ABSTRACT
Dengue Hemorrhagic Fever (DHF) in Indonesia has increased steadily with Bandung as a hyper-endemic area holding a high number of cases for years. This study aimed to identify cluster areas and their correlation with land use changes which was indicated by changes of Normalized Difference Vegetation Index (NDVI). Hospital surveillance of 28,327 cases during 2008-2013 was geo-coded into sub-district levels and analyzed to find cluster areas over time and space using SaTScan and ArcGIS. Spearman correlation was used to analyze NDVI with Incidence Rate (IR) in each area. IR of DHF cases tended to increase over 6 years during high precipitation period. Cases were concentrated in several cluster areas in 2009 then moved to eastern part of the city in 2013. NDVI had negative correlation with IR in 2008 (r = -0.258; p = 0.001) and positive correlation in 2012 (r = 0.193; p = 0.017). Clear geographical pattern by cluster identification overtime is beneficial for targeting appropriate vector-control program.
