Spatio-temporal assessment of hotspots and seasonally adjusted environmental risk factors of malaria prevalence

Abstract

While the changing bio-climate may impact regional malaria transmission, results may be conflicting. We, therefore, examined spatiotemporal hotspots and risk factors for malaria prevalence. Land surface temperature (LST), elevation (GDEM), and enhanced vegetation index (EVI) were obtained from the USGS portal. Mean air temperature and monthly mean rainfall amount were acquired from WorldClim, whereas stream density and distance from streams were computed from stream networks acquired from Diva GIS. Also, malaria cases were obtained from the DHIMS-2. We used the Getis-Ord G* statistic for hotspot analysis, while the seasonal bio-climatic impacts were analyzed using GWR. The spatial cluster of malaria prevalence has increased with time (Z = 1.853; p = 0.06 in 2012 to Z = 8.921; p < 0.001 in 2021). In the wet season, elevation (β = 0.176; se = 0.0057), rainfall (β = 0.022; se = 0.0083), EVI (β = 0.451; se = 0.0095), and streams density (β = 0.002; se = 0.0057) were positively associated with malaria. LST (β = −0.224; se = 0.008), distance from streams (β = −0.169; se = 0.007), and mean temperature (β = −0.006; se = 0.0071) were inversely related. For the dry season, elevation (β = −0.204; se = 0.0078), EVI (β = −0.331; se = 0.0068), rainfall (β = −0.121; se = 0.0059), mean temperature (β = −0.561; se = 0.0083), distance from the stream (β = −0.374; se = 0.0045), and LST (β = −0.472; se = 0.039) were inversely related to malaria. Our study provides two sets of information; malaria hotspots and seasonally adjusted environmental risk factors. These findings may be useful for policing.