Introduction
Dengue and Dengue Hemorrhagic Fever (DHF) is the most rapidly spreading vector borne infection all over the world. Its occurrence has been registered in 124 countries with 3.61 billion people are at risk for infection and 500 million people infected each year. It was estimated that nearly 36 million cases of dengue fever and 2.1 million cases of severe DHF occur annually, and nearly 21,000 deaths are likely attributable to dengue. It also exerts huge societal and economic costs to endemic countries, where mostly vector control strategies have been unsuccessful.
 
Dengue virus commonly transmitted by the adult female Aedes aegypti L. mosquitoes. However, the virus is sometime transmitted by Aedes albopictus (Skuse) in some locations around the globe (cf Strickman & Kittayapong 2002). Vector control remains the only way to prevent dengue transmission (Guzman & Kouri 2002; Deen 2004; Guzman et al. 2004). Entomological surveillance of dengue has been standardized on different indices based on the simple determination of the presence or absence of Aedes larvae either in each container or somewhat in each house (Tun-Lin et al. 1996; Focks 2003). The house index (HI, percentage of houses positive for larvae), container index (CI, percentage of containers positive for larvae) and Breteau index (BI, number of positive containers per 100 houses) have become the most widely used indices (Reiter & Gubler 1997).

Dengue is described as ‘endemic’ in many countries in the South East Asian region – which means that cases occur every year, although there is a significant variation between countries and within each country. In 2010, World Health Organization (WHO) stratified the current situation of Dengue/DHF in India under category A (earlier it was under category B up to 2009), which means a major public health problem, a leading cause of hospitalization and death among children, shows cyclical epidemics in urban centers spreading to rural areas with multiple virus serotypes circulating. In India, National Vector Borne Disease Control Program (NVBDCP) reported 28292 laboratory-confirmed dengue cases in 2010 from 31 out of 35 states in India (highest ever in a year). In 2011, 18,860 confirmed dengue cases were reported and 169 deaths and in 2012, total 37,070 cases and 227 deaths were attributable to dengue in India (NVBDCP 2012).

The northern states of India are badly affected by dengue, although the intensity of transmission varies every 2-4 years. Uttarakhand is one of the dengue prone states in northern India, which has a history of dengue epidemics since long with a heavy epidemic of dengue/DHF in 2010. Moreover, Uttarakhand (India) is the prime destination of tourists from all over the world. This also increases the traveler’s health risk. Cultural tradition, socio-environmental conditions, and household circumstances often result in the absence or in the irregularity of water supply for a significant part of the population in Uttarakhand, causing the need to store water in reservoirs (Singh et al. 2010). Moreover, due to a relatively dry climate in summer seasons, people use water coolers (a cooling device made up of metal/plastic having a water tank) in their houses. These coolers are the potential place for Aedes mosquito breeding. Furthermore, because of irregular garbage collection in many areas, other breeding sites like plastic container, tin containers, bottles, water storage tanks, tire, and many unusual breeding sites.

In this study, epidemiological characteristics of dengue/DHF and the entomological parameters were assessed during the epidemic of dengue in 2010 to determine the epidemic risk of dengue in Uttarakhand, India with the aim to establish better policies to control the dengue epidemics in these areas (e.g., improvement of water supply to minimize the requirement of storage water, which are found to the breeding place of Aedes aegypti; establish a local threshold of HI, BI & CI for forecast the outbreak etc.). The results of above study are presented in this manuscript.

Methods
Study Area and its eco-environmental settings
Initially, the study was conducted in different parts of Uttarakhand, however, finally, we focused our epidemiological and entomological surveillance in Nainital and Dehradun, the two districts badly hit by dengue epidemics in 2010. We analyzed the epidemiological history of dengue in the entire Uttarakhand state, and selected Nainital (exhibited a low-middle epidemics in September 2009 with 51 confirmed dengue cases) and Dehradun (exhibited an early signal of epidemics on August 2010 with 19 confirmed dengue cases) districts for investigations of entomological measures of risk of dengue. The localities within these two districts for the entomological survey were picked out based on dengue suspected cases reported during 2009-2010. Total 11 localities each in Nainital (viz. Forest compound, Kotdwar Rd., Khatyari, Gular Ghati, Perumdara, 25 acres Colony, CCPM colony, Nageena Colony, bungalow colony, Bindukhatta, Indiranagar) and Dehradun (viz. Patel Nagar, Vikas Nagar, Chandreshwar Nagar, Kailash gate, Shyampur, Aadarsh gram, Rani Pokhari, Reetha Mandi, Deep Nagar, Kedar Purum, Indra colony) districts were chosen.

Topography: Uttarakhand lies within the Himalayan region, with three distinct topographical belts - the Shivaliks within the sub Himalayan tract [300-600 m Above Mean Sea Level, (AMSL)], the Himachal ranges within the lower Himalayan region (1,500-2,700 m AMSL), and the Himadri ranges within the upper Himalayan region (4,800-6,000 m AMSL). The lower hilly region, ranging between 600 m-1,800 m, and it is within the region where the majority of the human populations live. Major parts of our study sites are located in this region.

General Climate: The climatic conditions change with the increase of altitude, extending from sub-zero temperatures at high altitudes in winter to moderate temperatures in the lower tracts in summer. The territory has three distinct seasons - monsoon (June - September), winter (October- February) and summer (March - May). Snowfall occurs between January and March within the upper and lower Himalayan ranges.

Temperature and rainfall: Monthly maximum, minimum, and average temperatures and rainfall data of Uttarakhand were collected from Indian Meteorological Department. In this article, we have included the temperature and rainfall data of Nainital and Dehradun districts only, as entomological surveys were carried in these two districts only. Temperature and rainfall data based on mean of the last 25 years (monthly arithmetic averages) are presented graphically in the result section.

Population: The Uttarakhand has a population of approximately 10,116,752 inhabitants. Nainital and Dehradun are having an estimated population of 955,128 (494,115 males and 461,013 females) and 1,698,560 (893,222 males & 805,338 females) inhabitants respectively (data source: Census India) (data source: Census of India 2011).

Dengue fever and Epidemiological data collection
We worked in collaboration with the Department of Health, Govt. of Uttarakhand for this study. The dengue case definition was adopted as per the recommendation of World Health Organization Regional Office (Southeast Asia Regional Office 1999). Only the serologically confirmed dengue cases, diagnosed using Rapid Detection kit (NS1 Ag Ab Combo) and enzyme-linked immunosorbent assay in 2010 by different sentinel laboratories of the Department of Health, Govt. of Uttarakhand were included in the analysis reported here. All duplicate notifications were removed before analysis. We have also tried our level best to identify cases with a history of travel within 10 preceding days of the onset of illness to states where dengue also occurred were classified as imported and were not included in the analysis. We have collected the epidemiological report of dengue cases in Uttarakhand for the last five years (2006 to 2010) from the Department of Health, Govt. of Uttarakhand. During epidemic in 2010, suspected cases were identified and a sero-epidemiologic survey was conducted in Nainital and Dehradun districts through state health services for laboratory-confirmation of suspected cases. The denominator for the calculation of total and sex wise incidence rates of reported indigenous cases of dengue were based on the estimated mid-year total population and male/female population obtained from the Office of Registrar General and Census Commissioner (Uttarakhand State), India.

Entomological Surveillance
We have chosen 11 localities each in Nainital and Dehradun as mentioned above for entomological survey. In these areas, we surveyed all potential breeding sites in monsoon and post-monsoon months from September to November 2010, immediately after confirmations of an early outbreak of dengue in August 2010 with 19 confirmed dengue cases. All non-hermetically closed containers containing any volume of water were considered as potential breeding sites. All water-holding containers were examined. Either the name/type of the containers (viz. Tire, cooler, fridge etc.) or construction materials (viz. Tin, cement tank, clay pots, and plastic container) classified the breeding sites. We recorded the number of houses inspected, positive containers (with Aedes pupae or larvae) and houses with ≥ 1 positive container.

Statistical Analysis
For all statistical analysis, we performed Kolmogorov- Smirnov (with Dallal- Wilkinson-Lilliefor P value), D'Agostino & Pearson Omnibus and Shapiro-Wilk normality test to see whether the data follow a Gaussian distribution and implemented appropriate statistical tests and calculated the corresponding P values. We calculated the total as well as sex wise incidence rates of reported indigenous cases of dengue (combined of Nainital & Dehradun) based on the estimated mid-year total population and male/female population respectively. We analyzed larval survey data of different breeding sites based on the name/type of the containers (viz. Tire, cooler, fridge, etc.); construction materials (viz. Tin, cement tank, clay pots, and plastic container) and three sigma limits were applied to judge the significance of chance of getting Aedes larvae. The three sigma limits were applied to alert vector control operation process regarding potential dangerous container for mosquito breeding following statistical quality measurement. For all areas, we calculated different entomological indices, viz. HI, CI, and BI and performed the Spearman rank correlation coefficient between the different indices in all inspected areas. This statistical test was aimed to verify whether the ranks of variables (HI, CI & BI) co-vary. The entomological indices from Nainital and Dehradun districts were transformed to approximately normal distribution (by using square root transformation) for calculating means, standard deviation (SD), and 95% confidence intervals (CI). After performing the descriptive statistical test on a transformed variable, we have back-transformed the outcome statistics (mean, SD and 95% CI) into a count by squaring it to get a feel for the precision of the magnitude. Statistical comparisons in the distribution of entomological indices (HI, CI, & BI) in Nainital and Dehradun districts were made by performing pair-wise comparisons with Mann-Whitney U non -parametric test. This statistical test was aimed to verify whether the ranks of variables (HI, CI & BI) are same in these two districts and the differences we observed are just coincidence.

Results
The history of dengue epidemics in last five years (2006 to 2010) in Uttarakhand is presented in Table 1 and Figure 1, which show an epidemic of dengue during 2010 with total 4140 laboratory-confirmed dengue cases, which means about 5347% increase of dengue cases (Table 1). We observed that both numbers of dengue cases and numbers of deaths due to dengue are gradually increasing from 2006 to 2010. However, percentage death rates due to dengue exhibit a fluctuating trend (Table 1), although, Figure 1 shows the direction towards the actual flow of dengue epidemics in Uttarakhand during the past few years.

Figure 1 Dengue cases and deaths in Uttarakhand (India) during last five years. Both the number of cases and death has been increased during last five years. However, the percentage of deaths has decreased.

Table 1 Dengue cases and deaths in Uttarakhand in last five years (2006 to 2010)

During the 2010 epidemic, our survey revealed that out of six dengue prone districts of Uttarakhand state, Nainital (862 cases) and Dehradun (2913 cases) were the worst dengue affected districts during the epidemics followed by Haridwar (257 cases), Pouri Garwal (73 cases), Udham Singh Nagar (55 cases) and Tehri Garwal (04 cases) (Figure 2). Total 3775 laboratory-confirmed dengue cases were registered from Nainital and Dehradun districts in 2010, an incidence rate of 142.5 per 100000 populations (Table 2). It shows about 91.19% of total dengue cases in Uttarakhand were reported from these two districts. Table 2 summarizes the age and sex-specific distribution of dengue cases in Nainital and Dehradun (combined) during 2010. The incidence rate in male (169.9 per 100000) is significantly higher than in female (112.3 per 100000). Dengue cases are predominant in the age group 21-30 years (Table 2).

Figure 2 Month wise distribution of dengue cases in Uttarakhand during epidemics in 2010. The dengue epidemic rises during late August and continues up to November

Table 2 Age and sex specific distribution of dengue cases in Nainital and Dehradun (combined) in 2010

Monthly weather data in Nainital and Dehradun districts (Figure 3) shows that temperature variation during the year is moderate; Nainital has a relatively colder climate than Dehradun. However, rainfall variation throughout the year is great. Although the rainy season begins each year in late May or early June, most rainfall occurs in July and August and gradually decreases in September and October. February and November–January was dry, with very little or no rain at all. The warmest minimum and maximum temperatures were recorded in June and the lowest in January.

Figure 3 Summary of weather data from Nainital and Dehradun districts of Uttarakhand. Temperature (°C) and rainfall (mm) data based on mean of the last 25 years (monthly arithmetic averages).

The classical approach of probability shows that Dehradun is more dengue prone area than Nainital as 77% of cases are recorded from Dehradun, whereas Nainital registered 23% cases, but large amount of containers with mosquito breeding were found in Nainital (Table 3). We determined that 41% and 23% of total containers surveyed are found positive for mosquito breeding in Nainital and Dehradun respectively (Table 3). We classified breeding sites in seven different types (e.g., tires, coolers, tin-containers, cement tanks, clay pots, plastic containers, and Fridge vessels). Among these different containers, tires (95%), tin containers (51%), cement tanks (44%), and plastic containers (41%) show the maximum potential for breeding in Nainital, whereas, in the Dehradun maximum breeding of Aedes mosquitoes were found in fridge vessels (43%) followed by coolers (27%), plastic containers (25%) and cement tanks (23%) (Table 3). Distribution of seven types of positive containers in different surveyed localities is summarized in Table 4 and Table 5 for Nainital and Dehradun respectively. In terms of mosquito breeding in different localities, Khatyari, Gular Ghati and Forest compound in Nainital, on the other hand, Reetha Mandi and Deep Nagar in Dehradun are the high alert areas (Table 4 and Table 5).

Table 3 Summary of distribution of observed breeding sites for Aedes larvae and dengue cases in Nainital and Dehradun

Table 4 Percentage distribution of positive container and influence of types of containers for breeding of Aedes mosquitoes at different localities surveyed in Nainital

Table 5 Percentage distribution of positive container and influence of types of containers for breeding of Aedes mosquitoes at different localities surveyed in Dehradun

We calculated that the overall mean values of all entomological indices are relatively higher in Nainital (HI = 32.33, CI = 25.92 and BI = 124.99) than Dehradun (HI = 28.68, CI = 17.93 and BI = 77.33) (Table 6). Distribution of these indices in different localities is presented in Figure 4. A pairwise comparison of these indices (HI, CI, and BI) with Mann-Whitney U nonparametric test reveals that the observed ranking difference of these indices in Nainital and Dehradun is not statistically significant (p>0.05).

Figure 4 Comparisons of entomological indices (HI, CI, and BI) in different localities in Nainital and Dehradun

Table 6 Mean House Index (HI), Container Index (CI) and Breteau Index (BI) in Nainital and Dehradun during 2010 dengue epidemics

We observed a very high correlation between overall HI and BI values (r ≥ 0.92, p<0.01). In most positive houses, more than one container with Aedes larvae/pupae was found.

Discussion
Dengue has been on the rise since the year 2010 with epidemics reported in as many as 50 countries in 2010 (WHO-South East Asian Regional Office 2010). In this article, we analyzed the dengue transmission dynamics in Uttarakhand in the last few years, investigating the dengue/DHF epidemic in 2010 and exploring how different types of containers act on the mosquito proliferation and how all this knowledge may help in establishing better policies to control the dengue epidemics. Apparently, the effective vector control may be possible but not easy. In many developing countries limited infrastructural facilities in programs to control the Aedes mosquitoes has resulted in a huge expansion of dengue vectors in urban as well as rural habitats (Tauil 2002; Barcellos et al. 2005; Medronho et al. 2009).

Here, we should acknowledge that technically the present study might have few missing values and limitations.
The existence of detailed entomological surveillance data before, during, and after the dengue epidemic in study sites may offer a unique opportunity to analyze cause and effect relationship. The entomological data collection has been limited to water collections in artificial containers, which may ignore the natural breeding sites of Ae. albopictus. Moreover, we may not able to identify all imported dengue cases, or detect asymptomatic dengue cases. However, our study successfully describes the dengue endemicity in Uttarakhand, characterizes the dengue epidemics in 2010, and investigates the entomological measures of risk of dengue.

We observed that both numbers of dengue cases and numbers of deaths due to dengue are gradually increasing from 2006 to 2010. This increase may be associated with several reasons like increasing population, unhygienic living, climate change, developed disease reporting systems, better diagnostic facilities etc.  Interestingly, it is evident from the month wise distribution of dengue cases (Figure 2) that a dengue epidemic was more severe from September to November, just after the highest rainfall, probably reflecting the cultural habit of collecting rainwater, and the abundance of unmanaged water sources. These findings are consistent with those of Strickman and Kittayapong (2002).

We studied the prevalence of Aedes larvae in context to dengue epidemics in Uttarakhand. During our survey, all non-hermetically closed containers containing any volume of water were considered potential for mosquito breeding; the highest percentage of positive containers for larvae/pupae has shown up in tires, tin containers, Fridge vessels, Cement tanks and Plastic containers. Interestingly, the majority of these positive containers is related to water supply and garbage collections. These findings are in line with those reported by Focks & Chadee (1997) and Medronho et al. (2009). It is worth noting that in some areas we found 100% of surveyed containers are positive for mosquito breeding (Table 4).

We found very high entomological indices (HI, CI, and BI) for all areas. However, it is difficult to give a precise threshold level of entomological indices in epidemiological context. Dengue outbreak had previously been recorded even in very low entomological indices in Cuba (Pelaez et al. 2004), and Singapore (cf Sanchez et al. 2006). In this present study, we found that Dehradun (77% of cases) is more dengue prone area than Nainital (23% cases) (Table 3). However, all entomological indices are relatively higher in Nainital than Dehradun (Table 6, and Figure 4). Moreover, Mann-Whitney U nonparametric test revealed that the observed differences of these indices (HI, CI, and BI) between Nainital and Dehradun is not statistically significant (p>0.05). Therefore, we hypothesize that the epidemic intensity or severity of 2010-dengue epidemic in Nainital and Dehradun are independent of the rate of entomological indices. However, the layout of housing structures differs between these two districts; population density is higher or more clustered in Dehradun than the Nainital, which could also make Dehradun more dengue prone than Nainital. There is always the possibility that other factors may influence the relationship between entomological indices and intensity of dengue epidemics. These factors may be vector competence, vector bionomics and composition, virulence of the virus, immunity status of the population as whole or local health authorities in reporting correct epidemic status. Some studies have shown that  entomological indices do not seem to reliably assess dengue transmission risks (Bang & Pant 1972; Kay et al. 1987; Reiter & Gubler 1997; Focks et al. 2000), other programs have used them successfully (Pontes et al. 2000) or continue to recommend their use (WHO 1999).

We believe that where it is difficult to generalize the interpretation of the threshold indices at the global level or even country level, the threshold level of entomological indices for an outbreak of dengue should be set up locally at district or even city level since all vector borne diseases are focal and local in nature. However, entomological indices are very useful to understand the actual man-mosquito contact, abundance of adult mosquitoes, monitoring vector control programs and use as a tool for forecasting outbreak, pinpointing and mapping high-risk areas.

We observed a very high correlation between overall HI and BI values. A high correlation and divergence of HI and BI were also observed in West Indies (Thongcharoen 1993), where it was also noticed that at low rate of Aedes infestation, the HI and BI were nearly the same, whereas at higher infestation rate, a divergence between the indices was observed. Therefore, the high correlation and divergence between HI and BI in the present study indicated high rates of Aedes infestation, which may ultimately responsible for the dengue epidemics in these areas.

Finally, Dengue/DHF is a burning health problem throughout India, and cases and deaths were reported from 31 states out of 35 during 2010. The results of this study reinforce the practical importance of entomological surveillance in measuring the risk of dengue epidemics. Moreover, we have successfully analyzed the epidemic characteristics of dengue/DHF, which indicates the association of dengue with the cultural habit of collecting rainwater due to irregular water supply and the abundance of unmanaged water sources due to the irregular garbage collection. Therefore, improving these public services can improve the dengue situation.

Acknowledgment
We are thankful to LS Chauhan, Director NCDC for giving us the opportunity to carry out the study. We are also thankful to the Director, Health Services, Uttarakhand for providing all logistic and administrative support during the study. We also acknowledge the effort of all field staff of IDSP (SSU, Uttarakhand) for their help during the fieldwork.

Funding: Ministry of Health and Family Welfare (Govt. of India), Delhi, India Competing interests: None Ethical approval: Not required An authorship statement: MS conceived the study, carried out experiments, analyzed all data, and drafted the manuscript. KK conceived the study, supervised the study, analyzed data and revising the manuscript critically for intellectual content. AKG contributed in data collection, compilation, analysis and revising manuscript. All authors read and approved the manuscript. KK, AKG and MS are guarantors of the paper.

Reference
Bang Y.H., and Pant C.P., 1972, A field trial of Abate larvicide for the control of Aedes aegypti in Bangkok, Thailand, Bulletin of World Health Organization, 46: 416–425

Barcellos C., Pustai A.K., Weber M.A., and Brito M.R.V., 2005, Identification of places with potential transmission of dengue fever in Porto Alegre using Geographical Information Systems, Revista da Sociedade Brasileira de Medicina Tropical, 38: 246–250
http://dx.doi.org/10.1590/S0037-86822005000300008

Census of India (Uttarakhand) in Hindi, 2011, Available at: http://censusindia.gov.in/2011-prov-results/data_files/uttarakhand/ppt_figures_press_rel.pdf; 2011 [accessed 03.02.2012]

Deen J.L., 2004, The challenge of dengue vaccine development and introduction, Tropical Medicine and International Health, 9: 1–3.
http://dx.doi.org/10.1046/j.1365-3156.2003.01159.x

Focks D.A., and Chadee D.D., 1997, Pupal survey: an epidemiologically significant surveillance method for Aedes aegypti: an example using data from Trinidad, American Journal of Tropical Medicine and Hygiene, 56: 159-167

Focks D.A., 2003, A review of entomological sampling methods and indicators for dengue vectors, Switzerland: World Health Organization, Geneva

Focks D.A., Brenner R.J., Hayes J., and Daniels E., 2000, Transmission thresholds for dengue in terms of Aedes aegypti pupae per person with discussion of their utility in source reduction efforts, American Journal of Tropical Medicine and Hygiene, 62: 11–8

Guzman M.G., and Kouri G., 2002, Dengue: an update, Lancet Infectious Disease, 2: 33–42http://dx.doi.org/10.1016/S1473-3099(01)00171-2

Guzman M.G., Mune M., and Kouri G., 2004, Dengue vaccine: priorities and progress, Expert Review of Anti-Infective Therapy, 2: 895–911
http://dx.doi.org/10.1586/14789072.2.6.895

Kay B.H., Barker-Hudson P., Hapgood G.D., McCurley J.O., Lyons G.C. and Ives W., 1987, Aedes aegypti and dengue in the Townsville area, 1982–1985, General and Applied Entomology, 19: 2–10.

Medronho R.A., Macrini L., Novellino D.M., Lagrotta M.T.F., Câmara V.M., and Pedreira C.E., 2009, Aedes aegypti Immature Forms Distribution According to Type of Breeding Site, American Journal of Tropical Medicine and Hygiene, 80(3): 401–404

NVBDCP, National Vector Borne Disease Control Programme, 2012, Available at: http://nvbdcp.gov.in/den-cd.html [accessed 13.12.2012]

Pelaez O., Guzman M.G., Kouri G., Perez R., San Martin J.L., Vazquez S. et al., 2004, Dengue 3 epidemics, Havana, 2001, Emerging Infectious Diseases, 10: 719-722
http://dx.doi.org/10.3201/eid1004.030271

Pontes R.J.S., Freeman J., Oliveira-Lima J.W., Hodgson J.C., and Spielman A., 2000, Vector densities that potentiate dengue outbreaks in a Brazilian city, American Journal of Tropical Medicine and Hygiene, 62: 378–383

Reiter P., and Gubler D.J., 1997, Surveillance and control of urban dengue vectors, In: Gubler DJ, Kuno G, (eds), Dengue and dengue hemorrhagic fever, New York: CAB International, pp.425–62

Sanchez L., Vanlerberghe V., Alfonso L., Marquetti M.C., Guzman M.G., Bisset J., and Stuyft P., 2006, Aedes aegypti larval indices and risk for dengue epidemics, Emerging Infectious Diseases, 12 (5): 800-806.
http://dx.doi.org/10.3201/eid1205.050866

Singh R.K., Dhiman R.C., Dua V.K., and Joshi B.C., 2010, Entomological investigations during an outbreak of dengue fever in Lal Kuan town, Nainital district of Uttrakhand, India, Journal of Vector Borne Disease, 47: 189-192.

Strickman D., and Kittayapong P.K., 2002, Dengue and its vectors in Thailand: Introduction to the study and seasonal distribution of Aedes larvae, American Journal of Tropical Medicine and Hygiene, 67(3): 247-259

Tauil P.L., 2002, Critical aspects of dengue control in Brazil. Rep Public Health, 18: 867–871
Thongcharoen P., 1993, Monograph on Dengue/ Dengue Haemorrhagic Fever, Vol No. 32, New Delhi: WHO SEARO

Tun-Lin W., Kay B.H., Barnes A., and Forsyth S., 1996, Critical examination of Aedes aegypti indices: correlations with abundance. American Journal of Tropical Medicine and Hygiene, 54: 543–7
WHO-South East Asian Regional Office, 2010, Available at: http://www.searo.who.int/LinkFiles/Dengue_Dengue_update_SEA_2010.pdf; 2010 [accessed 13.12.2011]

World Health Organization, 1999, Dengue/dengue haemorrhagic fever prevention and control programme in Thailand, WHO Project: ICP CTD 001, SEA/Haem, Fev./68, SEA/VBC/68, Geneva: World Health Organization, p.1–47