An information value based analysis of physical and climatic factors affecting dengue fever and dengue haemorrhagic fever incidence

BACKGROUND: Vector-borne diseases are the most dreaded worldwide health problems. Although many campaigns against it have been conducted, Dengue Fever (DF) and Dengue Haemorrhagic Fever (DHF) are still the major health problems of Thailand. The reported number of dengue incidences in 1998 for the Thailand was 129,954, of which Sukhothai province alone reported alarming number of 682. It was the second largest epidemic outbreak of dengue after 1987. Government arranges the remedial facilities as and when dengue is reported. But, the best way to control is to prevent it from happening. This will be possible only when knowledge about the relationship of DF/DHF with climatic and physio-environmental agents is discovered. This paper explores empirical relationship of climatic factors rainfall, temperature and humidity with the DF/DHF incidences using multivariate regression analysis. Also, a GIS based methodology is proposed in this paper to explore the influence of physio-environmental factors on dengue incidences. Remotely sensed data provided important data about physical environment and have been used for many vector borne diseases. Information Values (IV) method was utilised to derive influence of various factors in the quantitative terms. Researchers have not applied this type of analysis for dengue earlier. Sukhothai province was selected for the case study as it had high number of dengue cases in 1998 and also due to its diverse physical setting with variety of land use/land cover types. RESULTS: Preliminary results demonstrated that physical factors derived from remotely sensed data could indicate variation in physical risk factors affecting DF/DHF. A composite analysis of these three factors with dengue incidences was carried out using multivariate regression analysis. Three empirical models ER-1, ER-2 and ER-3 were evaluated. It was found that these three factors have significant relation with DF/DHF incidences and can be related to the forecast expected number of dengue cases. The results have shown significantly high coefficient of determination if applied only for the rainy season using empirical relation-2 (ER-2). These results have shown further improvement once a concept of time lag of one month was applied using the ER-3 empirical relation. ER-3 model is most suitable for the Sukhothai province in predicting possible dengue incidence with 0.81 coefficient of determination. The spatial statistical relationship of various land use/land cover classes with dengue-affected areas was quantified in the form of information value received from GIS analysis. The highest information value was obtained for the Built-up area. This indicated that Built-up area has the maximum influence on the incidence of dengue. The other classes showing negative values indicate lesser influence on dengue epidemics. Agricultural areas have yielded moderate risk areas based on their medium high information values. Water bodies have shown significant information value for DF/ DHF only in one district. Interestingly, forest had shown no influence on DF/DHF. CONCLUSION: This paper explores the potential of remotely sensed data and GIS technology to analyze the spatial factors affecting DF/DHF epidemic. Three empirical models were evaluated. It was found that Empirical Relatrion-3 (ER-3) has yielded very high coefficient of determination to forecast the number of DF/DHF incidence. An analysis of physio-environmental factors such as land use/ land cover types with dengue incidence was carried out. Influence of these factors was obtained in quantitative terms using Information Value method in the GIS environment. It was found that built-up areas have highest influence and constitute the highest risk zones. Forest areas have no influence on DF/DHF epidemic. Agricultural areas have moderate risk in DF/DHF incidences. Finally the dengue risk map of the Sukhothai province was developed using Information Value method. Dengue risk map can be used by the Public Health Department as a base map for applying preventive measures to control the dengue outbreak. Public Health Department can initiate their effort once the ER-3 predicts a possibility of significant high dengue incidence. This will help in focussing the preventive measures being applied on priority in very high and high-risk zones and help in saving time and money.