Chitti Chansang. Application of geographic information system (GIS) and remote sensing (RS) for evaluating dengue risk in Thailand . Doctoral Degree(Biology). Mahidol University. : Mahidol University, 2005.
Application of geographic information system (GIS) and remote sensing (RS) for evaluating dengue risk in Thailand
Abstract:
Dengue hemorrhagic fever (DHF) remains a public health problem. In the absence of vaccines, studies on methods of vector surveillance, finding risk factors, and predicting risk areas of DHF should be useful for integrating vector control efforts.
Global Positioning System (GPS) unit was used to record the locations of the survey houses and the abundance of Aedes immatures in Plaeng Yao District, Chachoengsao Province, Thailand. The resulting maps from Geographic Information System (GIS) showed high Aedes densities in the middle of the village. The relationship between the complete count and the sampling methods of Aedes immature stages was determined. The two surveillance methods were highly correlated. From the linear regression analysis, results suggested that the sampling method could be used for surveying Aedes immatures and the data obtained could be used to create the GIS map which then could be used in planning effective vector control.
To determine risk factors for dengue, the study area was chosen based on a serosurvey of anti-dengue immunoglobulin(Ig)G/IgM in primary school children. The focal areas of dengue transmission were designated to be the areas within a 100 meter radii of the 23 houses that had IgG/IgM-positive cases in 13 villages of the study area. Both entomological and environmental surveys were conducted in these focal areas and in the control areas where there were no IgG/IgM-positive cases. Comparison of the parameters between the treatment and control areas was made by using the Odds ratios, and 3 risk factors were identified. These factors were sanitation, distribution of houses, and standard water jars with or without larvae. The focal areas of dengue transmission had lower larval density per positive container than the control areas. In contrast, female mosquito wing lengths in the treatment areas were larger than those in the control areas. These results suggested that Aedes control should not only reduce the number of larvae but should eliminate all larvae from breeding places, since low larval density may produce larger mosquitoes which were better in vector competence.
To predict risk areas of DHF, the GPS was used to record the locations of houses of DHF cases between 1997-2002 in the study area. A map of DHF clusters was created using Kernel estimation. A case-control study using multiple logistic regression was designed. Local environmental conditions were characterized from the Landsat 7 satellite image. The best model developed from GIS, RS and statistical modelling predicted that DHF risk areas decreased with increasing greenness and increased with increasing thermal output. The model was used to create the risk map and areas where DHF risk were highly focal. Results suggest that the satellite image is useful in identifying local areas at risk for DHF and can be used to target vector control in these areas.