The Poisson inverse Gaussian regression model in the analysis of dengue hemorrhagic fever case in Central Java.

One of Indonesia's efforts to solve health problems is to deal with neglected tropical diseases and water-borne diseases, one of which is dengue hemorrhagic fever (DHF). Every year cases of dengue fever have increased. In addition to considering the number of cases, the case fatality rate is also a measure of the severity of dengue cases. Central Java is the province with the first highest case fatality rate, whereas DHF cases in Central Java are the eight lowest in Indonesia. In order to reduce the mortality rate, the modeling of the factors causing DHF was carried out. Because the data on DHF cases is in the form of counts, the modeling method can use Poisson regression. However, because the variance is greater than the mean (overdispersion), the assumption is not fulfilled. Therefore, the mixed Poisson method is used, namely the Poisson Inverse Gaussian (PIG). Several studies have shown that the application of the PIG regression method produces higher accuracy compared to other regression methods such as Negative Binomial. Therefore, the application of the PIG regression method for DHF cases in Central Java is an innovation that provides more accurate and relevant results. Parameter estimation using Maximum Likelihood Estimation (MLE) method with hypothesis testing using Maximum Likelihood Ratio Test (MLRT). Based on data analysis conducted using R software, the altitude of the area significantly affects the number of dengue fever in Central Java in 2021 by 0,205. [ABSTRACT FROM AUTHOR]