Am. J. Trop. Med. Hyg., 82(6), 2010, pp. 1024–1029
Copyright © 2010 by The American Society of Tropical Medicine and Hygiene
Vector-borne disease control programs have been a major
global challenge, especially with the emergence of insecticide
resistance among insect vectors, which resulted in the lim-
ited field success of single-method vector control programs. 1, 2
Integrated vector control programs, which involve combi-
nation strategies have been successful in some locations in
reducing the impact of vector-borne diseases, 3, 4 but there are
insufficient studies in different tropical settings where many
vector-borne diseases reside.
In Singapore, a tropical city-state in South-East Asia, malaria
was a main vector-borne disease resulting in substantial mor-
bidity and mortality in the early 20th Century. Singapore was
declared malaria-free in 1982 after meeting the World Health
Organization (WHO) assessment of having 1) a comprehen-
sive and efficacious case detection mechanism; 2) reliable
microscopic diagnosis of blood smears; 3) thorough epide-
miological investigations and a satisfactory epidemiological
situation; 4) adequate preventive and remedial actions upon
detection of cases; 5) adequate general health services, effec-
tive system of case notification, and epidemiological follow-up
for prevention of re-establishment of malaria. 5
Although Singapore was declared malaria-free, Tekong
Island (a forested island of approximately 5,900 acres located
to the north-east of Singapore that houses a military train-
ing facility, Figure 1 ) remained malaria-receptive because of
the presence of Anopheles sp. mosquitoes. From 1996 to 2003,
there were 19 cases of malaria on the island attributable to
secondary transmission on the island (5 in May/June 1996,
8 in July to September 1997, and 6 in August to October 2003),
despite the systematic use of anti-malaria prophylaxis for visi-
tors to the island since the 1980s. Because of the side effects of
anti-malaria chemoprophylaxis regimes 6– 8 and the operational
costs of chemoprophylaxis, an attempt was made through a
comprehensive integrated vector-borne disease management
program to reduce the mosquito population and negate the
need for chemoprophylaxis. This work describes the program’s
success and provides additional evidence for the use of combi-
nation strategies against mosquito-borne diseases in forested
To reduce the mosquito population and need for chemo-
prophylaxis, which had been in place since the 1980s, the
Singapore Armed Forces (SAF) performed an integrated vec-
tor-borne disease management program on Tekong Island from
December 2006. Surveillance programs were simultaneously
set up to determine the effectiveness of the interventions.
Tekong Island is located about 2 miles from the main island
of Singapore and is only accessible by sea or air ( Figure 1 ).
The island is forested, with mangrove swamps in the coastal
areas—ideal for the breeding of Anopheles mosquitoes. The
resident human population comprises several thousand con-
script military personnel who reside in the south of the island
and use the island for training. In addition, tens of thousands
of SAF military personnel visit the island annually for field
and jungle training, and thousands of foreign workers work on
construction projects on the island.
FOUR RINGS OF PREVENTION
To reduce the threat from malaria to individuals on the
island, we developed an integrated program to address the
different aspects of disease transmission. This is encapsulated
in the four prevention rings shown in Figure 2 , with the aim of
reducing the risk of local transmission of malaria to a negligi-
ble level where chemoprophylaxis can be removed and train-
ing can continue without substantial risk for malaria infections.
These comprise the following measures.
Prevent importation. A restriction period was imposed on
all visitors to the island who had visited malaria endemic areas
over the past 8 weeks (the long end of the average incubation
period for malaria). All foreign workers who work on
construction projects on the island would also be screened for
Elimination of Malaria Risk through Integrated Combination Strategies in a
Tropical Military Training Island
Vernon J. Lee ,* Samuel Ow , Harold Heah , Meng Yaw Tan , Patrick Lam , Lee-Ching Ng ,
Sai Gek Lam-Phua , Abdul Qadir Imran , and Benjamin Seet
Biodefence Center, Singapore Armed Forces, Singapore; Headquarters Army Medical Services, Singapore Armed Forces,
Singapore; Environmental Health Institute, National Environmental Agency, Singapore; Headquarters Medical Corps,
Singapore Armed Forces, Singapore
Abstract. On the military training facility of Tekong Island, Singapore, a comprehensive vector-borne disease control
program was started in end-2006 to reduce mosquito populations and negate the need for anti-malaria chemoprophylaxis.
The program was based on 1) preventing importation of malaria through screening of visitors, 2) preventing human-to-
mosquito transmission through early case detection and mosquito control, 3) preventing mosquito-to-human transmission
through personal protection, and 4) contingency plans. Systematic environmental works were performed to reduce breed-
ing sites, and insecticide use targeted both adult mosquitoes and larvae. Mosquito populations declined from 103 mosqui-
toes per sampling site in January 2007 to 6 per site by March 2007 ( P < 0.001). The proportion of positive ovitraps declined
from 93% in January 2007–2% in March 2007 ( P < 0.001). There were no malaria cases on the island despite chemopro-
phylaxis termination, showing that comprehensive combination vector-control strategies were effective in reducing the
risk of malaria.
* Address correspondence to Vernon Lee, 701 Transit Road, #04-01,
Singapore 778910. E-mail: firstname.lastname@example.org
COMPREHENSIVE STRATEGIES REDUCE MALARIA RISK
malaria using reverse transcription-polymerase chain reaction
(RT-PCR) after the quarantine period. In addition, any military
serviceman with fever of unknown origin would be screened
using real-time fluorescence-based RT-PCR for malaria
parasites performed by a certified laboratory, DSO National
Laboratories (Singapore). Positive cases were referred for
clinical management and prevented from entry to the island.
Early detection of human cases. All servicemen and
medical staff on the island and military camps in Singapore
were educated on malaria and the need for early reporting
of symptoms. Cases of fever (≥ 37.5°C) with no localizing
symptoms were tested for malaria and other common vector-
borne diseases in Singapore, and all confirmed malaria cases
were temporarily removed from the island.
Mosquito control program. The vector-control program
consists of environmental works to reduce mosquito breeding
habitats and an intense insecticide regimen. Malaria vectors
in Singapore were previously identified as Anopheles
sundaicus and Anopheles maculates. 9 Anopheles sinensis has
also recently been suspected to be a vector. Their respective
habitats, marshland with brackish water, streams in hilly areas,
and small pools, were thus specially targeted.
Systematic environmental and infrastructural improvements
were made to reduce breeding sites. They include shoreline
works to reduce the mixing of salt and fresh water, improve-
ments to drainage around built-up areas to reduce water pool-
ing, maintenance of drains, including clearance of vegetation
and undergrowth to allow accessibility across the island, and
filling up pools of water. Clearance of vegetation and under-
growth was maintained on a 3-monthly cycle to prevent rapid
re-growth in a jungle habitat. Residents were also encouraged
to reduce mosquito breeding habitats within the built-up areas
through education programs.
Insecticide use targeted mosquito larvae through the use
of Bacillus thuringiensis israelensis ( Bti ) larvicide by ultra-
low volume mist fogging. Bacillus thuringiensis israelensis has
been shown to be effective in sustained reduction of mosquito
populations in the control of malaria. 10– 12 This strategy used
a combination of vehicle-mounted ultra-low volume mist-
foggers for treatment of large areas along roads/tracks, and
man-packed mist-foggers for reaching deep forested areas
inaccessible to vehicles. Before the application of Bti , con-
ventional fogging using a combination of s -bioallethrin , per-
methrin, and piperonyl butoxide (Resigen, Bayer Cropscience,
Kuala Lumpur, Malaysia) and cypermethrin (New Cyper,
Agrolex Agrochemicals Manufacturing, Jiangsu, China) was
done daily for 3 consecutive days to reduce the adult mosquito
population. Bacillus thuringiensis israelensis was then admin-
istered across the island covering 200 g per acre of the entire
island at least four times per month. Handheld Global posi-
tional systems (GPS) were used to ensure systematic cover-
age of the entire island. In addition, indoor residual spraying
with alphacypermethrin 1.5 sc (Fendona, BASF, Malaysia) or
Resigen was applied by mist blowers in the existing buildings
to reduce exposure of the resident population to mosquitoes.
Personal protection. Existing personal protection measures
were continued to prevent mosquito-to-human transmission.
These included the use of bed nets, pre-treatment of uniforms
with permethrin, and application of N , N -Diethyl- meta -
toluamide (DEET) insect repellent regularly during jungle
Malaria contingency plan. A response plan involving
multiple agencies was developed in the event of an outbreak
of malaria (defined as a single case of malaria on the island)
despite the previous measures. In addition to routine clinical
management, this plan comprised active surveillance for
malaria cases and intensified vector surveillance. This was
done in conjunction with the DSO National Laboratories,
tertiary hospitals in Singapore, the Ministry of Health (MOH),
and the National Environment Agency (NEA).
To monitor the outcomes of reduction potential mosquito vec-
tors, which are most likely anthropophillic and anthrozoophillic,
Figure 2. The four rings of prevention of the malaria control program on Tekong Island.
Figure 1. Map of Singapore and Pulau Tekong.
LEE AND OTHERS
human bare leg catch was conducted by a commercial pest
control contractor without any soldiers being involved. A sys-
tematic surveillance at 60 fixed sites across the island was con-
ducted on a weekly basis using 20 men from 1800 to 0700 hrs
the following day, from November 2006 to April 2007. This was
reduced to fortnightly intervals from April 2007, and the num-
ber of sites was then reduced to 40 from June 2007 when the
initial vector-control target of 90% reduction was achieved.
Ovitraps were also placed in 50 fixed sites and any immatures
(ova or larvae) found in the trap would classify the trap as
positive for breeding.
This operational program, including use of the human
bait method for collection of mosquitoes, was approved by
the Singapore Armed Forces Medical Review Committee.
The program cost about SGD1 million per year (1 SGD =
US$0.664 in 2007).
The χ 2 and Fisher’s exact tests were used to compare the
categorical outcomes for bivariate analysis. The non-paramet-
ric Mann-Whitney test was used for two-group comparisons
continuous outcomes. Statistical analyses were performed
using Stata (version 9.0, Stata Corp., College Station, TX), and
all tests were conducted at the 5% level of significance.
By the end of 2008, 8,303 foreign workers had been screened
by DSO National Laboratories by RT-PCR before entry to the
island (100% compliance rate based on documented entry into
the island), and seven individuals tested positive for malaria.
They were barred from entering the island pending successful
treatment and evaluation.
The coastal construction work was completed by January
2007 with 12.5 km of silted and earthen drains cleared and
concretized, respectively, and 187 large permanent water
pools identified and filled across the island. For the insecti-
cide program, after one round of adulticide fogging, an initial
9 weekly cycles of Bti were applied across the entire island
from December 2006 to February 2007, and maintained at two
cycles per month thereafter.
Two months after program initiation, the mosquito catch
rate declined from a mean of 103 mosquitoes per human bait-
ing sampling site in January to 6 per site in March 2007 ( P <
0.001), and maintained at between 2 and 6 per site through
December 2008 ( Figure 3 and Figure 4 ). This represents a
decline of more than 94% of initial mosquito catch rate before
the program was instituted in December 2006. The proportion
of ovitraps that were positive declined from 93% in January
2007 to a nadir of 2% in March 2007 ( P < 0.001), and main-
tained at a median of 12% through December 2008 ( Figure 5 ).
The proportion of Anopheles mosquitoes (the main vector
of concern) remained low and did not change significantly
throughout the treatment period—from 1.63% in end 2006
and early 2007 to 1.32% in 2008 ( P = 0.41). This suggests a
similar reduction in the Anopheles population.
Subjective surveys on visitors to the island suggested a
substantial reduction in the number of mosquitoes and mos-
quito bites on the island compared with previous visits before
December 2006 when the program was started.
Because of the successful vector control efforts and strin-
gent malaria screening policies for visitors, the requirement
for malaria chemoprophylaxis on the island was lifted from
April 2007. Since April 2007 to the end of 2009, there had not
been any reported case of malaria on Tekong Island despite
the absence of chemoprophylaxis.
We have shown that a comprehensive combination approach
can reduce the overall mosquito population in a densely
Figure 3. Mean number of mosquitoes trapped by live human baiting per site from December 2006 to December 2008 (sixty sites from Dec
2006 to Jun 2007, 40 sites after Jun 2007).
COMPREHENSIVE STRATEGIES REDUCE MALARIA RISK
forested area, and this success can reasonably be maintained
across time. Although our circumstances and geography are
unique, different portions of the program can be extrapolated
in other forested areas where vector-borne disease control is
desirable. Elimination of breeding sites in forested areas is chal-
lenging, but we have done so through a systematic approach
using traditional methods together with more advanced tech-
nologies such as vehicle-mounted Bti application and GPS
mapping. The application of Bti together with environmental
works gave the program success in achieving control. Results
of our program show a likely systematic reduction of all mos-
quitoes before and after intervention.
Previous studies have also showed the efficacy of biological
larvicides alone in the control of malaria vector populations,
established the treatment doses, and frequencies of Bti . 10 This
program, which led to a significant reduction in mosquito pop-
ulations, is a translation of these results to show how Bti can
be incorporated into a comprehensive vector-control program
Figure 4. Adult mosquito surveillance through human baiting catches—at the start of the vector control program in January 2007 and in
Figure 5. Percentage of ovitraps that was positive for mosquito ova from January 2007 to December 2008.
LEE AND OTHERS
and effectively mitigate vector-borne disease risk. Many other
countries have also used Bti as part of vector-control pro-
grams to control various mosquito species such as Anopheles
gambiae , Anopheles sundaicus , and Aedes agypti , with success
rates of 50–99%. 11– 24
Bacillus thuringiensis israelensis treatment, although effec-
tive in reducing mosquito populations, did not totally eradicate
them and additional measures, such as screening for malaria
cases, was effective in identifying previously undetected cases.
Other combination programs were effective in different envi-
ronments. One was in the Region of Peel, Ontario, North
America, a region struck by multiple West Nile virus epidem-
ics. 25 In 2004, several initiatives were implemented including
the treatment of water with larvicide ( Bti and methoprene),
using Larvasonic devices with acoustic energy to kill mos-
quito larvae, and using fathead minnow fish for vector lar-
vae control. 25 The post-treatment pupal monitoring indicated
the methoprene pellet’s and Larvasonic device’s efficacy rate
was 90% each. 25 For treatment of stagnant water using Bti
products, pre- and post-treatment counts of mosquito larva
revealed an efficacy of 100% at one site, and an average effi-
cacy of 76.5% in nine other monitored sites. Between 1998 and
2004, the incidence of vector-borne malaria cases in Eriteria
was reduced substantially with the implementation of insec-
ticide-impregnated mosquito nets, indoor residual spraying,
and larval control measures. 16, 26 As part of the WHO initia-
tive, Intergrated vector management (IVM) programs are cur-
rently being implemented in multiple countries in Africa, and
early results of mosquito eradication and disease incidence
reduction appears promising. 27
From a public health policy standpoint, this comprehensive
program not only reduced the potential threat of mosquito-
borne diseases but also the need for anti-malaria chemopro-
phylaxis. Systematic chemoprophylaxis of a large number of
individuals with drugs that have substantial side-effects will
result in constant morbidity and occasional severe effects. Side
effects for mefloquine include neurological and cardiac distu-
bances; for doxycycline include gastrointestinal disturbances
and hypersensitivty reactions; for malarone include gastroin-
testinal disturbances and blood abnormalities. Reducing the
need for chemoprophylaxis was therefore a desirable out-
come of the program. The four rings allowed for the entire
program to work even if individual rings are not 100% effec-
tive. For example, although the mosquito population has been
significantly reduced, mosquitoes are still present. The addi-
tional screening program, which identified human malaria
cases, helped to reduce the risk of parasite introduction to
the island. The importance of program maintenance must
be emphasized. Bacillus thuringiensis israelensis needs to be
continually applied to all breeding and potential breeding
sites at regular intervals and environmental works must be
regularly maintained. Our results have also shown that low
vector populations are sustainable with a consistent, com-
prehensive, and multi-faceted program that is maintained
Compared with the occasional clusters of locally transmit-
ted malaria despite anti-malaria chemoprophylaxis previously,
we have not had a case with this program in more than 2 years
with the increased surveillance. The seven individuals iden-
tified to have possible malaria infection may have resulted
in outbreaks if they were allowed to enter the island and if
mosquito populations were at previous levels. However, the
number of human infections of malaria attributable to Tekong
Island during the period of chemoprophylaxis were low and
longer term follow-up is necessary to determine if the preven-
tive measures and mosquito population is below the threshold
required for such rare outbreaks to occur.
There were some limitations encountered in this program.
The substantial vector control and environmental works
may have affected the eco-system of the surrounding areas.
However, the combination strategy with Bti (which has less
ecological impact) and other larvicides may have helped to
reduce the impact of excessive use of any single method. The
use of the human baiting method for surveillance also meant
that the mosquitoes caught were mainly of the human-biting
variety. However, these are the mosquito species that are most
relevant to the spread of vector-borne diseases. Finally, the pro-
gram used multiple layers of protection to ensure the reduction
of risk from vector-borne diseases, and may not be easily repro-
duced in its entirety in rural regions. However, components of
the program such as visitor screening, Bti spraying, and per-
sonal protection can be combined for use in various settings.
We have shown that comprehensive combination vector-
control strategies are effective in reducing and maintaining
low mosquito populations and reduce the risk of disease trans-
mission. These lessons can be extrapolated in other forested
environments to reduce the threat of vector-borne infectious
Received September 16, 2009. Accepted for publication February 11,
Acknowledgments: We thank the staff of the Headquarters Army
Medical Services; Biodefence Centre, Singapore Armed Forces; DSO
National Laboratories; and Environmental Health Institute, National
Environmental Agency for their support for this study.
Disclaimer: The authors do not have any conflicts of interest, financial
or otherwise, in this study.
Authors’ addresses : Vernon J. Lee, Biodefence Center, Singapore Armed
Forces, Singapore, Centre for Health Services Research, Yong Loo
Lin School of Medicine, National University of Singapore, Singapore,
and Department of Epidemiology and Public Health, Yong Loo Lin
School of Medicine, National University of Singapore, Singapore,
E-mail: email@example.com. Samuel Ow, Patrick Lam, and Abdul
Qadir Imran, Biodefence Center, Singapore Armed Forces, Singapore,
E-mails: firstname.lastname@example.org, lam_hong_yeong_patrick_damian@
starnet.gov.sg, and email@example.com. Harold Heah, Meng
Yaw Tan, and Benjamin Seet, Headquarters Army Medical Services,
Singapore Armed Forces, Singapore, E-mails: firstname.lastname@example.org,
email@example.com, and seet_hun_yew_benjamin@starnet.
gov.sg. Lee-Ching Ng and Sai Gek Lam-Phua, Environmental Health
Institute, National Environmental Agency, Singapore, E-mails: ng_lee_
firstname.lastname@example.org and email@example.com.
1. Hargreaves K , Hunt RH , Brooke BD , 2003 . Anopheles arabiensis
and An. quadriannulatus resistance to DDT in South Africa .
Med Vet Entomol 17: 417 – 422 .
2. Hargreaves K , Koekemoer LL , Brooke BD , 2000 . Anopheles
funestus resistant to pyrethroid insecticides in South Africa .
Med Vet Entomol 7 : 181 – 189 .
3. Filliner U , Kannady K , William G , 2008 . A tool box for operational
mosquito larval control: preliminary results and early lessons
from the Urban Malaria Control Programme in Dar es Salaam,
Tanzania . Malar J 7: 20 .
COMPREHENSIVE STRATEGIES REDUCE MALARIA RISK
4. Kawaguchi I , Sasaki A , Mogi M , 2004 . Combining zooprophylaxis
and insecticide spraying: a malaria-control strategy limiting the
development of insecticide resistance in vector mosquitoes .
Proc Biol Sci 271: 301 – 309 .
5. Goh KT , 1983 . Eradication of malaria from Singapore . Singapore
Med J 25: 255 – 268 .
6. Tee AK , Oh HM , Wee IY , 2004 . Dapsone hypersensitivity syn-
drome masquerading as a viral exanthem: three cases and a
mini-review . Ann Acad Med Singapore 33 : 375 – 378 .
7. Teo RY , Tay YK , Tan CH , 2006 . Presumed dapsone-induced drug
hypersensitivity syndrome causing reversible hypersensitivity
myocarditis and thyrotoxicosis . Ann Acad Med Singapore 53 :
385 – 391 .
8. Hussien OA , 2007 . Antimalarial drug toxicity: a review .
Chemotherapy 53: 385 – 391 .
9. Lee YC , Tang CS , Ang LW , Han HK , James L , Goh KT , 2009 .
Epidemiological characteristics of imported and locally-acquired
malaria in Singapore . Ann Acad Med Singapore 38 : 840 – 849 .
10. Lancey LA , 2007 . Bacillus thuringiensis serovariety israelensis and
Bacillus sphaericus for mosquito control . J Am Mosq Control
Assoc 23 : 133 – 163 .
11. Fillinger U , Knols BGJ , Becker N , 2003 . Efficacy and efficiency of
new Bacillus thuringiensis var. israelensis and Bacillus sphaeri-
cus formulations against Afrotropical anophelines in Western
Kenya . Trop Med Int Health 8 : 37 – 47 .
12. Majambere , Lindsay S , Green C , Kandeh B , Fillinger U , 2007 .
Microbial larvicides for malaria control in The Gambia . Malar J
6 : 76 .
13. Pepperell C , Rau N , Krajden S , 2003 . West Nile virus infection in
2002: morbidity and mortality among patients admitted to hos-
pital in south central Ontario . CMAJ 168 : 1399 – 1405 .
14. Kroeger A , Horstick O , Riedl C , 1995 . The potential for malaria
control with the biological larvicide Bacillus thuringiensis israel-
ensis ( Bti ) in Peru and Ecuador . Acta Trop 60: 47 – 57 .
15. Brown M , Carter J , Watson TM , 2001 . Evaluation of liquid Bacillus
thuringiensis var. israelensis products for control of Australian
Aedes arbovirus vectors . J Am Mosq Control Assoc 17 : 8 – 12 .
16. Russell TL , Brown MD , Purdie DM , 2003 . Efficacy of VectoBac
( Bacillus thuringiensis variety israelensis ) formulations for mos-
quito control in Australia . J Econ Entomol 96 : 1786 – 1791 .
17. Nyarango PM , Gebremaeskel T , Mebrahtu G , 2006 . A steep
decline of malaria morbidity and mortality trends in Eritrea
between 2000 and 2004: the effect of combination of control
methods . Malar J 5 : 33 .
18. Gunasekaran K , Doss PS , Vaidyanathan K , 2004 . Laboratory and
field evaluation of Teknar HP-D, a biolarvicidal formulation of
Bacillus thuringiensis spp. israelensis , against mosquito vectors .
Acta Trop 92 : 109 – 118 .
19. Zhakhongirov ShM , Yarbaboyev MKh , Khamraeyeva ASh ,
Bekker N , Lebedeva NI , Ponomarev IM , Muminov MS , Tsoı ˘
EG , 2004 . Testing the efficiency of Bacillus thuringien israelesis
against mosquito larvae in Uzbekistan . Med Parazitol (Mosk)
28 – 31 .
20. Sharma SN , Shukla RP , Mittal PK , 2003 . Efficacy of a new formu-
lation of Bacillus thuringiensis var. israelensis ( Bti ) in labora-
tory and field conditions of Kumaun foothills of Uttaranchal,
India . J Commun Dis 35: 290 – 299 .
21. Teng HJ , Lu LC , Wu YL , 2005 . Evaluation of various control agents
against mosquito larvae in rice paddies in Taiwan. J Vector Ecol
30 : 126 – 132 .
22. Lee YW , Zairi J , 2006 . Field evaluation of Bacillus thuringiensis
H-14 against Aedes mosquitoes . Trop Biomed 23: 37 – 44 .
23. Setha T , Chantha N , Socheat D , 2007 . Efficacy of Bacillus thuringi-
ensis israelensis , VectoBac WG and DT, formulations against
dengue mosquito vectors in cement potable water jars in
Cambodia . Southeast Asian J Trop Med Public Health 38:
261 – 268 .
24. Becker N , 1997 . Microbial control of mosquitoes: management of
the upper rhine mosquito population as a model programme .
Parasitol Today 13: 485 – 487 .
25. Region of Peel , 2004 . West Nile Virus in the Region of Peel 2004
Report . Available at: http://www.peelregion.ca/health/westnile/
resources/reports.htm #report2004. Accessed April 9, 2009.
26. Graves PM , Osgood DE , Thomson MC , 2003 . Effectiveness of
malaria control during changing climate conditions in Eritrea,
1998 . Trop Med Int Health 13: 218 – 228 .
27. Manga L , Toure A , Shililu J , 2004 . Implementation of Integrated
Vector Management in the WHO-African Region: Progress
Report 2000–2003 . Washington, DC : Environmental Health
Project, U.S. Agency for International Development , 1 – 26 .