Content uploaded by Joshua Longbottom
Author content
All content in this area was uploaded by Joshua Longbottom on Jul 13, 2018
Content may be subject to copyright.
www.thelancet.com Published online July 12, 2018 http://dx.doi.org/10.1016/S0140-6736(18)31224-8
1
Articles
Vulnerability to snakebite envenoming: a global mapping of
hotspots
Joshua Longbottom, Freya M Shearer, Maria Devine, Gabriel Alcoba, Francois Chappuis, Daniel J Weiss, Sarah E Ray, Nicolas Ray, David A Warrell,
Rafael Ruiz de Castañeda, David J Williams, Simon I Hay, David M Pigott
Summary
Background Snakebite envenoming is a frequently overlooked cause of mortality and morbidity. Data for snake
ecology and existing snakebite interventions are scarce, limiting accurate burden estimation initiatives. Low global
awareness stunts new interventions, adequate health resources, and available health care. Therefore, we aimed to
synthesise currently available data to identify the most vulnerable populations at risk of snakebite, and where
additional data to manage this global problem are needed.
Methods We assembled a list of snake species using WHO guidelines. Where relevant, we obtained expert opinion
range (EOR) maps from WHO or the Clinical Toxinology Resources. We also obtained occurrence data for each snake
species from a variety of websites, such as VertNet and iNaturalist, using the spocc R package (version 0.7.0). We
removed duplicate occurrence data and categorised snakes into three groups: group A (no available EOR map or
species occurrence records), group B (EOR map but <5 species occurrence records), and group C (EOR map and
≥5 species occurrence records). For group C species, we did a multivariate environmental similarity analysis using
the 2008 WHO EOR maps and newly available evidence. Using these data and the EOR maps, we produced
contemporary range maps for medically important venomous snake species at a 5 × 5 km resolution. We subsequently
triangulated these data with three health system metrics (antivenom availability, accessibility to urban centres, and
the Healthcare Access and Quality [HAQ] Index) to identify the populations most vulnerable to snakebite morbidity
and mortality.
Findings We provide a map showing the ranges of 278 snake species globally. Although about 6·85 billion people
worldwide live within range of areas inhabited by snakes, about 146·70 million live within remote areas lacking
quality health-care provisioning. Comparing opposite ends of the HAQ Index, 272·91 million individuals (65·25%) of
the population within the lowest decile are at risk of exposure to any snake for which no eective therapy exists
compared with 519·46 million individuals (27·79%) within the highest HAQ Index decile, showing a disproportionate
coverage in reported antivenom availability. Antivenoms were available for 119 (43%) of 278 snake species evaluated
by WHO, while globally 750·19 million (10·95%) of those living within snake ranges live more than 1 h from
population centres. In total, we identify about 92·66 million people living within these vulnerable geographies,
including many sub-Saharan countries, Indonesia, and other parts of southeast Asia.
Interpretation Identifying exact populations vulnerable to the most severe outcomes of snakebite envenoming at a
subnational level is important for prioritising new data collection and collation, reinforcing envenoming treatment,
existing health-care systems, and deploying currently available and future interventions. These maps can guide future
research eorts on snakebite envenoming from both ecological and public health perspectives and better target future
estimates of the burden of this neglected tropical disease.
Funding Bill & Melinda Gates Foundation.
Copyright © 2018 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license.
Published Online
July 12, 2018
http://dx.doi.org/10.1016/
S0140-6736(18)31224-8
See Online/Comment
http://dx.doi.org/10.1016/
S0140-6736(18)31328-X
Big Data Institute, Li Ka Shing
Centre for Health Information
and Discovery
(J Longbottom MSc,
F M Shearer BSc, M Devine MSc,
D J Weiss PhD), and Nuffield
Department of Clinical
Medicine (D A Warrell FMedSci),
University of Oxford, Oxford,
UK; Centre for Health
Informatics, Computing and
Statistics, Lancaster Medical
School, Lancaster University,
Lancaster, UK (J Longbottom);
Department of Vector Biology,
Liverpool School of Tropical
Medicine, Liverpool, UK
(J Longbottom); Division of
Tropical and Humanitarian
Medicine, University Hospitals
of Geneva, Geneva,
Switzerland (G Alcoba MD,
F Chappuis MD); Division of
Tropical Medicine and
Neglected Tropical Diseases,
Médecins Sans Frontières,
Geneva, Switzerland
(G Alcoba); Institute for Health
Metrics and Evaluation,
University of Washington,
Seattle, WA, USA (S E Ray BS,
Prof S I Hay FMedSci,
D M Pigott DPhil); EnviroSPACE
Lab, Institute for
Environmental Sciences
(N Ray PhD), and Institute of
Global Health, Faculty of
Medicine
(R Ruiz de Castañeda PhD,
N Ray), University of Geneva,
Geneva, Switzerland; and
Australian Venom Research
Unit, Department of
Pharmacology and
Therapeutics, University of
Melbourne, Melbourne, VIC,
Australia (D J Williams PhD)
Introduction
Snakebite envenoming is a frequently overlooked
cause of mortality and morbidity, responsible for
81 000–138 000 deaths annually,1,2 and between 421 000 and
1·2 million envenomings.3 Contact from venomous
snakes, spiders, and scorpions contribute to 1·2 million
years of life lived with disability annually.4 The burden
remains poorly characterised because of under-reporting;
as snakebite is rarely notifiable, existing estimates are
typically derived from extrapolated hospital records
and community surveys.5 Snakebite primarily aects the
poor rural communities of Asia and sub-Saharan Africa,
where socioeconomic status and agricultural and other
practices contribute to increased snake–human inter-
action.6 Venomous snakebites can also inflict a heavy
burden on livestock, creating economic hardship for
already impoverished communities.7 Medically important
snake species, however, have a cosmopolitan distribution,
making snakebite a global challenge.3
In June, 2017, snakebite envenoming was classified as a
category A neglected tropical disease,8,9 and was
the subject of a resolution passed by the World Health
Articles
2
www.thelancet.com Published online July 12, 2018 http://dx.doi.org/10.1016/S0140-6736(18)31224-8
Correspondence to:
Prof Simon I Hay, Institute for
Health Metrics and Evaluation,
University of Washington,
Seattle, WA 98121, USA
sihay@uw.edu
or
Mr Joshua Longbottom,
Department of Vector Biology,
Liverpool School of Tropical
Medicine, Liverpool L3 5QA, UK
joshua.longbottom@lstmed.
ac.uk
Assembly in May, 2018. Consequently, there is a renewed
impetus to accurately assess the burden and distribution
of snakebite to ensure appropriate prevention and control
interventions are implemented, and that adequate
resources and funding are allocated nationally and
subnationally.10,11 For other neglected tropical diseases,
substantive global targets exist: Sustainable Development
Goal target 3·3 aims to “end the epidemics” of these
diseases by 2030,12,13 with routine reporting, surveillance,
and notification architecture in place. As a new neglected
tropical disease, snakebite monitoring and evaluation
should reflect these objectives.
Data for the presence of venomous snakes and
occurrence of snakebites are sparse and incomplete
at the global level, making estimation challenging.14,15
Although some countries have done household-level
surveys to determine the incidence of snakebites,14,15
the global magnitude of this disease remains poorly
characterised. Snakebite envenoming represents an
interesting One Health challenge requiring clinical,
ecological, and public health expertise. Consequently, this
issue can be approached by considering vulnerability to
snakebite envenoming as a nexus of ecological contexts
and public health weaknesses, to provide an evidence
base for targeting future quantitative studies.
Clinical challenges involve appropriate case diagnosis
and adequate provisioning of care whether supportive
(such as ventilators) or direct treatment with antivenom,
which might not be available at any given point of
care.2,16 Ascertainment of the correct antivenom can be
challenging,17 and current diagnostics can be expensive
and slow.18,19 Furthermore, nearly half of venomous snakes
do not have antivenoms available as tracked by WHO.2,20
To comprehensively address snakebites, these clinical
challenges need to be considered within an ecological
context by understanding snake behaviour and life-history
traits that contribute to the frequency and geographical
distribution of snakebites. Therefore, by contextualising
contemporary knowledge about snake distributions with
indicators of the quality of health-care provisioning,21
the accessibility of these resources,22 and antivenom
availability,20 we aimed to identify populations vulnerable
to the worst health outcomes of an envenoming event.
Methods
Study overview
We evaluated range maps for 278 snakes to consider
their presence at a 5 × 5 km (grid cell) resolution. To
identify the most vulnerable populations, this ecological
information was paired with three key metrics: the
Research in context
Evidence before this study
Snakebite envenoming is a category A neglected tropical disease
of particular public health importance in tropical areas of Africa,
Asia, Latin America, and Papua New Guinea. It is estimated that
up to 1·2 million people are envenomed annually, resulting in
81 000–138 000 fatalities. Although effective therapies exist to
treat envenoming by some snakes of highest medical
importance, there are many species without such treatments.
The global distribution of venomous snakes and vulnerable
populations remains inadequately characterised; therefore, the
lack of knowledge of subnational disease burden might impede
production of antivenom supplies and distribution efforts
among populations currently at risk. To investigate this further,
we searched for articles on PubMed published before
March 1, 2017, using the search terms “snakebite”, “distribution”,
and “burden”. Contemporary studies have investigated
venomous snake distributions and snakebite risk at national
levels (several countries in Latin America) or subnational levels
(India, Nigeria, and Sri Lanka), but these studies did not
encompass all medically important snake species and are limited
in both geographical extent and spatial resolution. A more
recent analysis mapped the distribution of venomous snakes in
Central America and Latin America but was restricted to widely
studied species with ample occurrence data. Although an
important start, no study has coupled global ecological
information about snake distributions with measures relating to
public health capabilities to hone in on populations most
vulnerable to this cause of mortality and morbidity.
Added value of this study
We identified populations most vulnerable to 278 medically
important snake species by using expert opinion, species’ ranges
refined by publicly available occurrence data and multivariate
analyses, information about effective therapies, and metrics of
health-care quality and accessibility. Although a large proportion
of the world’s population live in areas where such snakes could be
present, proxy metrics such as the Healthcare Access and Quality
Index and urban accessibility paired with broad-scale information
about market antivenom availability provide a subnationally
resolved yet globally comprehensive picture of vulnerability,
highlighting populations that could be most affected.
Implications of all the available evidence
We highlight locations where the combination of the presence
of a variety of venomous snakes, inequalities in health care and
accessibility, and possible absence of effective therapy might
contribute toward increased vulnerability of snakebite
envenoming. Our analyses can be used to inform the
positioning of local-scale household surveys to assess the true
risk of snakebite in areas where such estimates are currently
inadequate. This study highlights the importance of continuing
to iterate, improve, and re-evaluate existing geographical
assessments of snake distributions, and the need to incorporate
spatially heterogeneous risk within future burden estimation
efforts. This work is a first step in trying to identify and assist
the most neglected populations of this newly designated
neglected tropical disease.
Articles
www.thelancet.com Published online July 12, 2018 http://dx.doi.org/10.1016/S0140-6736(18)31224-8
3
market availability of antivenom therapies as reported
by WHO,20 accessibility to urban centres as a proxy for
access to health care,22 and the Healthcare Access and
Quality (HAQ) Index as a proxy for adequacy and
ecacy of medical interventions at health-care centres.21
Figure 1 shows conceptually how populations lacking in
all these measures should be seen as the most vulnerable
populations, and how these measures could vary
geographically.
Global list of snake species
We assembled a list of snake species, using WHO
guidelines for venomous snake species of medical
importance (hereafter referred to as snakes),23 which
define two tiers of medical importance that reflect
both ecological knowledge on propensity to interact
with humans and clinical grading of toxicity. Category
one species are common or widespread snakes that
result in high morbidity, disability, or mortality. Cate-
gory two species are snakes capable of causing mor-
bidity, disability, or death, or for which epidemiological
or clinical data are missing or are less frequently
implicated.
Where relevant, expert opinion range (EOR) maps
were obtained from WHO blood products online
database or the Clinical Toxinology Resources data-
base.20,24 Occurrence data for each species were obtained
from the Global Biodiversity Information Framework,
VertNet (version 2016-09-29), iNaturalist, iDigBio, and
Ecoengine, using the spocc R package (version 0.7.0) on
May 29, 2017.25 Duplicate records based upon shared
collection year and latitude or longitude and those
missing latitude or longitude were removed.
Given the availability of data, we placed snakes into
three groups: group A (no available EOR map or species
occurrence records), group B (EOR map but <5 species
occurrence records), and group C (EOR map and
≥5 species occurrence records). Group A species (n=9)
were excluded from this analysis because of the absence
of geographical information, reducing our species
inclusion list to 278 (99 group B species and 179 group C
species; appendix 1).
Multivariate environmental similarity surface
generation and species’ ranges
For group C species with sucient occurrence records,
potential updates to the EOR maps were assessed. EOR
maps were updated with data that has become publicly
accessible since publication of the WHO EOR maps in
2008. Multivariate environmental similarity method was
applied to the occurrence records obtained for group C
species, situated within the EOR, allowing for rapid
classification of occurrence records outside of the EOR
within the environmental range of other records (ie,
interpolation) or beyond these limits (ie, extrapolation).
Multivariate environmental similarity surfaces (MESS)
measure the similarity between new environments
(records outside of the EOR) and those in the training
sample (records within the EOR), by identifying the
maximum and minimum values of environmental data
within the training sample, with respect to a set of
predictor variables (covariates).26 We fitted species-specific
MESS using occurrence records within the EOR, and
eight bioclimatic covariates thought to influence snake
distribution (appendix 2 provides information about the
MESS parameters and covariate specifics).
Figure 1: Conceptual overview of vulnerability to snakebite envenoming
(A) Vulnerability can be considered as the intersection of populations who live
within the range of venomous snakes that have no antivenoms available,
cannot easily access health care, and have poor quality health care in delivery of
antivenoms or ensuring necessary stocks. The intersection of all three defines
the most vulnerable populations. (B) These factors vary in space. By overlaying
these features, the most vulnerable populations can be identified spatially
(represented here by the boxes outlined in black).
No
antivenom
Poor quality
health care
Inaccessible
health care
Suitable habitat for snakes
No antivenom
Poor quality health care
Inaccessible health care
Suitable habitat for snakes
A
B
For the Global Biodiversity
Information Framework see
https://www.gbif.org/
For more on VertNet see
http://vertnet.org/index.html
For more on iNaturalist see
https://www.inaturalist.org
For more on iDigBio see
https://www.idigbio.org/
For more on Ecoengine see
https://ecoengine.berkeley.edu/
See Online for appendix 1
See Online for appendix 2
Articles
4
www.thelancet.com Published online July 12, 2018 http://dx.doi.org/10.1016/S0140-6736(18)31224-8
Occurrence records outside of the currently accepted
EOR were overlaid on top of each species-specific threshold
MESS. Records located within cells of environ mental
interpolation (termed MESS-positive) were considered
valid records of species occurrence. Proposed ranges
were developed to encompass all valid MESS-positive
records, generated by applying a buer radius of 0·898°
(approx imately 100 km at the equator) to each MESS-
positive record to address potential movement of species,
and possible geopositioning errors.27,28 Buered locations
were masked by the threshold MESS to remove areas
of environmental extrapolation, and merged with the
currently accepted EOR to produce a proposed con-
temporary range.
Global distribution of snakes
To reflect the geographical diversity of the snakes
studied, we aggregated the ranges of dierent species.
Modified (ie, group C species with MESS-positive
records [n=96]) or original EOR surfaces (group B
and group C species with no MESS-positive records
[n=182]) were converted into 5 × 5 km raster (gridded)
files. They were then stacked by summing overlapping
cell values, resulting in three composite output layers:
a count of the number of unique category one or
category two species per cell, or both; a count of the
number of unique category one species per cell; and a
count of the number of unique category two species
per cell.
Figure 2: Ranges of venomous snake species and number of medically important venomous snake species per 5 × 5 km location for which no effective therapy is currently listed by WHO
(A) Counts range from low (n=1) to high (n=13). The light grey areas represent locations where no medically important venomous snake species are present. (B) Counts range from low (n=1) to high
(n=7). The light grey areas represent locations where snake species present have effective therapies listed by WHO, and the dark grey areas represent locations where no medically important venomous
snake species are present.
Number of category one and two
snake species
High
Low
A
B
High
Low
Number of c
ategory one and two
snake species with no effective
therapy
Articles
www.thelancet.com Published online July 12, 2018 http://dx.doi.org/10.1016/S0140-6736(18)31224-8
5
Pairing ecological measures with health system metrics
To identify the extent to which snakebites could vary
globally as a public health problem, we evaluated three
key dimensions: existence of any marketed antivenom
therapy, quality of health care and treatment options
available, and geographical accessibility to health care.
Of the 278 snakes considered, the WHO antivenoms
database documents that any form of antivenom (either
monospecific or polyvalent) exists for 159 species.20
Coupling this availability information with each species’
range we identified the geographical distribution of
species with no listed antivenoms, stratified by WHO
category.
To address dierences in health-care quality and
therefore identify populations to whom treatment
options might not be available or eectively deployed,
we categorised countries or regions to identify
populations living within each decile of a composite
indicator measure of health care (ie, the HAQ Index).21
The HAQ Index provides a metric for national levels of
personal health-care access and quality, drawing from
mortality rates from 32 causes that are amenable to
health care. The Index uses risk-standardised cause-
specific mortality rates derived from the Global Burden
of Disease 2016 study,29 scaled to a common 0–100 value,
and aggregated using weights derived from a principal
component analysis. To construct deciles, countries
were ranked on the basis of the HAQ Index score, and
threshold values splitting countries into ten equally
sized groups were identified. Because of variable
numbers of administrative units, subnational locations
were not used to construct decile thresholds;
subnationals for which HAQ values were estimated
were assigned to the corresponding nationally derived
decile on the basis of their value. To evaluate the
appropriateness of the HAQ Index as a proxy metric for
severe snakebite-related outcomes, we analysed the
relationship between published estimates of snakebite-
specific mortality numbers and the index, mimicking
analyses undertaken on other development indices and
mortality outcomes.6
To reflect relative geographical isolation from health
care, we coupled mortality data with a contemporary
surface of accessibility to major population centres.
Habib and Abubakar30 identified that, for a Nigerian
cohort of cases, each hour delay between envenomation
and antivenom administration was associated with
an increased mortality outcome of 1·01% (95% CI
1·00–1·02).30 A contemporary surface of accessibility to
high-density urban locations (travel time in minutes to
locations with a population >50 000) was used to identify
remote populations and compared with the mortality
statistics above.22 To evaluate the suitability of a
population–centre-based metric versus a health-care-
focused measure, we did a sensitivity analysis using
published data for African health-care facilities.31
Populations living within these geographical regions of
vulnerability were enumerated using the most recent
gridded population estimates from WorldPop, producing
estimates at the 5 × 5 km pixel level, and aggregated to
each country’s second-level administrative division
aiding government interpretation.32
Data sharing
All codes used throughout this study are available at
https://github.com/joshlongbottom/snakebite.
Figure 3: Average travel time to nearest major city for populations living within snake ranges
The light grey areas represent locations without the presence of medically important venomous snake species.
Average travel time
≥
24 h
0
h
Articles
6
www.thelancet.com Published online July 12, 2018 http://dx.doi.org/10.1016/S0140-6736(18)31224-8
Role of the funding source
The funder of the study had no role in study design, data
collection, data analysis, data interpretation, or writing of
the report. The corresponding authors had full access to
all the data in the study and had final responsibility for
the decision to submit for publication.
Results
Through the combination of publicly available data, we
provide a surface showing the ranges of 278 snake
species per 5 × 5 km area globally. Our MESS validation
method resulted in range amendments of 96 species
(appendix 1). Given the broad distribution of snakes,
approximately 6·85 billion people live within the range
of one or more of the species considered (figure 2A).
When filtered by medical classification, 5·80 billion
people live within range of category one species and
5·53 billion people live within range of category
two species (appendix 2 pp 13, 14). Hotspots of venomous
snake diversity include the Congo Basin, southeast Asia,
and Latin America.
Using the only openly available database for antivenom
availability, we identified 119 (43%) of the 278 mapped
species with no specific therapy. Of these identified
species, 24 (20%) were category one importance and
95 (80%) were category two importance. Hotspots of
species with no listed antivenom occur throughout
west Africa (eg, Ghana has ≤7 species per cell),
central Africa (eg, Cameroon has ≤7 species per cell),
South America (eg, Colombia has ≤7 species per cell), and
south Asia (eg, India has ≤6 species per cell; figure 2B;
appendix 2 pp 15, 16). Among category one snake species
with no therapy, Myanmar and Bangladesh have the
highest number (≤3 species per cell), with areas in
west Africa (Mali, Senegal, and Guinea) and Namibia
having up to two therapy-naive species per cell.
Populations living within these ranges of snake species
vary greatly in terms of accessibility to population centres
and presumed health care. Although antivenoms are
deployed in health facilities of some countries with very
small communities, this deployment is not universal, and
in the absence of exact data for antivenom access, we
were required to approximate the influence of travelling
time to health-care facilities via a proxy of distance to
centres with more than 50 000 inhabitants. Our time-
delay surface highlights that should envenoming
occur in large areas of Sudan, Algeria, Indonesia,
Papua New Guinea, Colombia, and Peru, the time taken
to travel to a city in which we might expect to find available
treatment could worsen mortality outcomes by more than
25%, assuming linear scaling of the statistic from Habib
and Abubakar30 (figure 3). For instance, 2 531 665 people
(78·71% of the population) live within ranges of any snake
species within South Sudan and 624 204 people
(89·89% of the population) in Papua New Guinea live
more than 1 h from locations with 50 000 people or more
(appendix 2 pp 17–26); globally, 750·19 million people
(10·95% of the population) potentially at risk from
snakebite envenoming live more than 1 h from high-
density urban areas, increasing the likelihood of delay-
based mortality outcomes after envenoming.
Sensitivity analyses using African hospitals versus cities
similarly showed consistent results (appendix 2 pp 34, 35).
Separating the populations living within ranges of
species by HAQ Index deciles reveals large dieren ces
across the sociodemographic spectrum (figure 4).
Such dierences are best highlighted when analysing
populations by medical classification: approximately
389·58 million individuals (92·91%) of the population
within the lowest HAQ Index decile are at risk of exposure
to a category one snake compared with approximately 1·61
billion individuals (86·27%) within the highest decile
(appendix 2 p 16). Furthermore, 272·91 million individuals
(65·25%) of the population within the lowest decile are at
risk of exposure to any snake for which no eective
therapy exists compared with 519·46 million individuals
(27·79%) within the highest HAQ Index decile (figure 4).
Vulnerable populations (ie, people in geographical reg-
ions living within the range of any snake species who also
lived more than 3 h away from major urban centres, had
health systems that scored within the lowest three deciles
Figure 4: Proportion of populations living within range of snake species by each HAQ Index decile
(A) Populations living within the range of one or more medically important venomous snake species (either
category one or two). (B) Populations living within the range of one or more medically important venomous snake
species (either category one or two), for which no effective therapy is listed. HAQ=Healthcare Access and Quality.
1
0
Population (%)
HAQ Index decile HAQ Index decile
AB
20
40
60
80
100
2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
Outside of rangeWithin range
Figure 5: Hotspots of vulnerable populations to medically important
venomous snake species
Hotspots are defined as people living in areas within the range of one or more
medically important venomous snake species, and more than 3 h away from
major urban centres with Healthcare Access and Quality Index deciles of 1–3.
(A) Pixel-level vulnerability surface (ie, vulnerability to all species of medically
important snakes). (B) Aggregated second administrative level vulnerability to
all species of medically important venomous snakes, as measured by the
absolute number of people. (C) Aggregated second administrative level
vulnerability to only those species for which no effective therapy is currently
listed by WHO, as measured by the absolute number of people.
Articles
www.thelancet.com Published online July 12, 2018 http://dx.doi.org/10.1016/S0140-6736(18)31224-8
7
C
0
1–50
000
50
001–500
000
500
001–1
000
000
>1
000
001
Vulnerable populations
B
0
1–50
000
50
001–500
000
500
001–1
000
000
>1
000
001
Vulnerable populations
A
Vulnerable populations
Not vulnerable populations
Articles
8
www.thelancet.com Published online July 12, 2018 http://dx.doi.org/10.1016/S0140-6736(18)31224-8
of the HAQ Index, and were further stratified by the
presence or absence of a WHO listed antivenom) are
highlighted in figure 5. Vulnerability estimates for HAQ
Index deciles of 1–10 are provided in the table. Within the
lowest three deciles, we highlighted regions where about
92·66 million vulnerable individuals live (table), with
Angola, Pakistan, Indonesia, Ethiopia, and the Democratic
Republic of the Congo ranking as the highest locations
in absolute numbers. The majority of countries across
Africa, many of which have some of the lowest scores on
the HAQ Index, have vulnerable populations present.
When excluding infor mation about the existence of
antivenoms, 146·70 million people live within remote
areas lacking quality health care provisioning (deciles 1–3;
appendix 2 pp 28–31).
Discussion
Understanding the distribution of venomous snakes and
their potential burden on health systems at national,
regional, and global levels is important for eective re-
duction and control of snakebite.8 By combining species’
range maps, available information about antivenoms, and
measures of quality of and distance to health care, this
study provides global contemporary maps of vulnerable
populations to snakebite and its clinical complications.
This analysis, therefore, provides a means of identifying
communities in greatest need of support from herpe-
tologists, clinicians, and public health experts, and to
prioritise new data-collection activities.
Although this analysis is not a substitute for a full global
burden estimation, there is overlap between vulnerable
communities and existing burden estimates, with vulner-
able countries such as Nigeria, Benin, Congo (Brazzaville),
Myanmar, and Papua New Guinea identified as burden-
some in country-specific estimates,1 and south Asia and
sub-Saharan Africa as regions with considerable mort-
ality and morbidity.3 A post-hoc analysis of national
envenoming and death burden values, shows that, for
vulnerable countries, such values were more likely to be
estimates as opposed to data-driven numbers (χ² test at
90% significance level, p=0·0476 for envenoming and
p=0·0517 for deaths).3 Chippaux33 similarly shows that
where data are available in sub-Saharan countries, they
are not necessarily contemporary information. These
Decile 1 Decile 2 Decile 3 Decile 4 Decile 5 Decile 6 Decile 7 Decile 8 Decile 9 Decile 10
Afghanistan 281 586 ·· ·· ·· ·· ·· ·· ·· ·· ··
Algeria ·· ·· ·· ·· 74 397 ·· ·· ·· ·· ··
Angola ·· 3 652 123 ·· ·· ·· ·· ·· ·· ·· ··
Argentina ·· ·· ·· ·· ·· 78 462 ·· ·· ·· ··
Armenia ·· ·· ·· ·· ·· ·· 27 064 ·· ·· ··
Azerbaijan ·· ·· ·· ·· ·· 116 150 ·· ·· ·· ··
Bangladesh ·· ·· ·· 359 780 ·· ·· ·· ·· ·· ··
Belize ·· ·· ·· ·· 14 532 ·· ·· ·· ·· ··
Benin 97 491 ·· ·· ·· ·· ·· ·· ·· ·· ··
Bhutan ·· ·· ·· 114 385 ·· ·· ·· ·· ·· ··
Bolivia ·· ·· ·· 1 307 831 ·· ·· ·· ·· ·· ··
Botswana ·· ·· ·· 323 599 ·· ·· ·· ·· ·· ··
Brazil ·· ·· ·· ·· 4 107 300 2296 619 ·· ·· ··
Brunei ·· ·· ·· ·· ·· ·· ·· 9630 ·· ··
Burkina Faso 699 570 ·· ·· ·· ·· ·· ·· ·· ·· ··
Burundi 10 597 ·· ·· ·· ·· ·· ·· ·· ·· ··
Cambodia ·· ·· 43 972 ·· ·· ·· ·· ·· ·· ··
Cameroon ·· 1 279 030 ·· ·· ·· ·· ·· ·· ·· ··
Central African
Republic
1 081 841 ·· ·· ·· ·· ·· ·· ·· ·· ··
Chad 5814 ·· ·· ·· ·· ·· ·· ·· ·· ··
China ·· ·· ·· 167 314 6 787 216 6 793 121 11 323 668 8 414 197 14 142 ··
Colombia ·· ·· ·· ·· ·· 6 277 835 ·· ·· ·· ··
Congo
(Brazzaville)
·· 521 800 ·· ·· ·· ·· ·· ·· ·· ··
Costa Rica ·· ·· ·· ·· ·· ·· 138 011 ·· ·· ··
Côte d’Ivoire 379 448 ·· ·· ·· ·· ·· ·· ·· ·· ··
Democratic
Republic of the
Congo
22 586 819 ·· ·· ·· ·· ·· ·· ·· ·· ··
Djibouti ·· 73 050 ·· ·· ·· ·· ·· ·· ·· ··
(Table continues on next page)
Articles
www.thelancet.com Published online July 12, 2018 http://dx.doi.org/10.1016/S0140-6736(18)31224-8
9
Decile 1 Decile 2 Decile 3 Decile 4 Decile 5 Decile 6 Decile 7 Decile 8 Decile 9 Decile 10
(Continued from previous page)
Ecuador ·· ·· ·· ·· 552 572 ·· ·· ·· ·· ··
Egypt ·· ·· ·· ·· 37 780 ·· ·· ·· ·· ··
Equatorial
Guinea
·· ·· ·· 242 345 ·· ·· ·· ·· ·· ··
Eritrea 905 464 ·· ·· ·· ·· ·· ·· ·· ·· ··
Ethiopia 10 422 734 ·· ·· ·· ·· ·· ·· ·· ·· ··
Gabon ·· ·· 499 707 ·· ·· ·· ·· ·· ·· ··
Georgia ·· ·· ·· ·· ·· 111 973 ·· ·· ·· ··
Ghana ·· ·· 354 713 ·· ·· ·· ·· ·· ·· ··
Greece ·· ·· ·· ·· ·· ·· ·· ·· 7327 ··
Guatemala ·· ·· ·· 533 186 ·· ·· ·· ·· ·· ··
Guinea 427 253 ·· ·· ·· ·· ·· ·· ·· ·· ··
Guinea-Bissau 120 745 ·· ·· ·· ·· ·· ·· ·· ·· ··
Guyana ·· ·· ·· 138 904 ·· ·· ·· ·· ·· ··
Honduras ·· ·· ·· 58 882 ·· ·· ·· ·· ·· ··
India ·· 231 656 1 488 560 4 623 346 ·· 276 547 ·· ·· ·· ··
Indonesia ·· ·· 10 454 226 ·· ·· ·· ·· ·· ·· ··
Iran ·· ·· ·· ·· ·· ·· 341 174 ·· ·· ··
Iraq ·· ·· ·· 49 668 ·· ·· ·· ·· ·· ··
Japan ·· ·· ·· ·· ·· ·· ·· ·· 38 825 6288
Jordan ·· ·· ·· ·· ·· ·· 91 671 ·· ·· ··
Kazakhstan ·· ·· ·· ·· ·· ·· 2 494 396 ·· ·· ··
Kenya ·· ·· 1 825 765 ·· ·· ·· ·· ·· ·· ··
Kyrgyzstan ·· ·· ·· ·· 495 765 ·· ·· ·· ·· ··
Laos ·· ·· 18 010 ·· ·· ·· ·· ·· ·· ··
Liberia ·· 664 940 ·· ·· ·· ·· ·· ·· ·· ··
Malawi ·· 133 687 ·· ·· ·· ·· ·· ·· ·· ··
Malaysia ·· ·· ·· ·· ·· 1 790 903 ·· ·· ·· ··
Mali ·· 2 373 844 ·· ·· ·· ·· ·· ·· ·· ··
Mauritania ·· ·· 5129 ·· ·· ·· ·· ·· ·· ··
Mexico ·· ·· ·· ·· 229 259 384 600 275 250 ·· ·· ··
Morocco ·· ·· ·· ·· 93 028 ·· ·· ·· ·· ··
Mozambique 3 206 555 ·· ·· ·· ·· ·· ·· ·· ·· ··
Myanmar ·· ·· 2 544 010 ·· ·· ·· ·· ·· ·· ··
Namibia ·· ·· 752 476 ·· ·· ·· ·· ·· ·· ··
Nepal ·· ·· 2 665 443 ·· ·· ·· ·· ·· ·· ··
Nicaragua ·· ·· ·· ·· 73 046 ·· ·· ·· ·· ··
Niger 15 45 113 ·· ·· ·· ·· ·· ·· ·· ·· ··
Nigeria ·· ·· 2 067 928 ·· ·· ·· ·· ·· ·· ··
Oman ·· ·· ·· ·· ·· ·· ·· 84 388 ·· ··
Pakistan ·· ·· 4 425 880 ·· ·· ·· ·· ·· ·· ··
Panama ·· ·· ·· ·· ·· 137 946 ·· ·· ·· ··
Paraguay ·· ·· ·· ·· 227 331 ·· ·· ·· ·· ··
Peru ·· ·· ·· ·· ·· 2 674 949 ·· ·· ·· ··
Philippines ·· ·· ·· 1 517 133 ·· ·· ·· ·· ·· ··
Russia ·· ·· ·· ·· ·· ·· ·· 1 205 085 ·· ··
Rwanda ·· ·· 32 081 ·· ·· ·· ·· ·· ·· ··
Saudi Arabia ·· ·· ·· ·· ·· ·· ·· 2 392 280 ·· ··
Senegal ·· 467 056 ·· ·· ·· ·· ·· ·· ·· ··
Sierra Leone 154 596 ·· ·· ·· ·· ·· ·· ·· ·· ··
Somalia 2 140 834 ·· ·· ·· ·· ·· ·· ·· ·· ··
(Table continues on next page)
Articles
10
www.thelancet.com Published online July 12, 2018 http://dx.doi.org/10.1016/S0140-6736(18)31224-8
maps collate ecological and public health metrics, and
identify opportunities where substantial improvements
and refinements can be undertaken to move from
broad vulnerability assessment to a more nuanced
and accurate description of the most burdensome
populations.
Our study had key limitations, and future eorts can
focus on addressing some of these limitations: the relative
contribution of dierent snake species must be quantified,
the factors influencing snake–human interactions and
subsequent likelihood of envenoming events must be
identified, and snakebite-specific measures of local
preparedness, eectiveness, and coverage of existing
clinical countermeasures must be taken. Paucity of data
available at the global scale, despite comprehensive
coverage in several high-income countries, remains one
of the largest limitations throughout this study. Ultimately,
quantifying these additional components will allow for
estimates to be based on a bottom-up data synthesis,
rather than dependence on global-level datasets and
correlations.
Mapping snake species’ locations to reflect variations in
snake presence is also important. Fine-scale maps, such as
those of American venomous snake species,34 should be
extended globally—this current analysis identifies
216 species requiring updated assessments of current
ranges given the quality and quantity of records available.
This study also establishes a systematic prioritisation
based on medical importance (appendix 2 pp 3–11).
Although species occurrence surveys can be formally
done by public health initiatives or during ecological
assessments, citizen science has a complementary role
in facilitating broad-scale data collection.35
This assessment considers the presence of any one
venomous snake as a prerequisite for vulnerability;
however, dierent species contribute dierently to
envenoming risk. Species with a very high incidence of
envenoming events might be the dominant cause of high
snakebite burden in a locality,36 regardless of the presence
of other species,37 as reported for Echis ocellatus,38 Daboia
russelii,39 and others.40 Identifying and quantifying, at a
local scale, important species, risky human practices,
and ongoing changes to subsequent interactions given
climatic and socioeconomic change, are necessary.41
Future vulnerability assessments can explicitly leverage
interspecies’ dierences and weigh their relative
Decile 1 Decile 2 Decile 3 Decile 4 Decile 5 Decile 6 Decile 7 Decile 8 Decile 9 Decile 10
(Continued from previous page)
South Africa ·· ·· ·· 289 322 ·· ·· ·· ·· ·· ··
South Sudan 601 410 ·· ·· ·· ·· ·· ·· ·· ·· ··
Sri Lanka ·· ·· ·· ·· ·· ·· 62 951 ·· ·· ··
Sudan ·· ·· ·· 542 183 ·· ·· ·· ·· ·· ··
Suriname ·· ·· ·· 94 635 ·· ·· ·· ·· ·· ··
Swaziland ·· ·· 226 ·· ·· ·· ·· ·· ·· ··
Syria ·· ·· ·· ·· ·· 48 738 ·· ·· ·· ··
Tajikistan ·· ·· ·· 301 870 ·· ·· ·· ·· ·· ··
Tanzania ·· 4 950 775 ·· ·· ·· ·· ·· ·· ·· ··
Thailand ·· ·· ·· ·· ·· ·· 77 295 ·· ·· ··
The Gambia ·· 3245 ·· ·· ·· ·· ·· ·· ·· ··
Togo ·· 14 488 ·· ·· ·· ·· ·· ·· ·· ··
Trinidad and
Tobago
·· ·· ·· ·· ·· 6520 ·· ·· ·· ··
Turkey ·· ·· ·· ·· ·· ·· 48 653 ·· ·· ··
Uganda ·· 371 059 ·· ·· ·· ·· ·· ·· ·· ··
Ukraine ·· ·· ·· ·· ·· ·· ·· 4956 ·· ··
USA ·· ·· ·· ·· ·· ·· ·· ·· 4728 ··
Uzbekistan ·· ·· ·· ·· 805 059 ·· ·· ·· ·· ··
Venezuela ·· ·· ·· ·· ·· 3 3465 88 ·· ·· ·· ··
Vietnam ·· ·· ·· ·· 175 519 ·· ·· ·· ·· ··
Yemen ·· ·· 3 017 147 ·· ·· ·· ·· ·· ·· ··
Zambia 2 125 572 ·· ·· ·· ·· ·· ·· ·· ·· ··
Zimbabwe ·· 936 801 ·· ·· ·· ·· ·· ·· ·· ··
Total 46 793 442 15 673 554 30 195 273 10 664 383 13 672 804 22 046 628 14 880 752 12 110 536 65 022 6288
Country-level count of vulnerable people living within the range of one or more medically important venomous snake species, for which no effective therapy exists, and with a travel time of more than 3 h from
urban centres with a population of 50 000 people or more provided per HAQ Index decile (ranging from 1 [low] to 10 [high]). Appendix 2 shows the vulnerability estimates not incorporating antivenom
availability. HAQ=Healthcare Access and Quality.
Table: Vulnerable population count per HAQ Index decile
Articles
www.thelancet.com Published online July 12, 2018 http://dx.doi.org/10.1016/S0140-6736(18)31224-8
11
contribution as a function of species-specific envenoming
risk and associated burden. The transition of the
WHO resource20 into a living database documenting
contemporary antivenom availability, species taxonomic
changes, higher-resolution distribution data, and other
information will substantially aid in this eort.42,43
Areas where snakes are present can be further evaluated
to determine the true incidence of envenoming events.
Local-scale household surveys assessing incidence of
snakebite have been done in several countries.11,14,15,44
Questions relating to snakebite could also be nested within
existing demographic and health surveys,45 minimising
associated costs and informing current data-poor
estimates. By integrating preventive measures with
existing management systems for neglected tropical
diseases, many logistical obstacles to eective intervention
might be overcome.46 Corresponding quantification of key
risky behaviours will help reflect fine-scale population
heterogeneity to exposure. Surveys such as the World Bank
Living Standards Measurement Survey series could be
used to obtain local-scale information about agricultural
practices,47 further aiding the identification of communities
most at risk and increasing understanding of the public
health consequences of dierent land use. Through these
steps, eorts to prevent envenoming events can be tailored
to the specifics of any given population.
In many low-income and middle-income countries, a
multitude of barriers influence snakebite outcomes
including health care, transport, and communications
infrastructure, along with adequacy of and access to safe,
eective, and aordable antivenom supplies, medical
sta proficiency and training, and public health policy.
When considering antivenom availability, this method is
constrained to listings as reported by WHO.20 Since
initial compilation, new antivenoms have become
available (eg, EchiTAb-Plus-ICP),48 while others have
ceased production (eg, Fav-Afrique by Sanofi).42 Market
availability of antivenom products does not translate to
in-field availability and ecacy; further information
regarding country-specific, contemporary stockpiles, and
the positioning of antivenom holding centres is required.
Given that some of the countries with the lowest
HAQ Index deciles have the largest proportions of the
population living in areas with snakes for which no
antivenom is currently reported, documented socio-
economic dierences might amplify inequalities in care.6
Although health system indicators and accessibility
metrics act as generalised correlates for a location’s
ability to respond to cases, these measures will possibly
underestimate or overstate local vulnerabilities in some
settings. Existing analyses of health systems show
variation both nationally and subnationally in treatment-
seeking behaviours,49,50 quality of primary point of care
visits and referrals,51 and general practitioner knowledge
about the condition.52 However, the external validity of
these existing surveys is unknown. This vulnerability
analysis provides a foundation for the identification of
locations where further surveys of treatment-seeking
behaviours, quality of care, and existing coverage of
antivenom stockpiles and supply chains need to be
assessed.
The global burden of snakebite can be assessed through
an approach that integrates ecological information, human
behavioural data, and snakebite-specific health system
functioning. The impetus to reduce and control the burden
of snakebite envenoming, a thorough cataloguing of snake
presence and abundance, species-specific interaction
profiles with humans, and detailed understanding of
logistical hurdles to intervention delivery should be
long-term objectives. Contemporary assessments, such as
the resources presented, provide an immediate means of
identifying key hotspots and most vulnerable communities
where the need for such investigations is greatest.
Contributors
JL, SIH, and DMP conceived and planned the study. JL wrote the
computer code, and designed and did the analyses with input from
FMS and DMP. DJWe constructed the accessibility covariate data layer.
JL produced all output figures. NR, DAW, RRdC, and DJWi provided
intellectual input into aspects of this study. All authors contributed to the
interpretation of the results. JL wrote the first draft of the manuscript
and all authors contributed to subsequent revisions.
Declaration of interests
We declare no competing interests.
Acknowledgments
This study is funded by the Bill & Melinda Gates Foundation. SIH is
funded by grants from the Bill & Melinda Gates Foundation
(OPP1132415), the Wellcome Trust (209142 and Senior Research
Fellowship 095066), and the Fleming Fund. The Bill & Melinda Gates
Foundation grant OPP1093011 supports JL and DMP. The Fleming
Fund also supports MD. FMS is supported by a scholarship from
The Rhodes Trust. GA, FC, NR, and RRdC are partly supported by a
grant from the Swiss National Science Foundation (315130_176271).
DJWi is supported by a Doherty Biomedical Postdoctoral Fellowship
from the Australian National Health and Medical Research Council.
We thank C A Design Services for assistance with expert opinion range
map digitisation.
References
1 Chippaux JP. Snake-bites: appraisal of the global situation.
Bull World Health Organ 1998; 76: 515–24.
2 Gutiérrez JM, Calvete JJ, Habib AG, Harrison RA, Williams DJ,
Warrell DA. Snakebite envenoming. Nature 2017; 3: 17063.
3 Kasturiratne A, Wickremasinghe AR, de Silva N, et al. The global
burden of snakebite: a literature analysis and modelling based on
regional estimates of envenoming and deaths. PLoS Med 2008;
5: e218.
4 GBD 2015 Disease and Injury Incidence and Prevalence
Collaborators. Global, regional, and national incidence, prevalence,
and years lived with disability for 310 diseases and injuries,
1990–2015: a systematic analysis for the Global Burden of Disease
Study 2015. Lancet 2016; 388: 1545–602.
5 Gutiérrez JM, Warrell DA, Williams DJ, et al. The need for full
integration of snakebite envenoming within a global strategy to
combat the neglected tropical diseases: the way forward.
PLoS Negl Trop Dis 2013; 7: e2162.
6 Harrison RA, Hargreaves A, Wagsta SC, Faragher B, Lalloo DG.
Snake envenoming: a disease of poverty. PLoS Negl Trop Dis 2009;
3: e569.
7 Rodriguez C, Estrada R, Herrera M, et al. Bothrops asper
envenoming in cattle: clinical features and management using
equine-derived whole IgG antivenom. Vet J 2016; 207: 160–63.
8 WHO. Report of the tenth meeting of the WHO Strategic and
Technical Advisory Group for neglected tropical diseases. Geneva:
World Health Organization, 2017.
Articles
12
www.thelancet.com Published online July 12, 2018 http://dx.doi.org/10.1016/S0140-6736(18)31224-8
9 WHO. Neglected tropical diseases. Geneva: World Health
Organization, 2017. http://www.who.int/neglected_diseases/
diseases/en/ (accessed July 9, 2017).
10 Hansson E, Sasa M, Mattisson K, Robles A, Gutiérrez JM.
Using geographical information systems to identify populations in
need of improved accessibility to antivenom treatment for snakebite
envenoming in Costa Rica. PLoS Negl Trop Dis 2013; 7: e2009.
11 Habib AG. Public health aspects of snakebite care in west Africa:
perspectives from Nigeria. J Venom Anim Toxins Incl Trop Dis 2013;
19: 2 7.
12 Fitzpatrick C, Engels D. Leaving no one behind: a neglected tropical
disease indicator and tracers for the Sustainable Development
Goals. Int Health 2016; 8 (suppl 1): i15–18.
13 UN. Sustainable Development Goal 3: ensure healthy lives and
promote well-being for all at all ages. 2015. http://www.un.org/
sustainabledevelopment/health/ (accessed Sept 20, 2017).
14 Kasturiratne A, Pathmeswaran A, Wickremasinghe AR, et al.
The socio-economic burden of snakebite in Sri Lanka.
PLoS Negl Trop Dis 2017; 11: e0005647.
15 Vaiyapuri S, Vaiyapuri R, Ashokan R, et al. Snakebite and its
socio-economic impact on the rural population of Tamil Nadu,
India. PLoS One 2013; 8: e80090.
16 WHO. WHO guidelines for the production, control and regulation
of snake antivenom immunoglobulins. Geneva: World Health
Organization, 2010.
17 Warrell DA. Guidelines for the management of snake-bites.
New Delhi: World Health Organization, 2010.
18 Theakston RDG, Laing GD. Diagnosis of snakebite and the
importance of immunological tests in venom research. Toxins 2014;
6: 1667–95.
19 Sharma SK, Kuch U, Höde P, et al. Use of molecular diagnostic
tools for the identification of species responsible for snakebite in
Nepal: a pilot study. PLoS Negl Tro Dis 2016; 10: e0004620.
20 WHO. Venomous snakes and antivenoms search interface.
Geneva: World Health Organization, 2017. http://apps.who.int/
bloodproducts/snakeantivenoms/database/
(accessed May 29, 2017).
21 GBD 2016 Healthcare Access and Quality Collaborators. Measuring
performance on the Healthcare Access and Quality Index for
195 countries and territories and selected subnational locations:
a systematic analysis from the Global Burden of Disease Study
2016. Lancet 2018; published online May 23. https://doi.org/10.1016/
S0140-6736(18)30994-2.
22 Weiss DJ, Nelson A, Gibson HS, et al. A global map of travel time to
cities to assess inequalities in accessibility in 2015. Nature 2018;
553: 333–36.
23 WHO. WHO guidelines for the production, control and regulation
of snake antivenom immunoglobulins: post ECBS version.
Geneva: World Health Organization, 2016.
24 Women’s and Children’s Hospital, University of Adelaide.
Clinical toxinology resources: snakes search page. 2017.
http://www.toxinology.com/fusebox.cfm?fuseaction=main.snakes.
search (accessed May 29, 2017).
25 Chamberlain S, Ram K, Hart T. Spocc: interface to species
occurrence data sources (version 0.7.0). 2017. https://cran.r-project.
org/web/packages/spocc/index.html (accessed May 29, 2017).
26 Elith J, Kearney M, Phillips S. The art of modelling range-shifting
species. Methods Ecol Evol 2010; 1: 330–42.
27 Macartney JM, Gregory PT, Larsen KW. A tabular survey of data
on movements and home ranges of snakes. J Herpetol 1988;
22: 61–73.
28 Soberon J, Peterson AT. Biodiversity informatics: managing and
applying primary biodiversity data.
Philos Trans R Soc Lond B Biol Sci 2004; 359: 689–98.
29 GBD 2016 Causes of Death Collaborators. Global, regional,
and national age-sex specific mortality for 264 causes of death,
1980–2016: a systematic analysis for the Global Burden of Disease
Study 2016. Lancet 2017; 390: 1151–210.
30 Habib AG, Abubakar SB. Factors aecting snakebite mortality in
north-eastern Nigeria. Int Health 2011; 3: 50–55.
31 Ouma PO, Maina J, Thuranira PN, et al. Access to emergency
hospital care provided by the public sector in sub-Saharan Africa in
2015: a geocoded inventory and spatial analysis.
Lancet Glob Health 6: e342–50.
32 Stevens FR, Gaughan AE, Linard C, Tatem AJ. Disaggregating
census data for population mapping using random forests with
remotely-sensed and ancillary data. PLoS One 2015; 10: e0107042.
33 Chippaux JP. Estimate of the burden of snakebites in
sub-Saharan Africa: a meta-analytic approach. Toxicon 2011;
57: 586–99.
34 Yanez-Arenas C, Peterson AT, Mokondoko P, Rojas-Soto O,
Martinez-Meyer E. The use of ecological niche modeling to infer
potential risk areas of snakebite in the Mexican state of Veracruz.
PLoS One 2014; 9: e100957.
35 Geneviève LD, Ray N, Chappuis F, et al. Participatory approaches
and open data on venomous snakes: a neglected opportunity in the
global snakebite crisis? PLoS Negl Trop Dis 2018; 12: e0006162.
36 McGain F, Limbo A, Williams DJ, Didei G, Winkel KD.
Snakebite mortality at Port Moresby General Hospital,
Papua New Guinea, 1992–2001. Med J Aust 2004; 181: 687–91.
37 Hung D-Z. Taiwan’s venomous snakebite: epidemiological,
evolution and geographic dierences.
Trans R Soc Trop Med Hyg 2004; 98: 96–101.
38 Ademola-Majekodunmi FO, Oyediran FO, Abubakar SB.
Incidence of snakebites in Kaltungo, Gombe State and the ecacy
of a new highly purified monovalent antivenom in treating
snakebite patients from January 2009 to December 2010.
Bull Soc Pathol Exot 2012; 105: 175–78.
39 Kularatne SA. Epidemiology and clinical picture of the Russell’s
viper (Daboia russelii russelii) bite in Anuradhapura, Sri Lanka:
a prospective study of 336 patients.
Southeast Asian J Trop Med Public Health 2003; 34: 855–62.
40 Warrell DA. Redi award lecture: clinical studies of snake-bite in
four tropical continents. Toxicon 2013; 69: 3–13.
41 Yanez-Arenas C, Townsend Peterson A, Rodriguez-Medina K,
Barve N. Mapping current and future potential snakebite risk in
the new world. Clim Change 2016; 134: 697–711.
42 Sanofi Pasteur. Sanofi Pasteur’s information on Fav-Afrique. 2016.
https://www.sanofipasteur.com/media/Project/One-Sanofi-Web/
sanofipasteur-com/en/media-room/docs/SpecificInformation_
FavAfrique_EN.pdf (accessed Aug 14, 2017).
43 Williams DJ, Gutierrez JM, Calvete JJ, et al. Ending the drought:
new strategies for improving the flow of aordable, eective
antivenoms in Asia and Africa. J Proteomics 2011; 74: 1735–67.
44 Mohapatra B, Warrell DA, Suraweera W, et al. Snakebite mortality
in India: a nationally representative mortality survey.
PLoS Negl Trop Dis 2011; 5: e1018.
45 Corsi DJ, Neuman M, Finlay JE, Subramanian SV. Demographic
and health surveys: a profile. Int J Epidemiol 2012; 41: 1602–13.
46 Kabatereine NB, Malecela M, Lado M, Zaramba S, Amiel O,
Kolaczinski JH. How to (or not to) integrate vertical programmes
for the control of major neglected tropical diseases in
sub-Saharan Africa. PLoS Negl Trop Dis 2010; 4: e755.
47 Binswanger-Mkhize HP, Savastano S. Agricultural intensification:
the status in six African countries. Food Policy 2017; 67: 26–40.
48 EchiTAb-Plus-ICP. Instructions for the administration of
EchiTAb-Plus-ICP for the treatment of envenomings by snakebites
in sub-Saharan Africa. 2017. http://www.echitabplusicp.org/
product/description (accessed Dec 20, 2017).
49 Ediriweera DS, Kasturiratne A, Pathmeswaran A, et al. Health
seeking behavior following snakebites in Sri Lanka: results of an
island wide community based survey. PLoS Negl Trop Dis 2017;
11: e0006073.
50 Vongphoumy I, Phongmany P, Sydala S, Prasith N, Reintjes R,
Blessmann J. Snakebites in two rural districts in Lao PDR:
community-based surveys disclose high incidence of an invisible
public health problem. PLoS Negl Trop Dis 2015; 9: e0003887.
51 Shahmy S, Kularatne SAM, Rathnayake SS, Dawson AH.
A prospective cohort study of the eectiveness of the primary
hospital management of all snakebites in Kurunegala district of
Sri Lanka. PLoS Negl Trop Dis 2017; 11: e0005847.
52 Michael GC, Grema BA, Aliyu I, et al. Knowledge of venomous
snakes, snakebite first aid, treatment, and prevention among
clinicians in northern Nigeria: a cross-sectional multicentre study.
Trans R Soc Trop Med Hyg 2018; 112: 47–56.
