examining the impact of climate change on dengue...

EXAMINING THE IMPACT OF CLIMATE CHANGE ON

DENGUE TRANSMISSION IN THE ASIA-PACIFIC

REGION

Shahera Banu Doctor of Veterinary Medicine, Master of Science

A thesis submitted for the degree of Doctor of Philosophy

School of Public Health and Social Work

Faculty of Health

Queensland University of Technology

November 2013

Examining the impact of climate change on dengue transmission in the Asia-Pacific Region i

Keywords

Dengue fever

Climate change

Scan statistics

Space-time cluster

Distributed Lag Model

Socio-environmental factors

Geographic Information System

ii Examining the impact of climate change on dengue transmission in the Asia-Pacific Region

Abstract

Dengue fever (DF) is a serious public health concern in many parts of the

world. An increase in DF incidence has been observed globally over the past

decades. Multiple factors including urbanisation, increased international travels and

global climate change are thought to be responsible for increased DF. However, little

research has been conducted in the Asia-Pacific region about the impact of these

changes on dengue transmission. The overarching aim of this thesis is to explore the

spatiotemporal pattern of DF transmission in the Asia-Pacific region and project the

future risk of DF attributable to climate change.

Annual data of DF outbreaks for sixteen countries in the Asia-Pacific region

over the last fifty years were used in this study. The results show that the geographic

range of DF in this region increased significantly over the study period. Thailand,

Vietnam and Laos were identified as the highest risk areas and there was a southward

expansion observed in the transmission pattern of DF which might have originated

from Philippines or Thailand. Additionally, the detailed DF data were obtained and

the space-time clustering of DF transmission was examined in Bangladesh. Monthly

DF data were used for the entire country at the district level during 2000-2009.

Dhaka district was identified as the most likely DF cluster in Bangladesh and several

districts of the southern part of Bangladesh were identified as secondary clusters in

the years 2000-2002.

In order to examine the association between meteorological factors and DF

transmission and to project the future risk of DF using different climate change

scenarios, the climate-DF relationship was examined in Dhaka, Bangladesh. The

results show that climate variability (particularly maximum temperature and relative

humidity) was positively associated with DF transmission in Dhaka. The effects of

climate variability were observed at a lag of four months which might help to

potentially control and prevent DF outbreaks through effective vector management

and community education. Based on the quantitative assessment of the climate-DF

Examining the impact of climate change on dengue transmission in the Asia-Pacific Region iii

relationship, projected climate change will likely increase mosquito abundance and

activity and DF in this area. Assuming a temperature increase of 3.3oC without any

adaptation measures and significant changes in socio-economic conditions, the

consequence will be devastating, with a projected annual increase of 16,030 cases in

Dhaka, Bangladesh by the end of this century. Therefore, public health authorities

need to be prepared for likely increase of DF transmission in this region.

This study adds to the literature on the recent trends of DF and impacts of

climate change on DF transmission. These findings may have significant public

health implications for the control and prevention of DF, particularly in the Asia-

Pacific region.

iv Examining the impact of climate change on dengue transmission in the Asia-Pacific Region

Table of Contents

Keywords .................................................................................................................................................. i

Abstract ................................................................................................................................................... ii

Table of Contents ................................................................................................................................... iv

List of Figures ....................................................................................................................................... viii

List of Tables ........................................................................................................................................... ix

List of Abbreviations ................................................................................................................................ x

Statement of Original Authorship .......................................................................................................... xi

Acknowledgements ............................................................................................................................... xii

CHAPTER 1: INTRODUCTION........................................................................................................ 1

1.1 Background .................................................................................................................................. 1

1.2 Research Aims .............................................................................................................................. 4

1.3 Thesis Structure ........................................................................................................................... 4

CHAPTER 2: LITERATURE REVIEW ................................................................................................ 7

2.1 Characteristics and Ecology of DF ................................................................................................ 7

2.2 Determinants of DF transmission ................................................................................................ 9

2.3 Emergence and geographic spread of DF .................................................................................. 11

2.4 Climate variability and DF .......................................................................................................... 13

2.5 Global climate change and DF ................................................................................................... 15

Systematic review ................................................................................................................................. 16

Dengue transmission in the Asia-Pacific region: impact of climate change and socio-environmental factors 16

2.6 Summary .................................................................................................................................... 17

2.7 Introduction ............................................................................................................................... 18

2.8 Methods ..................................................................................................................................... 20

2.8.1 Inclusion criteria ............................................................................................................. 22

2.9 Results ........................................................................................................................................ 24

2.9.1 Association between climatic factors and DF ................................................................. 24

Examining the impact of climate change on dengue transmission in the Asia-Pacific Region v

2.9.2 Association between socio-environmental factors and DF ............................................ 32

2.10 Conclusion .................................................................................................................................. 35

2.11 Update of the systematic review ............................................................................................... 42

2.11.1 Association between climatic factors and DF ................................................................. 42

2.11.2 Association between socio-environmental factors and DF ............................................ 43

2.12 Knowlege Gaps........................................................................................................................... 44

CHAPTER 3: RESULTS PAPER 1 ................................................................................................... 45

Dynamic spatiotemporal trends of dengue transmission in the Asia-Pacific region during 1955–2004 45

3.1 Abstract ...................................................................................................................................... 46

3.2 Introduction ............................................................................................................................... 47

3.3 Methods ..................................................................................................................................... 48

3.3.1 Study area ....................................................................................................................... 48

3.3.2 Data collection ................................................................................................................ 49

3.3.3 Data Analyses.................................................................................................................. 50

3.4 Results ........................................................................................................................................ 56

3.4.1 Descriptive statistics ....................................................................................................... 56

3.4.2 Trends of DF transmission .............................................................................................. 56

3.4.3 Space-time clusters ......................................................................................................... 59

3.5 Discussion .................................................................................................................................. 59


Space-time clusters of dengue fever in Bangladesh ............................................................................. 67

4.1 Summary .................................................................................................................................... 68

4.2 Introduction ............................................................................................................................... 69

4.3 Methods ..................................................................................................................................... 70

4.3.1 Study area ....................................................................................................................... 70

4.3.2 Data collection ................................................................................................................ 70

4.3.3 Statistical analysis ........................................................................................................... 71

4.4 Results ........................................................................................................................................ 72

4.4.1 DF epidemics and outbreaks .......................................................................................... 72

vi Examining the impact of climate change on dengue transmission in the Asia-Pacific Region

4.4.2 Disease clusters .............................................................................................................. 72

4.4.3 Spatial dispersion of DF .................................................................................................. 77

4.5 Discussion .................................................................................................................................. 78

4.6 Conclusion .................................................................................................................................. 80


Projecting the impact of climate change on dengue transmission in Dhaka, Bangladesh. ................... 85

5.1 Abstract ...................................................................................................................................... 86

5.2 Introduction ............................................................................................................................... 87

5.3 Materials and methods .............................................................................................................. 89

5.3.1 Study area ....................................................................................................................... 89

5.3.2 Data collection ................................................................................................................ 89

5.3.3 Data analysis ................................................................................................................... 90

5.4 Results ........................................................................................................................................ 93

5.5 Discussion ................................................................................................................................ 101

5.6 Conclusion ................................................................................................................................ 105

CHAPTER 6: GENERAL DISCUSSION .......................................................................................... 110

6.1 Substantive discussion ............................................................................................................. 110

6.2 Implications of the research findings ....................................................................................... 115

6.3 Strengths of this thesis ............................................................................................................. 116

6.4 Limitations of this thesis .......................................................................................................... 117

6.5 Future research directions ....................................................................................................... 118

6.5.1 Better understand the ecology of DF ........................................................................... 118

6.5.2 Advance the risk assessment techniques ..................................................................... 118

6.5.3 Improve disease surveillance and monitoring .............................................................. 119

6.5.4 Evaluate the effectiveness of the public health interventions ..................................... 119

6.5.5 Translate research into policy and risk management practice ..................................... 119

6.6 Conclusions .............................................................................................................................. 120

BIBLIOGRAPHY 121

APPENDICES 137

Examining the impact of climate change on dengue transmission in the Asia-Pacific Region vii

Appendix A 137

Appendix B 138

Appendix C 139

viii Examining the impact of climate change on dengue transmission in the Asia-Pacific Region

List of Figures

Figure 2.1 Transmission cycle of dengue fever ....................................................................................... 8

Figure 2.2 Determinants of dengue transmission (Sutherst, 2004) ...................................................... 10

Figure 2.3 Countries at risk of DF transmission, 2008 (Source: World Health Organization, http://www.nathnac.org/pro/factsheets/images/clip_image002_015.jpg)........................ 12

Figure 2.4 Flow diagram of article selection process ............................................................................ 23

Figure 3.1 Total numbers of countries with DF outbreaks in the Asia-Pacific region, during 1955-2004. (Data source: WHO DengueNet) ....................................................................... 53

Figure 3.2 Cumulative incidence of DF in the Asia-Pacific countries. A: 1955-1964, B: 1965-1974, C: 1975-19854,D:1985-1994,E:1995-2004.The X and Y axis of the map show the longitude and latitude respectively. .............................................................................. 55

Figure 3.3 Space-time clusters of DF transmission in the Asia-Pacific region. A: 1955-1964, B: 1965-1974,C:1975-19854,D:1985-1994,E:1995-2004. The X and Y axis of the map show the longitude and latitude respectively. ..................................................................... 58

Figure 4.1 Monthly number of DF cases, deaths and districts with DF notification between January 2000 and December 2009 in Bangladesh. .............................................................. 73

Figure 4.2. Space-time clusters of DF identified in three different periods in Bangladesh. .................. 74

Figure 5.1 Auto-correlation function partial auto-correlation function and scatter plot of residuals for DLM. ................................................................................................................ 97

Figure 5.2 Association between climatic variables (maximum temperature and relative humidity) and DF at different lags. ....................................................................................... 98

Figure 5.3 Validated distributed lag model of climate variation in Dhaka, Bangladesh (validation period = Jan 2009 – Dec 2010 i.e., the cross validation period). ........................ 99

Figure 6.1 Framework of research findings in this thesis .................................................................... 112

Examining the impact of climate change on dengue transmission in the Asia-Pacific Region ix

List of Tables

Table 2.1 Search strategy ...................................................................................................................... 21

Table 2.2 Characteristics of studies on the association between climatic variables and DF transmission ......................................................................................................................... 25

Table 2.3 Characteristics of studies on the association between Socio-environmental factors and DF transmission ............................................................................................................. 34

Table 3.1 Summary statistics for annual number of DF cases in the Asia-Pacific countries during 1955-2004 ................................................................................................................. 52

Table 3.2 Space-time clusters of DF transmission in the Asia-Pacific region during 1955- 2004 .......... 57

Table 4.1 Space-time clusters of DF in Bangladesh, 2000-2009. ........................................................... 75

Table 5.1 Descriptive statistics of monthly climatic conditions and DF in Dhaka, Bangladesh, 2000-2010. ........................................................................................................................... 94

Table 5.2 Spearman’s correlation coefficients between monthly climatic variations in Dhaka, Bangladesh. .......................................................................................................................... 94

Table 5.3 Deviances for the relationship between DF and different temperature measures by DLM ...................................................................................................................................... 95

Table 5.4 Deviances for DLM using different covariates ....................................................................... 95

Table 5.5: Sensitivity and specificity of DLM for dengue occurrence. .................................................. 96

Table 5.6 Changes in annual DF incidence under different scenarios of temperature increase by 2100 in Dhaka, Bangladesh. .......................................................................................... 100

x Examining the impact of climate change on dengue transmission in the Asia-Pacific Region

List of Abbreviations

AIC Akaike Information Criterion

ARIMA Autoregressive Integrated Moving Average

BBS Bangladesh Bureau of Statistics

CIA Central Intelligence Agency

DF Dengue Fever

DHF Dengue Hemorrhagic Fever

DSS Dengue Shock Syndrome

DGHS Directorate General of Health Services

DLM Distributed Lag Model

ENSO EI Niño-Southern Oscillation

GIS Geographic Information System

IPCC Intergovernmental Panel on Climate Change

RR Relative Risk

SARIMA Seasonal Autoregressive Integrated Moving Average

SOI Southern Oscillation Index

WHO World Health Organization

QUT Verified Signature

xii Examining the impact of climate change on dengue transmission in the Asia-Pacific Region

Acknowledgements

I am grateful to my Principal Supervisor Professor Shilu Tong who provided

me with an opportunity to do a postgraduate study at QUT. I want to thank him for

his constant support, generosity, patience and unlimited time for help. It has been a

long journey but very pleasant. I would also like to thank my associate supervisors

Dr. Wenbiao Hu and Dr. Cameron Hurst for their kindness and helpful advice on

study design and research findings. Thanks to Dr. Yuming Guo for his assistance on

GIS techniques and R software.

I would also like to thank Dr. Su Naish, Dr. Wei Wei Yu, Dr. Lyle Turner, Mr.

Cunrui Huang, Ms. Xiaofang Ye, Ms. Xiaoyu Wang, Ms. Yan Bi, Mr. Xin Qi, Mr.

Zhiwei Xu, Ms. Jia Jia Wang and other colleagues for their help with my research.

Thanks to the Directorate General of Health Services, Bangladesh and

Bangladesh Meteorological Department for providing valuable data on dengue and

climate, respectively.

Thanks to the Faculty of Health for providing me with the Post Graduate

research Scholarship. I want to thank all staff members in the School of Public

Health (SPH), Research Student Centre and IT help desk who have provided an

excellent and convenient environment for postgraduate research at QUT.

Finally, I am grateful to my family for their love, inspiration and support,

without which I would not have gone so far as I did.

Chapter 1: Introduction 1

Chapter 1: Introduction

1.1 BACKGROUND

Dengue fever (DF) is one of the most important emerging arboviral diseases

worldwide (WHO, 2012a). The disease is caused by one of the four viral serotypes of

the genus flavivirus. Dengue virus transmits through the bite of container breeding

mosquitoes Aedes aegypti and Aedes albopictus, which occur in most tropical and

subtropical countries (Simmons et al., 2012, Gibbons, 2010). Symptoms of DF vary

from febrile illness to severe haemorrhagic fever (Guzman et al., 2010, Gubler,

2004b). Up to 100 million infections occur annually and 22,000 children die due to

DF globally (Guzman et al., 2010, Beatty et al., 2011, Gibbons, 2010). More than

50% of the world’s population live in over 100 countries which are currently at risk

of DF. According to the World Health Organization (WHO), DF transmission

increased by 30-fold worldwide over the last 50 years and the transmission

predominantly increased in urban and semi-urban areas (WHO, 2012a). DF has

become an increasing public health concern around the world due to its serious

health consequences including death and the lack of effective vaccine and specific

treatment.

The risk of DF transmission varies over space and time and the dynamics of

disease depend on individual, household and climatic factors (Mackenzie et al.,

2004). Weather variation, virus strains, mosquito densities, human activities and

movement and population immunity, all contribute to the transmission of DF

(Gubler, 2011). Transmission is higher in urban and semi-urban areas where

increased populations live in close contact with increased mosquito densities. DF

transmission is also higher in areas where two or more viral serotypes circulate

simultaneously (Rigau-Perez et al., 1998). International trade and increased air travel

potentially enhanced the dispersion of Aedes mosquitoes and the virus in recent

years.

2 Chapter 1: Introduction

DF typically follows a seasonal pattern which suggests a correlation between

weather and the pattern of DF (Reiter, 2001, Johansson et al., 2009a). Precipitation,

temperature and relative humidity are the most important weather factors attributed

to the growth and dispersion of mosquito vector and potential of DF outbreaks

(Chaturvedi et al., 2006, Chowell and Sanchez, 2006, Ram et al., 1998, Wu et al.,

2007). Temperature influences the life cycle of Aedes mosquitoes including

development rates and larval survival, timing of blood meal and the length of

reproductive cycle (Patz et al., 2005). Temperature can also affect the virus

replication, maturation and period of infectivity within the vector. Higher

temperature decreases the length of viral incubation, and thus increase the chance of

mosquitoes to become infected and infecting human with in their life span (Patz et

al., 1998). Precipitation contributes to the DF transmission via accumulation of rain

water which provides breeding grounds and stimulates egg hatching (Johansson et

al., 2009a). Higher levels of humidity are associated with an increase in biting

activity of mosquito (Rigau-Perez et al., 1998). Several studies have examined the

relationship between weather variability and DF epidemics and suggested that global

climate change will worsen the DF transmission (Hales et al., 2002, Hales et al.,

1999, Halide and Ridd, 2008, Thammapalo et al., 2005a). However, the magnitude of

the association between weather variability and DF fever varied with geographical

location and socio-environmental conditions (Thammapalo et al., 2007, Arcari et al.,

2007).


Mathematical modelling has been recently used to measure and predict the

impact of climate variation due to global climate change on DF transmission and

significant advances have been achieved in modelling approaches (McMichael et al.,

2006, Hu et al., 2010). Many countries around the world have developed different

models to predict the future distribution of DF in response to climate change (Hales

et al., 2002, Halide and Ridd, 2008, Johansson et al., 2009a, Johansson et al., 2009b).

Such projections can help to combat the increased risk of DF due to climate change

by taking necessary adaptation measures. However, only two studies have been

undertaken to identify the association between weather factors and DF transmission

in Bangladesh (Karim et al., 2012, Hashizume et al., 2012) and long term scenario-

based projections are yet to be developed. Therefore, it is important to further

examine the relationship between climate factors and DF in this location and develop

a better predictive model based on regional climate change scenarios.

Geographic information systems (GIS) have been widely used in vector-borne

disease epidemiology. GIS can visualize the spatial variation in disease risk and can

be used for disease mapping. Monitoring the space-time trends in disease occurrence

can highlight the changing patterns in risk and help to identify new risk factors

(Robertson and Nelson, 2010). Scan statistics are one of the most commonly used

approaches in spatial disease surveillance to explore high risk areas (Abrams and

Kleinman, 2007). Spatial analyses were used to identify the spatiotemporal pattern

and risk factors of DF transmission like socio-demographic and environmental

factors at both regional and country levels (Thai et al., 2010, Tran et al., 2004, Wen

et al., 2006, Wu et al., 2009, Mammen et al., 2008, Vanwambeke et al., 2006).

However, the trend of DF incidence and high risk areas for DF transmission at a

continental level remain to be determined.

4 Chapter 1: Introduction

1.2 RESEARCH AIMS

Given the limited knowledge on spatial distribution of DF and the imperative

to increase our knowledge on the impact of climate change on DF transmission, the

aims of this thesis are:

• To update and extend the current knowledge on the impact of climate

change and socio-environmental factors on DF transmission in the

Asia-Pacific region.

• To explore the spatiotemporal pattern and the high risk areas for DF

transmission in the Asia-Pacific region.

• To investigate the spatiotemporal distribution and to identify the high

risk areas for DF transmission in Bangladesh.

• To examine the association between climatic factors and DF

transmission in Bangladesh.

• To project the future risk of DF in Bangladesh based on different

climate change scenarios.

1.3 THESIS STRUCTURE

This thesis is presented in the conventional publication style, which consists of

four papers: Systematic Review, Results Paper 1, Results Paper 2 and Results Paper

3. Each paper was written for a particular journal. As each paper was designed to

stand alone, there was an inevitable degree of repetitiveness in their introduction,

methods and discussion sections. Chapter 2 presents the current knowledge on the

topic and critically reviews the relevant literature.

The four papers are presented in Chapter 2 through Chapter 5. Chapter 2

includes a systematic review of the related literature on climate change and DF in the

Asia-Pacific region which has been published in Tropical Medicine and

International Health (volume 16, issue 5, pages 598-607).


Chapter 3 (i.e., Results Paper 1) explored the spatial temporal pattern of DF

transmission in the Asia-Pacific region during 1955-2004. This paper has been

provisionally accepted by PLoS ONE.

Chapter 4 (i.e., Results Paper 2) examined the space-time distribution of DF in

Bangladesh during 2000- 2010 and identified the high risk cluster areas using scan

statistics. This paper has been published in Tropical Medicine and International

Health, (volume 17, issue 9, pages 1086-1091).

Chapter 5 (i.e., Results Paper 3) assessed the potential impact of climate

variability on the transmission of DF in Dhaka, Bangladesh and projected the future

DF risk based on different climate change scenarios. This paper has been accepted

for publication in Environment International.

Chapter 6 summarised the study findings across the three results papers and

discussed conclusions in relation to the overall aims of this study. This chapter

further discussed the study strengths, limitations, directions for future research and

public health implication of the research findings.

The references for each paper are presented at the end of their corresponding

chapter. A complete reference list (including all references cited in the manuscripts)

is provided at the end of the thesis.

Chapter 2: Literature Review 7

Chapter 2: Literature Review

This chapter presents a summary of the current knowledge on climate and dengue

research worldwide. It first describes the characteristics and ecology of DF and then

discusses the geographic distribution and the determinants of DF transmission followed

by a systematic review on the DF and global climate change. The final part of this

chapter is an update of the systematic review and list of knowledge gaps identified

through this literature review.

2.1 CHARACTERISTICS AND ECOLOGY OF DF

DF is the rapidly spreading vector borne disease endemic in tropical and

subtropical parts of the world (WHO, 2012a). More than 50% of the world’s population

live in over 100 countries are currently at risk of DF. Up to 100 million infections occur

annually and 22,000 children die due to DF globally (Guzman et al., 2010, Beatty et al.,

2011, Gibbons, 2010, Brady et al., 2012). The estimated number of disability-adjusted

life years (DALYs) is 1300 per million populations for the countries in Asia and

America (Guzman et al., 2010). The overall cost of a DF case was US$ 828 estimated

by a study in eight countries (Suaya et al., 2009). Thus, the aggregated annual economic

cost of DF was US$ 440 million during 2001-2005 in those countries (Suaya et al.,

2009). In America, the estimated cost is around 2.1 billion US dollars per year (Shepard

et al., 2011).

8 Chapter 2: Literature Review

DF is caused by one of the four serotypes of dengue virus which belong to the

genus flavivirus. Virus transmission occurs from viraemic to susceptible human mainly

by bites of Aedes aegypti and Aedes albopictus mosquitoes (Simmons et al., 2012,

Gibbons, 2010). Aedes aegypti is well established in most tropical and subtropical parts

of the world and is the principal vector for DF transmission. The Aedes albopictus is the

secondary vector of DF and continue to spread to new geographic areas of the tropical

and temperate climate (Wilder-Smith and Gubler, 2008). Typical disease manifestations

of DF range from an influenza like illness known as dengue fever (DF) to a severe

disease characterised by haemorrhage and shock, known as dengue hemorrhagic fever

(DHF) or dengue shock syndrome (DSS) (Gibbons, 2010, Guzman et al., 2010, Gubler,

2004b).

Figure 2.1 Transmission cycle of dengue fever


DF transmission is most common in urban areas as Aedes aegypti is a highly

domesticated mosquito and thrives in crowded cities (Wu et al., 2009, Gubler, 2011,

Gubler, 2004b). Transmission occurs when a female mosquito bites an infected

individual during the viraemic phase of the disease; the mosquito then becomes

infective after an incubation period of approximately 10 days (Guzman et al., 2010,

Tran and Raffy, 2006). At this stage the virus is capable of being transferred to a human

host when the mosquito probes the skin. After that, the mosquito remains infective for

the rest of its life. Aedes aegypti mosquitoes are most active during the day, with

highest levels of activity occurring during early morning and evening (Chowell et al.,

2007, Lyerla et al., 2000). The ideal temperature range for the transmission of DF is 18

to 33.2°C, with females feeding more frequently when temperatures are higher (Epstein,

2001). The risk of transmission is determined by a multitude of different climatic,

household and individual risk factors (Mackenzie et al., 2004). Transmission is also

higher in areas where two or more serotypes circulate simultaneously (Wilder-Smith

and Gubler, 2008, Rigau-Perez et al., 1998). Recovery from infection by one dengue

serotype provides lifelong immunity against that serotype, but not to the other serotypes

(Simmons et al., 2012, Gibbons, 2010, Guzman et al., 2010).

2.2 DETERMINANTS OF DF TRANSMISSION

There are many factors such as virus, vector, host, and environment that are

involved in the transmission cycles of DF and making its ecology complex. Weather

variation, virus strain, mosquito densities, survival and breeding, human activities and

movement, socioeconomic status, and population immunity, all contribute to the

transmission of DF (Gubler, 2011). Uncontrolled urbanization and concurrent

population growth have resulted in substandard housing and inadequate water, sewer

and waste disposal systems, all of which increase mosquito breeding. Thus, the

increased human populations living in intimate contact with increasingly high densities

of mosquito populations create ideal conditions for increased DF transmission (Ansari

and Shope, 1994, Sukri et al., 2003, Wilder-Smith and Gubler, 2008). Tourism and

travel have also become important mechanisms for facilitating the dispersion of DF and


its vectors (Rigau-Perez et al., 1998, Wilder-Smith and Gubler, 2008, Shang et al.,

2010, Tatem et al., 2006). The reinfestation of a region with Aedes aegypti or

introductions of new serotype are clear precursors of increased transmission. New viral

strain resulting from genetic changes and presence of multiple serotypes also increase

the risk of DF outbreak (Gubler, 1998, Leitmeyer et al., 1999, Rigau-Perez et al., 1998,

Simmons et al., 2012).

Figure 2.2 Determinants of dengue transmission (Sutherst, 2004)


Age of the population and their race can also be important determinants of DF

infection. It appears that young children and older adults are at higher risk of the

development of more severe DF infection (Guzman et al., 2002, Ashford, 2003, Lee et

al., 2006). It has also been reported that white persons are at higher risk of developing

DF compared to black among those with a history of asthma, diabetes, renal

insufficiency and hypertension (Sierra et al., 2007). A recent study in Thailand showed

that changes in population age structure might explain the changes in DF incidence in

absence of other demographic changes (Cummings et al., 2009, Simmons and Farrar,

2009).

Climate is an important determinant of temporal and spatial distribution of DF

vector and its pathogen (Kovats et al., 2001, Lafferty, 2009). Climate variability and

extreme events can influence the growth of the virus and mosquito as well as people’s

behaviour and therefore affect the pattern of DF transmission (Koopman et al., 1991, Bi

et al., 2001, Thai and Anders, 2011).

2.3 EMERGENCE AND GEOGRAPHIC SPREAD OF DF

DF is an old disease and documented as early as 1635 and 1699 in West Indies

and Central America, respectively (Wilder-Smith and Gubler, 2008, Mackenzie et al.,

2004). Even though the geographic origin of DF is still uncertain, the presence of all

viral serotypes in human and monkeys from Asia and the phylogenetic position of the

Asian sylvatic strain indicate the possible origin of DF in Asia rather than in Africa.

In Southeast Asia, DF has rapidly emerged as a public health concern after the

World War II (Gubler, 2011, Kyle and Harris, 2008). The major ecological and

demographic changes that occurred due to the war might facilitate the geographic

expansion of dengue viruses and the vectors. The economic development, unplanned

urbanization, modern transportation and lack of mosquito control activities after the war

were also responsible for the dramatic increase of DF epidemic in this area during


1950s and 1960s (Wilder-Smith and Gubler, 2008, Kyle and Harris, 2008, Tatem et al.,

2006, Mackenzie et al., 2004).

Figure 2.3 Countries at risk of DF transmission, 2008 (Source: World Health Organization, http://www.nathnac.org/pro/factsheets/images/clip_image002_015.jpg)

Before 1981, DF outbreaks in the American tropics were self-limiting due to the

presence of single serotype only (Guzman and Istúriz, 2010). However, the introduction

of new serotypes from Southeast Asia to the Caribbean, the South Pacific Islands and

the America during 1980s and 1990s made this region hyper endemic (presence of

multiple serotypes) and severe DF epidemic became frequent (Wilder-Smith and

Gubler, 2008, Gubler, 2011, Guzman and Istúriz, 2010). Imported DF is the most

common cause of febrile illness in returned travellers in Europe and Aedes albopictus

has already established in European countries including Albania, Italy, France, Spain,

Switzerland, Slovenia, Croatia, Bosnia and Herzegovina, Greece and Montenegro with

http://www.nathnac.org/pro/factsheets/images/clip_image002_015.jpg


increasing risk of DF transmission (Guzman and Istúriz, 2010, Semenza and Menne,

2009, Gasperi et al., 2012). The high adaptive capability of this species due to cold

hardiness, egg diapauses and the ability to mature eggs without a blood meal permits

them to spread and successfully establish in both tropical and temperate regions

(Gasperi et al., 2012, Delatte et al., 2009). In 2010 local DF transmissions were

documented in France and Croatia (WHO, 2012a). It has been predicted that the current

geographic range of Aedes albopictus might expand to new areas due to climate change

(Fischer et al., 2011, Caminade et al., 2012).

In the past 35 years, there has been a remarkable global re-emergence of DF

epidemic and the virus and its vectors expanded into new areas (Gubler, 2004a,

Mackenzie et al., 2004). More than 120 tropical countries are now endemic with DF

and the reported number of DF cases increased by fivefold over the last three decades

(Brady et al., 2012, Kyle and Harris, 2008). Between 2000 and 2007 at least eight

previously DF free countries became infected. Outbreaks of suspected DF were

recorded in Pakistan, Saudi Arabia, Yemen, Sudan and Madagascar during 2005 and

2006 (Guzman and Istúriz, 2010). However, the reason for this global pandemic is not

fully understood and there has been lack of empirical evidence about the spatial

temporal pattern of DF globally (Cummings et al., 2004, Wilder-Smith and Gubler,

2008).

2.4 CLIMATE VARIABILITY AND DF

DF is present either in epidemic cycles or seasonally which suggests a correlation

between climate and pattern of DF (Reiter, 2001). Rainfall, temperature and relative

humidity are thought as important factors attributing towards the growth and dispersion

of mosquito vector and potential of DF outbreaks (Chaturvedi et al., 2006, Chowell and

Sanchez, 2006, Ram et al., 1998, Wu et al., 2007). Temperatures influence the life cycle

of Aedes aegypti, including growth rate and survival of larval stage, time to first blood

meal, and the length of gonotropic cycle (Patz et al., 2005). Temperature is also capable


of affecting pathogen replication, maturation and period of infectivity. Higher

temperatures shorten the incubation period of the dengue virus, thus increasing the

proportion of mosquitoes that are infectious at any given time (Patz et al., 1998).

The humidity of a region is also a potential risk factor for DF transmission. It has

been suggested that the relative humidity is related to the hatching and activities of

mosquito vectors, with higher levels of humidity associated with a decreased incubation

time and an increase in activity levels (Rigau-Perez et al., 1998). A study by Hales et al.

(2002) suggested that rather than relative humidity, mean annual vapour pressure was

highly correlated with the incidence of DF. Precipitation has also been identified as an

important climatic risk factor contributing to the transmission of DF (Hales et al.,

2002). The accumulation of water that occurs by precipitation provides Aedes Aegypti

vectors with an increased number of available breeding grounds (Keating, 2001).

Observations have suggested that the greatest rate of increase in mosquito density is

associated with the onset of a rainy season, rather than the peak (Nagao et al., 2003).

El Niño/Southern Oscillation (ENSO) is a periodic variation in the atmospheric

condition and ocean surface temperatures of the tropical Pacific, which affects most

countries of the Pacific and Indian Oceans, bringing long drought and wet periods every

2-7 years (Hales et al., 1999). ENSO has an important influence on the timing and inter-

annual variability in DF transmission as it causes variation in local temperature and

precipitation. (Johansson et al., 2009a, Tipayamongkholgul et al., 2009, Cazelles et al.,

2005, Hales et al., 1999). A significant but non-stationary association between ENSO

and DF epidemic was observed in Thailand and Puerto Rico (Cazelles et al., 2005,

Johansson et al., 2009a). The non-stationary relationship implies that sometimes ENSO

plays role in the timing and synchrony of DF epidemic and sometimes it does not.


2.5 GLOBAL CLIMATE CHANGE AND DF

Climate change is a change in the statistical properties of the climate system when

considered over long periods of time, regardless of cause (IPCC, 2012). There is strong

evidence indicating that climate change is occurring (IPCC, 2007). Increases in global

population and corresponding increases in demand for energy resources have caused a

radical increase in energy consumption resulting in an increase in greenhouse gas

emissions. As a result global temperature has increased by 0.5 °C over the last thirty

years (IPCC, 2007). The rate of climate change is now faster than it has been in the

previous 1000 years; modelling has predicted that global temperatures will rise a further

1.5 – 4.5°C by 2100 (IPCC, 2007). As global temperatures continue to increase, it is

predicted that the endemic range of DF will expand geographically (Githeko et al.,

2000, Hopp and Foley, 2001, McMichael et al., 2006, Sutherst, 2004, Woodruff and

McMichael, 2004). Warmer temperatures will also allow for increased reproduction and

activity and decreased incubation time of larvae, resulting in an increased capacity for

producing offspring. Thus an increase in the transmission potential and prevalence of

DF seems likely (Jetten and Focks, 1997, McMichael et al., 2006).

The increasing temperatures associated with climate change could also increase

DF transmission by extending the season in which transmission occurs (Patz and

Reisen, 2001). Lengthy drought conditions in endemic areas without a stable drinking

water supply may encourage the storage of drinking water, thereby increasing the

number of developmental sites for the primary vector Aedes aegypti (Patz et al., 1998).

Conversely, high rainfall would ensure that small artificial containers used as larval

mosquito habitat, would remain flooded, thereby expanding adult mosquito population

(Patz et al., 1998). A continuous inter-annual variability in ENSO has been projected by

IPCC (IPCC, 2007). Local climate changes associated with ENSO may trigger an

increase in DF transmission in populated areas where the disease is endemic (Hales et

al., 1999). As climate change may have a significant impact on the transmission and

incidence of DF a better understanding of the association between climate and DF

transmission is imperative (Chan et al., 1999, Racloz et al., 2012).


Systematic review

Dengue transmission in the Asia-Pacific region: impact of climate change and socio-environmental factors

Published in of Tropical Medicine and International Health

Institute for Scientific Information (ISI) Impact Factor: 2.8

Citation:

BANU, S., HU, W., HURST, C. & TONG, S. 2011. Dengue transmission in the

Asia-Pacific region: impact of climate change and socio-environmental factors.

Tropical Medicine and International Health, 16, 598-607.

Authors Contribution:

Shahera Banu performed all literature searches and wrote the manuscript. Shilu

Tong and Wenbiao Hu supervised the study and assisted with writing the manuscript.

Cameron Hurst contributed to the manuscript in terms of providing feedback on initial

drafts.


2.6 SUMMARY

OBJECTIVE: To review the scientific evidence about the impact of climate

change and socio- environmental factors on dengue transmission, particularly in the


METHODS: A search of the published literature on PubMed, ISI web of

Knowledge and Google Scholar was conducted. Articles were included if an association

between climate or socio-environmental factors and dengue transmission was assessed

in any country of the Asia-Pacific region.

RESULTS: A total of twenty-two studies met the inclusion criteria. The weight of

the evidence indicates that global climate change is likely to affect the seasonal and

geographic distribution of dengue fever (DF) in the Asia-Pacific region. However,

current empirical evidence linking DF to climate change is inconsistent across

geographic locations and absent in some countries where DF is endemic.

CONCLUSION: Even though climate change may play an increasing role in the

transmission of DF, no clear evidence shows that such impact has already occurred.

More research is needed across countries to better understand the relationship between

climate change and DF transmission. Future research should also consider and adjust

for the influence of important socio-environmental factors in the assessment of the

climate change-related effects on DF transmission.


2.7 INTRODUCTION

Global climate is changing rapidly, due primarily to anthropogenic greenhouse

gas emissions (IPCC, 2007). Global mean temperature is projected to increase between

1.1–6.4oC by the end of this century relative to 1980-1999. There is growing evidence

that climate change has already affected, and will increasingly impact on human health

(Epstein, 2005, Haines et al., 2006, Costello et al., 2009, McMichael et al., 2006). The

changes in global temperature, precipitation and humidity that are expected to occur

due to projected climate change will affect the biology and ecology of disease vectors

and consequently the risk of vector-borne disease transmission (Gubler et al., 2001,

Tong et al., 2008, Shuman, 2010). Dengue fever (DF) has been recognised as the most

important arboviral disease in the world (Gubler, 2002a, WHO, 2006). About 2.5

billion people live in endemic areas and an additional 120 million people travel to

affected areas every year. The annual number of DF infections is estimated to be 50 to

100 million. Case fatality rates vary between 0.5%–3.5% in Asian countries (Guzman

and Kouri, 2002, Suaya et al., 2009, Halstead, 2007). Cases reported to the World

Health Organization (WHO) over the past four decades show an upward trend,

especially in urban areas (WHO, 2006).

Climate variability, lack of effective vector control programs, uncontrolled

urbanization and increased international travel are suggested to be the important risk

factors for increased activity of DF (Sutherst, 2004, Gubler, 2002a). Temperature,

rainfall and relative humidity are thought as important climatic factors contributing

towards the growth and dispersion of the mosquito vector and potential of DF outbreaks

(Patz et al., 2005). Temperature is also capable of affecting pathogen replication,

maturation and period of infectivity (Patz et al., 1998).


As global temperatures continue to increase, it has predicted that the endemic

range of DF will expand geographically (Hopp and Foley, 2001, Githeko et al., 2000,

Woodruff and McMichael, 2004). Warmer temperatures will also allow for increased

reproduction and activity and decreased incubation time of larvae, resulting in an

increased capacity for producing offspring. Thus an increase in the transmission

potential and prevalence of DF seems likely (Jetten and Focks, 1997, Barbazan et al.,

2010). The increasing temperatures could also increase DF transmission by extending

the season in which transmission occurs (Patz and Reisen, 2001). Lengthy drought

conditions in endemic areas without a stable drinking water supply may encourage the

storage of drinking water, thereby increasing the number of breeding sites for the

mosquito vector Aedes aegypti (Beebe et al., 2009). Conversely, high rainfall will

ensure that small artificial containers used as larval mosquito habitat, would remain

flooded, thereby expanding adult mosquito population (Patz et al., 1998). Moreover,

climate change associated with El Niño-Southern Oscillation (ENSO) may trigger

outbreak of DF in populated areas where the disease is endemic (Hales et al., 1999,

Johansson et al., 2009b).

As climate change may have a significant impact on the transmission and

incidence of DF, it is essential to examine the association between climatic variables

and DF epidemics. Recently a few studies have explored the impact of climate

variability and climate change on DF transmission (Hales et al., 2002, Johansson et al.,

2009a). In addition, several studies have assessed the influence of socio-ecological

factors on DF epidemics (Cummings et al., 2009, Johansson et al., 2009b, Bohra and

Andrianasolo, 2001, McBride et al., 1998, Thammapalo et al., 2005a). However, the

link between DF and climate change remains inconclusive because the potential

influence of socio-demographic factors on DF transmission has rarely been considered

seriously in previous research (Patz et al., 1998, Reiter, 2001, Semenza and Menne,

2009). In this paper, we reviewed the reported effects of climate change and other

socio-environmental factors on DF transmission in the Asia-Pacific region to draw


attention to the methodological challenges in this area. Additionally, we provided

recommendations for future research directions.

2.8 METHODS

A literature search using PubMed, ISI Web of Knowledge and Google scholar

was conducted in December 2009. Keywords such as ‘dengue , ‘dengue fever’ ,

‘dengue hemorrhagic fever’, ‘climate change*’ , ‘climate variability’ , ‘climate

model*’, ‘Risk factors’ and ‘socio-environmental factors’ were used in different

combination to identify potential articles and references. Search terms included MeSH

terms and free text terms. Limits were set for language (English) and publication year

(1990 to 2009). In Google Scholar, we limited our search by subject (Biology, Life

Science and Environmental science) as it gives huge number of irrelevant references.

However, we searched all fields for PubMed and Web of Science. Table 2.1 shows the

search results from each database. The initial search generated 1931 references

including duplicates. All titles and 466 abstracts were reviewed to identify potential

epidemiological studies. Then 290 full text articles were retrieved based on abstracts

reviewed and critically analysed by the first author.


Table 2.1 Search strategy

Search terms Pub Med ISI Web of Knowledge

Google Scholar

Total (including duplicates)

# 1 ('Dengue' OR 'Dengue fever' OR 'Dengue Hemorrhagic fever') AND ('climate change*' OR ' climate variability' OR 'climate model*')

310 134 167 611

# 2 ('Dengue' OR 'Dengue fever' OR 'Dengue Hemorrhagic fever') AND ('Risk factors' OR 'Socio-environmental factors')

325 216 584 1125

#1 AND #2 ('Dengue' OR 'Dengue fever' OR 'Dengue Hemorrhagic fever') AND ('climate change*' OR ' climate variability' OR 'climate model*') AND ('Risk factors' OR 'Socio-environmental factors')

47 19 129 195

Total (including duplicates) 1931


2.8.1 Inclusion criteria

Articles were included in the review if they (a) evaluated the effects of climate

variability or the influence of other socio-environmental factors on DF transmission (b)

conducted in countries of the Asia-Pacific region (c) used an observational study

design. Only studies conducted in the Asia-Pacific region were included because many

countries of this region are prone to DF transmission but no systematic review has

focused on this region in the literature yet. Observational study designs included time

series analysis, spatial or spatiotemporal analysis and descriptive study which applied to

identify the influence of climate variables (temperature, humidity, precipitation, wind

speed, El Niño events) and socioenvironmental factors (population density, urbanizatio

n, housing, family income, transport, vector control) on DF transmission.


Figure 2.4 Flow diagram of article selection process

22 articles found potentially relevant

18 modelled climatic factors

4 modelled environmental factors except climate variables

Inclusion criteria used were:

• From the Asia –Pacific region • Reports association between DF and climatic or

non-climatic factors • Observational study design • Published in English

1931 references were found after searching PubMed, ISI Web of knowledge and Google scholar databases

466 abstracts were read and 290 full text articles were retrieved

268 articles rejected due to irrelevant research question and non-observational study design

1465 references rejected (studies describing virological or serological features of DF and treatment or vaccine development)


2.9 RESULTS

Among 290 full text articles, 22 articles met the inclusion criteria and the major

findings of these studies were reviewed and summarized in Table 2.2 and Table 2.3. All

articles included in this review were published between 1998 and 2010. Eight studies

from Thailand, three each from Australia, Taiwan and Indonesia, two each from China

and India and one from Singapore. There was no study conducted in some countries of

the Asia-Pacific region where DF is endemic like Bangladesh, Sri Lanka, and

Myanmar.

2.9.1 Association between climatic factors and DF

Out of 22 published studies, eighteen evaluated the relationship between climatic

variables and DF incidence (Arcari et al., 2007, Bangs et al., 2006, Cazelles et al., 2005,

Chakravarti and Kumaria, 2005, Halide and Ridd, 2008, Hii et al., 2009, Hsieh and

Chen, 2009, Johansson et al., 2009b, Lu et al., 2009, Nagao et al., 2003, Nakhapakorn

and Tripathi, 2005, Thammapalo et al., 2005b, Tipayamongkholgul et al., 2009, Wu et

al., 2007, Yang et al., 2009b, Bi et al., 2001, Wu et al., 2009, Hu et al., 2010). Among

them, only three studies considered both climatic and social factors (Wu et al., 2009,

Nagao et al., 2003, Nakhapakorn and Tripathi, 2005). The main outcome variables

assessed in these studies were the number of DF cases and/ Aedes mosquito density.

Most studies collected data on temperature, amount of precipitation and relative

humidity as climatic variables (Cazelles et al., 2005, Nakhapakorn and Tripathi, 2005,

Thammapalo et al., 2005b, Wu et al., 2007). Several used El Niño -Southern Oscillation

(ENSO) as an independent variable (Johansson et al., 2009a, Tipayamongkholgul et al.,

2009, Cazelles et al., 2005). Only one study assessed the influence of extreme weather

event (Typhoon) on DF (Hsieh and Chen, 2009).


Table 2.2 Characteristics of studies on the association between climatic variables and DF transmission

Study Location (study period)

Statistical method

Risk factors Major findings Comments

Johansson et al. 2009

Thailand (1983-1996)

Wavelet analysis

ENSO Precipitation & temperature

The direct relationship between ENSO and DF was nonstationary and ENSO appeared to be associated with local temperature and precipitation.

Further research is needed to explain the nonstationarity in ENSO and DF incidence.

Tipayamongkholgul el al. 2009


Poisson regression

ENSO Temperature Relative humidity Wind speed

The effects of El Niño on DF transmission were varied according to geographical location within Thailand.

Only southern coastal and northern inland regions were included

Cazells et al. 2005 Thailand (1983-1997)

Wavelet analysis ENSO

The dynamics of El Niño were strongly associated with DF incidence but nonstationary existed in this relation which might influenced by the synchrony of previous DF epidemic.

The magnitude of the El Niño and DF relationship needs to be viewed within a wider context of socio-environmental variability.

Nakhapakorn and Tripathi 2005


Multiple regression

Rainfall Temperature Humidity Land use

Built-up area had highest risk of DF incidence and temperature, rainfall and humidity were likely to be key determinants of DF

Only one province was included

Thammapalo et al. 2005b


Linear least square regression

Monthly total rainfall Rain days Daily temperature Daily relative humidity

Increased temperature was positively associated with DF incidence in central and northern parts of Thailand where increased rainfall was negatively associated in southern Thailand.

Spatial variation was not examined


Table 2.2 (continued)


Statistical method Risk factors Major findings Comments

Nagao et al. 2003


Multiple regression

Temperature, Rainfall Water wells Tin houses

Larval abundance of Aedes mosquito was positively associated with house conditions, water supply and transport services. Increased rainfall in 2 months earlier and temperature were also correlated to larval indices.

Only 18 province of Northern Thailand were included

Hsieh and Chen 2009 Taiwan (2007)

Richards model Distributed lag model

Typhoon Temperature Precipitation

The multi-wave DF outbreak in Taiwan in 2007 was appeared to be influenced by rainfall and temperature variation due to two consecutive typhoons.

Further research is required to explore the relationship between extreme weather events and DF transmission

Wu et al. 2007 Taiwan (1998-2003)

ARIMA model Temperature Rainfall Relative humidity Vector density

The incidence of DF was negatively associated with monthly temperature variation and reversely with relative humidity at lags of 2 months.

Only one metropolitan city was included

Wu et al. 2009 Taiwan (1998-2002)

Spatial analysis Logistic regression

Temperature Rainfall Level of urbanization Percentage of elder population

Number of warm months and degree of urbanization were found to be associated with increasing risk of DF incidence at township level

Both climatic variables and socio-demographic factors were considered.




Halide and Ridd 2008

Indonesia (1998-2005)

Multiple regression

Temperature Relative humidity Rainfall

Relative humidity at 3-4 months lags and current number of DF cases appeared to be major determinants for prediction of DF outbreak in Indonesia.

Only one city was included

Arcari et al. 2007 Indonesia (1992-2001)

Multiple regression Temperature Rainfall Relative humidity SOI

Rainfall and temperature observed as important predictor of DF transmission in Indonesia, although the association differed across geographic region of the country.

Socio-environmental factors were not included

Bangs et al.2006 Indonesia (1997 &1998)

Descriptive analysis

Temperature Relative humidity Rainfall, ENSO House index

ENSO driven increased temperature exhibited greater impact on DF transmission by the vector population.

Only one city was included

Chakravarti and Kumaria 2005

India (2003)


Temperature Rainfall Relative humidity

Temperature, rainfall and relative humidity were major determinants for DF transmission and outbreak coincided in post monsoon period.

Only one outbreak in one city was considered.

Table 2.2 (continued)




Lu et al. 2009

China (2001-2006)

Poisson regression

Temperature, rainfall Relative humidity Wind velocity

Increase minimum temperature and decreased wind velocity were associated with increase DF incidence.

Only one province was included

Yang et al. 2009 China (July –October 2004)


Temperature, Precipitation Humidity Breteau index House index

DF incidence seemed to have no relationship with climatic factors. However, negatively associated with vector control.

Non endemic nature of DF in Cixi city may biased the relationship between climate variables and DF

Hii et al. 2009 Singapore (2000-2007)

Poisson regression Temperature Precipitation

Weekly mean temperature and total precipitation were related to increased DF incidence.

Socio-environmental factors were not included

Hu et al. 2010 Australia (1993-2005)

SARIMA model ENSO Decreased mean of SOI at lags of 3-12 months earlier was inversely associated with DF incidence in Queensland.

Only SOI was considered

Bi et al. 2001 Australia (1990-1994)

ARIMA model Temperature Precipitation Relative humidity

Monthly mean minimum temperature was the major contributor for DF outbreak in Townsville.

Only four years dataset used

Abbreviations: ENSO= El Niño-Southern Oscillation Index; ARIMA= Autoregressive Integrated Moving Average; SARIMA= Seasonal Autoregressive

Integrated Moving Average; SOI= Southern Oscillation Index.

Arranged by location of study


Different methods were used to evaluate the association between climatic

variables and DF incidence or mosquito density. Of the 18 studies reviewed, two

used spatial analyses (Wu et al., 2009, Nakhapakorn and Tripathi, 2005); thirteen

time series analyses (Arcari et al., 2007, Bi et al., 2001, Cazelles et al., 2005,

Halide and Ridd, 2008, Hii et al., 2009, Johansson et al., 2009a, Lu et al., 2009,

Thammapalo et al., 2005b, Tipayamongkholgul et al., 2009, Wu et al., 2007, Hu

et al., 2010, Hsieh and Chen, 2009, Nagao et al., 2003) and three descriptive

analyses (Bangs et al., 2006, Chakravarti and Kumaria, 2005, Yang et al., 2009b).

Various multiple regressions analyses like logistic regression and Poisson

regression were mostly used to analyse and predict DF transmission. These time

series regression analyses provide a way to establish the relationship between

changes of weather parameters, environmental factors and the occurrence of DF

cases which might be utilized for forecasting future DF outbreak based on model

scenarios (Wu et al., 2007). Both classical autoregressive integrated moving

average (ARIMA) model and seasonal ARIMA model were also applied to link

between weather variation and DF (Bi et al., 2001, Wu et al., 2007, Hu et al.,

2010). Spatial analysis of DF occurrence was used to identify the high risk area

for DF outbreaks (Wu et al., 2009). However, none of the previous studies

included both spatial and temporal data into their models. The identification of

high risk areas through spatiotemporal modelling may assist existing surveillance

and control efforts by permitting limited resources in areas where DF outbreaks

are most likely to occur.

Association between climatic factors and DF incidence or mosquito density

was reported in most studies (Arcari et al., 2007, Bangs et al., 2006, Bi et al.,

2001, Cazelles et al., 2005, Chakravarti and Kumaria, 2005, Halide and Ridd,

2008, Hii et al., 2009, Hsieh and Chen, 2009, Johansson et al., 2009a, Lu et al.,

2009, Nagao et al., 2003, Nakhapakorn and Tripathi, 2005, Thammapalo et al.,

2005b, Tipayamongkholgul et al., 2009, Wu et al., 2007, Hu et al., 2010).

However, the direction and intensity of this association varied by time and


location (Thammapalo et al., 2005b, Tipayamongkholgul et al., 2009, Arcari et

al., 2007). Temperature, rainfall and relative humidity were often found as major

determinants of DF transmission. There was no association between DF incidence

and climate variability in Cixi, China, which might due to the non-endemic nature

of DF in Cixi and the short study period (four months) (Yang et al., 2009b).

Several non-climatic factors such as urbanization, land use, vector-control

program, human movement, housing quality and population immunity were

identified as potential confounders for assessing the climate-DF relationship.

Among 18 studies reviewed, only one study was adjusted for socio-environmental

factors (Nagao et al., 2003). It was reported that the lack of relevant dataset makes

it difficult to control these variables (Wu et al., 2007, Arcari et al., 2007, Bi et al.,

2001).

Most studies examining the temperature influence on DF transmission or

mosquito density reported a positive association (Hii et al., 2009, Nagao et al.,

2003, Arcari et al., 2007, Thammapalo et al., 2005b, Lu et al., 2009, Bi et al.,

2001, Wu et al., 2009). Higher temperature favours virus replication, vector

proliferation and feeding frequency of mosquito (Patz et al., 1998, Jetten and

Focks, 1997). Therefore, rising temperature may enhance DF transmission.

However, the impact of increased temperature on DF outbreak was not

immediate. Various lag times were reported for this relationship, ranging from 4–

16 weeks (Arcari et al., 2007, Bi et al., 2001, Hsieh and Chen, 2009, Lu et al.,

2009, Wu et al., 2007, Hu et al., 2010). Bi et al. 2001 found that monthly mean

minimum temperatures impacted on the transmission of DF through a four month

lagged period which includes the period of replication and development of

mosquito, the extrinsic incubation period (time of replication of virus within

vector) and intrinsic incubation period of the virus (time of virus replication

within the host)(Bi et al., 2001). Therefore, observed lags were biologically

plausible.


Associations between rainfall and DF transmission are inconsistent across

geographical locations. DF outbreaks usually coincided with wet season and

positive associations between DF incidence and precipitation were reported in

many countries of the Asia-pacific region (Nagao et al., 2003, Nakhapakorn and

Tripathi, 2005, Arcari et al., 2007). Rainfall may increase the breeding sites of

Aedes mosquito and therefore, increase DF transmission. However, excess rainfall

was found negatively associated with DF incidence in Thailand, Indonesia and

Taiwan (Thammapalo et al., 2005b, Halide and Ridd, 2008, Wu et al., 2009). The

plausible explanation might be the mosquito larvae were washed away by heavy

rainfall (Arcari et al., 2007, Thammapalo et al., 2005b).

The El Niño Southern Oscillation (ENSO) is a periodic variation in the

atmospheric conditions and ocean surface temperatures of the tropical Pacific and

hypothesized to have influence on multiyear variation in DF incidence (Johansson

et al., 2009a). Among the studies reviewed, four examined the impact of ENSO

on the synchrony of DF epidemics in Thailand and Australia (Cazelles et al.,

2005, Johansson et al., 2009a, Tipayamongkholgul et al., 2009, Hu et al., 2010).

They reported a significant but non-stationary influence of ENSO on DF

transmission (Cazelles et al., 2005, Johansson et al., 2009a, Tipayamongkholgul

et al., 2009, Hu et al., 2010). Spatial heterogeneity and various lag effects were

also apparent in this relation (Tipayamongkholgul et al., 2009, Hu et al., 2010). In

Thailand, ENSO was associated with local temperature and precipitation changes.

It was suggested that decreased ENSO could result in increased temperature and

decreased rainfall leading to increased water storage, increased mosquito breeding

sites and therefore, increased DF transmission (Johansson et al., 2009a).

Moreover, further research is necessary to explain the non-stationarity and spatial

variation in the ENSO- DF relationship to effectively support the hypothesis of

inter -annual variation in DF transmission.


2.9.2 Association between socio-environmental factors and DF

Four studies have examined the impact of socio-environmental factors on

DF incidence in the Asia-Pacific region (Table 2.3) (McBride et al., 1998,

Thammapalo et al., 2005a, Bohra and Andrianasolo, 2001, Cummings et al.,

2009). The increasing trends in population growth, uncontrolled urbanization,

spread of mosquito vector and movement of virus through international travel

were suggested to be the major contributing factors for the recent DF expansion

in endemic areas (Kyle and Harris, 2008, Campbell-Lendrum and Reithinger,

2002, Gubler et al., 2001). Influences of different socio-environmental factors on

DF transmission were illustrated below from a global perspective as very little

evidence is available from the Asia-Pacific region.

Socio-economic condition

DF transmission might be influenced by people’s socio-economic status

(Gubler et al., 2001, Reiter et al., 2003). People in developed countries usually

live in better houses with glazed windows, piped water, insect screening and air-

conditioning than those in developing countries (Reiter, 2001). These facilities

may effectively reduce their contacts with vector mosquitoes and even if infected

mosquitoes gain entry to these buildings, the low ambient temperature and

artificially dry atmosphere may decrease their survival rate and reduce the risk of

transmission (Reiter, 2001). Besides these, many activities, particularly social

gatherings which occur in outdoor situations such as balconies, courtyards and

outdoor restaurants might facilitate contact with vector (Reiter, 2001, Gubler et

al., 2001). It has been suggested that the large difference in disease incidence

between U.S. and Mexico Border States is probably caused by differences in

living standards and human behaviours (Gubler et al., 2001).


A spatial analysis of DF in Thailand reported that the risk of DF was quite

geographically homogenous and associated with housing types and poor garbage

disposal (Thammapalo et al., 2007). Housing style was strongly related to the

number of water jars and discarded items. Town houses and slum houses in

Phuket had relatively low quantity of discarded wet containers than single houses

on rubber plantations. Therefore, single houses had 3-15 times higher risk of DF,

because these dwelling types have sufficient open space where discarded items

could be easily filled with rain water which can create mosquito breeding sites

(Thammapalo et al., 2005a).

International travel

International trade and transports were suggested to have potential influence

on the geographical distribution of vectors and pathogen (Gubler et al., 2001,

Reiter, 2001, Sutherst, 2004). Commercial shipping might be linked to the spread

of both A. aegypti and A. albopictus between regions (Romi et al., 1997). For

example, after the introduction of used tires, an American species of Aedes

albopictus has established in Italy (Romi et al., 1997). In addition, air-travel has

greatly increased the dissemination of dengue viruses via rapid transit of viraemic

individuals around the world (Kyle and Harris, 2008). As a result, the worldwide

movement of dengue virus has been greatly facilitated by air travel (Reiter, 2001).

Introduction of new dengue viruses by travellers was one of the important factors

for causing hyper-endemicity in Puerto Rico and increased severity of the disease

(Gubler and Trent, 1993, Gubler et al., 2001).


Table 2.3 Characteristics of studies on the association between Socio-environmental factors and DF transmission

Arranged by location of study



Cummings et al. 2009


Linear regression Wavelet analysis

Population data (age, birth rate, death rate, household size, education level, sanitation) Climatic data(rainfall)

Changes in the birth rate and death rate in Thailand can explain the changes in the age distribution of DF cases and its periodicity.

Both socio-demographic and climatic factors were considered.

Thammapalo et al. 2005a

Thailand (1999)

Logistic regression Housing style Window screen Ethnicity Education level Occupation Age, gender

Housing style, window screen and ethnicity seemed to affect the presence of Aedes mosquito larva in the house.

A complex relationship was revealed between Aedes larval abundance and socio- environmental factors

McBride et al. 1998

Australia (1993)

Logistic regression House hold cases, neighbour cases Window screen Travel history Mosquito control Presence of water tank

Viral serotypes, vector populations, unscreened houses and hard immunity emerged as important risk factors for DF increased DF incidence in Charter Tower in 1993.

Climatic factors were not included

Bohra and Andrianasolo 2001

India (1990)

Spatial analysis Multiple regression

Uncovered water container Presence of solid waste Mosquito control Housing pattern Water supply

Housing pattern, mosquito control facility, presence of solid waste and uncovered water container were most important socio-cultural risk factors for DF transmission.

Climatic factors were not considered.


Public Health Intervention

The health risk arising from climate change may differ between countries

depending on the quality of public health infrastructures (Githeko et al., 2000). In

Canada and USA, good surveillance and vector control programs limited endemic

transmission of DF, whereas in Mexico and other less developed countries the health

infrastructures are still ineffective (Githeko et al., 2000). The majority of DF

endemic countries in Asia do not have laboratory based active surveillance system

that can provide accurate early warning for epidemic DF (Gubler, 2002b). In the

pacific, only Australia, Tahiti and New Caledonia have good laboratory surveillance

(Gubler, 2002b). Recently, there have been considerable improvements in

surveillance and vector control in Australia (Ritchie et al., 2002). Therefore, it seems

unlikely that DF will become re-established as an endemic disease in Australia

(McMichael, 2003).

2.10 CONCLUSION

The review indicates that global climate change is likely to affect the seasonal

and geographic distribution of DF in the Asia-Pacific region. The complex nature of

the DF transmission which involves vector ecology, viral factors, population

immunity and socio-demographic changes makes it difficult to quantify how climate

change affects DF transmission. However, the impact of socio-environmental

changes on DF transmission is less well studied. Therefore, relative importance and

interaction between socio-environmental and climatic factors remain to be elucidated

and need further exploration. Biological judgement and caution are also necessary in

interpreting a direct relationship between climatic factors or non-climatic factors and

DF transmission. We hope this review inspires further work to elucidate the relative

importance and interaction between climatic or non-climatic factors in the

transmission of DF.


There is the lack of data in many countries where DF is endemic. In order to

demonstrate the regional variation in the climate-DF relationship, a multidisciplinary

approach incorporating region-specific predicted climate changes and non-climatic

factors could help to identify consistencies in interactions between DF and certain

climatic and non-climatic factors. Thus, a more sophisticated multivariable predictive

model may be constructed for control and prevention of DF. Furthermore, multi-

centre or multi-country studies using both climatic and socio-environmental data

could help to reveal the potential impact of climate change on DF transmission in the


Funding

SB is funded by a Queensland University of Technology Postgraduate

Scholarship. ST is supported by an NHMRC Research Fellowship (# 553043). WH is

funded by NHMRC Postdoctoral Training Fellowship (#519788).

Conflict of interest

We declare that we have no conflict of interest.


References

ARCARI, P., TAPPER, N. & PFUELLER, S. 2007. Regional variability in relationships between climate and dengue/DHF in Indonesia. Singapore Journal of Tropical Geography, 28, 251-272.

BANGS, M. J., LARASATI, R. P., CORWIN, A. L., WURYADI, S. & JAKARTA, I. 2006. Climatic factors associated with epidemic dengue in Palembang, Indonesia: implications of short-term meteorological events on virus transmission. Southeast Asian Journal of Tropical Medicine and Public Health, 37, 1103-1116.

BARBAZAN, P., GUISERIX, M., BOONYUAN, W., TUNTAPRASART, W., PONTIER, D. & GONZALEZ, J. P. 2010. Modelling the effect of temperature on transmission of dengue. Medical and Veterinary Entomology, 24, 66-73.

BEEBE, N. W., COOPER, R. D., MOTTRAM, P. & SWEENEY, A. W. 2009. Australia’s dengue risk driven by human adaptation to climate change. PLoS Neglected Tropical Diseases, 3, e429.

BI, P., TONG, S., DONALD, K., PARTON, K. A. & HOBBS, J. 2001. Climate variability and the dengue outbreak in Townsville,Queensland,1992-93. Enviornmental Health, 1, 54-60.

BOHRA, A. & ANDRIANASOLO, H. 2001. Application of GIS in modeling of dengue risk based on sociocultural data: case of Jalore, Rajasthan, India. Dengue Bulletin, 25, 92-102.

CAMPBELL-LENDRUM, D. & REITHINGER, R. 2002. Dengue and climate change. Trends in Parasitology, 18, 524.

CAZELLES, B., CHAVEZ, M., MCMICHAEL, A. J. & HALES, S. 2005. Nonstationary influence of EI Nino on the synchronous dengue epidemics in Thailand. PLoS Medicine, 2, 313-318.

CHAKRAVARTI, A. & KUMARIA, R. 2005. Eco-epidemiological analysis of dengue infection during an outbreak of dengue fever, India. Virology Journal, 2, 32.

COSTELLO, A., ABBAS, M., ALLEN, A., BALL, S., BELL, S., BELLAMY, R., FRIEL, S., GROCE, N., JOHNSON, A., KETT, M., LEE, M., LEVY, C., MASLIN, M., MCCOY, D., MCGUIRE, B., MONTGOMERY, H., NAPIER, D., PAGEL, C., PATEL, J., DE OLIVEIRA, J. A., REDCLIFT, N., REES, H., ROGGER, D., SCOTT, J., STEPHENSON, J., TWIGG, J., WOLFF, J. & PATTERSON, C. 2009. Managing the health effects of climate change: Lancet and University College London Institute for Global Health Commission. Lancet, 373, 1693-1733.


CUMMINGS, D. A. T., IAMSIRITHAWORN, S., LESSLER, J. T., MCDERMOTT, A., PRASANTHONG, R., NISALAK, A., JARMAN, R. G., BURKE, D. S. & GIBBONS, R. V. 2009. The impact of the demographic transition on dengue in Thailand: insights from a statistical analysis and mathematical modeling. PLoS Medicine, 6, e1000139.

EPSTEIN, P. R. 2005. Climate change and human health. New England Journal of Medicine, 353, 1433-1436.

GITHEKO, A. K., LINDSAY, S. W., CONFALONIERI, U. E. & PATZ, J. A. 2000. Climate change and vector-borne diseases: a regional analysis. Bulletin of the World Health Organization, 78, 1136-1147.

GUBLER, D. J. 2002a. Epidemic dengue/dengue hemorrhagic fever as a public health, social and economic problem in the 21st century. Trends in Microbiology, 10, 100-103.

GUBLER, D. J. 2002b. How effectively is epidemiological surveillance used for dengue programme planning and epidemic response? Dengue Bulletin, 26, 96-106.

GUBLER, D. J., REITER, P., EBI, K. L., YAP, W., NASCI, R. & PATZ, J. A. 2001. Climate variability and change in the United States: potential impacts on vector- and rodent-borne diseases. Environmental Health Perspectives, 109 Suppl 2, 223-233.

GUBLER, D. J. & TRENT, D. W. 1993. Emergence of epidemic dengue/dengue hemorrhagic fever as a public health problem in the Americas. Infectious Agents and Disease, 2, 383-393.

GUZMAN, M. G. & KOURI, G. 2002. Dengue: an update. Lancet Infectious Diseases, 2, 33-42.

HAINES, A., KOVATS, R. S., CAMPBELL-LENDRUM, D. & CORVALAN, C. 2006. Climate change and human health: impacts, vulnerability, and mitigation. Lancet, 367, 2101-2109.

HALES, S., DE WET, N., MAINDONALD, J. & WOODWARD, A. 2002. Potential effect of population and climate changes on global distribution of dengue fever: an empirical model. Lancet, 360, 830-834.

HALES, S., WEINSTEIN, P., SOUARES, Y. & WOODWARD, A. 1999. El Nino and the dynamics of vector-borne disease transmission. Environmental Health Perspectives, 107, 99-102.

HALIDE, H. & RIDD, P. 2008. A predictive model for dengue hemorrhagic fever epidemics. International Journal of Environmental Health Research, 18, 253-265.

HALSTEAD, S. B. 2007. Dengue. Lancet, 370, 1644-1652.


HII, Y. L., ROCKLOV, J., NG, N., TANG, C. S., PANG, F. Y. & SAUERBORN, R. 2009. Climate variability and increase in intensity and magnitude of dengue incidence in Singapore. Global Health Action, 2.

HOPP, M. J. & FOLEY, J. A. 2001. Global-scale relationships between climate and the dengue fever vector, Aedes aegypti. Climatic Change, 48, 441-463.

HSIEH, Y. H. & CHEN, C. W. S. 2009. Turning points, reproduction number and impact of climatological events for multi-wave dengue outbreaks. Tropical Medicine and International Health, 14, 628-638.

HU, W., CLEMENTS, A., WILLIAMS, G. & TONG, S. 2010. Dengue fever and El Niño/Southern Oscillation in Queensland, Australia: a time series predictive model. Occupational and Environmental Medicine, 67, 307-311.

IPCC 2007. Climate change 2007: the physical science basis: contribution of working group I to the fourth assessment report of the Intergovernmental Panel on Climate Change. Cambridge, UK and NY, USA: Cambridge University Press.

JETTEN, T. H. & FOCKS, D. A. 1997. Potential changes in the distribution of dengue transmission under climate warming. American Journal of Tropical Medicine and Hygiene, 57, 285-297.

JOHANSSON, M. A., CUMMINGS, D. A. T. & GLASS, G. E. 2009a. Multiyear climate variability and dengue - El Nino Southern Oscillation, weather, and dengue Incidence in Puerto Rico, Mexico, and Thailand: a longitudinal data analysis. PLoS Medicine, 6, e1000168.

JOHANSSON, M. A., DOMINICI, F. & GLASS, G. E. 2009b. Local and global effects of climate on dengue transmission in Puerto Rico. PLoS Neglected Tropical Diseases, 3, e382.

KYLE, J. L. & HARRIS, E. 2008. Global spread and persistence of dengue. Annual Review of Microbiology, 62, 71-92.

LU, L., LIN, H., TIAN, L., YANG, W., SUN, J. & LIU, Q. 2009. Time series analysis of dengue fever and weather in Guangzhou, China. BMC Public Health, 9, 395.

MCBRIDE, W. J., MULLNER, H., MULLER, R., LABROOY, J. & WRONSKI, I. 1998. Determinants of dengue 2 infection among residents of Charters Towers, Queensland, Australia. American Journal of Epidemiology, 148, 1111-1116.

MCMICHAEL, A. J. 2003. Global climate change: will it affect vector-borne infectious diseases? Internal Medicine Journal, 33, 554-555.

MCMICHAEL, A. J., WOODRUFF, R. E. & HALES, S. 2006. Climate change and human health: present and future risks. Lancet, 367, 859-869.


NAGAO, Y., THAVARA, U., CHITNUMSUP, P., TAWATSIN, A., CHANSANG, C. & CAMPBELL-LENDRUM, D. 2003. Climatic and social risk factors for Aedes infestation in rural Thailand. Tropical Medicine and International Health, 8, 650-659.

NAKHAPAKORN, K. & TRIPATHI, N. K. 2005. An information value based analysis of physical and climatic factors affecting dengue fever and dengue haemorrhagic fever incidence. International journal of health geographics, 4, 13.

PATZ, J. A., CAMPBELL-LENDRUM, D., HOLLOWAY, T. & FOLEY, J. A. 2005. Impact of regional climate change on human health. Nature, 438, 310-317.

PATZ, J. A., MARTENS, W. J. M., FOCKS, D. A. & JETTEN, T. H. 1998. Dengue fever epidemic potential as projected by general circulation models of global climate change. Environmental Health Perspectives, 106, 147-153.

PATZ, J. A. & REISEN, W. K. 2001. Immunology, climate change and vector-borne diseases. Trends in Immunology, 22, 171-172.

REITER, P. 2001. Climate change and mosquito-borne disease. Enviornmental Health Perspectives, 109, 141-161.

REITER, P., LATHROP, S., BUNNING, M., BIGGERSTAFF, B., SINGER, D., TIWARI, T., BABER, L., AMADOR, M., THIRION, J. & HAYES, J. 2003. Texas lifestyle limits transmission of dengue virus. Emerging Infectious Diseases, 9, 86-89.

RITCHIE, S. A., HANNA, J. N., HILLS, S. L., PIISPANEN, J. P., MCBRIDE, W. J. H., PYKE, A. & SPARK, R. L. 2002. Dengue control in north Queensland, Australia: case recognition and selective indoor residual spraying. Dengue Bulletin, 26, 7-13.

ROMI, R., SABATINELLI, G., SAVELLI, L. G., RARIS, M., ZAGO, M. & MALATESTA, R. 1997. Identification of a North American mosquito species, Aedes atropalpus (Diptera: Culicidae), in Italy. Journal of the American Mosquito Control Association, 13, 245-246.

SEMENZA, J. C. & MENNE, B. 2009. Climate change and infectious diseases in Europe. Lancet Infectious Diseases, 9, 365-375.

SHUMAN, E. K. 2010. Global climate change and infectious diseases. New England Journal of Medicine, 362, 1061-1063.

SUAYA, J. A., SHEPARD, D. S., SIQUEIRA, J. B., MARTELLI, C. T., LUM, L., TAN, L. H., KONGSIN, S., JIAMTON, S., GARRIDO, F. & MONTOYA, R. 2009. Cost of dengue cases in eight countries in the Americas and Asia: a prospective study. American Journal of Tropical Medicine and Hygiene, 80, 846-855.


SUTHERST, R. W. 2004. Global change and human vulnerability to vector-borne diseases. Clinical Microbiology Reviews, 17, 136-173.

THAMMAPALO, S., CHONGSUVIVATONG, V., GEATER, A. & DUERAVEE, M. 2007. Environmental factors and incidence of dengue fever and dengue haemorrhagic fever in an urban area, Southern Thailand. Epidemiology and Infection, 136, 135 -143.

THAMMAPALO, S., CHONGSUWIWATWONG, V., GEATER, A., LIM, A. & CHOOMALEE, K. 2005a. Socio-demographic and environmental factors associated with Aedes breeding places in Phuket, Thailand. Southeast Asian Journal of Tropical Medicine and Public Health, 36, 426-433.

THAMMAPALO, S., CHONGSUWIWATWONG, V., MCNEIL, D. & GEATER, A. 2005b. The climatic factors influencing the occurence of dengue hemorrhagic fever in Thailand Southeast Asian Journal of Tropical Medicine and Public Health, 36, 191-196.

TIPAYAMONGKHOLGUL, M., FANG, C. T., KLINCHAN, S., LIU, C. M. & KING, C. C. 2009. Effects of the El Nino-Southern Oscillation on dengue epidemics in Thailand, 1996-2005. BMC Public Health, 9, 422.

TONG, S., DALE, P., NICHOLLS, N., MACKENZIE, J. S., WOLFF, R. & MCMICHAEL, A. J. 2008. Climate variability, social and environmental factors, and Ross River virus transmission: research development and future research needs. Environmental Health Perspectives, 116, 1591-1597.

WHO 2006. Scientific working group report on Dengue, Geneva, Switzerland.

WOODRUFF, R. E. & MCMICHAEL, T. 2004. Climate change and human health: all affected bit some more than others. Social Alternatives, 23, 17-22.

WU, P. C., GUO, H. R., LUNG, S. C., LIN, C. Y. & SU, H. J. 2007. Weather as an effective predictor for occurrence of dengue fever in Taiwan. Acta Tropica, 103, 50-57.

WU, P. C., LAY, J. G., GUO, H. R., LIN, C. Y., LUNG, S. C. & SU, H. J. 2009. Higher temperature and urbanization affect the spatial patterns of dengue fever transmission in subtropical Taiwan. Science of the Total Environment, 407, 2224-2233.

YANG, T., LU, L., FU, G., ZHONG, S., DING, G., XU, R., ZHU, G., SHI, N., FAN, F. & LIU, Q. 2009. Epidemiology and vector efficiency during a dengue fever outbreak in Cixi, Zhejiang province, China. Journal of Vector Ecology, 34, 148-154.


2.11 UPDATE OF THE SYSTEMATIC REVIEW

As the systematic review was completed in December 2009, an updated

literature search was conducted in February 2013. We applied similar methods and

inclusion criteria as the systematic review to search databases and select relevant

articles. However, we modified the search terms by including weather, temperature,

rainfall and humidity in the keywords. The literature search was limited by

publication year (January 2010 to February 2013) to avoid duplication. A total of

twenty four articles met the inclusion criteria and included in the updated literature

review.

2.11.1 Association between climatic factors and DF

Of the 24 articles that met the inclusion criteria, 20 studies evaluated the

relationship between climatic factors and DF incidence. Time series analyses were

mostly used to explore the association between climate and DF and only two studies

used spatial analysis techniques. Among time series analyses, Poisson regression

model was largely used to examine and forecast the impact of climate variables on

DF. Most studies assessed the impact of temperature fluctuations, precipitation and

relative humidity on DF. Many of them reported positive association between

temperature variation and DF incidence (Pinto et al., 2011, Descloux et al., 2012,

Bannister-Tyrrell et al., 2013, Chen and Hsieh, 2012, Chen et al., 2010, Earnest et

al., 2012). One study examined the effects of sunshine duration on DF incidence and

reported that the risk of DF inversely associated with duration of sunshine, the

number of DF cases decreased as the sunshine increased (Pham et al., 2011). The

optimum time for forecast DF outbreak depending on climatic condition was

estimated between 1 to 5 months (Chen et al., 2010, Hii et al., 2012a, Hii et al.,

2012b). Two studies included both climatic and social factors when examining the

impact of climate variation on DF (Oki and Yamamoto, 2012). According to Oki and

Yamamoto 2012, the monthly mean temperature has increased in Singapore by 0.50C

in the past 30 years due to climate change, but the impact of this temperature change

on DF was minor comparative to the impact of fluctuations in population immunity

and hyperendemicity (Oki and Yamamoto, 2012, Sriprom et al., 2010). The original


systematic review showed that similar kind of study was absent in some endemic

countries like Bangladesh, Sri Lanka and Myanmar. However, in the updated

literature review, we identified a total of seven studies which conducted in those

countries over the last three years, two studies from Bangladesh (Karim et al., 2012,

Hashizume et al., 2012), one from Malaysia (Rohani et al., 2011), one from New

Caledonia (Descloux et al., 2012), two from Vietnam (Pham et al., 2011, Cuong et

al., 2011) and one from Myanmar (Oo et al., 2011).

2.11.2 Association between socio-environmental factors and DF

Among 24 articles, four studies were found which examined the impact of

different social and environmental factors on DF transmission in the Asia-Pacific

region during 2010 - 2013. A study in Vietnam explored that the risk of developing

symptomatic DF for both primary and secondary infection increased with patient’s

age. Therefore, older age group like adolescent and adults are more likely to develop

symptomatic DF than children (Thai et al., 2011). However, the risk of severe

complication including death is higher in children than adults (Anders et al., 2011).

The results of this study support the notion that the people of the South east Asia

would be at higher risk of DF due to the ageing population (Cummings et al., 2009).

Ethnicity and co-morbidity were also identified as important risk factors for DF. A

study in Singapore showed that the likelihood of Chinese patient developing DHF is

1.9 times higher than Indian or Malay (Pang et al., 2012). In addition, they found that

individuals with diabetes mellitus and hypertension have greater risk of developing

DHF compared with individuals with no diabetes mellitus and no hypertension.

Schmidt et al. (2011) have explored the role of population density and socio-

economic factors (water supply infrastructure) as risk factors for DF (Schmidt et al.,

2011). They showed that DF transmission may occur in remarkably low population

density with high vector-host ratio in absence of tap water supply. Lack of reliable

water supply in the immediate vicinity of a household requires constant storing of

water which provides breeding sites for Aedes mosquitoes and increases the risk of

DF transmission.


2.12 KNOWLEGE GAPS

There are a number of knowledge gaps according to the current literature

including:

• The spatiotemporal trends of DF remain unclear, although DF is the

most common arboviral disease in the Asia-Pacific region.

• Literature indicates that global climate change is likely to affect the

seasonal and geographical distribution of DF in the Asia-Pacific region.

However, empirical evidence linking DF to climate change is

inconsistent across geographical locations.

• Region-specific predictive models under different climate change

scenarios are yet to be developed for the surveillance and control of DF.

• The impact of socio-environmental changes on DF transmission is less

studied. Hence, relative importance and interaction between socio-

environmental and climatic factors remain to be elucidated and need

further exploration.

Chapter 3: Results Paper 1 45

Chapter 3: Results Paper 1

Dynamic spatiotemporal trends of dengue transmission in the Asia-Pacific region during 1955 -

2004

Provisionally accepted by PLoS ONE.


Authors: Shahera Banu1, Wenbiao Hu1, Yuming Guo2, Suchithra Naish1 and

Shilu Tong1

Affiliations:

1. School of Public Health and Social Work, Queensland University of Technology,

Brisbane, Australia.

2. School of Population Health, University of Queensland, Brisbane, Australia.


Shahera Banu performed all data analyses and wrote the manuscript. Shilu


Yuming Guo helped with software and Suchithra Naish assisted with data analysis

and provided feedback on initial drafts of manuscript.

46 Chapter 3: Results Paper 1

3.1 ABSTRACT

Background: Dengue fever (DF) is one of the most important emerging

arboviral diseases in humans. Globally, DF incidence has increased by 30-fold over

the last fifty years and the geographic range of the virus and its vectors has expanded

into new areas. The disease is now endemic in more than 120 countries in the

tropical and subtropical parts of the world. We examined the spatiotemporal trends of

DF transmission in the Asia-Pacific region during a 50-year study period and

identified the high risk cluster areas.

Methodology and Findings: We obtained annual numbers of DF cases for 16

countries of the Asia-Pacific region from the World Health Organization DengueNet

for the period 1955 to 2004. The fifty-year data were divided into five time periods,

with ten years in each time period, to investigate the trends of DF transmission.

Space-time cluster analyses were conducted using the scan statistics to detect the

high risk disease clusters. We observed an increasing trend in the spatiotemporal

distribution of DF in the Asia-Pacific region over the study period. Thailand,

Vietnam, Laos, Singapore and Malaysia were identified as the highest risk areas

(Relative risk =13.02) for DF transmission in this region during 1995 to 2004. This

study also indicates that the geographic expansion of DF transmission in the Asia-

Pacific region mostly occurred southward.

Conclusions: This study found that the spatiotemporal distribution of DF in

the Asia-Pacific region has increased in recent years. Thailand, Vietnam, Laos,

Singapore and Malaysia were identified as the highest risk areas for DF in this

region. This information may help to improve DF prevention and control strategies in

the Asia-Pacific region by prioritizing control efforts, where they are most needed.


3.2 INTRODUCTION

Dengue fever (DF) is one of the most important emerging arboviral diseases in

humans and is widespread and common in tropical and subtropical parts of the

world. It is estimated that about 3.6 billion of the global population and

approximately 120 million travellers are at risk of DF. The number of annual DF

cases is about 50-100 million and the mortality is approximately 2.5% (Halstead,

2007, Beatty et al., 2011, WHO, 2012a). DF incidence has increased 30-fold over the

last fifty years and the geographic range of the virus and its vectors has expanded

into new areas (Wilder-Smith, 2012). Only nine countries experienced DF epidemics

before 1970, but the disease is now endemic in more than 120 countries in Africa,

America, Eastern Mediterranean, South-east Asia and the Western Pacific (WHO,

2012a). Between 2000 and 2007 at least eight previously DF free countries became

infected. Outbreaks of suspected DF were recorded in Pakistan, Saudi Arabia,

Yemen, Sudan and Madagascar during 2005 and 2006 (Guzman and Istúriz, 2010).

In Asia, epidemic DF was common during the first half of the 20th century

(Gubler, 2004a). Severe DF epidemics first occurred in the Philippines and Thailand

during the 1950s. The recent geographic distribution of DF showed that the disease

has now spread out from Southeast Asian countries west to India, Sri Lanka,

Maldives and east to China. Several Pacific island nations such as the Cook Islands,

Tahiti, New Caledonia, Vanuatu, Niue, and Palau have also experienced DF

outbreaks (Gubler, 1998). Nearly 1.8 billion people living in the Asia-Pacific region

are currently at risk, which accounts for 70% of the global DF risk (WHO, 2012a).

The tropical climate of Asia and the Pacific is suitable for DF transmission. The

presence of four DF serotypes and high population density make this region

permissive for DF activity (Simmons et al., 2012).


Geographic information systems (GIS) have widely been used in vector borne

disease epidemiology. GIS used for disease mapping can visualize the spatiotemporal

pattern and variation in disease risk. Monitoring the spatiotemporal trends in disease

occurrence can highlight the changing patterns in risk and help to identify risk factors

(Robertson and Nelson, 2010). Spatial scan statistic is one of the most commonly

used approaches in spatial disease surveillance to explore high risk areas or disease

clusters (Abrams and Kleinman, 2007, Kulldorf, 2010). This method scans a larger

encompassing area for possible disease clusters without a priori specification of their

location and size. It identifies approximate location of clusters and performs

significance tests for each cluster (Kulldorff et al., 1997, Kulldorf, 2010). The scan

statistic has been widely used because of its following features (Kulldorff et al.,

1997, Ngui et al., 2013) : firstly, it adjusts for both inhomogeneous population

density and different confounding factors; secondly, searching for clusters without

specifying their size and location ameliorates the problem of pre-selection bias;

thirdly, the likelihood ratio-based test statistics takes multiple testing into account

and gives a single p value for the test of the null hypothesis; and finally, if the null

hypothesis is rejected, it is possible to specify the approximate location of the cluster

that caused the rejection (Kulldorf, 2010). GIS and spatial analyses were used to

identify geographic patterns and risk factors of DF transmission in different areas

(Thai et al., 2010, Tran et al., 2004, Wen et al., 2006, Wu et al., 2009, Mammen et

al., 2008, Hu et al., 2012). However, the spatial pattern of DF remains unexplored at

a continental level around the world. This study examined the spatiotemporal

patterns of DF in the Asia-Pacific region during 1955 to 2004 and identified DF

clusters I different periods of time. Such information is essential for improving DF

prevention and control strategies.

3.3 METHODS

3.3.1 Study area

The continents of Australasia were selected as the study area as the Asia and

Pacific regions are the most seriously affected by DF. Asia is the largest and most

populous continent in the world with approximately 3.9 billion people. The continent

is located in the eastern and northern hemispheres covering 44,579,000 km2 of the

Earth’s surface. Asia’s climate is moist across southeast region and dry across much


of the interior. The monsoon circulation dominates across the southern and eastern

regions, due to the presence of the Himalayas, forcing the formation of a thermal low

which draws in moisture during the summer. South-western parts of the continent are

hot.

The continent of Oceania includes Australia, New Zealand and a number of

widely scattered island nations across the Pacific Ocean. The total land area is

8,536,716 km2 with a population of 37 million. The climate of Oceania's islands is

tropical or subtropical, and ranges from humid to seasonally dry.

3.3.2 Data collection

We obtained the annual number of DF cases at a country level for 16 countries

of the Asia-Pacific region from the DengueNet data query managed by the World

Health Organization (WHO) (WHO, 2012b). DengueNet is an internet based

surveillance tool, established in 2005 to collect and provide current epidemiological

data and trends of DF globally. Currently DF statistics from 1955 onwards are

available in DengueNet. However, many countries did not report their DF outbreak

to WHO during the period of 2005 to 2012. Therefore, we restricted our study to the

period of 1955 to 2004. Among 82 countries of the Asia Pacific region, 22 countries

reported DF outbreaks to the WHO. Among these countries, we included 16

countries in our analyses, as other countries did not report their DF outbreak to WHO

for more than five years. To check the data consistency, we compared the retrieved

dataset for each country with historic DF data published in the literature. Location

information including coordinates, area and population size were collected from the

Central Intelligence Agency (CIA) World Factbook (CIA, 2012). The number of

population censuses varies with country. Therefore, we chose two periods of census,

one closest to the beginning of our study period and another to the end of our study

period. For times before the first census, the population size was set equal to the

population size at the first census time, while for times after the last census time, the

population was set equal to the population at the last census time. Then, linear

interpolation was used to estimate the population for times between censuses.


3.3.3 Data Analyses

To investigate the spatial and temporal patterns of DF transmission, the fifty

years data were divided into five time periods with ten years in each period, A: 1955-

1964, B: 1965-1974, C: 1975-1984, D: 1985-1994 and E: 1995-2004. Cumulative

incidence rates for each time period were mapped to visualize the temporal trends of

DF. To calculate the cumulative incidence for each country, we first calculated the

annual DF incidence. The annual DF incidence was calculated by dividing the

number of annual DF cases by the corresponding population and multiplied by

100,000. Then the annual DF incidences were aggregated for each ten year period to

estimate the cumulative incidence.

A disease cluster is an unusually high concentration of disease occurrence in a

region unlikely to occur by chance. To test for the presence of DF clusters, we used

the Kulldorff’s space-time scan statistic (SaTScan) (Kulldorf, 2010). In the analyses,

we assumed that the number of DF cases in each country to be Poisson distributed.

We then tested the null hypothesis that the number of DF cases is randomly

distributed in geographic space and time and the expected DF cases in each area are

proportional to its population (Kulldorf, 2010, Kulldorff et al., 1997). The space-time

scan statistic is defined by a cylindrical window with a circular geographic base and

with height corresponding to time. The cylindrical window is then moved in space

and time, so that we obtained an infinite number of overlapping cylinders of different

sizes and shapes, jointly covering the entire study region, where each cylinder

reflects a possible cluster (Kulldorf, 2010, Kulldorff et al., 1997). For each

cylindrical window, the scan statistic tests the null hypothesis against the alternative

hypothesis that there is an elevated risk of DF within window, compared to outside

window (Kulldorf, 2010). SaTScan detects potential clusters by calculating a

maximum likelihood ratio for each cylindrical window (Kulldorf, 2010). The

window with the maximum likelihood ratio is considered as the most likely cluster.

SaTScan also detects secondary clusters that have a significantly large likelihood

ratio but are not the most likely cluster. To evaluate the statistical significance of

both most likely and secondary clusters, SaTScan generates a large number of

random replication of the dataset under the null hypothesis to obtain the p-value

through Monte Carlo hypothesis testing. Then it compares the rank of the maximum


likelihood from the real dataset with the maximum likelihood from the random

dataset (Kulldorff et al., 1997). In our analyses, we used 9999 Monte Carlo

replications.

For cluster specification in space-time analyses, two parameters were set for

the maximum cluster size: the proportion of the population at risk and the proportion

of the study period. The population density in our study area (16 countries) varies

greatly and in disease surveillance, higher numbers of cases are usually expected in

urban areas compared to a similar sized area in the countryside due to the higher

population density. To adjust for this uneven population density and in line with

previous literature on mosquito-borne diseases, we decided to limit our spatial cluster

size to 15% of the population at risk (Naish et al., 2011, Chen et al., 2008). However,

analyses were conducted with maximum spatial cluster sizes of 50%, 40%, 30% and

20% of the population at risk to avoid the pre-selection bias, and the results were

very similar to the 15% population limit. We used a maximum of 50% of the study

period as a maximum cluster size in the temporal window.

We used the SaTScan software (version 9.1.1) for the space-time scan statistic

test (Kulldorf, 2010) and R software (version 2.12.0; R development Core Team

2009) to map all results. The “maptools” package of R was used to translate the

space-time outputs into maps and visualize the DF clusters


Table 3.1 Summary statistics for annual number of DF cases in the Asia-Pacific countries during 1955-2004

Countries (N=16) Minimum 25% Median 75% Maximum Mean Std. Deviation

Australia 0 0 0 44 868 88 189

Bangladesh 0 0 0 0 6,104 378 1,300

Cook Islands 0 0 0 25 2,256 126 437

India 0 0 0 773 16,517 1,552 3,609

Indonesia 0 0 6,449 21,552 78,690 14,948 19,258

Laos 0 0 0 1,733 17,690 1,553 3315

Malaysia 0 0 810 5,508 33,895 4,932 9,002

Maldives 0 0 0 0 2,054 99 388

Micronesia 0 0 0 0 700 30 134

Myanmar 0 0 1,795 4,854 16,047 3,177 4,053

Philippines 0 388 1,042 6,342 35,648 4,985 8,236

Singapore 0 91 273 1,268 9,459 1,105 1,797

Sri Lanka 0 0 1 679 15,408 942 2,621

Thailand 0 5,914 23,018 45,555 1,74,285 33,814 38,637

Tuvalu 0 0 0 0 811 17 114

Vietnam 0 40 27,306 49,668 3,54,517 41,819 63,532


Figure 3.1 Total numbers of countries with DF outbreaks in the Asia-Pacific region, during 1955-2004. (Data source: WHO DengueNet)


Figure 3.2 Cumulative incidence of DF in the Asia-Pacific countries. A: 1955-1964, B: 1965-1974, C: 1975-19854,D:1985-1994,E:1995-2004.The X and Y axis of the map show the longitude and latitude respectively.


3.4 RESULTS

3.4.1 Descriptive statistics

The annual number of DF for the selected countries ranged from zero to

3,54,517 cases during 1995-2004 (Table 3.1). When present, the lowest average

number of DF cases was reported in Tuvalu (17) and the highest was in Vietnam

(41,819). We observed that the number of countries with DF dramatically increased

over time (Figure 3.1). 22 (26%) countries of the Asia-Pacific region reported at least

one DF outbreak in the fifty years from 1955 to 2004.

3.4.2 Trends of DF transmission

Figure 3.2 shows that the DF endemic areas had expanded geographically in

the Asia-Pacific region over the 50-year study period. More and more countries were

affected by DF over time. On average at least two new countries experienced DF

outbreaks in each decade (Figure 3.2). Thailand, Vietnam, Singapore and Philippines

were affected by DF in the earliest years during 1955-1964 which suggest that any of

these countries can be the origin of DF transmission in this region. Figure 3.2 also

shows that the geographic expansion of DF mainly occurred southwardly in the Asia-

Pacific region. The countries in the south of Thailand or Philippines like Indonesia,

Malaysia, Australia and other Pacific Islands became infected by DF in recent years.

The highest DF incidence was observed in the Cook Islands (2,123/100,000 people)

during 1995 to 2004.


Table 3.2 Space-time clusters of DF transmission in the Asia-Pacific region during 1955- 2004

No.Obs, number of observed cases; No. Exp, number of expected cases; LLR, Log -likelihood Ratio.٭P<0.05; †Most likely cluster

Cluster Countries Radius (km) Time frame No. Obs. No. Exp. Relative risk LLR٭

1955-1964 1† Thailand 0 1960/1/1 to 1964/12/31 18337 527.21 96.13 54814.34 2 Philippines 0 1960/1/1 to 1964/12/31 3092 571.13 5.95 2819.03

1965-1974

1† Singapore, Malaysia, Thailand, Vietnam 1711.16 1971/1/1 to 1974/12/31 76393 6143.11 24.05 142940.013

2 Philippines 0 1966/1/1 to 1966/12/31 9384 528.09 18.88 18411.19 3 Myanmar 847.81 1974/1/1 to 1974/12/31 2477 1755.02 1.42 133.25

1975-1984 1† Vietnam, Laos, Thailand 812.13 1980/1/1 to 1984/12/31 510942 38105.18 31.06 1024627.77 2 Cook Islands 0 1980/1/1 to 1980/12/31 357 1.31 273.01 1646.82 3 Malaysia 0 1982/1/1 to 1982/12/31 3052 1114.03 2.75 1140.04 4 Indonesia 0 1983/1/1 to 1984/12/31 26585 23294.07 1.15 228.63

1985-1994 1† Vietnam, Laos, Thailand 812.13 1987/1/1 to 1991/12/31 1034416 79277 27.18 2009818.49 2 Cook Islands 0 1991/1/1 to 1991/12/31 1776 2.54 699.17 9858.32 3 Indonesia 0 1988/1/1 to 1988/12/31 44573 23048 1.96 7995.79 4 Maldives 0 1988/1/1 to 1988/12/31 2054 30.41 67.61 6630.25 5 Philippines 0 1991/1/1 to 1991/12/31 11317 8290 1.37 497.68

1995-2004

1† Singapore, Malaysia,

Thailand, Vietnam, Laos

1872.04 1995/1/1 to 1998/12/31 852301 95356.45 13.02 1243215.57

2 Philippines 0 2001/1/1 to 2004/12/31 94651 45564.40 2.12 21013.62 3 Sri Lanka, Maldives 983.29 2002/1/1 to 2004/12/31 29895 8771.22 3.44 15623.32


Figure 3.3 Space-time clusters of DF transmission in the Asia-Pacific region. A: 1955-1964, B: 1965-1974,C:1975-19854,D:1985-1994,E:1995-2004. The X and Y axis of the map show the longitude and latitude respectively.


3.4.3 Space-time clusters

Table 3.2 shows the results of the space-time cluster analysis, stratified in five

time periods. Using a maximum cluster size of 15% of the population at risk,

Thailand (RR=96.13) was identified as the most likely cluster by SaTScan and the

Philippines (RR=5.96) was the secondary cluster during 1955-1964. The most likely

cluster covered four countries (Singapore, Malaysia, Thailand and Vietnam) with a

radius of 1,711.16 km detected in 1965-1974. Thus, the DF cluster areas

substantially increased during 1965-1974 compared to the previous ten years and this

trend continued in the following years. In the recent decade (1995-2004), eight

countries were identified as a statistically significant cluster for DF. The most likely

cluster includes Singapore, Malaysia, Thailand, Vietnam and Laos (radius =1,872.04

km, RR=13.02). Overall, we observed that the DF cluster areas in the Asia-Pacific

region had expanded over time (Figure 3.3).

3.5 DISCUSSION

The results of this study indicate that the geographical range of DF

transmission in the Asia-Pacific region had expanded during 1955 to 2004. On

average, at least two countries entered into the DF cluster areas every ten years.

There are many factors which can be responsible for this geographic expansion of

DF in this region during the 20th century: for example, unprecedented population

growth, unplanned urbanization, lack of effective vector control and international

travel (Tatem et al., 2006, Gubler, 2011). The movement of troops and war materials

during World War II may have played a crucial role in the dissemination of Aedes

mosquitoes and the virus (Gubler, 2002a, Gubler, 2011). Another important factor

can be the major economic growth in Southeast Asia, which occurred after World

War II (Gubler, 2004a, Kyle and Harris, 2008). This economic growth leads to

unplanned urbanization where millions of people live in shanty towns with

inadequate housing, lack of water supplies and waste management facilities. This

overcrowded community with a large mosquito population creates ideal conditions

for DF transmission (Arunachalam et al., 2010, Gubler, 2004b, Schmidt et al., 2011).

A recent study in Taiwan showed that urbanization and increased temperature due to

climate change are the most important risk factors for DF transmission (Wu et al.,


2009). In addition, increased use of modern transportation due to globalization is

responsible for the importation of the dengue virus through viremic individuals and

dispersal of exotic mosquitoes into new areas. It has been suggested that Aedes

albopictus was introduced into many Pacific islands through modern container ships

(Tatem et al., 2006, Wilder-Smith and Gubler, 2008).

Global climate change has also been suggested to be an important factor for the

range expansion of DF in Asia (Benitez, 2009). Climatic factors, including

temperature, rainfall and humidity, have direct and indirect impacts on mosquito

survival, their life span and reproductive rate, which can influence the geographic

distribution of vectors and the virus (Patz et al., 2005). An association between DF

incidence and rainfall was reported in many countries of the Asia-Pacific region and

the outbreaks usually coincided with the rainy season (Banu et al., 2011). Rainfall

can potentially increase the number of mosquito breeding sites and thus can increase

the chance of DF transmission (Arcari et al., 2007).

Thailand, Vietnam, Laos, Singapore and Malaysia were identified as the most

likely cluster in the most recent years (1995 - 2004). DF transmission in this area

follows a cyclical pattern with highest incidence in the hot and rainy seasons from

May to October (Barbazan et al., 2002, Cuong et al., 2011). We know that DF

infection in travellers depends on destination, season of travel, duration of stay and

epidemic activity. Therefore, travellers to these countries may need to take

precaution like avoiding the monsoon season and shortening the duration of their

stay if a DF outbreak occurs. This awareness may significantly reduce the risk of DF

transmission to the non-endemic areas.

Our results also suggest that the geographic expansion of DF in the Asia-

Pacific region might originate from the Philippines or Thailand. These two countries

were identified as DF clusters as early as 1960. Many other studies also suggested

that Philippine or Thailand can be the origin of DF transmission in Asia (Gubler,

1998). Historically, the first severe DF outbreak occurred in Manila and Philippines


in 1953 followed by Bangkok and Thailand (Gubler, 2011, Guzman et al., 2010).

The geographic expansion of DF in the Asia-pacific region mostly occurred

southwardly. Global climate change might explain this southward expansion of DF

to some extent. In the past 100 years, mean surface temperature has increased by 0.3-

0.8oC across the continent, (IPCC, 2007) which could create climatic conditions

suitable for dengue mosquito vector and facilitated DF transmission in this region

(Benitez, 2009, Kyle and Harris, 2008, Hales et al., 2002). A southward expansion in

DF transmission was also observed in Argentina and Australia (Woodruff et al.,

2006, Hu et al., 2011, Carbajo et al., 2012). However, it is still debated whether the

southward expansion is real or an artefact (Russell, 2009).

This study has several strengths. This is the first empirical study exploring the

spatiotemporal pattern of DF transmission in the Asia-pacific region. DF data from

16 countries for 50 years were used in this study. It indicates the necessity for future

research assessing important determinants of DF emergence and rapid geographic

expansion in this region. It also suggests the importance of exploring the DF

transmission pattern within countries.

The main limitations of this study include the low resolutions of the DF

dataset. We only obtained annual country level data for DF. It would be better if we

could have higher resolution such as monthly or weekly data in this study which

would also show the seasonal variation in DF transmission, but this information is

not available for most countries in the WHO DengueNet. Additionally, we did not

include some DF endemic countries in the Asia-Pacific region like Taiwan and China

in this study due to the unavailability of data. Inclusion of these data might increase

the DF cluster area and help us to justify the southward expansion of DF in this

region. There are also some issues in the quality of the WHO DengueNet data. Under

reporting is possible when some countries did not report DF outbreak information to

the DengueNet for years which might bias our results. Over reporting is also possible

as some countries use only clinical diagnosis rather than serological diagnosis which

cannot differentiate DF from other diseases like chikungunya.


In summary, this study found that the spatial and temporal distribution of DF in

the Asia-Pacific region increased over the 50-year study period. Social, ecological

and demographic changes that have occurred in recent years are thought to be

responsible for the geographic expansion of DF. Global climate change can further

expand the geographic limit of DF. Thailand, Vietnam, Laos, Singapore and

Malaysia were identified as the most likely cluster for DF in this region. This

information may help to improve DF prevention and control strategies in the Asia-

Pacific region by prioritizing control efforts, where they are most needed.

Financial support

The work was supported by QUT postgraduate scholarships (SB), NMHRC

postdoctoral training fellowship (WH) and a NMHRC research fellowship (ST). The

funders had no role in study design, data collection and analysis, decision to publish,

or preparation of the manuscript.




References

ABRAMS, A. M. & KLEINMAN, K. P. 2007. A SaTScan macro accessory for cartography (SMAC) package implemented with SAS software. International journal of health geographics, 6, 6.


ARUNACHALAM, N., TANA, S., ESPINO, F., KITTAYAPONG, P., ABEYEWICKREME, W., WAI, K. T., TYAGI, B. K., KROEGER, A., SOMMERFELD, J. & PETZOLD, M. 2010. Eco-bio-social determinants of dengue vector breeding: a multicountry study in urban and periurban Asia. Bulletin of The World Health Organization, 88, 173-183.

BANU, S., HU, W., HURST, C. & TONG, S. 2011. Dengue transmission in the Asia-Pacific region: impact of climate change and socio-environmental factors. Tropical Medicine and International Health, 16, 598-607.

BARBAZAN, P., YOKSAN, S. & GONZALEZ, J. P. 2002. Dengue hemorrhagic fever epidemiology in Thailand: description and forecasting of epidemics. Microbes Infect, 4, 699-705.

BEATTY, M. E., BEUTELS, P., MELTZER, M. I., SHEPARD, D. S., HOMBACH, J., HUTUBESSY, R., DESSIS, D., COUDEVILLE, L., DERVAUX, B., WICHMANN, O., MARGOLIS, H. S. & KURITSKY, J. N. 2011. Health economics of dengue: a systematic literature review and expert panel's assessment. American Journal of Tropical Medicine and Hygiene, 84, 473-488.

BENITEZ, M. A. 2009. Climate change could affect mosquito-borne diseases in Asia. Lancet, 373, 1070-1070.

CARBAJO, A. E., CARDO, M. V. & VEZZANI, D. 2012. Is temperature the main cause of dengue rise in non- endemic countries? The case of Argentina. International journal of health geographics, 11, 26.

CHEN, J., ROTH, R. E., NAITO, A. T., LENGERICH, E. J. & MACEACHREN, A. M. 2008. Geovisual analytics to enhance spatial scan statistic interpretation: an analysis of U.S. cervical cancer mortality. International journal of health geographics, 7, 57.

CIA. 2012. CIA World Factbook [Online]. Available: https://www.cia.gov/library/publications/the-world-factbook/ [Accessed 12 May 2012].

CUONG, H. Q., HIEN, N. T., DUONG, T. N., PHONG, T. V., CAM, N. N., FARRAR, J., NAM, V. S., THAI, K. T. & HORBY, P. 2011. Quantifying the emergence of dengue in hanoi, Vietnam: 1998-2009. PLoS Neglected Tropical Diseases, 5, e1322.

http://www.cia.gov/library/publications/the-world-factbook/


GUBLER, D. J. 1998. Dengue and dengue hemorrhagic fever. Clinical Microbiology Reviews, 11, 480-496.

GUBLER, D. J. 2002. Epidemic dengue/dengue hemorrhagic fever as a public health, social and economic problem in the 21st century. Trends Microbiol, 10, 100-103.

GUBLER, D. J. 2004a. The changing epidemiology of yellow fever and dengue, 1900 to 2003: full circle? Comparative Immunology, Microbiology and Infectious Diseases, 27, 319-330.

GUBLER, D. J. 2004b. Cities spawn epidemic dengue viruses. Nature Medicine, 10, 129-130.

GUBLER, D. J. 2011. Dengue, urbanization and globalization: the unholy trinity of the 21(st) century. Tropical Medicine and Health, 39, 3-11.

GUZMAN, A. & ISTÚRIZ, R. E. 2010. Update on the global spread of dengue. International Journal of Antimicrobial Agents, 36, S40-42.

GUZMAN, M. G., HALSTEAD, S. B., ARTSOB, H., BUCHY, P., FARRAR, J., GUBLER, D. J., HUNSPERGER, E., KROEGER, A., MARGOLIS, H. S., MARTINEZ, E., NATHAN, M. B., PELEGRINO, J. L., SIMMONS, C., YOKSAN, S. & PEELING, R. W. 2010. Dengue: a continuing global threat. Nature Reviews. Microbiology, 8, S7-S16.



HU, W., CLEMENTS, A., WILLIAMS, G. & TONG, S. 2011. Spatial analysis of notified dengue fever infections. Epidemiology and infection, 139, 391-399.

HU, W., CLEMENTS, A., WILLIAMS, G., TONG, S. & MENGERSEN, K. 2012. Spatial patterns and socioecological drivers of dengue fever transmission in Queensland, Australia. Environmental Health Perspectives, 120, 260-266.


KULLDORF, M. 2010. SaTScanTM User Guide for Version 9.0. Available: http://www.satscan.org/ [Accessed 20 October 2010].

KULLDORFF, M., FEUER, E. J., MILLER, B. A. & FREEDMAN, L. S. 1997. Breast cancer clusters in the northeast United States: a geographic analysis. Am J Epidemiol, 146, 161-70.

http://www.satscan.org/



MAMMEN, M. P., PIMGATE, C., KOENRAADT, C. J. M., ROTHMAN, A. L., ALDSTADT, J., NISALAK, A., JARMAN, R. G., JONES, J. W., SRIKIATKHACHORN, A., YPIL-BUTAC, C. A., GETIS, A., THAMMAPALO, S., MORRISON, A. C., LIBRATY, D. H., GREEN, S. & SCOTT, T. W. 2008. Spatial and temporal clustering of dengue virus transmission in Thai villages. PLoS Med, 5, e205.

NAISH, S., HU, W., MENGERSEN, K. & TONG, S. 2011. Spatial and temporal clusters of Barmah Forest virus disease in Queensland, Australia. Tropical Medicine and International Health, 16, 884-893.

NGUI, A. N., APPARICIO, P., FLEURY, M.-J., LESAGE, A., GRÉGOIRE, J.-P., MOISAN, J. & VANASSE, A. 2013. Spatio-temporal clustering of the incidence of schizophrenia in Quebec, Canada from 2004 to 2007. Spatial and Spatio-temporal Epidemiology, 6, 37-47.


ROBERTSON, C. & NELSON, T. 2010. Review of software for space-time disease surveillance. International journal of health geographics, 9, 16.

RUSSELL, R. C. 2009. Mosquito-borne disease and climate change in Australia: time for a reality check. Australian Journal of Entomology, 48, 1-7.

SCHMIDT, W. P., SUZUKI, M., DINH THIEM, V., WHITE, R. G., TSUZUKI, A., YOSHIDA, L. M., YANAI, H., HAQUE, U., HUU THO, L., ANH, D. D. & ARIYOSHI, K. 2011. Population density, water supply, and the risk of dengue fever in Vietnam: cohort study and spatial analysis. PLoS Medicine, 8, e1001082.

SIMMONS, C. P., FARRAR, J. J., VAN VINH CHAU, N. & WILLS, B. 2012. Dengue. New England Journal of Medicine, 366, 1423-1432.

TATEM, A. J., HAY, S. I. & ROGERS, D. J. 2006. Global traffic and disease vector dispersal. Proc Natl Acad Sci U S A, 103, 6242-6247.

THAI, K. T. D., NAGELKERKE, N., PHUONG, H. L., NGA, T. T. T., GIAO, P. T., HUNG, L. Q., BINH, T. Q., NAM, N. V. & DE VRIES, P. J. 2010. Geographical heterogeneity of dengue transmission in two villages in southern Vietnam. Epidemiology and infection, 138, 585-591.

TRAN, A., DEPARIS, X., DUSSART, P., MORVAN, J., RABARISON, P., REMY, F., POLIDORI, L. & GARDON, J. 2004. Dengue spatial and temporal patterns, French Guiana, 2001. Emerging Infectious Diseases, 10, 615-621.

WEN, T.-H., LIN, N. H., LIN, C.-H., KING, C.-C. & SU, M.-D. 2006. Spatial mapping of temporal risk characteristics to improve environmental health risk


identification: A case study of a dengue epidemic in Taiwan. Science of the Total Environment, 367, 631-640.

WHO. 2012a. Dengue and severe dengue [Online]. World Health Organization. Available: http://www.who.int/mediacentre/factsheets/fs117/en/index.html [Accessed 27 September 2012].

WHO. 2012b. DengueNet data query [Online]. World Health Organization. Available: http://apps.who.int/globalatlas/DataQuery/default.asp.

WILDER-SMITH, A. 2012. Dengue infections in travellers. Paediatr Int Child Health, 32 Suppl 1, 28-32.

WILDER-SMITH, A. & GUBLER, D. J. 2008. Geographic expansion of dengue: the impact of international travel. Medical Clinics of North America, 92, 1377-1390.

WOODRUFF, R. E., MCMICHAEL, T., BUTLER, C. & HALES, S. 2006. Action on climate change: the healthrisks of procastinating. Australian and New Zealand Journal of Public Health, 30, 567-571.


http://www.who.int/mediacentre/factsheets/fs117/en/index.html

http://apps.who.int/globalatlas/DataQuery/default.asp



Space-time clusters of dengue fever in Bangladesh

Published in Tropical Medicine and International Health


Citation:

BANU, S., HU, W., HURST, C., GUO, Y., ISLAM, M. Z. & TONG, S. 2012. Space-

time clusters of dengue fever in Bangladesh.Tropical Medicine and International Health, 17,

1086-1091.


Shahera Banu performed all data analyses and wrote the manuscript. Shilu Tong and

Wenbiao Hu supervised the study and assisted with writing the manuscript. Cameron Hurst

contributed to the manuscript in terms of providing feedback on initial drafts. Yuming Guo

helped with software and Mohammad Zahirul Islam provided boundary dataset for mapping.


4.1 SUMMARY

OBJECTIVE: To examine the space-time clustering of dengue fever (DF) transmission

in Bangladesh using geographical information system and spatial scan statistics (SaTScan).

METHODS: We obtained data on monthly suspected DF cases and deaths by district in

Bangladesh for the period of 2000-2009 from the Directorate General of Health Services.

Population and district boundary data of each district were collected from national census

managed by Bangladesh Bureau of Statistics. To identify the space-time clusters of DF

transmission a discrete Poisson model was performed using SaTScan software.

RESULTS: The results indicate that space-time distribution of DF transmission was

clustered during three different periods 2000-2002, 2003-2005 and 2006-2009. Dhaka was

identified as the most likely cluster for DF in all three periods. Several other districts were

identified as significant secondary clusters. However, the geographic range of DF

transmission appears to have declined in Bangladesh over the last decade.

CONCLUSION: There were significant space-time clusters of DF in Bangladesh over

the last decade. Our results would prompt future studies to explore how social and ecological

factors may affect DF transmission and would also be useful for improving DF control and

prevention programs in Bangladesh.


4.2 INTRODUCTION

Dengue fever (DF) is one of the most important emerging arboviral diseases worldwide

(WHO, 2000). The virus is transmitted through the bite of container breeding mosquitoes

Aedes aegypti and Aedes albopictus which are present in most tropical and subtropical

countries (Rigau-Perez et al., 1998). Symptoms of DF infections in human vary from mild flu

like DF to life threatening dengue hemorrhagic fever (DHF) or dengue shock syndrome

(DSS)(Gubler, 2002a). It has been estimated that about 50 million people worldwide become

infected with the dengue viruses annually. Globally, the geographic range of DF transmission

has increased dramatically in recent years (WHO, 2000). It has been hypothesized that many

social and demographic changes such as population growth, urbanization, air travel and

climate change contribute to the increased incidence and geographical expansion of DF

transmission (Gubler, 2002a, Wu et al., 2007).

DF has become a serious public health concern in Bangladesh after the first large scale

outbreak occurred in 2000. However, evidence suggests that sporadic DF outbreaks occurred

in Bangladesh between 1964 and 1999 (Hossain et al., 2003, Rahman et al., 2002). Since

2000, DF cases have been reported every year in all major cities of Bangladesh. More than

23,872 cases were reported to the Directorate General of Health Services (DGHS) and 233

were fatal between 2000 and 2009. The worst outbreak was in 2002 with 6,132 cases and 58

deaths. Both Aedes aegypti and Aedes albopictus were identified as potential vectors for DF

transmission in Bangladesh (Ali et al., 2003). A national guideline based on the WHO

protocol was developed by DGHS in 2000 to control DF transmission and reduce its

morbidity and mortality (DGHS, 2000).

In the absence of effective vaccine and specific treatment, vector control is the only

way to prevent DF transmission. Previous studies suggest that the risk of DF transmission

varies over space and time (Mammen et al., 2008, Tran et al., 2004, Thai et al., 2010).

Therefore, an identification of high risk areas can be useful for prioritizing DF surveillance

and vector control efforts in areas where they are most needed (Ali et al., 2003). In this study,

we investigated the spatial and temporal distribution of DF at a district level in Bangladesh

during 2000 - 2009. Our aim was to identify high risk clustering areas for DF transmission in

Bangladesh using space-time scan statistics and geographic information system.


4.3 METHODS

4.3.1 Study area

Bangladesh is located in South Asia between latitude of 20-27oN and longitude of 88-

93oE. It is a low-lying, riverine country with a large marshy jungle coastline of 710 km on the

northern littoral of the Bay of Bengal. It is one of the most densely populated countries

(density 964/km2) in the world and covers 147,570 km2. Bangladesh is divided into seven

administrative divisions and these are subdivided into districts. There are 64 districts in

Bangladesh, each further subdivided into Upazila or Thana. Dhaka is the capital and largest

city of Bangladesh. Other major cities include Chittagong, Khulna, Rajshahi, Sylhet and

Barisal. Bangladesh has a tropical monsoon climate characterized by wide seasonal variation

in rainfall, temperatures and humidity. Regional climatic differences in this flat country are

minor. Four seasons are generally recognized; hot (21.7 - 35.5 0C), muggy summer from June

to August; humid and rainy autumn from September to November; cool (11.7 - 26.8 0C) and

dry winter from December to February; and warm spring from March to May. About 80 % of

Bangladesh's rain falls during the wet monsoon season from June to November.


As DF is a notifiable disease in Bangladesh, any DF case detected based on the

“clinical case definitions” of National guidelines for clinical management of dengue

syndrome must be reported to the DGHS through the Civil Surgeon of the district. A DF

suspected case is defined by the presence of acute fever accompanied by any two of the

following clinical symptoms such as headache, myalgia, arthralgia, rash, positive tourniquet

test and leucopenia, absence of any other febrile illness and high index of suspicion based

on period, population and place (DGHS, 2000). We obtained computerized datasets

containing the number of monthly suspected DF cases and deaths by district in Bangladesh

for the period of 1 January 2000 to 31 December 2009 from DGHS. There has been no

significant change in the dengue reporting system since 2000 in Bangladesh. Relevant

population data and electronic boundaries of each district were retrieved from the national

census database managed by Bangladesh Bureau of Statistics (BBS). District information

includes district name, code, area (km2), digital boundaries and base maps. The district

population were generated from the 2001 and 2010 census data and for the remaining years,

the district population were estimated based on linear interpolation (Kulldorf, 2010). This


study was approved by the Human Research Ethics Committee, Queensland University of

Technology.

4.3.3 Statistical analysis

A space-time statistical analysis was applied to detect high risk clusters of DF using

SaTScan software (version 9). Case files generated using monthly aggregated numbers of DF

cases for each district was used in data analysis. Population and coordinates data were also

used as inputs in SaTScan. We fit a discrete Poisson regression model to identify space-time

clusters after adjustment for the uneven geographical density of district population (Kulldorf,

2010, Kulldorf et al., 1998).

For cluster specification in space-time analyses, three parameters were set for cluster

size: the maximum circle radius in the spatial window, maximum temporal window and the

proportion of the population at risk. Of the 64 districts 99% of them were within a 40 km

radius. Thus, 40 km was chosen as the maximum circle radius for all analyses. Space-time

analyses were also performed using 10 km, and 20 km of maximum circle radius and the

results obtained were very similar to those using 40 km. The analyses were conducted using a

maximum spatial cluster size of 50% of the population at risk in the spatial window and a

maximum of 50% of the study period in the temporal window. Because of the differences in

population densities across Bangladesh, it was decided to limit the spatial cluster size to 50%

of the population at risk. The spatial clusters were also defined to cover less than 25% and

10% of total population at risk and similar results were obtained to the 50% of the population

limit, which suggests that the maximum cluster size is adequate to be limited by 50% of the

population at risk. The most likely and secondary likely clusters were detected through the

likelihood ratio test. Significance of the clusters was evaluated with Monte Carlo simulation

which was set to 9999. MapInfo Professional (version 10.0) was used to display space-time

clusters.


4.4 RESULTS

4.4.1 DF epidemics and outbreaks

The epidemic pattern of DF fluctuated from 2000 to 2009 with major outbreaks in 2000

and 2002 (Figure 4.1). The monthly number of DF cases ranged from 0 to 3,281 (mean =

198, standard deviation = 454) and the highest number of cases was reported in August 2002.

DF outbreaks in Bangladesh after 2002 showed a decreasing trend. Although DF outbreaks

were regular in Bangladesh, a one year inter-epidemic period between outbreaks was

apparent. The monthly number of DF deaths ranged from 0 to 33 and no death was recorded

after 2006. Figure 4.1 also shows a striking variation in monthly numbers of districts with DF

infection from 2000 to 2009. Peaks in DF cases generally coincided with high monthly

numbers of districts with DF.

4.4.2 Disease clusters

The cluster analyses show that the space-time distribution of DF was clustered during

three periods 2000-2002, 2003-2005 and 2006-2009. Using the maximum spatial cluster size

of 50% of the population at risk and a circle radius of 40 km, Dhaka district was identified as

the most likely cluster and Khulna and Chittagong as secondary clusters (Figure 4.2).

Another two districts (Barisal and Jhenaidah) were identified as secondary clusters in 2000

during August to October but those clusters were not found in later years (Figure 4.2). All

clusters were identified between June - November which represents a rainy monsoon season

in Bangladesh.


Figure 4.1 Monthly number of DF cases, deaths and districts with DF notification between January 2000 and December 2009 in Bangladesh.

0

5

10

15

20

25

30

35

0

500

1000

1500

2000

2500

3000

3500

Jan-

00A

pr-0

0Ju

l-00

Oct

-00

Jan-

01A

pr-0

1Ju

l-01

Oct

-01

Jan-

02A

pr-0

2Ju

l-02

Oct

-02

Jan-

03A

pr-0

3Ju

l-03

Oct

-03

Jan-

04A

pr-0

4Ju

l-04

Oct

-04

Jan-

05A

pr-0

5Ju

l-05

Oct

-05

Jan-

06A

pr-0

6Ju

l-06

Oct

-06

Jan-

07A

pr-0

7Ju

l-07

Oct

-07

Jan-

08A

pr-0

8Ju

l-08

Oct

-08

Jan-

09A

pr-0

9Ju

l-09

Oct

-09

Num

ber

of d

engu

e de

aths

/ nu

mbe

r of

dis

tric

st w

ith d

engu

e ca

ses

Num

ber

of d

engu

e c

ases

Dengue cases Districts with dengue cases Dengue deaths


Figure 4.2. Space-time clusters of DF identified in three different periods in Bangladesh.


Table 4.1 Space-time clusters of DF in Bangladesh, 2000-2009.

No.Obs, number of observed cases; No. Exp, number of expected cases; LLR, Log -likelihood Ratio.٭P<0.05; †Most likely cluster.

Cluster Name of District

Radius (km)

Area (Sq km)

Time frame No. Obs. No. Exp. Relative risk LLR٭

2000-2002 1† Dhaka

0 307.42 2001/7/1 to 2002/10/31 8037

415.15 43.48 18865.32

2 Khulna 0 56.83

2000/8/1 to 2000/11/30 590 28.39 21.63 1239.71

3 Chittagong 0 773.57

2000/8/1 to 2000/9/30 590 39.40 15.58 1057.01

4 Jhenaidah 0 227.23

2000/8/1 to 2000/8/31 24 4.80 5.00 19.44

5 Barisal 0 515.38

2000/8/1 to 2000/10/31 2.63 21.33 2.63 19.43

2003-2005

1† Dhaka

0 307.42 2004/6/1 to 2005/11/30 4887 181.22 233.31 14773.74

2 Chittagong 0 773.57

2003/10/1 to 2003/11/30 39 15.32 2.56 12.81

2006-2009

1† Dhaka 0 307.42 2006/6/1 to 2006/11/30 2144 35.44 119.80 7326.35

2 Khulna 0 56.83

2006/7/1 to 2006/9/30 46 4.87 9.53 62.32


Table 4.1 shows the names of districts included in each cluster, radius (km),

observed cases, expected cases, Relative Risk (RR) and log-likelihood ratio. The

highest risk for the most likely cluster (Dhaka) was found in 2004-2005 during June -

November (RR= 233.3, p< 0.05) and the risk was attenuated in 2006 during June -

November (RR=119.8, p< 0.05) and was lowest in 2000-2002 (RR= 43.48, p< 0.05).

For secondary clusters the RR ranged from 2.56 to 21.63 and the highest risk was

identified for Khulna in 2000 during August - November (RR= 21.64, p< 0.05)

which reduced to 9.53 in 2006 - 2009.

4.4.3 Spatial dispersion of DF

We attempted to identify whether changes in DF transmission varied with

latitude and longitude of district centroids in the periods 2000-2002, 2003-2005 and

2006-2009. A logistic regression model was constructed with the dichotomous

outcome variable defined as whether or not an increase of DF cases occurred in each

district between the three periods. Longitude and latitude of district centroids were

entered as explanatory variables. The results indicate that changes of DF

transmission were not significantly associated with geographic variation (i.e latitude

and longitude) during 2000-2002, 2003-2005 and 2006-2009.


4.5 DISCUSSION

The results of this study indicate a significant variation in spatiotemporal

distribution of DF in Bangladesh. The geographic extent of notified DF cases has

declined in Bangladesh over the study period. Dhaka was identified as the highest

risk area for DF transmission in Bangladesh.

Space-time cluster analysis is a valuable tool to examine how spatial patterns

change over time. This study shows that DF transmission in Bangladesh was

clustered in three different periods and provides a clear pattern of DF clustering

within this country. Dhaka was identified as the most likely cluster for DF

transmission. It may be because Dhaka is the largest city in Bangladesh with a high

population density. DF can be easily transmitted by mosquitoes in this area. Several

other districts (Chittagong and Khulna) in the southern part of the country were

identified as secondary clusters (Figure 4.2). These are also the fastest growing cities

located close to seaport and airport. Apart from population density, the movement of

infected individuals and/or mosquitoes into this region from overseas through air

travel or by ship may increase the chance of DF transmission (Sutherst, 2004).

However, rainwater collection for domestic purposes is not a common practice in

Bangladesh, about 36% household of the southwest coastal region of the country use

rainwater due to arsenic contamination in the ground water and high salinity problem

(Ferdausi and Bolkland, 2000). The storage of rainwater in uncovered containers in

this area might increase the risk of DF transmission by providing suitable mosquito

breeding habitats (Beebe et al., 2009). Climatic variation and socio-economic factors

may also play an important role in promoting DF transmission (Hu et al., 2011, Hu et

al., 2012). Further investigation in these high risk areas is required to understand the

dynamics of DF transmission and to discover the role of biological, social and

environmental factors in the transmission of DF.


We found that the geographic range of DF clusters has declined in Bangladesh

over the last decade (Figure 4.2). We speculate that the effective management of DF

patients according to the national dengue guidelines and huge public awareness after

the first outbreak in 2000 might have helped to reduce DF transmission in

Bangladesh (Rahman et al., 2002). There is no routine dengue vector control

programme in Bangladesh. Though, irregular mosquito control activities exist in

some cities in Bangladesh which is not specifically for Aedes mosquitoes

(Chepesiuk, 2003). According to Dhaka City Corporation (DCC), they spray

adulticide and larvicide in every summer in the areas with high mosquito population.

DCC also sprays insecticide indoor and outdoor in the notified area when a DF

outbreak occurs. Besides vector control measures, socio-economic changes,

population immunity and changes in viral strain and fitness may also contribute to

the decreased DF incidence in Bangladesh (Cummings et al., 2009, Podder et al.,

2006). Further research is needed to identify possible reasons for declining trend of

DF in Bangladesh.

Dhaka was constantly identified as the most likely cluster and very few

districts were identified as secondary clusters for DF transmission in Bangladesh in

recent years. We believe that patients detected with DF in other districts might have

acquired infections during their visits to Dhaka and symptoms manifested when they

returned to their home districts. Although virus was transmitted to other districts

through infected patients, it might not continue its transmission due to the reduced

number of vectors and the reduced number of virus populations in the dry season

(Aaskov et al., 2006, Williams et al., 2010, Cummings et al., 2004). However, we

were unable to identify the origin of DF transmission in Bangladesh and its spreading

direction due to the lack of information on patient’s location, their movement,

demographics and mosquito control activity. This should be a priority for future DF

research in Bangladesh.


This is the first study to examine the spatiotemporal pattern of DF in

Bangladesh. It has clearly demonstrated the heterogeneity of DF risk at the district

level in Bangladesh and revealed the spatiotemporal pattern of DF across the

country. As cluster analysis could be an important tool for decision makers to

prioritize areas where more surveillance and disease prevention efforts are required,

our findings can be useful for the Bangladesh health authority to further improve DF

control and prevention strategies. Additionally, the method developed in this study

may have wider applications in the field of disease surveillance and risk

management.

Our study has three key limitations. Firstly, in our analysis, we used reported

cases aggregated at the district level, which prohibits analysis at a higher spatial

resolution and may lead to important local clusters being missed. Secondly, there is

likely variation in the quality of the DGHS dengue surveillance data. Underreporting

is possible in the DGHS data when people infected by DF had subclinical infection

and did not seek for medical attention. Finally, we only identified potential DF

clusters in this preliminary study, but did not explore possible risk factors associated

with clustering. However, our future research will focus on the investigation of

various climatic, ecological, and socio-demographic determinants of DF in clustered

areas.

4.6 CONCLUSION

In summary, this study demonstrates that DF transmission in Bangladesh was

clustered in different spatial and temporal settings and the geographic distribution of

DF appears to have contracted over recent years. The impact of socio-demographic

changes and climatic factors on DF transmission in clustered areas remains to be

determined. Our findings can be useful for the Bangladesh health authority to further

improve DF control and prevention strategies. Additionally, the cluster methods

developed in this study may have wider applications in the field of disease

surveillance and risk management.


Acknowledgments

We thank to Director General of Health Services, Dhaka and Bangladesh

Bureau of Statistics for providing DF case record and census data respectively.

Financial support

The work was supported by QUT postgraduate scholarships (SB, YG),

NMHRC postdoctoral training fellowship (WH) and NMHRC research fellowship

(ST).




References

AASKOV, J., BUZACOTT, K., THU, H. M., LOWRY, K. & HOLMES, E. C. 2006. Long-term transmission of defective RNA viruses in humans and Aedes mosquitoes. Science, 311, 236-238.

ALI, M., WAGATSUMA, Y., EMCH, M. & BREIMAN, R. F. 2003. Use of a geographic information system for defining spatial risk for dengue transmission in Bangladesh: role for Aedes albopictus in an urban outbreak. American Journal of Tropical Medicine and Hygiene, 69, 634-640.


CHEPESIUK, R. 2003. Mosquito mismanagement? Environmental Health Perspectives, 111, A636.

CUMMINGS, D. A. T., IAMSIRITHAWORN, S., LESSLER, J. T., MCDERMOTT, A., PRASANTHONG, R., NISALAK, A., JARMAN, R. G., BURKE, D. S. & GIBBONS, R. V. 2009. The impact of the demographic transition on dengue in Thailand: insights from a statistical analysis and mathematical modeling. Plos Medicine, 6, e1000139.

CUMMINGS, D. A. T., IRIZARRY, R. A., HUANG, N. E., ENDY, T. P., NISALAK, A., UNGCHUSAK, K. & BURKE, D. S. 2004. Travelling waves in the occurrence of dengue haemorrhagic fever in Thailand. Nature, 427, 344-347.

DGHS. 2000. National guidelines for clinical management of dengue syndrome. Available: http://www.sdnbd.org/sdi/issues/health/dengue/other/dng.PDF [Accessed March 26 2009].

FERDAUSI, S. A. & BOLKLAND, M. W. Rainwater harvesting for application in rural Bangladesh. 26th WEDC conference, water, sanitation and hygiene : challenges of the millennium, 2000 Dhaka, Bangladesh. 16-19.

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, 100-103.

http://www.sdnbd.org/sdi/issues/health/dengue/other/dng.PDF


HOSSAIN, M. A., KHATUN, M., ARJUMAND, F., NISALUK, A. & BREIMAN, R. F. 2003. Serologic evidence of dengue infection before onset of epidemic, Bangladesh. Emerging Infectious Diseases, 9, 1411-1414.

HU, W., CLEMENTS, A., WILLIAMS, G. & TONG, S. 2011. Spatial analysis of notified dengue fever infections. Epidemiology and Infection, 139, 391-399.



KULLDORF, M., ATHAS, W., FEUER, E., MILLER, B. & KEY, C. 1998. Evaluating cluster alarms:a space-time scan statistic and brain cancer in Los Alamos, New Mexico. American Journal of Public Health, 88, 1377-1380.

MAMMEN, M. P., PIMGATE, C., KOENRAADT, C. J. M., ROTHMAN, A. L., ALDSTADT, J., NISALAK, A., JARMAN, R. G., JONES, J. W., SRIKIATKHACHORN, A., YPIL-BUTAC, C. A., GETIS, A., THAMMAPALO, S., MORRISON, A. C., LIBRATY, D. H., GREEN, S. & SCOTT, T. W. 2008. Spatial and temporal clustering of dengue virus transmission in Thai villages. Plos Medicine, 5, e205.

PODDER, G., BREIMAN, R. F., AZIM, T., THU, H. M., VELATHANTHIRI, N., KYMLOWRY, L. Q. M. & AASKOV, J. G. 2006. Short report: origin of dengue type-3 viruses associated with the dengue outbreak in Dhaka, Bangladesh,in 2000 and 2001. American Journal of Tropical Medicine and Hygiene, 74, 263-265.

RAHMAN, M., RAHMAN, K., SIDDQUE, A. K., SHOMA, S., KAMAL, A. H., ALI, K. S., NISALUK, A. & BREIMAN, R. F. 2002. First outbreak of dengue hemorrhagic fever, Bangladesh. Emerging Infectious Diseases, 8, 738-740.

RIGAU-PEREZ, J. G., CLARK, G. G., GUBLER, D. J., REITER, P., SANDERS, E. J. & VORNDAM, A. V. 1998. Dengue and dengue haemorrhagic fever. Lancet, 352, 971-977.




THAI, K. T. D., NAGELKERKE, N., PHUONG, H. L., NGA, T. T. T., GIAO, P. T., HUNG, L. Q., BINH, T. Q., NAM, N. V. & DE VRIES, P. J. 2010. Geographical heterogeneity of dengue transmission in two villages in southern Vietnam. Epidemiology and Infection, 138, 585-591.


WHO 2000. Dengue Hemorrhagic Fever:Diagnosis,Treatment ,Prevention and Control, WHO,Geneva, Switzerland.

WILLIAMS, C. R., BADER, C. A., KEARNEY, M. R., RITCHIE, S. A. & RUSSELL, R. C. 2010. The extinction of dengue through natural vulnerability of its vectors. PLoS Neglected Tropical Diseases, 4, e922.




Projecting the impact of climate change on dengue transmission in Dhaka, Bangladesh.

This paper has been accepted for publication in Environment International.


Authors: Shahera Banu1, Wenbiao Hu2, Yuming Guo2, Cameron Hurst3 and

Shilu Tong1

Affiliations:

1. School of Public Health and Social Work, Queensland University of

Technology, Brisbane, Australia.

2. School of Population Health, University of Queensland, Brisbane, Australia.

3. Clinical Epidemiology Unit, Khon Kaen University, Khon Kaen, Thailand.


Shahera Banu performed all data analyses and wrote the manuscript. Shilu


Cameron Hurst contributed to statistical support and Yuming Guo helped with the R

software and related packages.


5.1 ABSTRACT

Weather variables, mainly temperature and humidity influence vectors, viruses,

human biology, ecology and consequently the intensity and distribution of the

vector-borne diseases. There is evidence that warmer temperature due to climate

change will influence the dengue fever (DF) transmission. However, long term

scenario-based projections are yet to be developed. Here, we assessed the impact of

weather variability on DF transmission in a megacity of Dhaka, Bangladesh and

projected the future DF risk attributable to climate change. Our results show that

weather variables particularly temperature and humidity were positively associated

with DF transmission. The effects of weather variables were observed at a lag of four

months. We projected that assuming a temperature increase of 3.3oC without any

adaptation measure and changes in socio-economic condition, there will be a

projected increase of 16,030 DF cases in Dhaka by the end of this century. This

information might be helpful for the public health authorities to prepare for the likely

increase of DF due to climate change. The modelling framework used in this study

may be applicable to DF projection in other cities.


5.2 INTRODUCTION

According to the Intergovernmental Panel on Climate Change (IPCC), the

global temperature increased significantly over the 20th century (IPCC, 2007). Recent

trends in anthropogenic emissions and their modelled impacts of global climate

strongly suggest that both emissions and warming trends will continue to affect the

atmospheric process in the 21st century. It has been predicted that the global mean

temperature will increase by 1.1-6.4 oC by the end of this century (IPCC, 2007).

Annual average temperature for the South Asia region has been projected to rise by

3.3oC (range, 2 - 4.4 oC) in 2100, with summer temperature increases of 2.7 oC

(IPCC, 2007). A growing body of literature suggests that warmer temperatures will

enhance the transmission rate for mosquito-borne disease and will widen its

geographical distributions (McMichael et al., 2006, Hales et al., 2002, Kan et al.,

2012, Jetten and Focks, 1997).

Dengue fever (DF) is one of the most important mosquito-borne disease of

humans, and has emerged as a global public health concern throughout the tropical

and subtropical regions of the world (Gubler, 1998). DF transmission in these areas

typically follows a seasonal pattern which reflects the influence of weather on the

transmission cycle (Johansson et al., 2009b). DF is weather sensitive due to its

mosquito vector, which requires standing water to breed and warm ambient

temperature for larval development and virus replication (Patz et al., 1998, Hopp and

Foley, 2001, Banu et al., 2011). The incidence of DF has increased significantly in

last 35 years and various factors including urbanization, globalization and climate

change are thought to be the major contributors (Hopp and Foley, 2001, Gubler,

2011).


Several recent studies have demonstrated an association between weather

variability and DF (Chen and Hsieh, 2012, Hii et al., 2009, Wu et al., 2007, Hu et al.,

2012). Temperature, rainfall and humidity were found to be associated with DF

transmission (Wu et al., 2007, Bangs et al., 2006, Ram et al., 1998, Karim et al.,

2012). However, the magnitude of the association between weather and DF varied

with geographical location and socio-environmental conditions (Thammapalo et al.,

2007, Arcari et al., 2007). It is also evident that El Niño events have strongly

associated with DF epidemics, although spatial heterogeneity exists in this relation

(Cazelles et al., 2005, Hu et al., 2010). Mathematical modelling has recently been

used to measure and predict the impact of weather variation on DF and significant

advances have been achieved in modelling approaches (McMichael et al., 2006, Hu

et al., 2010). Many studies around the world have developed different models to

predict the future distribution of DF in response to climate change (Hales et al., 2002,

Hopp and Foley, 2003, Patz et al., 1998). Such projections can help to combat the

increased risk of DF due to climate change by taking necessary adaptation measures.

However, very few studies were conducted to identify the association between

weather variables and DF transmission in the South Asian region and long term

scenario-based projections are yet to be developed (Banu et al., 2011, Karim et al.,

2012, Oo et al., 2011, Chakravarti and Kumaria, 2005). In this study, we examined

the effects of weather variability on DF transmission and projected the potential

impact of climate change on the pattern of DF in the megacity of Dhaka.


5.3 MATERIALS AND METHODS

5.3.1 Study area

This study carried out in Dhaka, the capital of Bangladesh. Our previous study

showed that Dhaka is the highest risk area for DF transmission in Bangladesh and the

underlying cause of increased risk of DF in this location remains unknown, which

requires further investigation (Banu et al., 2012). Dhaka is located in central

Bangladesh at 23°42′ north latitude and 90°22′ east longitude with an area of

1,464 square kilometres. Dhaka along with its metropolitan area had a population of

11.8 million (2011 census), making it the biggest city in Bangladesh. Dhaka has a

hot, wet and humid tropical climate. The city is within the monsoon climate zone,

with an annual average temperature of 25 °C and monthly means varying between 18

°C in January and 29 °C in August. Nearly 80% of the annual average rainfall of

1,854 millimetres occurs between May and September.


Data on the monthly number of notified DF cases in Dhaka city were obtained

from the Directorate General of Health Services (DGHS) from January 2000 to

December 2010. As DF is a notifiable disease in Bangladesh, any case detected

based on the World Health Organization (WHO) clinical criteria must report to the

DGHS by the hospital. According to the WHO clinical criteria, a DF case was

defined by the presence of acute fever accompanied by any two of the following

clinical symptoms such as headache, myalgia, arthralgia, rash, positive tourniquet

test and leucopenia (WHO, 2000). We also obtained monthly weather data on

maximum, mean and minimum temperature, relative humidity and rainfall from

Bangladesh Meteorological Department (Dhaka, Bangladesh) between January 2000

and December 2010. We used monthly total number of DF cases, monthly mean

maximum temperature, mean minimum temperature and total rainfall for data

analyses. Population data were collected from Bangladesh Bureau of Statistics

(BBS).

.


5.3.3 Data analysis

We used Spearman’s correlation coefficients to summarize the relationships

between independent variables. A Poisson time series model combined with

distributed lag model (DLM) was used to estimate the effects of weather on DF

transmission. The observed number of DF cases followed a quasi-Poisson

distribution and the model allows for over dispersion.

𝑌𝑡 = Poisson(𝜇𝑡), t=1,.........., n

log(𝜇𝑡) = 𝛼 + �𝛽0�𝑇𝑡,𝑙�𝐿

𝑙=1

+ �𝛽1�𝐻𝑡,𝑙�𝐿

𝑙=1

+ �𝛽2�𝑅𝑡,𝑙�𝐿

𝑙=1

+ log(𝑁𝑡) + 𝑠(𝑡, 𝜆) + Ɛ𝑡

Where t is the month of the observation; Yt is the observed monthly DF counts

in month t; α is the intercept; Tt,l ,Ht,l and Rt,l are the matrices obtained by applying

the DLM to temperature, humidity and rainfall, respectively; l is the lag months; L is

the maximum lag; β0, β1 and β2 are the coefficients for Tt,l ,Ht,l and Rt,l, respectively,

Nt is an offset to control for population using a linear function of time based on the

2001 and 2011 census. The s(t,λ) is the natural cubic spline smoothing function of

time with assigned λ of 2 degrees of freedom per year to control for seasonal pattern.

We used a DLM that modelled the main effects as a linear function and the

delayed effects as a polynomial function. The selection of maximum lag was

conducted using model residual checking and we checked maximum lag up to 6

months. We used second order quadratic polynomial smoothing for the lag. The

mean value of each weather variable was used as a baseline (referring value) to

measure the relative risks. We plotted relative risks against weather variables and

lags to show the entire relationship between weather conditions and DF.


The climate and DF relationship were examined using different temperature

measures (maximum, mean and minimum temperature) in the DLM. The deviance

was used to choose the best model. Model including maximum temperature was

associated with lower deviance value (Table 5.3). We also compared the deviance for

the association between each weather variables and DF using DLM. The deviance

was also lower compare to other models when maximum temperature and relative

humidity were included (Table 5.4). The goodness-of-fit was performed to check the

model adequacy using auto-correlation functions of residuals and normality of the

residuals. Figure 5.1 shows that there was no significant auto-correlation between

residuals at different lags in the DLM when maximum temperature and humidity

were used as predictor variables. The scatter plot shows that the residuals in the

model fluctuated randomly around zero with no obvious trend. Thus the goodness-

of-fit analyses show that the model fits the data reasonably well. Therefore, we

selected the model including maximum temperature and relative humidity as the best

model and used it to estimate the effects of weather variation on DF transmission

The constructed model was then validated by dividing the data file into two

data sets. The data between January 2000 and December 2008 were used to develop

a DLM and those between January 2009 and December 2010 were used to validate

the model. The validation indicates that the model had reasonable accuracy as the

observed and predicted values were mostly coincided (Figure 5.3). In addition,

adequacy of the model predicting outbreak (≥ 168) was evaluated by sensitivity

analyses. For sensitivity analyses, the monthly number of DF cases was transformed

into a categorical variable (i.e., outbreak/non-outbreak). Then the sensitivity or true

positivity rate (predicted number of months with DF outbreak/observed number of

months with DF outbreak) and specificity or true negativity rate (predicted number

of months without DF outbreak/observed number of months without DF outbreak) of

the predictive model were calculated.


The results of the validated model were then applied to future climate change

situations to generate projections for DF risk in the 2100. We used IPCC regional

climate change projection for South Asia, which results in an increase of 3.3 oC in

annual mean temperature between 1980 to 1999 and 2080 to 2099 (IPCC, 2007). We

assumed that warming will be similar to south Asia in Dhaka. We estimated the

future monthly temperature in Dhaka by combining recorded baseline data with

projection. We added 1, 2 and 3.3 oC to the observed monthly temperature in 2010 to

simulate monthly temperatures in 2100. We assumed that there will be no adaptation

to climate change in Dhaka. We calculated the projected temperature related DF risk

in 2100 after adjusting for the 1.3% increase in population, which is the current

population growth rate in Dhaka (population census 2011).

All statistical tests were two-sided and the p <0.01 were considered statistically

significant. We used R software (version 2.12.0; R development Core Team 2009) to

fit all models, with its “dlnm” package to create the DLM (Gasparrini and

Armstrong, 2011).


5.4 RESULTS

There were 25,059 DF cases in the Dhaka during the study period. The average

monthly number of DF cases was 168 with an incidence rate of 1.8 per 100,000

populations. Descriptive statistics for each independent and dependent variable are

shown in Table 5.1. The monthly mean minimum and maximum temperature,

rainfall and relative humidity were 21.9oC, 30.7oC, 180.2 mm and 72.77%,

respectively, between 2000 and 2010 in Dhaka. Spearman correlation coefficients

between climatic variables show that all climatic variables were strongly correlated

to each other except the correlation between maximum temperature and relative

humidity (Table 5.2).

The three dimensional plots show the entire relationship between mean

maximum temperature and relative humidity with DF incidence at different lags

(Figure 5.2). The estimated effects of weather variables on DF incidence were linear

in current months and were nonlinear along lags. Temperature and humidity were

positively associated with DF incidence and the highest effects observed at two

months lag. The sensitivity analyses indicate that the overall model agreement was

89%, sensitivity was 84% and specificity was 91% (Table 5.5).

Table 5.6 reveals the estimated DF cases associated with the variation in

temperature due to climate change by the year 2100. We estimated 377 DF cases

attributable to temperature variation in 2010. Assuming a 1oC temperature increase

in 2100, we projected an increase of 583 DF cases. For a 2 oC increase, we projected

an increase of 2,782 DF cases. If temperature increase by 3.3 o C as the IPCC

projected, the consequence will be devastating, with a projected increase of 16,030

cases by the end of this century.


Table 5.1 Descriptive statistics of monthly climatic conditions and DF in Dhaka, Bangladesh, 2000-

2010.

Table 5.2 Spearman’s correlation coefficients between monthly climatic variations in Dhaka,

Bangladesh.

Climatic variables Minimum

Temperature Mean Temperature

Maximum Temperature

Rainfall

Mean Temperature 0.97**

Maximum Temperature 0.72** 0.84**

Rainfall 0.80** 0.73** 0.51**

Humidity 0.62** 0.53** 0.14 0.73**

** P < 0.01

Variables N Mean Standard

Deviation

Minimum Maximum

DF 132 168 394 0 3155

Minimum Temperature (oC) 132 21.9 4.3 11.7 26.8

Mean Temperature (oC) 132 26.3 3.5 16.7 30.7

Maximum Temperature (oC) 132 30.7 2.9 21.7 35.5

Rainfall (mm) 132 180.2 195.1 0 839

Relative Humidity (%) 132 72.8 8 53 85


Table 5.3 Deviances for the relationship between DF and different temperature measures by DLM

Temperature measure Deviance for DLM

Maximum Temperature 279275.9

Mean Temperature 303311.6

Minimum Temperature 319501.4

Table 5.4 Deviances for DLM using different covariates

DLM Models Deviance

Temperature and Humidity 279275.9

Humidity and Rainfall 327820.9

Temperature and Rainfall 352719.7


Table 5.5: Sensitivity and specificity of DLM for dengue occurrence.

Predicted Observed

Total Outbreak Non-outbreak

Outbreak 27 9 36

Non-outbreak 5 91 96

Total 32 100 132

Sensitivity, 27/32 = 84%; Specificity, 91/100 = 91%; Crude agreement or accuracy, (27+91)/132 =

89%.


Figure 5.1 Auto-correlation function partial auto-correlation function and scatter plot of residuals for DLM.


Figure 5.2 Association between climatic variables (maximum temperature and relative humidity) and DF at different lags.


Figure 5.3 Validated distributed lag model of climate variation in Dhaka, Bangladesh (validation period = Jan 2009 – Dec 2010 i.e., the cross validation period).


Table 5.6 Changes in annual DF incidence under different scenarios of temperature increase by 2100 in Dhaka, Bangladesh.

Climate change scenarios Projected annual DF incidence

Changes in annual DF incidence

Baseline 377

10C increase 960 583

2 0C increase 3,159 2,782

3.3 0C increase 16,407 16,030


5.5 DISCUSSION

Our results show that the monthly temperature and humidity were significantly

associated with the monthly DF incidence in Dhaka, with highest lag effects of two

months. These results are consistent with findings of other studies and may assist to

forecast DF outbreaks in different regions (Hii et al., 2009, Hsieh and Chen, 2009,

Descloux et al., 2012, Johansson et al., 2009b). Temperature and humidity are the

most important weather factors in the growth and dispersion of mosquito vector and

potential predictors of DF outbreaks (Wu et al., 2007, Chen et al., 2010).

Temperature influences the life cycle of Aedes mosquitoes including growth rate and

larval survival and the length of reproductive cycle (Patz et al., 2005, Hopp and

Foley, 2001). Maximum mosquito survival rate of 88-93% were observed between

temperature ranges of 20-30 oC (Tun-Lin et al., 2000). Temperature also affects the

virus replication, maturation and period of infectivity. Higher temperature decreases

the length of viral incubation within the vector, and thus increases the chance of

mosquitoes to become infective in their life span (Patz et al., 1998, Yang et al.,

2009a, Hopp and Foley, 2001). Adult mosquito survival also depends on humidity

(Patz et al., 1998, Hopp and Foley, 2001). Given the relationship between

temperature and DF, the projected change in temperature due to climate change may

exacerbate disease transmission in Dhaka. According to the IPCC, the annual mean

temperature increase will be 3.3 oC by the end of the 21st century in Dhaka. The

projected warming will occur both in summer and winter (IPCC, 2007). As summer

will be warmer than before, it is likely that warmer condition will enhance disease

transmission and will increase DF incidence. In previous years, there were few

reported DF cases in Dhaka during winter season. If the winter temperature increases

as projected, it may become more favourable for DF transmission and extend the

outbreak season. Hence DF outbreak may become more intense in future, if the

climate change happens.


There has been a significant emergence of DF in Dhaka during the last decade;

the reasons for this can be multiple. Both climatic and non-climatic factors like

socio-ecological changes, viral serotypes, herd immunity and mosquito control can

influence the risk of DF transmission (Gubler, 2011). Rapid urbanization around

Dhaka city can deteriorate the environmental condition and increase the DF

incidence through enhancing the mosquito breeding habitats. Increased air travel can

facilitate the introduction of new dengue serotypes from neighbouring endemic

countries and can make this region hyper endemic (presence of all four dengue virus

serotypes), which will obviously increase the likelihood of DF epidemic (Tatem et

al., 2006, Karim et al., 2012). Additionally, effective vector control can reduce the

vector density and can decrease the DF transmission.

Temperature and humidity affect the DF occurrence in several subsequent

months. We found that monthly maximum temperature and relative humidity were

associated with DF transmission through a 4-months lag period (highest effects in

two months) which includes the time of replication and development of mosquito

and the incubation period of the virus (time of replication both in vector and host).

Therefore, observed lag effects were biologically plausible and consistent with the

findings of other studies (Wu et al., 2007, Arcari et al., 2007, Hii et al., 2012a). A

previous study in Dhaka reported the positive association between maximum

temperature, relative humidity and DF which is consistent with our findings (Karim

et al., 2012). They also observed the highest lag effects at two months which is

similar to our observation. An accurate early warning system to predict DF

epidemics and enhance the effectiveness of preventive measures largely relies on the

sufficient lag time. Thus, four months lag time could be sufficient to warn people

about the possible disease outbreak and take necessary measures to prevent the

epidemic.


Different emissions scenarios were developed by IPCC which have been

widely used in the analysis of possible climate change impacts and options to

mitigate climate change. Each emission scenario represents different demographic,

social, economic, technological and environmental developments which are driving

forces of greenhouse gas emissions. “The A1 scenario family describes a future

world of rapid economic growth, global population that peaks at mid-century and

declines thereafter, and the rapid introduction of new and more efficient

technologies”(IPCC, 2000). The A1B emissions scenarios is one of the A1 group

scenarios which assumes the balanced use of energy system like fossil fuel and non-

fossil energy system. The B2 scenario families focus on local and regional

environmental protection and social equity where global population will increase

continuously with comparatively lower rate with intermediate level of economic

development and less rapid and more diverse technological developments than A1 or

B1. Climate change projections for all continents and sub continental regions of the

world were provided by IPCC (IPCC, 2007). These projections were generated using

multi model data set (MMD) and three emission scenarios B1, A1B and A2.

However, the results of most projections were presented and discussed by IPCC on

the basis of A1B scenario as the global mean surface temperature responses in the

ensemble mean of the MMD model follows a ratio of 0.69:1:1.7 for B1: A1B:A2

scenarios. The local temperature responses for almost all regions also follow the

same ratio. Similar to the IPCC regional climate projections, we used the MMD-A1B

projection scenario to predict and discuss the future temperature related DF risk in

Dhaka.

To the best of our knowledge, this is the first study to project the impact of

climate change on DF transmission in Dhaka. We showed that DF incidence will

increase by more than 40 times in Dhaka in the year 2100 relative to 2010, if the

ambient temperatures increase by 3.3 oC according to the IPCC regional climate

projection. It will have devastating consequences for the already stretched public

health systems in Dhaka due to the population ageing and increased burden of

disease (including chronic disease, infectious disease and injury). Human adaptation

to climate change may influence the likelihood of DF transmission. People may

adapt to higher temperatures through improved building design with glazed


windows, piped water, insect screening and air-conditioning. These facilities may

effectively reduce their contacts with vector mosquitoes and even if infected

mosquitoes gain entry to these buildings, the low ambient temperature and artificially

dry environment may decrease their survival rate and reduce the risk of disease

transmission (Reiter, 2001). On the other hand, water storing for domestic purposes

in summer months or during droughts may provide increase number of breeding sites

for mosquitoes and increase the risk of DF transmission (Beebe et al., 2009).

However, there is no information available on how people will adapt to climate

change in Dhaka. Therefore, in our study, we assumed that there will be little

adaptation to climate change in the study site.

The weaknesses of this study must be acknowledged. This is a large scale,

ecologic assessment of the relationship between climate and the DF transmission at a

city level. For a comprehensive and systematic risk assessment, more detailed risk

assessment at a community and individual level is required. Inclusion of other factors

such as mosquito density, population immunity, viral factors and human behaviours

may improve the model. Due to the lack of seroprevalance and entomological data,

these variables could not be included into our model. Adaptation to climate change

and changes in socio-economic trends might influence the likelihood of disease

occurrence. However, we have not accounted for all possible socio-economic

features and climate adaptation behaviour. Underreporting bias is inevitable in the

surveillance data to some extent as people infected with subclinical DF infection did

not seek for medical advice. This model is only applicable to Dhaka and areas with a

similar socio-ecologic background as local data were used in the construction of this

model.


5.6 CONCLUSION

This study shows that maximum temperature and relative humidity were best

predictors among the major determinants of DF transmission in Dhaka for the period

of 2000-2010. Projected climate change will increase mosquito activity and DF

transmission in this area. Assuming a temperature increase of 3.3 oC by 2100 as

projected by IPCC, there would be a substantial increase in DF incidence in Dhaka.

Therefore, public health authorities need to be well prepared for likely increases of

DF transmission in this region.

Acknowledgments

We thank to Director General of Health Services, Dhaka and Bangladesh

Bureau of Statistics for providing DF case record and census data respectively. We

would also like to thank Bangladesh Meteorological Department, Dhaka for

providing climate data. We thank to Associate Professor Adrian Barnett and Dr.

Weiwei Yu for their valuable comments on the manuscript.

Financial support

The work was supported by QUT postgraduate scholarships (SB), NMHRC

postdoctoral training fellowship (WH) and NMHRC research fellowship (ST).




References


BANGS, M. J., LARASATI, R. P., CORWIN, A. L., WURYADI, S. & JAKARTA, I. 2006. Climatic factors associated with epidemic dengue in Palembang, Indonesia: implications of short-term meteorological events on virus transmission. Southeast Asian Journal of Tropical Medicine and Public Health, 37, 1103-1116.

BANU, S., HU, W., HURST, C., GUO, Y., ISLAM, M. Z. & TONG, S. 2012. Space-time clusters of dengue fever in Bangladesh. Tropical Medicine and International Health, 17, 1086-1091.


BEEBE, N. W., COOPER, R. D., MOTTRAM, P. & SWEENEY, A. W. 2009. Australia’s dengue risk driven by human adaptation to climate change. Plos Neglected Tropical Diseases, 3, e429.



CHEN, S. C. & HSIEH, M. H. 2012. Modeling the transmission dynamics of dengue fever: implications of temperature effects. Science of the Total Environment, 431, 385-391.

CHEN, S. C., LIAO, C. M., CHIO, C. P., CHOU, H. H., YOU, S. H. & CHENG, Y. H. 2010. Lagged temperature effect with mosquito transmission potential explains dengue variability in southern Taiwan: Insights from a statistical analysis. Science of the Total Environment, 408, 4069-4075.

DESCLOUX, E., MANGEAS, M., MENKES, C. E., LENGAIGNE, M., LEROY, A., TEHEI, T., GUILLAUMOT, L., TEURLAI, M., GOURINAT, A. C., BENZLER, J., PFANNSTIEL, A., GRANGEON, J. P., DEGALLIER, N. & DE LAMBALLERIE, X. 2012. Climate-based models for understanding and forecasting dengue epidemics. PLoS Neglected Tropical Diseases, 6, e1470.

GASPARRINI, A. & ARMSTRONG, B. 2011. Distributed lag non-linear model in R: the package dlnm. Available:


cran.r.project.org/web/packages/dlnm/vignettes/dlnmOverview.pdf [Accessed 20 November 2011].





HII, Y. L., ROCKLOV, J., WALL, S., NG, L. C., TANG, C. S. & NG, N. 2012. Optimal lead time for dengue forecast. PLoS Neglected Tropical Diseases, 6, e1848.

HOPP, J. M. & FOLEY, J. A. 2003. Worldwide fluctuations in dengue fever cases related to climate variability. Climate Research, 25, 85-94.


HSIEH, Y. H. & CHEN, C. W. S. 2009. Turning points, reproduction number and impact of climatological events for multi-wave dengue outbreaks. Tropical Medicine & International Health, 14, 628-638.



IPCC. 2000. IPCC Special report: emissions scenarios. Available: https://www.ipcc.ch/pdf/special-reports/spm/sres-en.pdf [Accessed 6/9/2013].


JETTEN, T. H. & FOCKS, D. A. 1997. Potential changes in the distribution of dengue transmission under climate warming. American journal of tropical medicine and hygiene, 57, 285-297.

http://www.ipcc.ch/pdf/special-reports/spm/sres-en.pdf


JOHANSSON, M. A., DOMINICI, F. & GLASS, G. E. 2009. Local and global effects of climate on dengue transmission in Puerto Rico. PLoS Neglected Tropical Diseases, 3, e382.

KAN, H., CHEN, R. & TONG, S. 2012. Ambient air pollution, climate change, and population health in China. Environment International, 42, 10-19.

KARIM, M. N., MUNSHI, S. U., ANWAR, N. & ALAM, M. S. 2012. Climatic factors influencing dengue cases in Dhaka city: a model for dengue prediction. Indian Journal of Medical Research, 136, 32-39.


OO, T. T., STORCH, V., MADON, M. B. & BECKER, N. 2011. Factors influencing the seasonal abundance of Aedes (Stegomyia) aegypti and the control strategy of dengue and dengue haemorrhagic fever in Thanlyin Township, Yangon City, Myanmar. Tropical Biomedicine, 28, 302-311.



RAM, S., KHURANA, S., KAUSHAL, V., GUPTA, R. & KHURANA, S. B. 1998. Incidence of dengue fever in relation to climatic factors in Ludhiana, Punjab. Indian Journal of Medical Research, 108, 128-133.


TATEM, A. J., HAY, S. I. & ROGERS, D. J. 2006. Global traffic and disease vector dispersal. Proc Natl Acad Sci U S A, 103, 6242-6247.

THAMMAPALO, S., CHONGSUVIVATONG, V., GEATER, A. & DUERAVEE, M. 2007. Environmental factors and incidence of dengue fever and dengue haemorrhagic fever in an urban area, Southern Thailand. Epidemiol. Infect., 136, 135 -143.

TUN-LIN, W., BURKOT, T. R. & KAY, B. H. 2000. Effects of temperature and larval diet on development rates and survival of the dengue vector Aedes aegypti in north Queensland, Australia. Medical and Veterinary Entomology, 14, 31-37.

WHO 2000. Dengue Hemorrhagic Fever: Diagnosis,Treatment , Prevention and Control, WHO, Geneva, Switzerland.



YANG, H. M., MACORIS, M. L., GALVANI, K. C., ANDRIGHETTI, M. T. & WANDERLEY, D. M. 2009. Assessing the effects of temperature on the population of Aedes aegypti, the vector of dengue. Epidemiol Infect, 137, 1188-1202.

110 Chapter 6: General Discussion

Chapter 6: General Discussion

Chapters 3-5 present the main findings, as well as discussion and conclusions

of each study in turn. The connection and major features of three results papers are

shown in Figure 6.1. This Chapter discusses the key findings of this thesis and the

overall strengths and limitations. This chapter also makes recommendations for

future research directions and then draws conclusions.

6.1 SUBSTANTIVE DISCUSSION

This thesis investigated the spatiotemporal pattern of DF in the Asia-Pacific

region and projected the future risk of DF transmission under different climate

change scenarios in Dhaka, Bangladesh. Although a number of previous studies

identified the association between climate variation and DF transmission (Halide and

Ridd, 2008, Thammapalo et al., 2005a, Chen and Hsieh, 2012, Hii et al., 2009, Wu et

al., 2007, Hu et al., 2012), knowledge on long term impacts of climate change on DF

is limited, particularly in the Asia-Pacific region. There is a need to better understand

the spatiotemporal pattern of DF and the potential impact of climate change on DF

transmission in this region.

I examined the dynamic spatiotemporal pattern of DF transmission in the Asia-

Pacific region over the fifty years from 1955 to 2004 using spatial analysis

techniques (Chapter 3). Fifty years data of DF outbreaks for sixteen countries of this

region were used in this study. The results of the spatiotemporal analysis indicate

that there was a remarkable variation in the spatial distribution of DF in the Asia-

Pacific region. The geographic range of DF was expanded over the last 50 years and

the disease was clustered in high risk areas. Socio-ecological changes occurred in

this region during the 20th century are thought to be responsible for this geographic

expansion (Gubler, 2011, Tatem et al., 2006). World War II might have influenced

the DF transmission in this region through ecological disruption and troop’s

Chapter 6: General Discussion 111

movement. The movement of troop and war materials helped the dissemination of

Aedes mosquitoes and the virus between countries (Gubler, 2002a). Besides this,

massive urbanization, rapid economic growth, increased air travel and climate

change may also contribute to the increased DF transmission and its expansion in the

Asia-Pacific region (Arunachalam et al., 2010, Gubler, 2004b, Schmidt et al., 2011,

Wu et al., 2009, Wilder-Smith and Gubler, 2008). Further research is required to

confirm the contribution of these factors to the pattern of DF transmission.

Thailand, Vietnam and Laos were identified as the highest risk countries for

DF in the Asia-Pacific region and there was a southward expansion in the geographic

distribution, which might originate from Philippines and Thailand. Historically, the

first severe DF outbreak occurred in Manila, Philippines in 1953 followed by

Bangkok, Thailand (Gubler, 2011, Guzman et al., 2010). A southward expansion in

DF transmission was also observed in Argentina and Australia (Woodruff et al.,

2006, Carbajo et al., 2012, Hu et al., 2011). Little is known about the causes of this

southward spread of DF. Global climate change was mentioned as one of the reasons

for this southward expansion (Hu et al., 2011). In the past 100 years, mean surface

temperature has increased by 0.3-0.8 oC across the continent (IPCC, 2007), which

could create climatic conditions suitable for dengue mosquito vector and facilitated

DF transmission in this region (Benitez, 2009, Kyle and Harris, 2008, Hales et al.,

2002).


Figure 6.1 Framework of research findings in this thesis

Results Paper 3 Results Paper 2 Results Paper 1

Spatiotemporal pattern of DF

High risk areas for DF

Asia-Pacific region

Most likely clusters

Secondary clusters

Increased incidence and increased number of affected countries

Transmission of DF

Bangladesh Dhaka

Spatiotemporal pattern of DF

High risk areas for DF

Most likely clusters

Dhaka

Decreased incidence and decreased number of affected districts

Predictive model

Projecting DF risk due to future climate change situations

Temperature Humidity

Increase DF incidence


As my literature review shows, there has been limited research conducted on

spatial epidemiology of DF in Bangladesh. I examined the trends and space-time

clustering of DF transmission in Bangladesh in Chapter 4. I used district level DF

data for the entire country during 2000-2009. Dhaka district was identified as the

most likely DF cluster in Bangladesh. Several districts of the southern part of

Bangladesh were identified as secondary clusters in the years 2000-2002. Even

though there was a substantial increase in DF transmission in the Asia-Pacific region,

an opposite trend was observed in Bangladesh. The prevalence of DF declined in

Bangladesh over the last decade. It has speculated that the effective management of

DF patients according to the national dengue guidelines and huge public awareness

after the first outbreak in 2000 might have helped to reduce the DF transmission

(Rahman et al., 2002). Besides vector control measures, socio-economic changes,

population immunity and changes in viral strains and fitness may also contribute to

the decreased DF incidence (Cummings et al., 2009, Podder et al., 2006). Further

research is needed to identify possible reasons for the declining trend of DF. Dhaka

was constantly identified as the highest risk cluster for DF transmission in

Bangladesh over the study period. Dhaka is the capital and the largest city in

Bangladesh with a high population density. The suitable environmental conditions

for mosquito breeding due to the high population density and the movement of

infected individuals and/or mosquitoes into this region from overseas through air

travel may directly and/or indirectly increase the chance of DF transmission

(Sutherst, 2004).

Recently, several studies examined the spatiotemporal clustering of DF

transmission. Focal nature of DF transmission was reported by previous studies

which was also observed in my study (Hu et al., 2011, Li et al., 2012, Mammen et

al., 2008, Honorio et al., 2009, Hu et al., 2012, Wen et al., 2006). Studies conducted

in Australia and China found that the geographic range of DF transmission expanded

significantly in both countries (Hu et al., 2011, Li et al., 2012). I noticed similar

pattern in DF transmission in the Asia-Pacific pacific region. Like Dhaka district, DF

transmission was clustered in highly urbanized areas in other countries such as China

(Li et al., 2012).


In order to examine the association between climatic factors and DF

transmission and to project the future risk of DF in Bangladesh based on different

climate change scenarios, I examined the effects of climate variability on DF

transmission and projected the future climate-related DF risk in Dhaka, Bangladesh

(Chapter 5). I selected Dhaka as the study location as it was identified as the highest

risk area for DF in Bangladesh in the previous chapter, but until now no projection

has been made to consider the impact of climate change on DF transmission in this

location.

The results suggest that temperature and relative humidity played crucial role

in the transmission of DF in Dhaka. Temperature influences the breeding, survival

and longevity of Aedes mosquitoes and the dengue virus replication and thereby

affects the transmission of DF (Patz et al., 2005, Hopp and Foley, 2001). Humidity

influences the incubation period of virus and biting activity of mosquitoes which

ultimately impact the likelihood of DF transmission (Patz et al., 1998). The results of

this study also show that these variables can be well used for forecasting outbreaks of

DF in Dhaka. An accurate early warning system to predict DF epidemics and

enhance the effectiveness of preventive measures largely relies on the sufficient lead

time. Therefore, it is important to understand the lag effects of climate on the DF

transmission. This study showed that monthly maximum temperature and relative

humidity affected the DF transmission through a lag of up to four months. Four

months are sufficient to warn the people about the possible disease outbreak and take

necessary measures to prevent the epidemic. Several studies have also documented

the optimal lead time to forecast DF ranged from 1to 5 months based on the climatic

condition (Hii et al., 2012a, Chen et al., 2010, Halide and Ridd, 2008).

Previous study in Dhaka reported a linear association between maximum

temperature, relative humidity and DF which is consistent with my findings (Karim

et al., 2012). They also observed the highest lag effects at two months which is

similar to my observation. However, Karim et al., (2012) did not account for the lag

effects of climatic factors as non-linear in their study. In my study, I found that the

lag effects of climatic variables on DF transmission are polynomial in nature.

Furthermore, I projected that DF incidence in Dhaka city will increase by more than


40 times in the year 2100 compare to 2010 if the ambient temperatures increase by

3.3oC according to the IPCC regional climate projections (IPCC, 2007). I projected

the temporal risk of DF due to climate change where some studies projected the

spatial risk of DF due to changes in climatic and socio-economic factors (Astrom et

al., 2013, Hales et al., 2002). Hales et al., (2002) showed that the current

geographical limit of DF transmission can be explained with 89% accuracy on the

basis of average vapour pressure (a measure of humidity) and about 50-60% of the

projected global population will be at risk of DF by 2085 if the climate change

continues. Besides climate change, socio-economic development will influence the

spatial distribution of DF(Astrom et al., 2013). If the global climate changes as

projected and gross domestic product per capita (GDPpc) remains constant, still

global population at risk of DF will increase substantially (Astrom et al., 2013).

Adaptation to climate change might influence the likelihood of DF transmission.

Since projections for adaptation to climate change and GDPpc in Dhaka were

unavailable, I was unable to account for all these changes in my model.

6.2 IMPLICATIONS OF THE RESEARCH FINDINGS

The findings of this study may have significant implications for the planning of

public health interventions and development of DF surveillance and control policies.

The GIS technique used in this study may help the surveillance of DF and other

infectious diseases through monitoring the disease pattern over different periods of

time across different places. This technique can also help to identify high risk

communities and/or population groups and could be an important tool for the

decision makers to prioritise surveillance and disease prevention efforts into the areas

where attention is most needed. Residents and travellers to the high risk areas should

also take additional precaution to protect them from DF infection. In this study, a

scenario based projection was developed for DF in Dhaka city, which suggests that

climate change will substantially increase the DF transmission risk in Bangladesh

where the public health systems are already stretched due to the population ageing

and increased burden of disease. Therefore, it is vital information for policy makers

who want to understand the potential effects climate change on infectious diseases,

and to set priorities for mitigation and adaptation.


6.3 STRENGTHS OF THIS THESIS

This thesis has three major strengths as illustrated below:

Firstly, to my knowledge this is the first study to examine the dynamic

spatiotemporal pattern of DF at a continental level. In disease control programmes,

several factors involved in the estimation of disease burden, monitoring disease

trend, identification of risk factors and planning and allocating of resources etc(Hu et

al., 2011). A common tool involved in this process is spatiotemporal trend analysis

which helps to highlight the changing disease patterns and to identify new risk

factors (Abrams and Kleinman, 2007). The risk of DF transmission varies in space

and time as mosquito density and longevity depend on a number of environmental

and ecological factors (e.g., temperature, precipitation and mosquito breeding

habitat) (Githeko, 2012). Therefore, it is important to identify the spatiotemporal

pattern of DF transmission and its determinants to predict the onset and severity of

epidemics.

Secondly, I explored high risk cluster areas for DF transmission in both the

Asia-Pacific region and Bangladesh using spatial scan statistics. Scan statistic is one

of the most commonly used approaches in spatial disease surveillance to explore

high risk areas (Kulldorff et al., 2005). Identification of high risk areas is essential

for improving disease prevention and control strategies in places where they are

mostly needed. Hence, it has wide applicability in surveillance and risk management

of DF.

Finally, I predicted the future risk of DF in Dhaka under different climate

change scenarios. This is the first attempt to make such projections in an Asian

country. There has been a need to develop an early warning system that can identify

and quantify the risk of DF outbreaks (Githeko, 2012). An early epidemic prediction

tool is critical for evaluating the risk of an outbreak that enables decision making for


public health interventions. Thus, early identification of disease risk allows early

interventions and preventions of an epidemic.

6.4 LIMITATIONS OF THIS THESIS

Several limitations of this research must also be acknowledged.

Firstly, I used country level DF data for spatial analyses in the Asia-Pacific

region, which is unable to show the risk at higher resolution. Higher resolution data

(monthly administrative level 1 data) could able to show the transmission pattern

within country settings and seasonal variation.

Secondly, underreporting is likely to occur for the DF outbreak data because

the dengue notification system does not include asymptomatic patients and those

who had clinical symptom but did not seek for medical advice. This underreporting

can cause information bias in my study and under estimate the actual risk of DF

transmission. The quality of DF notification data may also vary depending on time

and location due to the difference in disease diagnosis and surveillance system. Some

developing countries in the Asia-pacific region like Vietnam, Bangladesh use only

symptomatic assessment or serological test to confirm a DF case when developed

countries (Australia) use more specific molecular test to diagnose and report a DF

case. This variation in DF reporting might cause selection bias in our study.

Finally, this study is solely based on the available routinely collected data. We

could not take into account a number of environmental factors that could influence

the relationship between climate and DF transmission like urbanization, mosquito

species and density, population immunity. Detail data on these factors are

unavailable for Dhaka city. Finally, the projection of the climate impact on DF was

only made for one city which may restrict the generalisability of the results.


6.5 FUTURE RESEARCH DIRECTIONS

DF is one of the most important public health concerns in this century globally

(WHO, 2012a). The relationship between climate and DF is complex and further

research in this field is clearly warranted. This thesis has attempted to fill some

knowledge gaps but there are still many research questions should be addressed in

future research. Based on the major findings of this thesis, I would like to make the

following recommendations for future research:

6.5.1 Better understand the ecology of DF

It has been hypothesized that changes in socio-demographic features and

climate might be the possible reasons for the identified trends in DF transmission in

the Asia-Pacific region and in Bangladesh. However, how socio-economic factors

interact with climate change to influence the geographic distribution and dynamic

pattern of DF in this region remains unclear. Further research should focus on this

issue. It is also desirable to link meteorological data with long-term DF data in

different areas to examine the impact of climate and climate change on DF. If

possible, further data should be collected on important factors, such as mosquito

species and density, population immunity and mosquito control activities.

6.5.2 Advance the risk assessment techniques

Previous studies in the Asia-Pacific region and in this study, empirical

approach relying on statistical models were used to quantify the future burden of DF

due to climate change. Although, development of DF predictive model involves a

multi-disciplinary approach as climate itself cannot explain the total variation in the

disease trends (Githeko, 2012). In compare to statistical models, process-based

models, explicitly describe every aspect of disease transmission and its underlying

biological mechanisms (Bannister-Tyrrell et al., 2013). This type of models can

incorporate meteorologic information with virological, demographic and

epidemiologic data. Thus, the DF model may further improve by taking process

based approach.


6.5.3 Improve disease surveillance and monitoring

The quality of DF surveillance data varies with country and time. Some

countries do not have long-term consistent surveillance data. The current surveillance

should be strengthened to increase the accuracy of surveillance data. The likelihood

of under-reporting and over-reporting bias in the dataset should be examined

rigorously.

6.5.4 Evaluate the effectiveness of the public health interventions

The effectiveness of public health interventions can be improved by using

spatiotemporal analyses technique, which enables to identify and monitor of high

risk areas for disease transmission and targeting education campaign and vector

control at specific locations. Providing training to the medical practitioners and

public health professionals of the high risk areas can also be effective to control DF

outbreaks. Therefore, it is necessary to evaluate the effectiveness of public health

intervention for DF considering the spatiotemporal techniques.

6.5.5 Translate research into policy and risk management practice

Research to date mainly focused on the identification and quantification of the

effects of climate variation on DF, with less focus on how this information can help

to manage the future risk of DF transmission. This should be addressed urgently

specially for the vulnerable communities and population groups.


6.6 CONCLUSIONS

In this study, the spatiotemporal pattern and the high risk areas for DF

transmission were explored in the Asia-Pacific region. In addition, the space-time

clustering of DF transmission was examined in Bangladesh. Results show that the

geographic range of DF in the Asia-Pacific region increased significantly over the

study period. Thailand, Vietnam and Laos were identified as the highest risk areas

and there was a southward movement observed in the transmission pattern. In

Bangladesh, Dhaka was identified as the highest risk area for DF transmission.

Unlike the Asia-Pacific region, the geographic range of DF transmission in

Bangladesh reduced in recent years. The effects of climate variability on DF

transmission were examined and the future climate-related DF risks in Dhaka were

projected. The results show that climate variability (particularly maximum

temperature and relative humidity) was positively associated with DF transmission in

Dhaka. The delayed effects of climatic factors were observed at four months, which

may be applicable for warning people about the possible disease outbreak and take

necessary measures to prevent the epidemic. Assuming a temperature increase of 3.3 oC without any adaptation measures and constant socio-economic conditions by

2100, the consequence will be devastating, with a projected annual increase of

16,030 cases in Dhaka. This information might be helpful for improving DF

surveillance and control through effective vector management and community

education programs in Bangladesh and other countries of the Asia-Pacific region

with a similar situation.

Bibliography 121

Bibliography

AASKOV, J., BUZACOTT, K., THU, H. M., LOWRY, K. & HOLMES, E. C. 2006. Long-term transmission of defective RNA viruses in humans and Aedes mosquitoes. Science, 311, 236-238.

ABRAMS, A. M. & KLEINMAN, K. P. 2007. A SaTScan macro accessory for cartography (SMAC) package implemented with SAS software. International journal of health geographics, 6, 6.

ALI, M., WAGATSUMA, Y., EMCH, M. & BREIMAN, R. F. 2003. Use of a geographic information system for defining spatial risk for dengue transmission in Bangladesh: role for Aedes albopictus in an urban outbreak. American Journal of Tropical Medicine and Hygiene, 69, 634-640.

ANDERS, K. L., NGUYET, N. M., CHAU, N. V., HUNG, N. T., THUY, T. T., LIEN LE, B., FARRAR, J., WILLS, B., HIEN, T. T. & SIMMONS, C. P. 2011. Epidemiological factors associated with dengue shock syndrome and mortality in hospitalized dengue patients in Ho Chi Minh City, Vietnam. American Journal of Tropical Medicine and Hygiene, 84, 127-134.

ANSARI, M. & SHOPE, R. 1994. Epidemiology of arboviral infections. Public Health Review, 22, 1-26.


ARUNACHALAM, N., TANA, S., ESPINO, F., KITTAYAPONG, P., ABEYEWICKREME, W., WAI, K. T., TYAGI, B. K., KROEGER, A., SOMMERFELD, J. & PETZOLD, M. 2010. Eco-bio-social determinants of dengue vector breeding: a multicountry study in urban and periurban Asia. Bulletin of the World Health Organization, 88, 173-183.

ASHFORD, D. A. 2003. Outbreak of dengue fever in Palau, Western Pacific: risk factors for infection. American Journal of Troical Medicine and Hygiene, 69, 135-140.

ASTROM, C., ROCKLOV, J., HALES, S., BEGUIN, A., LOUIS, V. & SAUERBORN, R. 2013. Potential Distribution of dengue fever under scenarios of climate change and economic development. EcoHealth, 14, 14.

BANGS, M. J., LARASATI, R. P., CORWIN, A. L., WURYADI, S. & JAKARTA, I. 2006. Climatic factors associated with epidemic dengue in Palembang, Indonesia: implications of short-term meteorological events on virus

122 Bibliography

transmission. Southeast Asian Journal of Tropical Medicine and Public Health, 37, 1103-1116.

BANNISTER-TYRRELL, M., WILLIAMS, C., RITCHIE, S. A., RAU, G., LINDESAY, J., MERCER, G. & HARLEY, D. 2013. Weather-driven variation in dengue activity in Australia examined using a process-based modeling approach. American Journal of Tropical Medicine and Hygiene, 88, 65-72.

BANU, S., HU, W., HURST, C., GUO, Y., ISLAM, M. Z. & TONG, S. 2012. Space-time clusters of dengue fever in Bangladesh. Tropical Medicine and International Health, 17, 1086-1091.


BARBAZAN, P., GUISERIX, M., BOONYUAN, W., TUNTAPRASART, W., PONTIER, D. & GONZALEZ, J. P. 2010. Modelling the effect of temperature on transmission of dengue. Medical and Veterinary Entomology, 24, 66-73.

BARBAZAN, P., YOKSAN, S. & GONZALEZ, J. P. 2002. Dengue hemorrhagic fever epidemiology in Thailand: description and forecasting of epidemics. Microbes and Infection, 4, 699-705.

BEATTY, M. E., BEUTELS, P., MELTZER, M. I., SHEPARD, D. S., HOMBACH, J., HUTUBESSY, R., DESSIS, D., COUDEVILLE, L., DERVAUX, B., WICHMANN, O., MARGOLIS, H. S. & KURITSKY, J. N. 2011. Health economics of dengue: a systematic literature review and expert panel's assessment. American Journal of Tropical Medicine and Hygiene, 84, 473-488.


BENITEZ, M. A. 2009. Climate change could affect mosquito-borne diseases in Asia. Lancet, 373, 1070-1070.

BI, P., TONG, S., DONALD, K., PARTON, K. A. & HOBBS, J. 2001. Climate variability and the dengue outbreak in Townsville,Queensland,1992-93. Enviornmental Health, 1, 54-60.

BOHRA, A. & ANDRIANASOLO, H. 2001. Application of GIS in modeling of dengue risk based on sociocultural data: case of Jalore, Rajasthan, India. Dengue Bulletin, 25, 92-102.

Bibliography 123

BRADY, O. J., GETHING, P. W., BHATT, S., MESSINA, J. P., BROWNSTEIN, J. S., HOEN, A. G., MOYES, C. L., FARLOW, A. W., SCOTT, T. W. & HAY, S. I. 2012. Refining the global spatial limits of dengue virus transmission by evidence-based consensus. PLoS Neglected Tropical Diseases, 6, e1760.

CAMINADE, C., MEDLOCK, J. M., DUCHEYNE, E., MCINTYRE, K. M., LEACH, S., BAYLIS, M. & MORSE, A. P. 2012. Suitability of European climate for the Asian tiger mosquito Aedes albopictus: recent trends and future scenarios. Journal of the Royal Society Interface, 9, 2708-2717.

CAMPBELL-LENDRUM, D. & REITHINGER, R. 2002. Dengue and climate change. Trends in Parasitology, 18, 524.

CARBAJO, A. E., CARDO, M. V. & VEZZANI, D. 2012. Is temperature the main cause of dengue rise in non- endemic countries? The case of Argentina. International journal of health geographics, 11, 26.



CHAN, N., EBI, K., SMITH, F., WILSON, T. & SMITH, A. 1999. An integrated assessment framework for climate change and infestious diseases. Environmental Health Perspectives, 107, 329 - 337.

CHATURVEDI, U. C., RACHNA, N. & RICHA, S. 2006. Dengue and dengue haemorrhagic fever: implications of host genetics. FEMS Immunology and Medical Microbiology, 47, 155-166.

CHEN, J., ROTH, R. E., NAITO, A. T., LENGERICH, E. J. & MACEACHREN, A. M. 2008. Geovisual analytics to enhance spatial scan statistic interpretation: an analysis of U.S. cervical cancer mortality. International journal of health geographics, 7, 57.

CHEN, S. C. & HSIEH, M. H. 2012. Modeling the transmission dynamics of dengue fever: implications of temperature effects. Science of the Total Environment, 431, 385-391.

CHEN, S. C., LIAO, C. M., CHIO, C. P., CHOU, H. H., YOU, S. H. & CHENG, Y. H. 2010. Lagged temperature effect with mosquito transmission potential explains dengue variability in southern Taiwan: Insights from a statistical analysis. Science of the Total Environment, 408, 4069-4075.

124 Bibliography

CHEPESIUK, R. 2003. Mosquito mismanagement? Environmental Health Perspectives, 111, A636.

CHOWELL, G., DIAZ-DUENAS, P., MILLER, J. C., ALCAZAR-VELAZCO, A., HYMAN, J. M., FENIMORE, P. W. & CASTILLO-CHAVEZ, C. 2007. Estimation of the reproduction number of dengue fever from spatial epidemic data. Mathematical Biosciences, 208, 571-589.

CHOWELL, G. & SANCHEZ, F. 2006. Climate-based descriptive models of dengue fever: the 2002 epidemic in Colima, Mexico. Journal of Environmental Health, 68, 40-44, 55.

CIA. 2012. CIA World Factbook [Online]. Available: https://www.cia.gov/library/publications/the-world-factbook/ [Accessed 12 May 2012].

COSTELLO, A., ABBAS, M., ALLEN, A., BALL, S., BELL, S., BELLAMY, R., FRIEL, S., GROCE, N., JOHNSON, A., KETT, M., LEE, M., LEVY, C., MASLIN, M., MCCOY, D., MCGUIRE, B., MONTGOMERY, H., NAPIER, D., PAGEL, C., PATEL, J., DE OLIVEIRA, J. A., REDCLIFT, N., REES, H., ROGGER, D., SCOTT, J., STEPHENSON, J., TWIGG, J., WOLFF, J. & PATTERSON, C. 2009. Managing the health effects of climate change: Lancet and University College London Institute for Global Health Commission. Lancet, 373, 1693-1733.

CUMMINGS, D. A. T., IAMSIRITHAWORN, S., LESSLER, J. T., MCDERMOTT, A., PRASANTHONG, R., NISALAK, A., JARMAN, R. G., BURKE, D. S. & GIBBONS, R. V. 2009. The impact of the demographic transition on dengue in Thailand: insights from a statistical analysis and mathematical modeling. PLoS Medicine, 6, e1000139.

CUMMINGS, D. A. T., IRIZARRY, R. A., HUANG, N. E., ENDY, T. P., NISALAK, A., UNGCHUSAK, K. & BURKE, D. S. 2004. Travelling waves in the occurrence of dengue haemorrhagic fever in Thailand. Nature, 427, 344-347.

CUONG, H. Q., HIEN, N. T., DUONG, T. N., PHONG, T. V., CAM, N. N., FARRAR, J., NAM, V. S., THAI, K. T. & HORBY, P. 2011. Quantifying the emergence of dengue in hanoi, Vietnam: 1998-2009. PLoS Neglected Tropical Diseases, 5, e1322.

DELATTE, H., GIMONNEAU, G., TRIBOIRE, A. & FONTENILLE, D. 2009. Influence of temperature on immature development, survival, longevity, fecundity, and gonotrophic cycles of Aedes albopictus, vector of chikungunya and dengue in the Indian Ocean. Journal of Medical Entomology, 46, 33-41.

DESCLOUX, E., MANGEAS, M., MENKES, C. E., LENGAIGNE, M., LEROY, A., TEHEI, T., GUILLAUMOT, L., TEURLAI, M., GOURINAT, A. C.,

http://www.cia.gov/library/publications/the-world-factbook/

Bibliography 125

BENZLER, J., PFANNSTIEL, A., GRANGEON, J. P., DEGALLIER, N. & DE LAMBALLERIE, X. 2012. Climate-based models for understanding and forecasting dengue epidemics. PLoS Neglected Tropical Diseases, 6, e1470.

DGHS. 2000. National guidelines for clinical management of dengue syndrome. Available: http://www.sdnbd.org/sdi/issues/health/dengue/other/dng.PDF [Accessed March 26 2009].

EARNEST, A., TAN, S. B. & WILDER-SMITH, A. 2012. Meteorological factors and El Nino Southern Oscillation are independently associated with dengue infections. Epidemiology and Infection, 140, 1244-1251.

EPSTEIN, P. R. 2001. Climate change and emerging infectious diseases. Microbes and Infection, 3, 747-754.

EPSTEIN, P. R. 2005. Climate change and human health. New England Journal of Medicine, 353, 1433-1436.

FERDAUSI, S. A. & BOLKLAND, M. W. Rainwater harvesting for application in rural Bangladesh. 26th WEDC conference, water, sanitation and hygiene : challenges of the millennium, 2000 Dhaka, Bangladesh. 16-19.

FISCHER, D., THOMAS, S. M., NIEMITZ, F., REINEKING, B. & BEIERKUHNLEIN, C. 2011. Projection of climatic suitability for Aedes albopictus Skuse (Culicidae) in Europe under climate change conditions. Global and Planetary Change, 78, 54-64.

GASPARRINI, A. & ARMSTRONG, B. 2011. Distributed lag non-linear model in R: the package dlnm. Available: cran.r.project.org/web/packages/dlnm/vignettes/dlnmOverview.pdf [Accessed 20 November 2011].

GASPERI, G., BELLINI, R., MALACRIDA, A. R., CRISANTI, A., DOTTORI, M. & AKSOY, S. 2012. A new threat looming over the Mediterranean basin: emergence of viral diseases transmitted by Aedes albopictus mosquitoes. PLoS Neglected Tropical Diseases, 6, e1836.

GIBBONS, R. V. 2010. Dengue conundrums. International Journal of Antimicrobial Agents, 36, S36-S39.

GITHEKO, A. K. 2012. Advances in developing a climate based dengue outbreak models in Dhaka, Bangladesh: Challenges & opportunities. Indian Journal of Medical Research, 136, 7-9.

GITHEKO, A. K., LINDSAY, S. W., CONFALONIERI, U. E. & PATZ, J. A. 2000. Climate change and vector-borne diseases: a regional analysis. Bulletin of the World Health Organization, 78, 1136-1147.

http://www.sdnbd.org/sdi/issues/health/dengue/other/dng.PDF

126 Bibliography


GUBLER, D. J. 2002a. Epidemic dengue/dengue hemorrhagic fever as a public health, social and economic problem in the 21st century. Trends in Microbiology, 10, 100-103.

GUBLER, D. J. 2002b. How effectively is epidemiological surveillance used for dengue programme planning and epidemic response? Dengue Bulletin, 26, 96-106.

GUBLER, D. J. 2004a. The changing epidemiology of yellow fever and dengue, 1900 to 2003: full circle? Comparative Immunology, Microbiology and Infectious Diseases, 27, 319-330.

GUBLER, D. J. 2004b. Cities spawn epidemic dengue viruses. Nature Medicine, 10, 129-130.


GUBLER, D. J., REITER, P., EBI, K. L., YAP, W., NASCI, R. & PATZ, J. A. 2001. Climate variability and change in the United States: potential impacts on vector- and rodent-borne diseases. Environmental Health Perspectives, 109 Suppl 2, 223-233.

GUBLER, D. J. & TRENT, D. W. 1993. Emergence of epidemic dengue/dengue hemorrhagic fever as a public health problem in the Americas. Infectious Agents and Disease, 2, 383-393.

GUZMAN, A. & ISTÚRIZ, R. E. 2010. Update on the global spread of dengue. International Journal of Antimicrobial Agents, 36, S40-42.

GUZMAN, M. G., HALSTEAD, S. B., ARTSOB, H., BUCHY, P., FARRAR, J., GUBLER, D. J., HUNSPERGER, E., KROEGER, A., MARGOLIS, H. S., MARTINEZ, E., NATHAN, M. B., PELEGRINO, J. L., SIMMONS, C., YOKSAN, S. & PEELING, R. W. 2010. Dengue: a continuing global threat. Nature Reviews. Microbiology, 8, S7-S16.

GUZMAN, M. G. & KOURI, G. 2002. Dengue: an update. Lancet Infectious Diseases, 2, 33-42.

GUZMAN, M. G., KOURI, G., BRAVO, J., VALDES, L., SUSANA, V. & HALSTEAD, S. B. 2002. Effect of age on outcome of secondary dengue 2 infections. International Journal of Infectious Diseases, 6, 118-124.

Bibliography 127

HAINES, A., KOVATS, R. S., CAMPBELL-LENDRUM, D. & CORVALAN, C. 2006. Climate change and human health: impacts, vulnerability, and mitigation. Lancet, 367, 2101-2109.


HALES, S., WEINSTEIN, P., SOUARES, Y. & WOODWARD, A. 1999. El Nino and the dynamics of vector-borne disease transmission. Environmental Health Perspectives, 107, 99-102.

HALIDE, H. & RIDD, P. 2008. A predictive model for dengue hemorrhagic fever epidemics. International Journal of Environmental Health Research, 18, 253-265.


HASHIZUME, M., DEWAN, A. M., SUNAHARA, T., RAHMAN, M. Z. & YAMAMOTO, T. 2012. Hydroclimatological variability and dengue transmission in Dhaka, Bangladesh: a time-series study. BMC Infectious Diseases, 12, 98.


HII, Y. L., ROCKLOV, J., WALL, S., NG, L. C., TANG, C. S. & NG, N. 2012a. Optimal lead time for dengue forecast. PLoS Neglected Tropical Diseases, 6, e1848.

HII, Y. L., ZHU, H., NG, N., NG, L. C. & ROCKLOV, J. 2012b. Forecast of dengue incidence using temperature and rainfall. PLoS Neglected Tropical Diseases, 6, e1908.

HONORIO, N. A., NOGUEIRA, R. M., CODECO, C. T., CARVALHO, M. S., CRUZ, O. G., MAGALHAES MDE, A., DE ARAUJO, J. M., DE ARAUJO, E. S., GOMES, M. Q., PINHEIRO, L. S., DA SILVA PINEL, C. & LOURENCO-DE-OLIVEIRA, R. 2009. Spatial evaluation and modeling of Dengue seroprevalence and vector density in Rio de Janeiro, Brazil. PLoS Neglected Tropical Disseases, 3, e545.

HOPP, J. M. & FOLEY, J. A. 2003. Worldwide fluctuations in dengue fever cases related to climate variability. Climate Research, 25, 85-94.


128 Bibliography

HOSSAIN, M. A., KHATUN, M., ARJUMAND, F., NISALUK, A. & BREIMAN, R. F. 2003. Serologic evidence of dengue infection before onset of epidemic, Bangladesh. Emerging Infectious Diseases, 9, 1411-1414.

HSIEH, Y. H. & CHEN, C. W. S. 2009. Turning points, reproduction number and impact of climatological events for multi-wave dengue outbreaks. Tropical Medicine and International Health, 14, 628-638.


HU, W., CLEMENTS, A., WILLIAMS, G. & TONG, S. 2011. Spatial analysis of notified dengue fever infections. Epidemiology and Infection, 139, 391-399.


IPCC. 2000. IPCC Special report: emissions scenarios. Available: https://www.ipcc.ch/pdf/special-reports/spm/sres-en.pdf [Accessed 6/9/2013].


IPCC 2012. Managing the risks of extreme events and disasters to advance climate change adaptation (SREX). In: FIELD, C. B., V. BARROS, T.F. STOCKER, D. QIN, D.J. DOKKEN, K.L. EBI, M.D. MASTRANDREA, K.J. MACH, & G.-K. PLATTNER, S. K. A., M. TIGNOR, AND P.M. MIDGLEY (eds.). Cambridge, ENGLAND: Cambridge University Press.

JETTEN, T. H. & FOCKS, D. A. 1997. Potential changes in the distribution of dengue transmission under climate warming. American Journal of Tropical Medicine and Hygiene, 57, 285-297.

JOHANSSON, M. A., CUMMINGS, D. A. T. & GLASS, G. E. 2009a. Multiyear climate variability and dengue - El Nino Southern Oscillation, weather, and dengue Incidence in Puerto Rico, Mexico, and Thailand: a longitudinal data analysis. PLoS Medicine, 6, e1000168.

JOHANSSON, M. A., DOMINICI, F. & GLASS, G. E. 2009b. Local and global effects of climate on dengue transmission in Puerto Rico. PLoS Neglected Tropical Diseases, 3, e382.

http://www.ipcc.ch/pdf/special-reports/spm/sres-en.pdf

Bibliography 129

KAN, H., CHEN, R. & TONG, S. 2012. Ambient air pollution, climate change, and population health in China. Environment International, 42, 10-19.

KARIM, M. N., MUNSHI, S. U., ANWAR, N. & ALAM, M. S. 2012. Climatic factors influencing dengue cases in Dhaka city: a model for dengue prediction. Indian Journal of Medical Research, 136, 32-39.

KEATING, J. 2001. An investigation into the cyclical incidence of dengue fever. Social Science and Medicine, 53, 1587-1597.

KOOPMAN, J. S., PREVOTS, D. R., MANN, M. A. V., DANTES, H. G., AQUINO, M. L. Z., LONGINI, I. M., JR. & AMOR, J. S. 1991. Determinants and predictors of dengue infection in Mexico. American Journal of Epidemiology, 133, 1168-1178.

KOVATS, R. S., CAMPBELL-LENDRUM, D. H., MCMICHAEL, A. J., WOODWARD, A. & COX, J. S. 2001. Early effects of climate change: do they include changes in vector-borne disease? Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 356, 1057-1068.


KULLDORF, M., ATHAS, W., FEUER, E., MILLER, B. & KEY, C. 1998. Evaluating cluster alarms:a space-time scan statistic and brain cancer in Los Alamos, New Mexico. American Journal of Public Health, 88, 1377-1380.

KULLDORFF, M., FEUER, E. J., MILLER, B. A. & FREEDMAN, L. S. 1997. Breast cancer clusters in the northeast United States: a geographic analysis. American Journal of Epidemiology, 146, 161-70.

KULLDORFF, M., HEFFERNAN, R., HARTMAN, J., ASSUNCAO, R. & MOSTASHARI, F. 2005. A space-time permutation scan statistic for disease outbreak detection. PLoS Medicine, 2, e59.


LAFFERTY, K. D. 2009. The ecology of climate change and infectious diseases. Ecology, 90, 888-900.

LEE, M. N., HAWANG, K.-P., CHEN, T.-C., LU, P.-N. & CHEN, T.-P. 2006. Clinical characteristics of dengue and dengue hemorrhagic fever in a medical center of southern Taiwan during the 2002 epidemic. Journal of Microbiology Immunology and Infection, 39, 121-129.


130 Bibliography

LEITMEYER, K. C., VAUGHN, D. W., WATTS, D. M., SALAS, R. & VILLALOBOS, I. 1999. Dengue virus structural differences that correlate with pathogenesis. Journal of Virology, 73, 4738-4747.

LI, Z., YIN, W., CLEMENTS, A., WILLIAMS, G., LAI, S., ZHOU, H., ZHAO, D., GUO, Y., ZHANG, Y., WANG, J., HU, W. & YANG, W. 2012. Spatiotemporal analysis of indigenous and imported dengue fever cases in Guangdong province, China. BMC Infectious Diseases, 12, 132.

LU, L., LIN, H., TIAN, L., YANG, W., SUN, J. & LIU, Q. 2009. Time series analysis of dengue fever and weather in Guangzhou, China. BMC Public Health, 9, 395.

LYERLA, R., RIGAU-PEREZ, J. G., VORNDAM, A. V., REITER, P., GEORGE, A. M., POTTER, I. M. & GUBLER, D. J. 2000. A dengue outbreak among camp participants in a Caribbean island, 1995. Journal of Travel Medicine, 7, 59-63.

MACKENZIE, J. S., GUBLER, D. J. & PETERSEN, L. R. 2004. Emerging flaviviruses: the spread and resurgence of Japanese encephalitis, West Nile and dengue viruses. Nature Medicine, 10, S98-S109.

MAMMEN, M. P., PIMGATE, C., KOENRAADT, C. J. M., ROTHMAN, A. L., ALDSTADT, J., NISALAK, A., JARMAN, R. G., JONES, J. W., SRIKIATKHACHORN, A., YPIL-BUTAC, C. A., GETIS, A., THAMMAPALO, S., MORRISON, A. C., LIBRATY, D. H., GREEN, S. & SCOTT, T. W. 2008. Spatial and temporal clustering of dengue virus transmission in Thai villages. PLoS Medicine, 5, e205.

MCBRIDE, W. J., MULLNER, H., MULLER, R., LABROOY, J. & WRONSKI, I. 1998. Determinants of dengue 2 infection among residents of Charters Towers, Queensland, Australia. American Journal of Epidemiology, 148, 1111-1116.

MCMICHAEL, A. J. 2003. Global climate change: will it affect vector-borne infectious diseases? Internal Medicine Journal, 33, 554-555.


NAGAO, Y., THAVARA, U., CHITNUMSUP, P., TAWATSIN, A., CHANSANG, C. & CAMPBELL-LENDRUM, D. 2003. Climatic and social risk factors for Aedes infestation in rural Thailand. Tropical Medicine and International Health, 8, 650-659.

Bibliography 131

NAISH, S., HU, W., MENGERSEN, K. & TONG, S. 2011. Spatial and temporal clusters of Barmah Forest virus disease in Queensland, Australia. Tropical Medicine and International Health, 16, 884-893.

NAKHAPAKORN, K. & TRIPATHI, N. K. 2005. An information value based analysis of physical and climatic factors affecting dengue fever and dengue haemorrhagic fever incidence. International journal of health geographics, 4, 13.

NGUI, A. N., APPARICIO, P., FLEURY, M.-J., LESAGE, A., GRÉGOIRE, J.-P., MOISAN, J. & VANASSE, A. 2013. Spatio-temporal clustering of the incidence of schizophrenia in Quebec, Canada from 2004 to 2007. Spatial and Spatio-temporal Epidemiology, 6, 37-47.

OKI, M. & YAMAMOTO, T. 2012. Climate change, population immunity, and hyperendemicity in the transmission threshold of dengue. PLoS ONE, 7, e48258.

OO, T. T., STORCH, V., MADON, M. B. & BECKER, N. 2011. Factors influencing the seasonal abundance of Aedes (Stegomyia) aegypti and the control strategy of dengue and dengue haemorrhagic fever in Thanlyin Township, Yangon City, Myanmar. TROPICAL BIOMEDICINE, 28, 302-311.

PANG, J., SALIM, A., LEE, V. J., HIBBERD, M. L., CHIA, K. S., LEO, Y. S. & LYE, D. C. 2012. Diabetes with hypertension as risk factors for adult dengue hemorrhagic fever in a predominantly dengue serotype 2 epidemic: a case control study. PLoS Neglected Tropical Diseases, 6, e1641.



PATZ, J. A. & REISEN, W. K. 2001. Immunology, climate change and vector-borne diseases. Trends in Immunology, 22, 171-172.

PHAM, H. V., DOAN, H. T., PHAN, T. T. & MINH, N. N. 2011. Ecological factors associated with dengue fever in a central highlands province, Vietnam. BMC Infectious Diseases, 11, 172.

PINTO, E., COELHO, M., OLIVER, L. & MASSAD, E. 2011. The influence of climate variables on dengue in Singapore. International Journal of Environmental Health Research, 1-12.

132 Bibliography

PODDER, G., BREIMAN, R. F., AZIM, T., THU, H. M., VELATHANTHIRI, N., KYMLOWRY, L. Q. M. & AASKOV, J. G. 2006. Short report: origin of dengue type-3 viruses associated with the dengue outbreak in Dhaka, Bangladesh,in 2000 and 2001. American Journal of Tropical Medicine and Hygiene, 74, 263-265.

RACLOZ, V., RAMSEY, R., TONG, S. & HU, W. 2012. Surveillance of dengue Fever virus: a review of epidemiological models and early warning systems. PLoS Neglected Tropical Diseases, 6, e1648.

RAHMAN, M., RAHMAN, K., SIDDQUE, A. K., SHOMA, S., KAMAL, A. H., ALI, K. S., NISALUK, A. & BREIMAN, R. F. 2002. First outbreak of dengue hemorrhagic fever, Bangladesh. Emerging Infectious Diseases, 8, 738-740.

RAM, S., KHURANA, S., KAUSHAL, V., GUPTA, R. & KHURANA, S. B. 1998. Incidence of dengue fever in relation to climatic factors in Ludhiana, Punjab. Indian Journal of Medical Research, 108, 128-133.


REITER, P., LATHROP, S., BUNNING, M., BIGGERSTAFF, B., SINGER, D., TIWARI, T., BABER, L., AMADOR, M., THIRION, J. & HAYES, J. 2003. Texas lifestyle limits transmission of dengue virus. Emerging Infectious Diseases, 9, 86-89.

RIGAU-PEREZ, J. G., CLARK, G. G., GUBLER, D. J., REITER, P., SANDERS, E. J. & VORNDAM, A. V. 1998. Dengue and dengue haemorrhagic fever. Lancet, 352, 971-977.

RITCHIE, S. A., HANNA, J. N., HILLS, S. L., PIISPANEN, J. P., MCBRIDE, W. J. H., PYKE, A. & SPARK, R. L. 2002. Dengue control in north Queensland, Australia: case recognition and selective indoor residual spraying. Dengue Bulletin, 26, 7-13.

ROBERTSON, C. & NELSON, T. 2010. Review of software for space-time disease surveillance. International journal of health geographics, 9, 16.

ROHANI, A., SUZILAH, I., MALINDA, M., ANUAR, I., MOHD MAZLAN, I., SALMAH MASZAITUN, M., TOPEK, O., TANRANG, Y., OOI, S. C., ROZILAWATI, H. & LEE, H. L. 2011. Aedes larval population dynamics and risk for dengue epidemics in Malaysia. TROPICAL BIOMEDICINE, 28, 237-248.

ROMI, R., SABATINELLI, G., SAVELLI, L. G., RARIS, M., ZAGO, M. & MALATESTA, R. 1997. Identification of a North American mosquito

Bibliography 133

species, Aedes atropalpus (Diptera: Culicidae), in Italy. Journal of the American Mosquito Control Association, 13, 245-246.

RUSSELL, R. C. 2009. Mosquito-borne disease and climate change in Australia: time for a reality check. Australian Journal of Entomology, 48, 1-7.

SCHMIDT, W. P., SUZUKI, M., DINH THIEM, V., WHITE, R. G., TSUZUKI, A., YOSHIDA, L. M., YANAI, H., HAQUE, U., HUU THO, L., ANH, D. D. & ARIYOSHI, K. 2011. Population density, water supply, and the risk of dengue fever in Vietnam: cohort study and spatial analysis. PLoS Medicine, 8, e1001082.

SEMENZA, J. C. & MENNE, B. 2009. Climate change and infectious diseases in Europe. Lancet Infectious Diseases, 9, 365-375.

SHANG, C. S., FANG, C. T., LIU, C. M., WEN, T. H., TSAI, K. H. & KING, C. C. 2010. The role of imported cases and favorable meteorological conditions in the onset of dengue epidemics. PLoS Neglected Tropical Diseases, 4, e775.

SHEPARD, D. S., COUDEVILLE, L., HALASA, Y. A., ZAMBRANO, B. & DAYAN, G. H. 2011. Economic impact of dengue illness in the Americas. American Journal of Tropical Medicine and Hygiene, 84, 200-207.

SHUMAN, E. K. 2010. Global climate change and infectious diseases. New England Journal of Medicine, 362, 1061-1063.

SIERRA, B. C., KOURI, G. & GUZMÁN, M. G. 2007. Race: a risk factor for dengue hemorrhagic fever. Archives of Virology, 152, 533-542.

SIMMONS, C. P. & FARRAR, J. 2009. Changing patterns of dengue epidemiology and implications for clinical management and vaccines. PLoS Medicine, 6, e1000129.

SIMMONS, C. P., FARRAR, J. J., VAN VINH CHAU, N. & WILLS, B. 2012. Dengue. New England Journal of Medicine, 366, 1423-1432.

SRIPROM, M., CHALVET-MONFRAY, K., CHAIMANE, T., VONGSAWAT, K. & BICOUT, D. J. 2010. Monthly district level risk of dengue occurrences in Sakon Nakhon Province, Thailand. Science of the Total Environment, 408, 5521-5528.

SUAYA, J. A., SHEPARD, D. S., SIQUEIRA, J. B., MARTELLI, C. T., LUM, L., TAN, L. H., KONGSIN, S., JIAMTON, S., GARRIDO, F. & MONTOYA, R. 2009. Cost of dengue cases in eight countries in the Americas and Asia: a prospective study. American Journal of Tropical Medicine and Hygiene, 80, 846-855.

134 Bibliography

SUKRI, N. C., LARAS, K., WANDRA, T., DIDI, S., LARASATI, R. P. & RACHDYMATMAKA, J. R. 2003. Transmission of epidemic dengue heamorrhagic fever in eastermost Indonesia. American Journal of Tropical Medicine and Hygiene, 68, 529-535.


TATEM, A. J., HAY, S. I. & ROGERS, D. J. 2006. Global traffic and disease vector dispersal. Proceedings of the National Academy of Sciences of the United States of America, 103, 6242-6247.

THAI, K. T. & ANDERS, K. L. 2011. The role of climate variability and change in the transmission dynamics and geographic distribution of dengue. Experimental Biology and Medicine, 236, 944-954.

THAI, K. T., NISHIURA, H., HOANG, P. L., TRAN, N. T., PHAN, G. T., LE, H. Q., TRAN, B. Q., NGUYEN, N. V. & DE VRIES, P. J. 2011. Age-specificity of clinical dengue during primary and secondary infections. PLoS Neglected Tropical Diseases, 5, e1180.

THAI, K. T. D., NAGELKERKE, N., PHUONG, H. L., NGA, T. T. T., GIAO, P. T., HUNG, L. Q., BINH, T. Q., NAM, N. V. & DE VRIES, P. J. 2010. Geographical heterogeneity of dengue transmission in two villages in southern Vietnam. Epidemiology and Infection, 138, 585-591.

THAMMAPALO, S., CHONGSUVIVATONG, V., GEATER, A. & DUERAVEE, M. 2007. Environmental factors and incidence of dengue fever and dengue haemorrhagic fever in an urban area, Southern Thailand. Epidemiology and Infection, 136, 135 -143.

THAMMAPALO, S., CHONGSUWIWATWONG, V., GEATER, A., LIM, A. & CHOOMALEE, K. 2005a. Socio-demographic and environmental factors associated with Aedes breeding places in Phuket, Thailand. Southeast Asian Journal of Tropical Medicine and Public Health, 36, 426-433.

THAMMAPALO, S., CHONGSUWIWATWONG, V., MCNEIL, D. & GEATER, A. 2005b. The climatic factors influencing the occurence of dengue hemorrhagic fever in Thailand Southeast Asian Journal of Tropical Medicine and Public Health, 36, 191-196.

TIPAYAMONGKHOLGUL, M., FANG, C. T., KLINCHAN, S., LIU, C. M. & KING, C. C. 2009. Effects of the El Nino-Southern Oscillation on dengue epidemics in Thailand, 1996-2005. BMC Public Health, 9, 422.

TONG, S., DALE, P., NICHOLLS, N., MACKENZIE, J. S., WOLFF, R. & MCMICHAEL, A. J. 2008. Climate variability, social and environmental

Bibliography 135

factors, and Ross River virus transmission: research development and future research needs. Environmental Health Perspectives, 116, 1591-1597.


TRAN, A. & RAFFY, M. 2006. On the dynamics of dengue epidemics from large-scale information. Theoretical Population Biology, 69, 3-12.

TUN-LIN, W., BURKOT, T. R. & KAY, B. H. 2000. Effects of temperature and larval diet on development rates and survival of the dengue vector Aedes aegypti in north Queensland, Australia. Medical and Veterinary Entomology, 14, 31-37.

VANWAMBEKE, S. O., VAN BENTHEM, B. H. B., KHANTIKUL, N., BURGHOORN-MAAS, C., PANART, K., OSKAM, L., LAMBIN, E. F. & SOMBOON, P. 2006. Multi-level analyses of spatial and temporal determinants for dengue infection. International journal of health geographics, 5, 5.

WEN, T.-H., LIN, N. H., LIN, C.-H., KING, C.-C. & SU, M.-D. 2006. Spatial mapping of temporal risk characteristics to improve environmental health risk identification: A case study of a dengue epidemic in Taiwan. Science of the Total Environment, 367, 631-640.

WHO 2000. Dengue Hemorrhagic Fever: Diagnosis,Treatment , Prevention and Control, WHO, Geneva, Switzerland.

WHO 2006. Scientific working group report on Dengue, Geneva, Switzerland.

WHO. 2012a. Dengue and severe dengue [Online]. World Health Organization. Available: http://www.who.int/mediacentre/factsheets/fs117/en/index.html [Accessed 27 September 2012].

WHO. 2012b. DengueNet data query [Online]. World Health Organization. Available: http://apps.who.int/globalatlas/DataQuery/default.asp.

WILDER-SMITH, A. 2012. Dengue infections in travellers. Paediatr Int Child Health, 32 Suppl 1, 28-32.

WILDER-SMITH, A. & GUBLER, D. J. 2008. Geographic expansion of dengue: the impact of international travel. Medical Clinics of North America, 92, 1377-1390.

http://www.who.int/mediacentre/factsheets/fs117/en/index.html

http://apps.who.int/globalatlas/DataQuery/default.asp

136 Bibliography

WILLIAMS, C. R., BADER, C. A., KEARNEY, M. R., RITCHIE, S. A. & RUSSELL, R. C. 2010. The extinction of dengue through natural vulnerability of its vectors. PLoS Neglected Tropical Diseases, 4, e922.

WOODRUFF, R. E. & MCMICHAEL, T. 2004. Climate change and human health: all affected bit some more than others. Social Alternatives, 23, 17-22.

WOODRUFF, R. E., MCMICHAEL, T., BUTLER, C. & HALES, S. 2006. Action on climate change: the healthrisks of procastinating. Australian and New Zealand Journal of Public Health, 30, 567-571.



YANG, H. M., MACORIS, M. L., GALVANI, K. C., ANDRIGHETTI, M. T. & WANDERLEY, D. M. 2009a. Assessing the effects of temperature on the population of Aedes aegypti, the vector of dengue. Epidemiology and Infection, 137, 1188-1202.

YANG, T., LU, L., FU, G., ZHONG, S., DING, G., XU, R., ZHU, G., SHI, N., FAN, F. & LIU, Q. 2009b. Epidemiology and vector efficiency during a dengue fever outbreak in Cixi, Zhejiang province, China. Journal of Vector Ecology, 34, 148-154.

Appendices 137

Appendices

Appendix A

Conferences

“23rd Conference of the International Society for Environmental Epidemiology (ISEE)”, 13-16 September 2011, Barcelona, Spain (Oral presentation/ presented by Shilu Tong: Space-Time clusters of dengue fever in Bangladesh.)

“2010 International Climate change Adaptation Conference” NCCARF and CSIRO, 29 June -1st July 2010, Gold Coast, (Poster presentation: Impact of climate change and socio-environmental factors on dengue transmission in the Asia-Pacific region)

“2009 Queensland Institute of Health Conference” 30 November-1st December, Townsville, Australia.

“IHBI inspire 2009 Post graduate student conference”, Queensland University of Technology, 17-18 November 2009, Brisbane, Australia.

138 Appendices

Appendix B

Publications

1. Banu S, Hu W, Hurst C, Guo Y, Islam Z M &Tong S, 2012. Space-time clusters of dengue fever in Bangladesh. Tropical Medicine and International Health 17(9): 1086-1091.

2. Banu S, Hu W, Hurst C &Tong S, 2011. Dengue transmission in the Asia-Pacific region: impact of climate change and socio-environmental factors. Tropical Medicine and International Health 16: 598-607.

Appendices 139

Appendix C

examining the impact of climate change on dengue...

Documents