Mosquitoes & Vulnerable Spaces: Mapping Local Knowledge of Sites for Dengue Control in Seremban and Putrajaya Malaysia

Abstract

Dengue is a mosquito-borne viral infection that continues to represent a significant health challenge in many tropical and subtropical regions. At a local scale, dengue prevention and control is a cooperative effort as favorable vector breeding sites may be found across residential, commercial and public spaces within a community. However, many vector control initiatives do not take into account local understanding of dengue risk and how this impacts the actions of residents to prevent dengue by eliminating breeding sources. The objective of this study was to use a participatory mapping approach to identify spatial perceptions of risk to dengue at a community scale. Four mapping groups were formed in two urban Malaysian communities that have experienced high dengue rates, divided into male and female groups to encourage gender equity. Participants were asked to draw a map of areas they associated with dengue and mosquito breeding in their communities, and to describe the important features on the map. Sketch map features were digitized into a GIS to create a georeferenced map of community knowledge, translating the outputs into formats accessible to stakeholders. Community spaces linked to dengue identified in the mapping exercises differed between the two communities, and included green spaces, construction projects, drainage networks and abandoned land areas. The findings indicated that resident perceptions of some vulnerable areas, such as green spaces, differed from the views of local public health staff, and could influence the actions of residents to adequately destroy breeding sites. This highlights the need to understand local knowledge of mosquito breeding in order to enhance co-operative efforts with vector-control workers, and increase the effectiveness of dengue prevention efforts at a local scale in Malaysia.

Full text

Mosquitoes & vulnerable spaces: Mapping local knowledge of sites
for dengue control in Seremban and Putrajaya Malaysia
Sarah K. Dickin
a
,
b
,
*
, Corinne J. Schuster-Wallace
a
,
b
, Susan J. Elliott
c
a
School of Geography and Earth Sciences, McMaster University, Hamilton, ON, Canada
b
United Nations University Institute for Water Environment and Health, Hamilton, ON, Canada
c
Faculty of Applied Health Sciences, University of Waterloo, Waterloo, Canada
Keywords:
Dengue
Participatory GIS
Community mapping
Risk perceptions
Vector-borne disease
Urban health
abstract
Dengue is a mosquito-borne viral infection that continues to represent a significant health challenge in
many tropical and subtropical regions. At a local scale, dengue prevention and control is a cooperative
effort as favorable vector breeding sites may be found across residential, commercial and public spaces
within a community. However, many vector control initiatives do not take into account local under-
standing of dengue risk and how this impacts the actions of residents to prevent dengue by eliminating
breeding sources. The objective of this study was to use a participatory mapping approach to identify
spatial perceptions of risk to dengue at a community scale. Four mapping groups were formed in two
urban Malaysian communities that have experienced high dengue rates, divided into male and female
groups to encourage gender equity. Participants were asked to draw a map of areas they associated with
dengue and mosquito breeding in their communities, and to describe the important features on the map.
Sketch map features were digitized into a GIS to create a georeferenced map of community knowledge,
translating the outputs into formats accessible to stakeholders. Community spaces linked to dengue
identified in the mapping exercises differed between the two communities, and included green spaces,
construction projects, drainage networks and abandoned land areas. The findings indicated that resident
perceptions of some vulnerable areas, such as green spaces, differed from the views of local public health
staff, and could influence the actions of residents to adequately destroy breeding sites. This highlights the
need to understand local knowledge of mosquito breeding in order to enhance co-operative efforts with
vector-control workers, and increase the effectiveness of dengue prevention efforts at a local scale in
Malaysia.
Ó2013 Elsevier Ltd. All rights reserved.
Introduction
Dengue is a vector-borne disease that presents an important
global health challenge, with 40% of the world’s population living in
areas where there is a transmission risk from Aedes mosquitoes. The
morbidity and mortality associated with dengue has widespread
implications, including loss of workers from the labor force and a
burden on health systems (Suaya et al., 2009). In Southeast Asia,
health costs associated with dengue are estimated to be US$950m
per year, and the disease burden (based on disability-adjusted life
years) is greater than many other conditions including Japanese
encephalitis, upper respiratory infections, and hepatitis B (Shepard,
Undurraga, & Halasa, 2013). Dengue hemorrhagic fever (DHF) and
dengue shock syndrome (DSS) are severe forms of the disease, with
Malaysia reporting a fatality rate of approximately 3.6%, compared
with 0.003% in neighboring Singapore (Beatty et al., 2010). A vac-
cine has yet to be developed in order to combat the illness, there-
fore the emphasis is on measures to control vector populations
(Gubler, 2002). In Malaysia, this involves education and awareness
campaigns, as well as insecticide fogging to reduce transmission
that is concentrated in areas with identified cases. In addition, a
WHO Asia-Pacific Dengue Strategic Plan (2008e2015) is underway
to reverse the rising trend of dengue in Southeast Asia by
strengthening capacity across country borders to prevent disease
transmission.
However, combatting dengue relies not only on the work of
health officials, but also on participation of residents and commu-
nity leaders to remove breeding sites from residential and com-
munity spaces. Insecticide fogging activities carried out by vector-
control staff can only target adult mosquito populations, so
*Corresponding author. School of Geography and Earth Sciences, General Science
Building Room 206, McMaster University, 1280 Main Street West, Hamilton, ON L8S
4K1, Canada. Tel.: þ1 905 667 5511.
E-mail addresses: dickinsk@mcmaster.ca,sarahdickin@gmail.com (S.K. Dickin).
Contents lists available at ScienceDirect
Applied Geography
journal homepage: www.elsevier.com/locate/apgeog
0143-6228/$ esee front matter Ó2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.apgeog.2013.11.003
Applied Geography 46 (2014) 71e79

breeding sources such as plastic containers, discarded tires, or
flower vases filled with water have implications for the surround-
ing neighborhood and are critical to dengue control. Despite this
important role played at a neighborhood level, vector control ini-
tiatives in Malaysia are generally top-down approaches that do not
take into account local perceptions of breeding sites. To improve
the effectiveness of control efforts, research is needed to better
understand these perceptions and how they impact the actions of
residents to protect themselves and their communities from
dengue. This study provides insight on perceptions of risk at the
local level through a participatory mapping approach carried out in
two areas experiencing high dengue rates in Malaysia.
Participatory mapping opportunities
Mapping tools are increasingly being used for disease surveil-
lance, vector control and health promotion activities (Eisen & Eisen,
2011; Hay et al., 2009). In particular, geographic information sys-
tems (GIS) have been introduced in operational control pro-
grammes for malaria (Martin, Curtis, Fraser, & Sharp, 2002) and
real-time surveillance of West Nile Virus (Shuai, Buck, Sockett,
Aramini, & Pollari, 2006). Some dengue control programs use a
GIS to monitor daily mosquito surveillance data (Ai-leen & Song,
2000).
While the use of GIS has been criticized as undemocratic due to
barriers that limit stakeholder involvement, innovative new tools
and methods are being developed that are both inexpensive and
engage communities in the process (Elwood, 2006; McCall & Dunn,
2012). Moreover, the increasing accessibility of high-resolution
satellite imagery, geospatial datasets and GIS platforms opens
new avenues for health applications of GIS. For example, Google
Earth imagery has been used in Nicaragua to create a base map in
order to monitor areas of mosquito infestation (Chang et al., 2009),
and to create maps of the Congo River for vaccination teams
(Kamadjeu, 2009). In rural Indonesia, inexpensive open source GIS
software was successfully trialled for mapping health indicators by
clinic staff (Fisher & Myers, 2011).
Participatory geographic information systems (PGIS) is an
approach that inputs community knowledge into a GIS platform,
and is increasingly being applied to involve stakeholders in delin-
eating local boundaries and prominent landmarks (Dunn, 2007;
Kyem, 2004). Within a health context, PGIS has been used to
create detailed maps of urban wards in Dar es Salaam, Tanzania,
designed to facilitate larval surveillance and control of malaria
(Dongus et al., 2007). In Bo, Sierra Leone, input from local elders
and long-term residents was used to create a detailed municipal
map to inform public policy applications, such as identifying hos-
pital catchments (Ansumana et al., 2010). In Bangladesh, commu-
nity involvement through PGIS was used in deep tubewell planning
to prevent arsenic poisoning (Hassan, 2005).
Participatory sketch mapping methods have been used to assess
perceptions of risk held by individuals and across different com-
munities, particularly in the context of disaster planning (Gaillard
et al., 2013). For example, this method has been used to identify
areas prone to flooding hazards in Masantol, Philippines, including
weak points along dikes and vulnerable populations, such those
living in poorly constructed houses and locations inhabited by
elderly residents (Cadag & Gaillard, 2012). However few research
projects have incorporated participatory approaches and
geographic methods such as GIS to increase understanding of
community perceptions of health risks (Beyer, Comstock, &
Seagren, 2010). In particular, combatting mosquito-borne diseases
such as dengue relies on cooperative efforts and communication
between health officials and residents, and maps can provide
useful planning and decision-making tools. This study used a
participatory mapping approach to collect local-level knowledge of
breeding sites and dengue risk across two different Malaysian
communities. This information was georeferenced to create visu-
alizations that can be used by residents and community vector
control staff to enhance two-way dialog, as well as inform
community-based and municipal decision-making on dengue
prevention.
Methods
Study location: Seremban and Putrajaya
The Malaysian municipalities of Putrajaya and Seremban are
situated in South-western Malaysia and are approximately 20 and
100 km away from Kuala Lumpur city centre, respectively. Both are
well-connected to Kuala Lumpur by road networks and train lines
used by commuters (Fig. 1). Located approximately 3

north of the
equator, the region is characterized by a humid tropical climate.
Putrajaya is a planned government city established in the late
1990s, located in a ‘Multimedia Super Corridor’connected to the
neighboring IT-hub Cyberjaya and the capital Kuala Lumpur. A large
proportion of housing in Putrajaya is owned and maintained by the
government and residents are mostly civil servants with above
average socioeconomic status (Moser, 2010). In contrast, Seremban
is a more characteristic Malaysian town consisting of an ethnically
diverse population, surrounded by oil-palm agriculture and tropical
forest. Putrajaya and Seremban are subdivided into neighborhoods
called ‘tamans’or ‘precincts’respectively, which were used as the
unit of analysis for the study. Both research sites are of interest
because dengue is an ongoing concern in the municipalities. Be-
tween 2009 and 2011, the average dengue rate in Seremban and
Putrajaya was approximately 200 cases per 100,000 persons, and
above the national target of less than 50 cases per 100,000 persons
(Malaysia Ministry of Health, 2010).
In Malaysia, the core dengue control activities carried out by the
public health department are household larval surveys to identify
and destroy breeding sites, and insecticide fogging to destroy adult
mosquitoes. The focus of these activities is on areas where dengue
cases have been identified. Media such as flyers, billboards and
radio advertisements are used to disseminate information on
Fig. 1. Map of study sites in peninsular Malaysia in reference to capital city Kuala
Lumpur.
S.K. Dickin et al. / Applied Geography 46 (2014) 71e7972

dengue to residents; the key message is ‘10 minutes a week,’
designed to remind residents to check for and remove breeding
sites from their homes weekly. Community dengue control pro-
grams called ‘COMBI’(Communication-for-Behavioural-Impact)
have been established in many towns because of the important role
residents play in combating dengue. COMBI programs involve local
volunteers and community leaders whose activities include
awareness campaigns, neighborhood clean-up events, and house-
hold visits to demonstrate how to remove mosquito larvae from
premises. In Malaysia, community knowledge and perceptions of
dengue have been under-researched, with many studies in the
Southeast Asia region focused on lower-income countries (eg.
Mayxay et al., 2013; Phuanukoonnon, Brough, & Bryan, 2006).
Study design
This study was guided by an ecohealth approach which seeks
to promote health and wellbeing by deepening understanding of
linkages between humans and the environment, as well as
emphasizing community participation in the research process
(Forget & Lebel, 2001; Waltner-Toews & Kay, 2005). Data gathering
involved resident focus groups where the mapping activities and
discussion of risk perceptions took place, and several semi-
structured interviews with key informants including community
leaders and public health staff. The United Nations University In-
ternational Institute for Global Health in Kuala Lumpur (UNU-
IIGH) assisted in applying for ethics approval, and a research as-
sistant from UNU-IIGH acted as an interpreter during the mapping
exercises as some participants preferred to communicate in
Bahasa Melayu (Malay language). Following analysis of commu-
nity maps, meetings were held with stakeholders involved in the
study to discuss study findings and share the digitized map
outputs.
Community leaders were recruited through the municipal
health departments dealing with vector control in both Seremban
and Putrajaya, and were invited to meetings to introduce the ob-
jectives and proposed methods of the study. Those interested in the
study facilitated contact and recruitment of residents from their
communities, involving two neighborhooods in Seremban and
Putrajaya: Taman Limau Manis (S1), Taman Marida (S2), Precinct 8
(P1) and Precinct 18 (P2) respectively. Residents in these neigh-
borhoods primarily live in terrace (joined) houses in Seremban and
government owned apartment buildings in Putrajaya. Community
leaders in both cities work closely with the local public health
department, often holding presentations and activities to promote
wellbeing such as information fairs and free aerobics classes, and
these types of events were used to advertise the study. The
researcher also participated in local activities to gain familiarity
with the study sites and build rapport with community members
prior to starting the mapping exercises, including COMBI clean-ups,
insecticide demonstrations and cultural events.
A total of 8 mapping groups were formed in Seremban and
Putrajaya with 8e10 adult participants in each. These consisted of 4
male groups and 4 female groups to encourage equitable partici-
pation. Participants were provided with materials (large sheets of
paper and markers), and asked to draw their local environment in
sketch map format, including elements such as streets, houses, or
other landmarks. They were asked to identify areas they considered
to be associated with dengue in their neighborhoods. During the
mapping exercise participants were asked to verbally describe the
local dengue situation as well as the features highlighted on their
maps. These descriptions were recorded with the permission of the
participants to enrich analyses of the maps. To compare an alter-
nate mapping approach, in one Putrajaya mapping group the par-
ticipants received satellite imagery from Google Maps as a paper
basemap. These participants were asked similarly to depict land-
marks and locations perceived to be problems for mosquitoes and
dengue, by overlaying this information on the satellite image of
their community.
In addition to the mapping groups, several key informant in-
terviews with community leaders and public health officials (n¼5)
were conducted to provide additional context on the dengue situ-
ation in both Putrajaya and Seremban and were recorded and
transcribed with the permission of the participants. A year later,
meetings were held in Seremban and in Putrajaya with community
leaders and vector-control staff involved in the study to present the
study findings, disseminate digitized maps, and gather feedback
during two sessions that were also recorded and transcribed.
Digitization and analysis of community sketch maps
A digitization process was designed to import and georeference
the sketch map information using ArcGIS (ESRI v. 10.1). Georefer-
encing refers to aligning spatial data to a map coordinate system,
and this process began by identifying and linking map items, such
as construction sites, with the features on satellite imagery in the
GIS (Fig. 2). Next, these items were labeled to create a georefer-
enced map with community generated information. The goal was
to create map outputs that could be shared with other stakeholders,
since community sketch maps generally contain elements of
importance to participants, rather than precise representations of
space. In the mapping group using Google imagery as a base map,
an advantage of the method was that features did not need to be
georeferenced as features were identified and labeled directly on
the base map.
An inductive coding approach was used to examine map fea-
tures associated with dengue and transcipts of participant re-
sponses. Thematic groupings were identified and included: green
spaces, abandoned properties, drainage networks, construction
projects, management of government property, waste accumula-
tion, public spaces, and community responsibility towards vector
control.
Results
Dengue ea community-based challenge
Different types of neighborhood spaces were linked to breeding
sources and dengue risk by residents in Putrajaya and in Seremban,
and are discussed with the map outputs in the following section.
However, one sentiment expressed strongly across all communities
was the conflicting attitudes of ‘gotong royong’meaning ‘mutual
cooperation’versus ‘tidak apa’meaning ‘don’t bother’or ‘it’s
nothing,’which are common Malay expressions.
Residents and community leaders reported that local partici-
pation and cooperation is needed to remove breeding sites from
their neighborhoods to prevent dengue. A community leader in
Seremban (S1) described the local situation, “To control this
disease is not an easy task, we have to have some community
participation to control dengue, and also the help of government
agencies.”Participants explained that,in the context of dengue,
gotong royong refers to cooperative clean-up events organized
by community leaders and local vector control staff to remove
neighborhood breeding sites including waste left in public spaces
or vegetation blocking drains. Residents in both Seremban and
Putrajaya reported participating in these events, one female
participant (P1) commented, “We have gotong royong, in our
area once a month, it is a community effort.”Onemaleresident
(P2) explained that some parts of the city had more activities
than others, such as more established neighborhoods, “Precinct
S.K. Dickin et al. / Applied Geography 46 (2014) 71e79 73

18 is new, so they don’t have programs, compared to 16 which is
more active.”A male resident in Seremban (S2) explained that as
part of the COMBI vector control program in the taman there are
volunteer groups that conduct voluntary checks for larvae, “We
look after our own area first, and then we organize groups that go
to every house once a month.”A male resident (S1) felt these
types of community activities had made a difference, “About 15
years I’ve stayed here. Before, every year we got dengue and they
came and sprayed [insecticides], but now, for a few years we
have less, because we are doing all these activities, we call up the
health department, they come and clean, and everybody partic-
ipates; from young to old, now it is better.”On the other hand, a
male resident described fatigue with the community programs
(S2), “They have gotong royong every week but they get tired,
people are not really participating.”
Despite the emphasis on cooperative efforts, participants
noted that many people are not concerned with dengue until it
affects them personally ethe ‘tidak apa’eattitude as described
by a male resident (P1), “Some members of the community are
very aware of this. But some members don’t care, they don’tjoin
us, they don’t want to get involved in any of our activities, so it is
not easy to ask them, to convince them, to join us to get involved
in this.”A community leader (P1) described a local resident’s
experience with dengue, “She got it, and now she realizes how
important it is, she says she had trees where mosquitoes can
develop, like a banana tree, that can collect water, and she cut
down all the trees.”Another community leader (P2) expressed
frustration that during routine checks for mosquito breeding sites
he carried out on properties, many residents continued to allow
mosquito breeding to occur, “People take it very lightly. People
are not serious about it, the only way to make people serious
about it is to let them know they are committing an offense, is to
have a punishment, issue them a fine.”Many participants brought
up the issue that while they take the time to clean their proper-
ties some people put others at risk by not playing their part. One
female participant expressed there was no point in being vigilant
in removing breeding sites when those in her neighborhood were
not (P2), “Everyone knows about dengue but they all just take it
easy. They keep clear water in the house, all of that. The problem
is that if we keep our house clean, what about our neighbors? If
the neighborhood doesn’t keep their house clean, what about us?
If our neighbors let the mosquitoes into the surrounding area,
they will bite, so there’s no point. If our neighbor keeps water in
their house and their property, there’s no point, if our neighbors’
drains are dirty, there’s no point.”
Vulnerable spaces: mapped perceptions of dengue risk
While the need for community involvement in dengue preven-
tion was identified across groups, differences emerged between
mapped perceptions of dengue risk in Seremban and in Putrajaya. In
Putrajaya participants raised concerns over maintenance of govern-
ment owned property and public spaces, including green space in
the city. In Seremban, concern stemmed from abandoned properties,
and waste accumulation in the neighborhood. These themes are
explored in more depth in the maps and responses documented
below. In two of the groups in Seremban (S1), participants preferred
to draw pictographs of their surroundings. Because these diagrams
fell outside the scope of this paper they were removed from analysis.
Due to frequent rainfall in the region, drainage issues emerged
as a common issue in maps of Putrajaya and Seremban. These
features can be observed in Figs. 3e5, where participants labeled
several representative drains in their neighborhoods which were
linked in discussions to stagnant water and mosquito breeding.
Participants in Putrajaya labeled drains around government-owned
apartment buildings and described problems with their mainte-
nance due to trapped vegetation which blocks water flow, as one
female participant described (P2), “Right now there is an issue with
maintenance, they are not being fully responsible, they are
responsible to clean up all the drains, but right now the contractor
doesn’t want to, you can see that the drains and the roof is very
messy.”Some specific areas where participants were concerned
rainwater could accumulate in Putrajaya were also mapped,
including a football field, underground parking lot and a wet fish
market area. However, participants stated that the city was clean
and waste did not contribute to drainage problems: “Garbage is not
a problem. Garbage is collected quickly, 2 or 3 times a week. It’s
very efficient”(Female participant, P1). In Seremban, participants
mapped areas in their taman with poor drainage, including open
drains in residential areas especially in back alleys between rows of
houses, as a male resident described (S2), “The areas in the back of
the house have risk for breeding, if the back area is clean, then there
is no risk of getting dengue”. Unlike Putrajaya, these risk areas were
associated with waste accumulation, and participants explained
that people routinely throw waste and containers in drains and on
the roadside instead of garbage bins: “The problem is we are
careless about handling our garbage and water is contained in our
garbage”(Female participant, S2). Participants associated areas in
their taman such as abandoned houses, a school yard and a small
stream with increased dengue risk (Fig. 5). Abandoned properties
in particular were linked to mosquito breeding due to lack of
Fig. 2. Community sketch map in Putrajaya showing links to satellite imagery. Arrows show which features are linked. Some labels are in Malay, such as Tasik, indicating the
lake.
S.K. Dickin et al. / Applied Geography 46 (2014) 71e7974

upkeep, waste accumulation, and legal barriers which stop health
officials from entering private property. A male resident (S2)
explained, “The most uncomfortable area is this shopping lot, most
of it has been abandoned, and there is a stream right behind it, this
area should be cleaned thoroughly because there are a lot of old
tires down there”.
Putrajaya is characterized by extensive landscaped gardens,
treed spaces and an artificial lake. These green spaces were iden-
tified by participants during the mapping exercise as being asso-
ciated with dengue, including treed areas lining the artificial lake
(Fig. 3), around apartment blocks and small recreation parks. Forest
land located at the edge of the built environment was also mapped,
and participants explained this environment might contain
standing water conducive to mosquitoes. Participants were con-
cerned that plants in Putrajaya could provide mosquito breeding
sites, and that vector control workers did not address these sites, as
two male residents described (P1), “There are too many plants in
Putrajaya that support the life of the mosquito”;“It is quite green in
Putrajaya, some of the leaves of plants that we have, they collect
clean water in the trees. It makes a breeding ground. The fogging is
always being done, but not in the forest.”
Putrajaya is a planned city still being built, so construction
projects were underway in some precincts. Several construction
sites were mapped by participants as features thought to contribute
to dengue risk due to potential breeding locations (Fig. 3), such as
leftover materials or barrels found on these sites as one male
Fig. 3. Georeferenced map of community generated information in Putrajaya in three areas A, B, C. Labeled cross-hatched polygons indicate areas linked with dengue or
mosquitoes.
Fig. 4. Google imagery showing neighborhood with areas associated with dengue labeled by participants (A), close-up view of apartment block (B).
S.K. Dickin et al. / Applied Geography 46 (2014) 71e79 75

participant explained (P2), “Construction, the biggest area is in
precinct 16. There is some big storage there, so it creates some
problems with water, and mosquitoes.”One male resident said
construction areas were not monitored for breeding sites (P1), “I
noticed the construction sites, the health department normally
checks our houses, but construction areas have plastic and mate-
rials that can collect water, but nobody checks, the department
never checks construction sites.”In Seremban, a sewerage system
construction project was linked to dengue risk by participants
(Fig. 5), who described that foreign workers involved in the project
were unaware of the dengue problems or were uncooperative, and
did not properly dispose of plastic containers that could accumulate
water as one male resident said (S2), “This area is being developed,
mostly by foreigner workers. They consume food in styrofoam, it’sa
very serious issue, because after they simply chuck the container.”
This highlights the differing attitudes between residents in Ser-
emban, who described the importance of personal responsibility in
removing breeding sources from property, with residents in
Putrajaya who believed municipal staff had a large role in main-
taining outdoor housing areas and drains within the community. In
addition to these differences, male and female mapping groups
emphasized different areas in their community. Male participants
focused on describing larger features in the surrounding commu-
nity such as a construction and shipping storage facility and
sewerage project. Female participants brought up the importance
of places inside and around the home, such as containers, vases and
trays that could be contributing to dengue transmission.
Stakeholder feedback
After analysis, georeferenced map outputs were distributed to
public health staff and community leaders in both Seremban and
Putrajaya and a discussion of the findings took place. Key in-
formants who were involved in the study were asked whether the
results were useful for their work in preventing dengue. In Ser-
emban dengue control was described as “just firefighting,”and
understanding local perceptions was discussed as a way to be more
proactive in removing breeding sites. In Putrajaya, public health
staff were surprised and concerned that the risk areas identified by
residents included green spaces which are designed to improve
quality of life in the city. These areas did not correspond with pri-
ority residential areas identified by vector control workers, and
stakeholders wondered how to change these perceptions. This
highlights a complex relationship between dengue and urban
planning, a perspective that has often been excluded from main-
stream urban development (Mulligan, Elliott, & Schuster-Wallace,
2012). The differences between the residents’perceptions of risk
and targeted breeding sites was recognized by the discussion par-
ticipants as a barrier to dengue prevention.
Discussion eimplications for dengue control
Dengue risk is characterized by complex interactions between
environmental, social and economic processes operating across
different scales (Machado-Machado, 2012). In Malaysia, dengue is
Fig. 5. Georeferenced map of community generated information in Seremban. Labeled cross-hatched polygons indicate areas associated with dengue or mosquitoes.
S.K. Dickin et al. / Applied Geography 46 (2014) 71e7976

an on-going challenge and at a local level residents play an
important role in removing potential breeding sites from their local
environment. Because Aedes mosquitoes fly distances of several
hundred meters, unmonitored areas in one part of the neighbor-
hood can have implications for the surrounding areas (Harrington
et al., 2005). In both Putrajaya and Seremban, community level
activities form a part of dengue control efforts. However, local
perceptions of mosquito breeding sites are not factored into the
programs, which can impact their effectiveness. In addition, the
gotong royong or tidak apa attitudes of residents towards their role
in preventing dengue within the community impacts the sustain-
ability of control programs.
The map outputs generated in Putrajaya and Seremban identi-
fied several areas participants associated with dengue risk
including construction sites, green space, abandoned property, and
drainage networks. Some of these areas, such as abandoned lots
and drains, have been found to be productive breeding sites in
Malaysia and other endemic countries (David, Lourenço-de-
Oliveira, & Maciel-de-Freitas, 2009; Teng & Singh, 2001) and were
consistent with the views of vector-control staff. Abandoned lots
can pose challenges to health inspectors who have limited ability to
enter buildings and remove breeding sites due to legal barriers,
although a large entomological survey in six Asian cities found that
eliminating or treating outdoor containers in these spaces can have
an important impact (Arunachalam et al., 2010). Many of the spaces
described by participants, such as public or intradomestic areas
brought up the question of who is responsible, and emphasized the
need for communication of roles between residents and local
public health staff. The different focus of male and female mapping
group participants on the larger neighborhood areas or peri-
dometic areas respectively, also highlighted the importance of
considering gender roles in community-level health promotion
messages communicated by vector-control staff.
In Seremban, construction sites and associated garbage left by
migrant workers were perceived to contribute to dengue, high-
lighting a challenge in a region where large numbers of foreign
workers reside. There are socio-cultural and language barriers to
creating prevention strategies and community dengue activities
that include these groups, and an influx of immigrant workers can
reduce the social cohesion required to cooperate in removing
breeding sources (Spiegel et al., 2005). Foreign workers also create
pathways for transmission through increased human movement
between endemic dengue regions. In addition, foreign workers in
other Southeast Asian countries have been identified as a high-risk
group for mosquito-borne infections because they often live in
temporary housing quarters conducive to the outdoor breeder
Aedes albopictus (Ho et al., 2011). Residents in Putrajaya perceived
construction sites as risk areas for different reasons, due to the
storage of materials which could collect rainwater, and which were
considered the responsibility of the public health department. In
the past, active and abandoned construction sites have previously
been identified as some of the most productive (high percentage of
positive samples) Aedes breeding sites in Malaysia (Teng & Singh,
2001). This has been observed particularly in non-active sites,
such as during the Asian economic downturn in 1997/1998 where
significant Aedes mosquito breeding was found in abandoned
construction sites in Malaysia (Kwa, 2008). While these sites are
monitored by vector control officials in Putrajaya for Aedes mos-
quito larvae, residents were not aware of these activities taking
place, which highlights an opportunity for increasing awareness of
the work done by vector-control staff.
Although the prominent green spaces in Putrajaya the ‘Intelli-
gent Garden City’are elements designed to improve quality of life
for residents, the mapping groups identified these as risk areas for
dengue. Participants further regarded vector control activities such
as insecticide fogging in these areas as the responsibility of
municipal health staff. In some regions with endemic dengue,
household waste in outdoor environments is known to provide
mosquito breeding sites because it is not removed regularly and
collects rainfall (Banerjee, Aditya, & Saha, 2013; Thammapalo,
Chongsuvivatwong, Geater, & Dueravee, 2008). However in Putra-
jaya, concerns emerged from outdoor breeding sites created in
vegetation rather than waste containers. Although Aedes Albopictus
mosquitoes breed and rest in outdoor or vegetated areas, they are a
secondary dengue vector (Lambrechts, Scott, & Gubler, 2010; Wee
et al., 2011), while Aedes aegypti, which breeds and rests indoors,
is considered to be the primary vector. Ae. Albopictus has be found
in Kuala Lumpur and Selangor state region where Putrajaya is sit-
uated (Chen et al., 2006), but mosquito-trapping studies would be
required identify its presence in Putrajaya. Local vector staff did not
consider green spaces or construction sites mapped by participants
to be priority areas for dengue control, and their foremost concern
is housing and peri-domestic areas where the primary vector Aedes
aegypti breeds. This disparity in the way green areas in Putrajaya
are perceived as vulnerable spaces by mapping participants but not
by control staff of indicates an opportunity for increased commu-
nication of vector-control priorities.
The range of sites mapped by residents in this study highlights
the complexity of preventing dengue in Malaysia, and the interplay
between risk perception, socio-cultural factors, and practical issues
associated with interventions. Residents have limited time to spend
monitoring their environments for breeding sites, and public health
staff have limited resources to reduce mosquito populations,
requiring evidence-based control measures that target priority
areas. This study showed that understanding local perceptions of
dengue risk provides an additional pillar of information to vector
control staff, and this can be utilized to better deliver health pro-
motion messages related to breeding source elimination. However,
fatigue with clean-up programs and the tidak apa attitude towards
dengue prevention is a challenge, especially during periods when
dengue transmission is low, and thus strategies to augment social
capital and sustain community participation are critical (Atkinson
et al., 2010). Participation could be encouraged by integrating
dengue messages with other health campaigns, such as free aero-
bics classes offered in Putrajaya, and at popular community gath-
erings, such as neighborhood fairs. Involving the communities in
mosquito surveys, program monitoring, and sharing of results
could build a sense ownership, and enhance acceptability of
behavior change communication to address misconceptions and
incorrectly identified breeding sites.
Visualizing risk perceptions
A large focus of this study was on portraying visual un-
derstandings of risk perceptions, an approach which has several
advantages and limitations. Visualization can enable quick assess-
ments of trends and relationships, and is useful for communicating
in cross-cultural environments (Fisher & Myers, 2011). In this study,
most participants demonstrated spatial understanding of the local
environment and were able to illustrate this in map format. The
process of drawing a sketch map highlighted features of interest to
community members, rather than containing all geographic ele-
ments and, as expected, the maps were generally not accurate to
scale. However in other contexts participants could require training
to obtain this level of geospatial knowledge. A limitation in this
study was that some features were more straightforward to map
than others, such as community landmarks compared with smaller
features like drains or alleys. In addition, the approach focused on
outdoor environments, and areas inside the home were not
included. However, domestic spaces are important contributors to
S.K. Dickin et al. / Applied Geography 46 (2014) 71e79 77

dengue transmission due to the adaptation of the A. aegypti mos-
quito to artificial breeding sites and need to be considered as part of
dengue prevention activities (Phillips, 2008). While community
sketch maps are commonly used by planners (Talen, 2000), geore-
ferencing this information in a GIS allowed the outputs to be more
easily understood by stakeholders involved in dengue prevention.
Conclusion
Community mapping and participatory GIS approaches are
increasingly being used to promote health. This study applied an
innovative approach to map and georeference local perceptions of
mosquito breeding sites and dengue risk in two municipalities in
Malaysia. Dengue control is a shared challenge among community
members and local public health staff in Malaysia, but top-down
approaches are used to communicate messages to residents. In
both Putrajaya and Seremban, participants mapped many sites in
their communities where mosquitoes could breed, including
blocked drains. However, different risk factors emerged, such as
vegetation causing blocked drains in Putrajaya, while waste mate-
rials blocked drains in Seremban. By georeferencing the community
generated maps, public health staff were able to visualize and
discuss community risk perceptions and identify spaces that
differed from the health department’s vector control priorities.
The range of breeding sites identified by residents also em-
phasizes the need for evidence-based control activities, including
entomological surveys to understand the distribution of mosquito
populations, as well as the importance of tapping potential com-
munity knowledge in identifying the most productive breeding
sites. Understanding local perceptions of mosquito sources could
improve vector control plans by identifying areas that may be
overlooked during regular control activities. Furthermore, under-
standing perceptions and can help public health workers create
education or awareness building strategies based on the level of
community knowledge. Visual approaches for assessing risk are
conducive to easy dissemination of findings to stakeholders and
vector-control staff. The different locations of breeding sites and
risk factors described in Putrajaya and Seremban also highlight the
need for tailored control programs to combat dengue even within
small geographic regions.
Acknowledgments
This research was funded by the Community of Practice in
Ecosystem Approaches to Health (CoPEH-Canada) and a SSHRC
Vanier Canada Graduate Scholarship. The United Nations University
International Institute for Global Health in Kuala Lumpur (UNU-
IIGH) assisted in applying for ethics approval and translation.
References
Ai-leen, G. T., & Song, R. J. (2000). The use of GIS in ovitrap monitoring for dengue
control in Singapore. Dengue Bulletin, 24,110e116.
Ansumana, R., Malanoski, A. P., Bockarie, A. S., Sundufu, A. J., Jimmy, D. H.,
Bangura, U., et al. (2010). Enabling methods for community health mapping in
developing countries. International Journal of Health Geographics, 9(1), 56.
Arunachalam, N., Tana, S., Espino, F., Kittayapong, P., Abeyewickreme, W., &
Wai, K. T. (2010). Eco-bio-social determinants of dengue vector breeding: a
multicountry study in urban and periurban Asia. Bulletin of the World Health
Organization, 88,173e18 3.
Atkinson, J. A. M., Fitzgerald, L., Toaliu, H., Taleo, G., Tynan, A., Whittaker, M., et al.
(2010). Research community participation for malaria elimination in Tafea
Province, Vanuatu: part I. Maintaining motivation for prevention practices in
the context of disappearing disease. Malaria Journal, 9,93.
Banerjee, S., Aditya, G., & Saha, G. K. (2013). Household disposables as breeding
habitats of dengue vectors: linking wastes and public health. Waste Manage-
ment, 33(1), 233e239.
Beatty, M. E., Stone, A., Fitzsimons, D. W., Hanna, J. N., Lam, S. K., Vong, S., &
Margolis, H. S. (2010). Best practices in dengue surveillance: a report from the
Asia-Pacific and Americas Dengue Prevention Boards. PLoS Neglected Tropical
Diseases, 4(11), e890.
Beyer, K. M. M., Comstock, S., & Seagren, R. (2010). Disease maps as context for
community mapping: a methodological approach for linking confidential
health information with local geographical knowledge for community health
research. Journal of Community Health, 35,635e644.
Cadag, J. R. D., & Gaillard, J. C. (2012). Integrating knowledge and actions in disaster
risk reduction: the contribution of participatory mapping. Area, 44,100e109.
Chang, A. Y., Parrales, M. E., Jimenez, J., Sobieszczyk, M. E., Hammer, S. M.,
Copenhaver, D. J., et al. (2009). Combining Google Earth and GIS mapping
technologies in a dengue surveillance system for developing countries. Inter-
national Journal of Health Geographics, 8,49.
Chen, C. D., Nazni, W. A., Lee, H. L., Seleena, B., Masri, S. M., Chiang, Y. F., et al. (2006).
Mixed breeding of Aedes aegypti (L.) and Aedes albopictus Skuse in four dengue
endemic areas in Kuala Lumpur and Selangor, Malaysia. Tropical Biomedicine,
23(2), 224e227.
David, M. R., Lourenço-de-Oliveira, R., & Maciel-de-Freitas, R. (2009). Container
productivity, daily survival rates and dispersal of Aedes aegypti females in a high
income dengue epidemic neighborhood of Rio de Janeiro: presumed influence
of differential urban structure on mosquito biology. Memórias do Instituto
Oswaldo Cruz, 104,927e932.
Dongus, S., Nyika, D., Kannady, K., Mtasiwa, D., Mshinda, H., Fillinger, U., et al.
(2007). Participatory mapping of target areas to enable operational larval
source management to suppress malaria vector mosquitoes in Dar es Salaam,
Tanzania. International Journal of Health Geographics, 6,37.
Dunn, C. E. (2007). Participatory GIS a people’s GIS? Progress in Human Geography,
31(5), 616e637.
Eisen, L., & Eisen, R. J. (2011). Using geographic information systems and decision
support systems for the prediction, prevention, and control of vector-borne
diseases. Annual Review of Entomology, 56,41e61.
Elwood, S. (2006). Critical issues in participatory GIS: deconstructions, re-
constructions, and new research directions. Transactions in GIS, 10(5), 693e708.
Fisher, R., & Myers, B. (2011). Free and simple GIS as appropriate for health mapping
in a low resource setting: a case study in eastern Indonesia. International Journal
of Health Geographics, 10,15.
Forget, G., & Lebel, J. (2001). An ecosystem approach to human health. International
Journal of Occupational and Environmental Health, 7,S3eS336.
Gaillard, J. C., Monteil, C., Perrillat-Collomb, A., Chaudhary, S., Chaudhary, M.,
Chaudhary, O., et al. (2013). Participatory 3-dimension mapping: a tool for
encouraging multi-caste collaboration to climate change adaptation and
disaster risk reduction. Applied Geography, 45,158e16 6.
Gubler, D. J. (2002). Epidemic dengue/dengue hemorrhagic fever as a public health,
social and economic problem in the 21st century. Trends in Microbiology, 10,
100 e103 .
Harrington,L.C.,Scott,T.W.,Lerdthusnee,K.,Coleman,R.C.,Costero,A.,Clark,G.G.,etal.(2005).
Dispersalofthe denguevectoraedesaegyptiwithinandbetweenruralcommunities.
AmericanJournalofTropicalMedicineandHygiene,72(2),209e220.
Hassan, M. M. (2005). Arsenic poisoning in Bangladesh: spatial mitigation planning
with GIS and public participation. Health Policy, 74,247e260.
Hay, S. I., Guerra, C. A., Gething, P. W., Patil, A. P., Tatem, A. J., Noor, A. M., et al.
(2009). A world malaria map: Plasmodium falciparum endemicity in 2007. PLoS
Medicine, 6(3), e1000048.
Ho, K., Ang, L. W., Tan, B. H., Tang, C. S., Ooi, P. L., James, L., et al. (2011). Epidemi-
ology and control of chikungunya fever in Singapore. Journal of Infection, 62(4),
263e270.
Kamadjeu, R. (2009). Tracking the polio virus down the Congo River: a case study
on the use of Google Earth in public health planning and mapping. International
Journal of Health Geographics, 8,4.
Kwa, B. H. (2008). Environmental change, development and vectorborne disease:
Malaysia’s experience with filariasis, scrub typhus and dengue. Environment,
Development and Sustainability, 10, 209e217.
Kyem, P. A. K. (2004). Power, participation, and inflexible institutions: an exami-
nation of the challenges to community empowerment in participatory GIS
applications. Cartographica, 38,5e17.
Lambrechts, L., Scott, T. W., & Gubler, D. J. (2010). Consequences of the expanding
global distribution of Aedes albopictus for dengue virus transmission. PLoS
Neglected Tropical Diseases, 4(5), e646.
Machado-Machado, E. A. (2012). Empirical mapping of suitability to dengue fever in
Mexico using species distribution modeling. Applied Geography, 33,82e93.
Malaysia Ministry of Health. (2010). Clinical practices guidelines: Management of
dengue infection in adults (Revised 2nd ed.). Putrajaya: Ministry of Health.
Martin, C., Curtis, B., Fraser, C., & Sharp, B. (2002). The use of a GIS-based malaria
information system for malaria research and control in South Africa. Health &
Place, 8,227e236.
Mayxay, M., Cui, W., Thammavong, S., Khensakhou, K., Vongxay, V., Inthasoum, L.,
et al. (2013). Dengue in peri-urban Pak-Ngum district, Vientiane capital of Laos:
a community survey on knowledge, attitudes and practices. BMC Public Health,
13(1), 434.
McCall, M. K., & Dunn, C. E. (2012). Geo-information tools for participatory spatial
planning: fullfilling the criteria for good governance. Geoforum, 43,81e94.
Moser, S. (2010). Putrajaya: Malaysia’s new federal administrative capital. Cities,
27(4), 285e297.
Mulligan, K., Elliott, S. J., & Schuster-Wallace, C. (2012). The place of health and the
health of place: dengue fever and urban governance in Putrajaya, Malaysia.
Health and Place, 18,613e620.
S.K. Dickin et al. / Applied Geography 46 (2014) 71e7978

Phillips, M. (2008). Dengue reborn: widespread resurgence of a resilient vector.
Environmental Health Perspectives, 116, A382eA388.
Phuanukoonnon, S., Brough, M., & Bryan, J. H. (2006). Folk knowledge about dengue
mosquitoes and contributions of health belief model in dengue control pro-
motion in Northeast Thailand. Acta Tropica, 99,6e14.
Shepard, D. S., Undurraga, E. A., & Halasa, Y. A. (2013). Economic and disease burden
of dengue in Southeast Asia. PLoS Neglected Tropical Diseases, 7(2), e2055.
Shuai, J., Buck, P., Sockett, P., Aramini, J., & Pollari, F. (2006). A GIS-driven integrated
real-time surveillance pilot system for national West Nile virus dead bird sur-
veillance in Canada. International Journal of Health Geographics, 5(1), 17.
Spiegel, J. M., Bennett, S., Hattersley, L., Hayden, M. H., Kittayapong, P., Nalim, S.,
et al. (2005). Barriers and bridges to prevention and control of dengue: the need
for a social-ecological approach. EcoHealth Journal, 2,273e290.
Suaya, J. A., Shepard, D. S., Siqueira, J. B., Martelli, C. T., Lum, L. C., Tan, L. H., et al.
(2009). Cost of dengue cases in eight countries in the Americas and Asia: a
prospective study. American Journal of Tropical Medicine and Hygiene, 80, 846e
855.
Talen, E. (2000). Bottom-up GIS ea new tool for individual and group expression in
participatoryplanning. Journalof the AmericanPlanning Association, 66(3),279e294.
Teng, A. K., & Singh, S. (2001). Epidemiology and new initiatives in the prevention
and control of dengue in Malaysia. Dengue Bulletin, 25,7e12.
Thammapalo,S.,Chongsuvivatwong,V.,Geater,A.,&Dueravee,M.(2008).Environmental
factorsandincidenceofdenguefeveranddenguehaemorrhagicfeverinanurbanarea,
SouthernThailand.EpidemiologyandInfection,136,135e143.
Waltner-Toews, D., & Kay, J. (2005). The evolution of an ecosystem approach: the
diamond schematic and an adaptive methodology for ecosystem sustainability
and health. Ecology and Society, 10,38.
Wee, L. K., Raduan, N., Wah, S. K., Ming, W. H., Shi, C. H., Rambli, F., et al. (2011).
Ovitrap surveillance of dengue and chikungunya vectors in several suburban
residential areas in Peninsular Malaysia. Dengue Bulletin, 35,153e160.
S.K. Dickin et al. / Applied Geography 46 (2014) 71e79 79