Open AccessGeographically Weighted Negative Binomial Regression Modeling of Tuberculosis Cases with Distribution Evaluation
Author(s) -
Choirun Nisa,
Muhammad Nur Aidi,
I Made Sumertajaya
Publication year - 2020
Publication title -
international journal of scientific research in science, engineering and technology
Language(s) - English
Resource type - Journals
eISSN - 2395-1990
pISSN - 2394-4099
DOI - 10.32628/ijsrset1207473
Subject(s) - negative binomial distribution , statistics , tuberculosis , mathematics , population , regression analysis , weighting , demography , veterinary medicine , medicine , environmental health , pathology , sociology , poisson distribution , radiology
Tuberculosis (TB) is contagious disease caused by the bacteria mycobacterium tuberculosis and is one of the top 10 causes of death in the world. Central Java is included as one of the three provinces with highest number of TB cases in Indonesia. End the TB epidemic by 2030 is the final goal of Sustainability Development Goal and Indonesia has set target for TB elimination by 2035. The number of TB cases is non-negative count data. The distribution pattern of the count data needs to be noticed in order to produce valid analysis. Based on calculation of the VMR (Variance Mean Ratio) value and suitability test, the data on the number of TB cases follows negative binomial distribution. The infection of TB disease tends to be clumped and is affected by geographical factors (environmental, social, and economic). This study aims to determine factors that affecting TB cases through Geographic Weighted Negative Binomial Regression (GWNBR) model approach which considering spatial aspects. Based on the ratio of AIC and BIC value, GWNBR model with an adaptive g aussian kernel weighting gives the best results. The affecting factors are the number of hospitals , the percentage of population with good water access , the population density , the percentage of household with a distance of drinking water sources and feces septic tank less than 10 meters.