Modeling tuberculosis in children under five using poisson and negative binomial regression

Tuberculosis is an infectious disease caused by Mycobacterium tuberculosis. Indonesia is the country with the second highest number of tuberculosis cases after India. The Ministry of Health stated that there has been a significant increase in cases of tuberculosis among children in Indonesia where the increase in cases of tuberculosis among children has reached more than 200 percent. The number of TB cases can be reduced if the factors that affect the number of TB patients are known. Therefore, efforts should be made to model the number of cases of Tuberculosis among children under five years of age to provide useful information to prevent and control Tuberculosis. The relationship between these factors and the number of people with Tuberculosis can be determined using Poisson regression analysis because the number of cases of Tuberculosis is calculated data. Tuberculosis data contain overdispersion, so another approach is used to overcome it, which is by using a negative binomial regression model. The best model obtained based on the AIC value is the Negative Binomial regression model with an AIC value of 184.095. For further research, it is suggested to test the spatial effect and modeling using the Negative Binomial geographic weighted regression method to find out whether the characteristics of one region and the other influence the geographic location on the model.