Available via license: CC BY 4.0
Content may be subject to copyright.
SUPPLEMENT ARTICLE
Typhoid Heterogeneity inNepal • CID 2020:71 (Suppl 3) • S205
Clinical Infectious Diseases
Correspondence: J.R. Andrews, Division of Infectious Diseases and Geographic Medicine,
Stanford University School of Medicine, 300 Pasteur Drive, Lane Bldg, Suite 143, Room 141,
Stanford, CA 94305 (jandr@stanford.edu).
Clinical Infectious Diseases® 2020;71(S3):S205–13
© 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/ciaa1319
Spatial Heterogeneity of Enteric Fever in 2 Diverse
Communities inNepal
Dipesh Tamrakar,1 Krista Vaidya,1 AlexanderT. Yu,2 Kristen Aiemjoy,2 ShivaRam Naga,1 Yanjia Cao,2 Caryn Bern,3 Rajeev Shrestha,1 Biraj M. Karmacharya,1
Sailesh Pradhan,4 Farah Naz Qamar,5 Samir Saha,6 Kashmira Date,7 Ashley T. Longley,7,8 Caitlin Hemlock,9 Stephen Luby,2 Denise O. Garrett,9 IsaacI. Bogoch,10
and JasonR. Andrews2,
1Dhulikhel Hospital, Kathmandu University Hospital, Kavrepalanchok, Nepal, 2Division of Infectious Diseases and Geographic Medicine, School of Medicine, Stanford University, Stanford, California, USA,
3Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, California, USA, 4Kathmandu Medical College and Teaching Hospital, Kathmandu, Nepal, 5Department
of Pediatrics and Child Health, Aga Khan University, Karachi, Pakistan, 6Child Health Research Foundation, Department of Microbiology, Dhaka Shishu (Children’s) Hospital, Dhaka, Bangladesh, 7Global
Immunization Division, Centers for Disease Control and Prevention, Atlanta, Georgia, USA, 8National Foundation for the Centers for Disease Control and Prevention, Atlanta, Georgia, USA, 9Applied
Epidemiology, Sabin Vaccine Institute, Washington, DC, USA, and 10Department of Medicine, University of Toronto, Toronto, Canada
Background. Typhoid fever is endemic in the urban Kathmandu Valley of Nepal; however, there have been no population-
based studies of typhoid outside of this community in the past 3 decades. Whether typhoid immunization should be prioritized in
periurban and rural communities has been unclear.
Methods. We performed population-based surveillance for enteric fever in 1 urban catchment (Kathmandu) and 1 periurban
and rural catchment (Kavrepalanchok) as part of the Surveillance for Enteric Fever in Asia Project (SEAP). We recruited individuals
presenting to outpatient and emergency departments at 2 study hospitals with suspected enteric fever and performed blood cultures.
Additionally, we conducted a household survey in each catchment area to characterize care seeking for febrile illness. We evaluated
spatial heterogeneity in febrile illness, care seeking, and enteric fever incidence.
Results. Between September 2016 and September 2019, we enrolled 5736 participants with suspected enteric fever at 2 study
hospitals. Among these, 304 (5.3%) were culture positive for Salmonella Typhi (249 [81.9%]) or Paratyphi A(55 [18.1%]). Adjusted
typhoid incidence in Kathmandu was 484 per 100000 person-years and in Kavrepalanchok was 615 per 100000 person-years. While
all geographic areas for which estimates could be made had incidence >200 per 100000 person-years, we observed spatial heteroge-
neity with up to 10-fold variation in incidence between communities.
Conclusions. In urban, periurban, and rural communities in and around Kathmandu, we measured a high but heterogenous
incidence of typhoid. ese ndings provide some support for the introduction of conjugate vaccines in Nepal, including outside
urban areas, alongside other measures to prevent enteric fever.
Keywords. typhoid; enteric fever; Salmonella; Nepal; geospatial.
Enteric fever, caused by Salmonella enterica subspecies enterica sero-
types Typhi and Paratyphi A, B, and C, is among the leading causes
of invasive bacterial infection in South Asia [1, 2]. The World Health
Organization has recommended the introduction of typhoid conju-
gate vaccine (TCV) in settings with high typhoid incidence or high
rates of antimicrobial-resistant typhoid [3]. Health policy makers
in Nepal, as in many other countries, have been faced with deciding
whether and where to deploy TCV vaccinations with limited typhoid
incidence data outside of focused areas. As part of the Surveillance of
Enteric Fever in Asia Project (SEAP), we sought to understand the
population incidence of enteric fever in diverse communities inNepal.
Studies over the past 2 decades have revealed a high degree
of endemicity in the Kathmandu Valley, the largest urban center
in the country [4, 5]. However, the majority of typhoid studies
conducted over the past 20years have been in a single urban
center in the Kathmandu Valley, Lalitpur [4–7], which has con-
sistently shown a high burden of typhoid in children (428 cases
per 100000 person-years [PY] in the control arm of the recent
vaccine trial). ere have been no population-based typhoid
studies in Nepal outside of this community since the 1980s,
when the Vi polysaccharide vaccine was rst evaluated [8].
Even elsewhere within the Kathmandu Valley, the only avail-
able data have been through retrospective analysis of passively
collected culture data, which do not provide population-level
incidence estimates [9].
As 80% of the population of Nepal lives outside of urban
areas [10], there is need for typhoid incidence estimates from
periurban and rural communities to determine whether the
pattern of enteric fever occurrence warrants introduction of
TCVs nationwide. While data on clinically diagnosed enteric
fever cases are collected through a national reporting system,
S206 • CID 2020:71 (Suppl 3) • Tamrakar etal
such diagnoses are not based on culture conrmation, and a
recent study in the country found poor correlation between
clinically reported typhoid and culture-conrmed disease
[11]. To address this knowledge gap, we undertook prospec-
tive, population-based surveillance using blood cultures in
urban, periurban, and rural communities in Kathmandu
and Kavrepalanchok, Nepal. Here, we report on spatial het-
erogeneity in enteric fever symptoms, care-seeking, and
culture-conrmed typhoid and paratyphoid disease in these
communities.
METHODS
StudyDesign
SEAP was a prospective, population-based study conducted
in communities in Nepal, Bangladesh, and Pakistan between
September 2016 and September 2019. The study utilized a “hy-
brid surveillance” approach to estimate incidence [12], com-
bining hospital-based surveillance to identify enteric fever cases
and a community-based healthcare utilization survey to esti-
mate the proportion of enteric fever cases captured by surveil-
lance hospitals. The clinical characteristics and outcomes were
reported separately [13, 14]. Here, we report on heterogeneity
in healthcare seeking and enteric fever incidence in Nepal.
Study Site and Population
This study was conducted in 2 settings in Nepal, defined by
the catchment area for the 2 main SEAP surveillance hos-
pitals: Kathmandu Medical College and Teaching Hospital
in Kathmandu Metropolitan City and Dhulikhel Hospital in
Kavrepalanchok District, which is approximately 30 km east of
Kathmandu. Kathmandu is the capital of Nepal and has a dense,
urban population of >20000 people per square kilometer. The
Kathmandu catchment area for this study was identified based
on review of the home addresses of 100 consecutive patients
meeting SEAP clinical enrollment criteria (fever on at least
3days within the past 7days). We selected 9 wards (6, 7, 8, 9, 10,
32, 33, 34, and 35)within Kathmandu, which accounted for the
home addresses of >60% of patients meeting the study defini-
tion. In Kavrepalanchok, which has periurban and rural popu-
lations, we identified the home address of the past 100 patients
with culture-confirmed enteric fever diagnosed at Dhulikhel
Hospital and selected 4 municipalities, which accounted for
>60% of cases. These municipalities consisted of Dhulikhel
(12.1 km2; density: 1066 persons/km2), Banepa (5.6 km2; 4050
persons/km2), Paanchkhal (19.1 km2; 387 persons/km2), and
Panauti (31.7 km2; 723 persons/km2). We utilized administra-
tive geographical units (eg, wards, municipalities) rather than
physical boundaries to facilitate the determination of whether
participants arriving at the hospital lived within the catchment
area. Because the most recent available population census was
8 years old, and there had been substantial population shifts
over the interim (particularly following the 2015 earthquake),
we estimated the population from the healthcare utilization
survey as described below.
Study Procedures and Definitions
Clinical Surveillance
We prospectively enrolled participants from the 2 main SEAP
surveillance hospitals. At each hospital, we screened all pa-
tients presenting to the outpatient departments (adult and pe-
diatric) and emergency department for history of fever. Those
reporting fever on at least 3 consecutive days within the past
7days who resided within the predefined catchment area were
invited to participate in the study. In the inpatient department,
we recruited all participants who were suspected of enteric fever
by clinicians. There were no age restrictions for enrollment.
Participants were recruited 6days per week (Sunday through
Friday); those who sought care in the emergency room or in-
patient department overnight or on Saturdays were recruited
the following day if still present. After obtaining informed con-
sent, we administered a standardized questionnaire to ascertain
demographic and clinical information. We collected 2–10mL
of peripheral blood, which were inoculated into BACTEC
Aerobic or Peds Plus bottles and incubated using a BACTEC
automated system for up to 5days. Bottles showing growth were
subcultured on sheep blood agar and MacConkey agar, and bi-
ochemical and antisera testing was performed to identify iso-
lates. We defined typhoid and paratyphoid cases as individuals
with positive blood cultures for Salmonella Typhi or Salmonella
Paratyphi A, B, or C. Additionally, we established a network
of 7 microbiology laboratories in Kathmandu (Bir Hospital,
Helping Hands Community Hospital, Nepal Medical College,
Kathmandu Model Hospital, Alka Hospital, Kanti Children’s
Hospital, and Nepal Police Hospital). We contacted all patients
identified at these laboratories with culture-confirmed enteric
fever cases for enrollment in the study.
Healthcare UtilizationSurvey
We conducted a household-based healthcare utilization survey
in both catchment areas [15, 16]. In brief, we used grid-based
random sampling to select geographic clusters, from which all
households were approached. Astandardized questionnaire was
administered to the head of all consenting households, with the
primary objective of determining the proportion of individuals
with a typhoid-like illness (fever ≥3days) who sought care at
the 2 surveillance hospitals. We inquired about any healthcare
seeking for fever ≥3days that occurred within the past 8 weeks,
as well as any hospitalization for fever within the past year, rea-
soning that recall accuracy would be longer for hospitalization.
Analytic Approach
We characterized enteric fever cases by demographic variables,
reporting median and interquartile range (IQR) for continuous
Typhoid Heterogeneity inNepal • CID 2020:71 (Suppl 3) • S207
variables and proportions for dichotomous variables. For the
healthcare utilization survey, we used mixed effects logistic re-
gression models with random effects for cluster to estimate the
proportion seeking care at the study site. We performed spatial
interpolation to estimate local density of fever and care seeking
using an inverse distance weighted model. We assessed spatial
risk factors for fever and care seeking using logistic regres-
sion, and investigated the effects of community water sources
and wealth. We defined improved water sources as municipally
supplied water or that purchased from a vendor, and unim-
proved water sources as surface waters, rainwater, public taps,
or groundwater. We created a household wealth index using
principal components analysis using the following assets: elec-
tricity, ownership of radio, television, landline telephone, mo-
bile phone, computer, watch, bicycle, motorcycle, car, and bank
account. We classified administrative areas (wards and munici-
palities) by population density as urban (≥5000 persons/km2),
periurban (1000–4999 persons/km2), or rural (<1000 persons/
km2).
We used OpenStreetMap (accessed in October 2019
using QGIS 3.8) to build a street network and estimated
road distance from household to surveillance hospital using
Origin-Destination Matrix in ArcGIS 10.7.1 (Esri, Redlands,
California). To assess the relationship between distance and
care seeking, we t both generalized linear models and gener-
alized additive models, both with logistic link functions, and
compared models by Akaike Information Criteria and χ
2 good-
ness oft.
e primary objective of this study was to estimate incidence
of typhoid and paratyphoid fever in various geographic areas.
We have previously described the overall analytic approach
used for this study [12]. We rst estimated the crude, culture-
conrmed typhoid and paratyphoid incidence by dividing all
culture-conrmed cases by the catchment population and du-
ration of the study. e catchment population and population
of each administrative area was estimated by dividing the meas-
ured population from the healthcare utilization survey by the
cluster sampling fraction and the response rate. In the crude
Figure 1. Population density in the Kathmandu wards and Kavrepalanchok municipalities comprising the study catchment area. Yellow crosses denote the location of study
surveillance hospitals. Abbreviation: SEAP, Surveillance for Enteric Fever in Asia Project.
S208 • CID 2020:71 (Suppl 3) • Tamrakar etal
estimates, we included all cases who resided in the catchment
area that were identied through surveillance at the hospital
and laboratory networksites.
We then undertook several adjustments to estimate the true
incidence of typhoid and paratyphoid. For the adjusted esti-
mates, we excluded cases enrolled from the laboratory net-
work, as systematic active surveillance was not undertaken at
those facilities. First, we adjusted for blood culture sensitivity,
estimating it to be 59% (95% condence interval [CI], 54%–
64%) [17], and dividing our crude estimate by this number.
Second, we adjusted to account for eligible patients who were
missed by surveillance, either because they presented during
evenings, weekends, or when the study sta were unavail-
able, consented to enrollment but did not have a blood cul-
ture obtained, or because they were approached but declined
to participate. We divided our estimate by the proportion of
cases that were captured by the surveillance study. Finally, we
adjusted for individuals who sought care at sites other than the
surveillance sites by dividing by the proportion seeking care
at the study site, stratied by each age group and geographical
unit. We propagated uncertainty in culture sensitivity and care
seeking through Monte Carlo sampling from distributions es-
timated for each of these parameters and generated 95% cred-
ible intervals for incidence estimates.
All analyses and geo-visualization were performed using
ArcGIS 10.7.1 (Esri) and R soware.
Ethics Statement
All participants provided informed consent. For participants
under age 18, a parent or guardian provided informed con-
sent, and those 7–17years of age provided assent. The study
was approved by the institutional review boards at Kathmandu
University School of Medical Sciences, Stanford University,
the Nepal Health Research Council, and through local ethics
boards at participating hospitals.
Figure 2. Proportion of individuals with fever in the past 8 weeks (top) and hospitalized for fever in the past year (bottom) for Kathmandu and Kavrepalanchok. Yellow
crosses denote the location of study surveillance hospitals.
Typhoid Heterogeneity inNepal • CID 2020:71 (Suppl 3) • S209
RESULTS
Between September 2016 and September 2019, we enrolled 5667
participants with suspected enteric fever at the 2 study hos-
pitals (Dhulikhel Hospital: 2434; Kathmandu Medical College
Hospital: 3233). Among these, 4827 (85.2%) were enrolled from
outpatient or emergency departments, and the remaining 841
were enrolled from inpatient wards. Overall, 304 (5.4%) of these
patients were culture positive for enteric fever. Additionally, we
identified 1296 cases from the laboratory network sites. Among
all culture-confirmed cases, 1366 (85.4%) were Salmonella Typhi
and 234 (14.6%) were Salmonella Paratyphi A. The majority
(58.9%) of cases occurred among males. The median age of S.
Paratyphi Apatients (21 [IQR, 17–26] years) was slightly higher
than that of Typhi patients (19.5 [IQR, 15–24] years; P = .0003).
Approximately one-quarter of typhoid cases (369/1369 [27.0%])
occurred among children <16 years of age, and more than
half (720/1369 [52.6%]) occurred among individuals aged
16–25years. Age distribution of typhoid cases did not differ be-
tween Kathmandu and Kavrepalanchok (P = .414).
We conducted healthcare utilization surveys continuously
for 24months, from January 2017 through December 2018. We
enrolled 16 744 households in Kathmandu (covering 50039
participants) and 8729 households in Kavrepalanchok (34041
participants) (Figure 1). Among those in Kavrepalanchok,
59% were in periurban areas and 41% were in rural areas. e
most common drinking water source in urban areas was de-
livery by truck (38.0%) followed by piped into house (23.9%).
In periurban areas, the most common drinking water source
was water piped into the household (61%), and in rural areas,
surface waters were the most common source of drinking
water (38.6%). Household toilet ownership was high (>95%)
in allareas.
Among enrolled households, 919 (5.4%) in Kathmandu and
1353 (15.5%) in Kavrepalanchok, respectively, reported at least
1 member with febrile illness within the past 8 weeks, and 188
Figure 3. Proportion of individuals with fever in past 8 weeks (top) or hospitalized in past month (bottom) who sought care at the study hospitals in Kathmandu and
Kavrepalanchok. Yellow crosses denote the location of study surveillance hospitals.
S210 • CID 2020:71 (Suppl 3) • Tamrakar etal
(1.1%) and 324 (3.7%) reported 1 member with hospitalization
for fever within the past year. We found spatial heterogeneity
in the proportion of households reporting febrile illness in the
past 8 weeks (Figure 2). Households in communities lacking
improved water (odds ratio [OR], 2.26; P < .0001) and lower
wealth quintiles (OR, 1.06 per quintile; P = .008) were more
likely to report a member having fever in the past 8weeks.
In Kathmandu, 6.6% of individuals with fever in the past
week and 4.2% of individuals hospitalized for fever in the past
year sought care at the study site. In Kavrepalanchok, 13.3% of
individuals with fever and 12.6% of individuals hospitalized for
fever sought care at the study site. We noted substantial heter-
ogeneity across each catchment area in the proportion seeking
care at the study site (Figure3). We tested how geodesic distance
and road length between households and surveillance hospitals
aected probability of seeking care at that study site. We found
that road length was a better predictor of care seeking than ge-
odesic distance for both catchment areas (P < .0001 for model
comparisons), and that the relationship was nonlinear (Figure4).
In both communities, participants in the rst quintile of road dis-
tance were far more likely to seek care at the study site compared
with those in the top 3 quintiles for distance (Kathmandu: 17.7%
vs 3.1%; P < .0001; Kavrepalanchok: 34.2% vs 7.0%; P < .0001).
e crude incidence of blood culture–conrmed typhoid
and paratyphoid fever was 31 (95% CI, 26–37) and 6 (95% CI,
4–9) cases per 100000 PY, respectively, in Kathmandu and 36
(95% CI, 24–51) and 7 (95% CI, 3–16) cases per 100000 PY,
respectively, in Kavrepalanchok. Aer adjusting for culture sen-
sitivity, enrollment capture, and care-seeking, we found a high
incidence of typhoid in both communities (484 [95% CI, 384–
612] per 100000 PY in Kathmandu; 615 [95% CI, 527–721] per
100000 PY in Kavrepalanchok). In 1 ward of Kathmandu (ward
6), which was furthest from the surveillance hospital, only 25
participants were enrolled in clinical surveillance and none had
typhoid. For the remaining 8 wards in Kathmandu and 4 muni-
cipalities in Kavrepalanchok, estimated typhoid incidence was
>200 per 100000 PY (Figure 5). e highest incidence was in
wards 7 and 33 (2510 per 100000 PY and 2661 per 100 000
PY), and the lowest incidence occurred in ward 34 (223 per
100000 PY) and Panauti municipality (309 per 100000 PY).
Paratyphoid incidence was 117 (95% CI, 93–148) per 100000
PY in Kathmandu and 105 (95% CI, 90–123) cases per 100000
PY in Kavrepalanchok, with substantial geographic heteroge-
neity ranging between 642 per 100000 ward 10 to 0 per 100000
in ward 33 of Kathmandu.
DISCUSSION
In this population-based study in 2 areas in Nepal, we found an
overall high incidence of enteric fever, >200 cases per 100000
Figure 4. Probability of seeking care at study site as a function of road distance between household and the study site for Kathmandu (top) and Kavrepalanchok (bottom).
Typhoid Heterogeneity inNepal • CID 2020:71 (Suppl 3) • S211
PY in all geographic areas for which we were able to make es-
timates. We also noted substantial heterogeneity in incidence,
ranging from just over 200 cases per 100 000 PY to >2000
cases per 100000 PY. Incidence of typhoid fever was 4–6 times
greater than that of paratyphoid fever, which was an average of
80–100 cases per 100000 PY. While individuals living far from
study sites were less likely to seek care for febrile illness, after
adjusting for this we did not detect a consistent relationship be-
tween population density and typhoid incidence. These find-
ings suggest that there is substantial burden of typhoid even in
periurban and rural areas outside of Kathmandu.
Over the past 20years, virtually all of the prospective en-
teric fever studies in Nepal have been conducted in Lalitpur, a
metropolitan city in the Kathmandu Valley [4–7]. e placebo
arm of the recent TCV trial conducted in this community
demonstrated a high incidence (428 per 100000 PY) among
children 6months to 15years of age [18]. Population-based
data outside of this community are lacking. One recent study
found that cases of clinically diagnosed enteric fever reported
through the national Health Management Information System
had grown over the past 15years, reaching reported rates of
1800 per 100000 PY, highest in the rural areas [11]. However,
in prospective surveillance at sites with blood culture ca-
pacity, only 4.1% of those clinically diagnosed with typhoid
had positive blood cultures for typhoidal Salmonella, and this
rate ranged between 0 and 2.8% in rural areas of Nepal [11].
Culture positivity in that study was strongly, positively asso-
ciated with population density. However, the earlier study in-
cluded fever patients from all rural areas in the catchment of
Dhulikhel Hospital. Within the more limited catchment pop-
ulation of the present study, we found no consistent dier-
ences in typhoid incidence between the urban (Kathmandu),
periurban (Banepa, Dhulikhel) and rural (Paanchkhal,
Panauti) communities. In other settings, some studies have
shown higher incidence in urban than rural areas [19, 20],
whereas others have found the converse [21].
The study catchment area for the periurban and rural area
communities was selected based on where enteric fever cases
Figure 5. Incidence (cases per 100000 person-years) of typhoid (A) and paratyphoid (B) by ward in Kathmandu and by municipality in Kavrepalanchok. Yellow crosses de-
note the location of study surveillance hospitals. Abbreviation: SEAP, Surveillance for Enteric Fever in Asia Project.
S212 • CID 2020:71 (Suppl 3) • Tamrakar etal
had been detected in the previous 2years, which biased to-
ward selection of higher-risk communities. Furthermore,
Nepal has highly diverse ecosystems, from subtropical com-
munities in the South to alpine villages in the Himalaya. The
communities selected for this study were all in 1 province
and are not likely to be representative of the typhoid risk in
areas throughout the country. Nevertheless, this study adds
to what was a sparse collection of population-based data on
enteric fever in the country. Amajor obstacle to generating
such data is the resource-intensive nature of population-
based surveillance systems; emerging approaches including
seroepidemiology and environmental surveillance may
enable more efficient typhoid risk mapping in resource-
constrained settings [22].
We observed moderate spatial heterogeneity in febrile ill-
ness in both catchment areas, with overall higher rates of fever
in the rural communities. As anticipated, care seeking for fe-
brile illness at study sites in both communities was strongly
predicted by road distance from the site. We found that most
of this eect occurred at smaller distances, and that aer a
point, distance did not further aect the probability of seeking
care at the site. Understanding these heterogeneities can be
important to the design and interpretation of surveillance
approaches.
The median age of enteric fever cases in our study was
20years, and nearly half of all cases occurred among indi-
viduals between the ages of 16 and 25years. This age distri-
bution is higher than reported in some neighboring South
Asian countries, but similar to what has been reported in
Lalitpur (median ages of 16 and 20years for typhoid and
paratyphoid, respectively) [4]. In retrospective data reported
from Nepal and India for phase 1 of the SEAP study, the
median age of typhoid was 19years in Nepal and 24years
in India [23]. In these settings, catch-up vaccination cam-
paigns among older children, adolescents, and young adults
may be important for addressing the burden of typhoid and
preventing transmission.
ese results should be interpreted within the context of the lim-
itations of the study design and available data. We used healthcare
utilization data to adjust for the proportion of patients with en-
teric fever who did not seek care at the study site; it is possible that
individuals who sought care at the study site diered from those
who did in ways that were related to their risk of typhoid. One ap-
proach to handle this confounding is inverse probability weighting
based on household characteristics; while we performed this ap-
proach for the overall area estimates, we were unable to do this at
ner geographical resolution due to lack of power. We made esti-
mates based on administrative boundaries (wards, municipalities),
because this address information was available for all participants
enrolled in clinical surveillance. ere are no conventional street
addresses in Nepal, which precluded more precise geocoding of
participants based on the available data. Our estimates are based on
the participant’s household address, but typhoid may be acquired
outside the household, so the geospatial estimates may not reect
where transmission occurs. We assumed a xed blood culture sen-
sitivity for all participants based on a previous meta-analysis [17]; in
reality, this may vary by age, blood volume, and prior antibiotic use.
We opted for more parsimonious and data-driven adjustments to
our incidencemodel.
In conclusion, we found a high incidence of typhoid fever
in urban, periurban, and rural communities in and nearby
to Kathmandu, with no clear relationship between typhoid
incidence and population density. e incidence in all com-
munities was above previously proposed thresholds at which
typhoid vaccination would be deemed cost-eective [24].
ese ndings support the introduction of TCV in Nepal,
alongside improved water and sanitation interventions, to
prevent both typhoid and paratyphoid fever in communities
throughout the country.
Notes
Acknowledgments. The authors are grateful for the efforts of the
Surveillance for Enteric Fever in Asia Project (SEAP) Nepal research
assistants, field team, laboratory staff, and laboratory networkteams.
Disclaimer. e ndings and conclusions in this study are those of the
authors and do not necessarily reect the position of the Centers for Disease
Control and Prevention.
Financial support. is work was supported by the Bill & Melinda Gates
Foundation (BMGF) (award number 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 all acknowledge grant support
from BMGF. K.D.reports a conditional gi agreement between the Sabin
Vaccine Institute (primary grantee for the SEAP grant from BMGF) and the
CDC Foundation. S.S.reports grants from the World Health Organization
(WHO) during the conduct of the study, and grants from the WHO, BMGF,
GlaxoSmithKline, Pzer, Sano Pasteur, and Edinburgh University, out-
side the submitted work. All other authors report no potential conicts
of interest. All authors have submitted the ICMJE Form for Disclosure of
Potential Conicts of Interest. Conicts that the editors consider relevant to
the content of the manuscript have been disclosed.
References
1. AndrewsJR, HarrisJB, RyanET. Typhoid fever, paratyphoid fever, and typhoidal
fevers. In: Mandell, Douglas, and Bennett’s principles and practice of infectious
diseases. Philadelphia, PA: Elsevier, 2020:1365–79.e3.
2. GBD 2017 Typhoid and Paratyphoid Collaborators. The global burden of typhoid
and paratyphoid fevers: a systematic analysis for the Global Burden of Disease
Study 2017. Lancet Infect Dis 2019; 19:369–81.
3. World Health Organization. Typhoid vaccines: WHO position paper—March
2018. Wkly Epidemiol Rec 2018; 93:153–72.
4. KarkeyA, Arjyal A, AndersKL, et al. The burden and characteristics of enteric
fever at a healthcare facility in a densely populated area of Kathmandu. PLoS One
2010; 5:e13988.
5. ZellwegerRM, BasnyatB, ShresthaP, etal. A 23-year retrospective investigation
of Salmonella Typhi and Salmonella Paratyphi isolated in a tertiary Kathmandu
hospital. PLoS Negl Trop Dis 2017; 11:e0006051.
6. Baker S, Holt KE, ClementsACA, et al. Combined high-resolution genotyping
and geospatial analysis reveals modes of endemic urban typhoid fever transmis-
sion. Open Biol 2011; 1:110008.
7. ThompsonCN, KarkeyA, DongolS, etal. Treatment response in enteric fever in
an era of increasing antimicrobial resistance: an individual patient data analysis
Typhoid Heterogeneity inNepal • CID 2020:71 (Suppl 3) • S213
of 2092 participants enrolled into 4 randomized, controlled trials in Nepal. Clin
Infect Dis 2017; 64:1522–31.
8. AcharyaIL, LoweCU, ThapaR, etal. Prevention of typhoid fever in Nepal with
the Vi capsular polysaccharide of Salmonella Typhi. Apreliminary report. N Engl
J Med 1987; 317:1101–4.
9. BajracharyaD, KhanMI, PachA 3rd, etal. 25years after Vi typhoid vaccine effi-
cacy study, typhoid affects significant number of population in Nepal. PLoS One
2014; 9:e77974.
10. World Bank. World Bank Open Data: free and open access to data about devel-
opment in countries around the globe. Available at: data.worldbank.org. Accessed
10 February 2014.
11. Andrews JR, VaidyaK, Bern C, et al. High rates of enteric fever diagnosis and
lower burden of culture-confirmed disease in peri-urban and Rural Nepal. J
Infect Dis 2018; 218:214–21.
12. AndrewsJR, BarkumeC, YuAT, etal. Integrating facility-based surveillance with
healthcare utilization surveys to estimate enteric fever incidence: methods and
challenges. J Infect Dis 2018; 218:268–76.
13. Aiemjoy K, TamrakarD, Saha S, et al. Diagnostic value of clinical features to
distinguish enteric fever from other febrile illnesses in Bangladesh, Nepal, and
Pakistan. Clin Infect Dis 2020; 71:S257–65.
14. LongleyAT, HemlockC, DateK, etal. Illness severity and outcomes among
enteric fever cases from Bangladesh, Nepal, and Pakistan: data from the sur-
veillance for enteric fever in Asia project, 2016–2019. Clin Infect Dis 2020;
71:S222–31.
15. YuAT, ShakyaR, AdhikariB, etal. A cluster-based, spatial-sampling method for
assessing household healthcare utilization patterns in resource-limited settings.
Clin Infect Dis 2020; 71:S239–47.
16. AndrewsJR, VaidyaK, SahaS, et al. Healthcare utilization patterns for acute fe-
brile illness in Bangladesh, Nepal and Pakistan: results from the surveillance for
enteric fever in Asia project. Clin Infect Dis 2020; 71:S248–56.
17. Antillon M, Saad NJ, BakerS, Pollard AJ, PitzerVE. The relationship between
blood sample volume and diagnostic sensitivity of blood culture for typhoid
and paratyphoid fever: a systematic review and meta-analysis. J Infect Dis 2018;
218:255–67.
18. Shakya M, Colin-Jones R, Theiss-Nyland K, et al; TyVAC Nepal Study Team.
Phase 3 efficacy analysis of a typhoid conjugate vaccine trial in Nepal. N Engl J
Med 2019; 381:2209–18.
19. SahaSK. Multiple modalities to explore typhoid among children: implication in
vaccine policy. Dhaka: 2013. Available at: http://www.coalitionagainsttyphoid.
org/sites/default/files/05.Saha_.8TC.pdf. Accessed 19 May 2013.
20. 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.
21. Cruz Espinoza LM, Nichols C, Adu-Sarkodie Y, et al. Variations of invasive
Salmonella infections by population size in Asante Akim North Municipal,
Ghana. Clin Infect Dis 2016; 62(Suppl 1):S17–22.
22. Andrews JR, Yu AT, Saha S, et al. Environmental surveillance as a tool for
identifying high-risk settings for typhoid transmission. Clin Infect Dis 2020;
71:71–8.
23. BarkumeC, DateK, SahaSK, etal. Phase Iof the surveillance for enteric fever in Asia
Project (SEAP): an overview and lessons learned. J Infect Dis 2018; 218:188–94.
24. Lo NC, Gupta R, Stanaway JD, et al. Comparison of strategies and incidence
thresholds for Vi conjugate vaccines against typhoid fever: a cost-effectiveness
modeling study. J Infect Dis 2018; 218:232–42.