteric fever surveil-
lance tool. Am J Trop Med Hyg 2019; 100:43–6.