Spatial Autocorrelation for Malaria Hotspots Identification in Mekong Basin, Ubon Ratchathani, Thailand
Jaruwan Wongbutdee MSc*, Wacharapong Saengnill MSc*
Affiliation : * College of Medicine and Public Health, Ubon Ratchathani University, Ubon Ratchathani, Thailand
Background : Malaria is a major public health problem in Thailand transmitted by the Anopheles mosquito. Spatial
autocorrelation effectively measures epidemiology, distribution of diseases, and is important in the understanding of the
distribution of malaria spatial patterns.
Objective : This study aimed to detect malaria hotspots in the Mekong basin in Ubon Ratchathani using spatial statistical
analyses between 2011 and 2014.
Material and Method: Data of malaria cases and incidence rates were collected during years 2011 to 2014. The study area
had 42, 130, 175, and 119 cases in 2011, 2012, 2013, and 2014, respectively. Moran’s I was used to calculate the correlation
of spatial autocorrelation. Getis-Ord Gi* identified the differentials of the patterns of hotspots.
Results : The incidence rates showed 59.38, 183.23, 253.31, and 160.27 per 10,000 population in years 2011, 2012, 2013, and
2014, respectively. The results in 2012, 2013, and 2014 were 0.347, 0.606, and 0.393, respectively of Moran’s I value.
Therefore, the Z-score statistics for these years showed cluster patterns of strong significance (p-value <0.01) while 2011 had
random patterns. The Getis-Ord General G* statistics showed that 2013 had the highest number of hotspots, followed by
2012 and 2014.
Conclusion : Spatial autocorrelation can estimate unknown located spatial data from the values of located attributed data
with reference to the trend and distribution of malaria. Therefore, the information can support and confirm incidences and
outbreaks of malaria cases. Spatial autocorrelation is utilized for decision-making in the support of plans, epidemiological
control, and surveillance of malaria.
Keywords : Malaria incidence, Spatial autocorrelation, Hotspots
