Geospatial disease risk modeling for the identification of potential areas of encephalitis in a subtropical region of India: a micro-level case study of Gorakhpur tehsil

Encephalitis is a complex disease with a high mortality, morbidity and disability ratio. It is basically a paediatric problem. Every year, from July to November, the outbreak of encephalitis causes many deaths and long-term disabilities among children and young adults. Geographical and seasonal distribution of encephalitis shows that the disease has a positive correlation with the environmental variables, such as rainfall, humidity, temperature and waterlogging) as well as with socio-economic variables like population below the poverty line, population up to 15 years of age, malnutrition, pig/cattle living near/inside the house and housing conditions, which may have a significant impact on the timing, incidence and magnitude of the disease outbreak and vulnerability of the population. Identification of encephalitis high-risk areas is an important and effective tool for policy makers to provide more effective preventive methods, health services, vaccination programmes, vector control, and other public health initiatives. Uttar Pradesh is a highly encephalitis affected state among all the states of India and its eastern part is a highly encephalitis-sensitive area. The study tries to create an Encephalitis Risk Model for the identification of potential areas of encephalitis in a subtropical region, located in the eastern part of Uttar Pradesh, state (Gorakhpur tehsil) of India. Factors selected to create risk model for encephalitis are inhabited/uninhabited (abandoned) villages, land use, pig population distribution, occurrence of encephalitis cases and population under age of 6. Every variable has been given due weightage and relative importance according to their impact on the occurrence and incidence rate of encephalitis. Encephalitis Risk Model for Gorakhpur tehsil classifies the area into five categories of risk zones. About 62.08% of the area of the tehsil is under low/no risk category and only 14.75% area constitutes the high-risk area. The rest of the 23.17% area is in the moderate-risk category. The result indicates that the disease risk areas are almost homogenously spread over the region with several small patches reaching the extreme value of risk and making them highly sensitive. The model is validated by using 45 sites of sample villages and wards. The multiple linear regression is calculated to know the correlation between the selected variables and their influence values and Pearson’s and Spearman’s correlations of coefficient are used for es