TY - JOUR
T1 - Modeling the Prevalence of Tuberculosis in Java, Indonesia
T2 - An Ecological Study Using Geographically Weighted Regression
AU - Putra, I. Gusti Ngurah Edi
AU - Rahmaniati, Martya
AU - Sipahutar, Tiopan
AU - Eryando, Tris
N1 - Funding Information:
We thank the Ministry of Health and Statistics Indonesia for making the reports, which this study was based on, publicly accessible.
Publisher Copyright:
© 2022
PY - 2022
Y1 - 2022
N2 - There is a paucity of studies investigating the spatial pattern and factors associated with the prevalence of tuberculosis (TB) in Java. This study aimed to identify spatial autocorrelation, clusters, and factors associated with TB prevalence in Java using districtor city-level data. This was an ecological study using data from 118 districts or cities across six provinces in Java. Spatial analyses (i.e., Global Moran’s I, Local Indicator Spatial Autocorrelation [LISA], and Geographically Weighted Regression [GWR]) were used. This study found positive spatial autocorrelation of TB prevalence in Java (Global Moran’s I = .45, p= .001). Statistically significant high-high clusters (p< .05) were identified in some districts within the capital city of Jakarta, Banten, and West Java provinces. The GWR model with the Bi-square Kernel weighting function was selected as the best model to predict the prevalence of TB (R 2 = 37.50%, AIC = -59.94%). Findings from the GWR model indicate that the average number of years in education, the percentages of households with floor space per capita<8 m2 and reporting easy access to health care facilities were associated with the prevalence of TB in some districts within West and Central Java provinces. Therefore, considering district differences in factors associated with TB prevalence, locally-focused interventions are worth considering.
AB - There is a paucity of studies investigating the spatial pattern and factors associated with the prevalence of tuberculosis (TB) in Java. This study aimed to identify spatial autocorrelation, clusters, and factors associated with TB prevalence in Java using districtor city-level data. This was an ecological study using data from 118 districts or cities across six provinces in Java. Spatial analyses (i.e., Global Moran’s I, Local Indicator Spatial Autocorrelation [LISA], and Geographically Weighted Regression [GWR]) were used. This study found positive spatial autocorrelation of TB prevalence in Java (Global Moran’s I = .45, p= .001). Statistically significant high-high clusters (p< .05) were identified in some districts within the capital city of Jakarta, Banten, and West Java provinces. The GWR model with the Bi-square Kernel weighting function was selected as the best model to predict the prevalence of TB (R 2 = 37.50%, AIC = -59.94%). Findings from the GWR model indicate that the average number of years in education, the percentages of households with floor space per capita<8 m2 and reporting easy access to health care facilities were associated with the prevalence of TB in some districts within West and Central Java provinces. Therefore, considering district differences in factors associated with TB prevalence, locally-focused interventions are worth considering.
KW - Geographical access
KW - household density
KW - Java
KW - spatial analysis
KW - tuberculosis
UR - http://www.scopus.com/inward/record.url?scp=85132535817&partnerID=8YFLogxK
U2 - 10.25133/JPSSv302022.041
DO - 10.25133/JPSSv302022.041
M3 - Article
AN - SCOPUS:85132535817
SN - 2465-4418
VL - 30
SP - 741
EP - 763
JO - Journal of Population and Social Studies
JF - Journal of Population and Social Studies
ER -