Quantitative Research On Dengue Disease: Kingdom Of Cambodia
Introduction
Dengue is one of the most rapidly spreading mosquito-borne viral disease in the world, and is caused by bites of infected Aedes mosquitoes, principally Aedes Aegypti mosquitoes. This disease outbreaks occur in cycles every 3–5 years and during dengue outbreaks the government does vector control activities to the entire affected area. This method spends a lot of money and human resources because we don't have a method to accurately predict which houses are at high risk for infestation of Aedes aegypti mosquitos. Therefore, this study is being conducted to help government or other health authorities to identify houses which are likely to have dengue vectors and reduce expenses and time required to control outbreaks. This research study will contribute to demonstrating the correlation between Premise Condition Index (PCI) and presence of Aedes aegypti mosquitos in each household in Kampong Cham province.
Methods
The cross-sectional study design was used in this study. Kampong Cham was purposively selected as they have one of the highest dengue incidence rates in Cambodia and their entomological and environmental characteristics are similar to most dengue-endemic areas of Cambodia. This study took place in 30 clusters from two operational districts in Kampong Cham province. The survey randomly selected 40 HH from each cluster. The survey included adult resting catches and larvae/pupae collection from jars and record the number and size of all water containers found in the household. The team also completed the PCI a rapid assessment tool. ResultAmong 1200 households, the total houses with PCI scores equaling 8 (OR=4. 79, P-value=0. 048) or 9 (OR=12. 75, P-value=0. 031), and number of residents (OR=1. 46, P-value=0. 003) were associated with presence of Aedes aegypti mosquitoes. Conclusion The present study showed the effectiveness of the PCI because worse conditions in houses showed a higher correlation with the presence of Aedes aegypti mosquitoes than in well maintained houses. The major advantage of the PCI is that it offers a rapid assessment method for selecting houses for survey or to predicting which houses should be targeted for vector control activities. In the case of dengue outbreaks, by having all representative house indices of the region, it will be much easier and less expensive to control the epidemic.
Introduction
Background
Dengue is the most rapidly spreading mosquito-borne viral disease in the world, and is caused by bites of infected Aedes mosquitoes, principally Aedes Aegypti. Dengue infection is a systemic and dynamic disease with a wide clinical spectrum that includes both severe and non-severe manifestations and in some cases can lead to death. With an estimated 3. 6 billion people in 124 countries at risk of contracting the disease and 390 million dengue infections occurring each year (of which 96 million are clinically apparent) the dengue virus has become a leading cause of illness and death in the tropics and subtropics. Asia records 70% of the global disease burden due to dengue, and Cambodia has one of the highest per-capita incidence rates in the region. Identified in Cambodia in 1966, a total of 194,726 dengue cases were reported to the National Dengue Control Program (NDCP) between 1980 and 2008. Between 2003 and 2008, annual dengue incidence ranged between 0. 7 and 3. 0 per 1,000 persons, the cost to society being between $3,327,284 and $14,429,513. Since most of this cost falls onto the family, it is estimated that 67% of affected households fall into debt to pay for medical bills.
However, it is likely that the real number of cases and cost to society is much greater, with some studies suggesting the real case numbers are between 3. 9 and 29. 0 times or 9. 1 times higher than those of the National Dengue Surveillance System. Kampong Cham is the one of the highest dengue incidence rates in Cambodia. In 2015, the incidence rate in Kampong Cham was 91. 8 per 100,000 population. Although a number of promising vaccine candidates are in preclinical and clinical development, and methods of genetic control of mosquitoes are being developed, they are years from operational roll-out in Cambodia and will be unlikely to provide universal protection. Without a cure or vaccine the best measures to prevent new infections of dengue are through vector control and the avoidance of mosquito bites. Several vector control methods have been studied in Cambodia including use chemical and biological substances (temephos, pyriproxfen, and Bacillus thuringiensis israelensis), jar covers, distribution of copepods and larvivorous fish.
Rationale
Dengue outbreaks occur in cycles every 3–5 years and during dengue outbreaks the government does vector control activities to the entire affected area. This method spends a lot of money and human resources because we don't have a method to accurately predict which houses are at high risk for infestation of Aedes aegypti mosquitos. Therefore, this study is being conducted to help government or other health authorities to identify houses which are likely to have dengue vectors and reduce expenses and time required to control outbreaks. This research study will contribute to demonstrating the correlation between PCI and presence of Aedes aegypti mosquitos in each household in Kampong Cham province.
Literature Review
A study in Australia suggests that the major advantage of the Premise Condition Index (PCI) is that it offers a rapid assessment method of selecting premises for vector control. It does not require identification or counting of larvae and containers. It can be done without entering people’s yards and therefore is fast. By scoring the conditions of the house and yard, as well as the degree of shade, this index offers a quantitative framework for streamlining and improving the accuracy of premise surveys. In north Queensland where outdoor breeding occurs, poorly maintained or dilapidated premises with plants and shrubbery were more likely to be positive for Aedes aegypti immatures (Tun-Lin et al. 1995). Similarly, study in Brazil by Andrighetti et al shows that the surveillance tool and associated data on resident housing characteristics were useful in identifying heavily infested premises. The index could thus provide a basis for prioritizing control measures within communities during routine vector control activities. Basker et al. showed that in India PCI helps in conducting rapid surveys to identify the presence of Aedes larvae with a minimum number of staff for both inspection and treatment of Aedes larvae during an epidemic. Nogueira, and Gushi, et al. suggest PCI relate to conditions of property, such as houses and yards, and the degree of shade with the occurrence of Aedes sp. oviposition, and is calculated as scores from 3 to 9. The lowest score indicates property in good condition and an unfavorable breeding environment, while the highest score indicates property at high risk for infestation by Aedes sp. PCI was shown effective because worse conditions in houses showed a higher correlation with the presence of Aedes mosquitoes than in well maintained houses.
Data Source
In this study, secondary data from an entomology survey conducted in 30 clusters from two operational districts in Kampong Cham province were used.
Study Design and Setting
A cross-sectional study design was used in this study. The study were take place in 30 clusters (containing one or more villages) in two operational districts (one operational district includes the jurisdiction of 10 health centers or roughly 100,000-200,000 people) within Kampong Cham province. Each village has approximately 200 households or 1,000 individuals. Kampong Cham is located on the central lowlands of the Mekong River. The rainy season runs from April to November, and the peak dengue season is from May-July. Kampong Cham was purposively selected as it has one of the highest dengue incidence rates in Cambodia and the environmental characteristics are similar to most dengue-endemic areas of Cambodia (personal communication, Ngan Chantha, 2014).
Sampling and Sample size
As of 2013, it is subdivided into 10 districts which in turn are subdivided into 109 communes and further divided into 916 villages with a total population of about 1,264,000. There are a variety of water supply sources including wells, water tanks, earthen jar and tap water schemes. However, the majority of residents were collect rain water and store it in 200-400 liters cement jars or water tanks. Every house has numerous cement jars regardless of whether they have piped water or get water from a well or rain. It is estimated that about 80% of mosquito breeding takes place in water jars and water tanks in suburban and rural households in Cambodia. Kampong Cham were purposively selected as they have one of the highest dengue incidence rates in Cambodia and their entomological and environmental characteristics are similar to most dengue-endemic areas of Cambodia. The clusters were selected based on availability of entomological data from previous surveys, and being isolated from nearby villages. Isolation was considered to avoid contamination and was interpreted as being at least 200 meters from the nearest households as Aedes in this region have an average flight range of 50-100m. We selected 30 clusters and a sample size of 10 clusters per arm and 40 HHs per cluster for the survey was devised using the method developed by Hemming and Marsh. The calculation assumed a mean of 0. 1 adult resting Aedes per household in the intervention arms compared to 0. 25 in the control arm for each collection. The households were randomly selected each collection. This assumption was based on the results from the earlier World Health Organization/Asian Development Bank guppy fish project in the same province and to be conservative assumed no impact from the PPF in arm 1. The intracluster correlation was assumed to be 0. 01 based on previous studies, and the coefficient of variation in cluster sample size was assumed to be 0. 1. The latter value is expected be small as we plan to sample the same number of houses from each cluster. Power was set at 90%, and significance set at 0. 05.
Data collection
The survey included adult resting catches and larvae/pupae collection from jars. The survey teams were consisting of experienced government staff who received three days of training before the start of activities. All tools and materials were done pre-tested at the entomology training. The same team used to complete each entomology survey. Houses within each cluster were selected using a random number generator applied to the village list kept by the village head. The adult resting catch completed using a battery-powered, portable aspirator (Camtech, Phnom Penh, Cambodia) for 10 minutes in the bedrooms and living spaces in each house, starting in the bedroom and aspirating up and down the wall (from floor to 1. 5 m) around the home in a clockwise manner. The mosquitos kept in a cold box and transported to the provincial laboratory for identification to the species level for Aedes, otherwise to genus. After identification they stored in a -20 C freezer and taken to NAMRU-2 for PCR analysis.
The larvae and pupae collection completed using the five sweep net method for large containers. The size of the net is 20 cm by 33 cm. We turn in anti-clockwise manner 5 times, then wait 1 minute and do one sweep from the bottom. This method can sample around 35% of larvae and 31% of pupae in round ceramic jars, and the total number is estimated by an adjustment factor. For smaller containers, we empty all the water into the sweep net and collect all the larvae/pupae. All containers around randomly selected houses were inspected. All pupae and 10 larvae were put in a plastic Ziploc bag, labeled (with date and ID code), and taken back to the laboratory for identification. There were a quality control check at NAMRU-2 for both immature and adult mosquitos. The survey team also record the number and size of all water containers found in the household. The team also complete a rapid assessment tool called premise condition index (PCI) [26] to try and understand how effectively the scores can predict mosquito populations in dengue-endemic areas of Cambodia. If proven useful as an indicator of risk for Aedes aegypti infestation PCI could be used to streamline surveys to inform dengue program activities at least where Ae. aegypti breeds outdoors, and possibly reduce program costs.
Data Management
Double data entry was entered into Epidata and completed by an experienced data entry company. Restrictions on which values can be entered into each field were placed in the data entry frame. The process of data cleaning were handled by the experienced data management team at Malaria Consortium. The original forms were kept in a locked cabinet in the Malaria Consortium Phnom Penh office, and will be available if needed during the data cleaning phase. Participant files were maintained in storage for a period of two years after completion of the study.
Study Goal and Objective
The goal of the study is to create evidence for new methods of predicting which houses should be targeted for vector control activities in the face of restricted funding. The main objective in this study is to determine the correlation between the PCI and the presence of adult Aedes aegypti mosquitoes in each household.
Conceptual Framework
Figure 1 shows the casual diagram of correlation between PCI and presence of adult Aedes aegypti mosquitoes. This diagram was drawn using a software called “DAGitty v2. 0” [30] which specifically created for casual diagram drawing in epidemiological studies. Using this software, the colors automatically display between exposure, outcome, potential confounders, ancestors of outcome, and ancestor of exposure. As it can be seen in this diagram PCI (green circle) is the main exposure of interest while presence of adult Aedes aegypti mosquitoes is outcome of interest (blue circle). The potential confounders (dark pink circle) include; number of wet container, number of residents, house elevation, house type, exterior of home painted, and interior of home painted.
PCI is an approach that may help in providing guidelines for locating positive houses as an alternative index. By scoring the conditions of the house and yard, as well as the degree of shade, this index offers a quantitative framework for streamlining and improving the accuracy of premise surveys. Prior to collecting the details of PCI variables field teams were familiarized with their definition. House was considered where humans live. Its condition is being determined by the periodical maintenance of house, yard (this is some space available either in front of the house or at the back of the house) and the degree of shadow is determined based on the vegetation in and around the house.
Methods
Data Analysis
The STATA statistical program, version 12. 0 was used for statistical analysis. We used descriptive statistics to describe the general characteristics of the houses sampled. The logistic regression analyses were used to examine the correlation between PCI and presence of adult Aedes aegypti mosquitoes while controlling for the number of containers with water, number of residents, house elevation, house type, exterior of home painted, and interior of home painted. The significance of individual coefficients in the logistic regression models were determined by the 95% confidence limits for odds ratios.
Results
Among 1200 households sampled 90. 31% were on stilts and 76. 36% made from woods. Nineteen percent of houses had exterior paint and 6. 59% had interior paint. Fifty percent of households had five and more individuals that slept in the house the night before. Furthermore, 73. 92% of households had three or more containers with water. Additionally, 29. 08% of households had Adult Aedes aegypti infestation. The majority of households had a PCI score of 5 (29%) or 6 (42%).
