Implementation of Chi Square Automatic Interaction Detection (CHAID) Method to Identify Type 2 Diabetes Mellitus in Tuberculosis Patient. A Case Study in Cipto Mangunkusumo Hospital

A Pulmonary Tuberculosis (pulmonary TB) is a chronic infectious disease caused by Mycobacterium tuberculosis. Chronic infection cause the body in a state of oxidative stress. In the state of stress, stress hormone production increases and can affect the increase of blood sugar levels which then trigger the occurrence of diabetes mellitus (DM). The purpose of this study is to determine the factors associated with the emergence of DM Type 2 and make a classification to characterize DM Type 2 in patients with pulmonary TB. In this research, the data used are secondary data of pulmonary TB patients obtained from Cipto Mangunkusumo Hospital (RSCM) for six years, 2012 - 2017. Chi Square Automatic Interaction Detection (CHAID) method is used to classify categorical data by dividing the data set into subgroups. The dependent variable is type 2 DM status, and the independent variables are gender, age, body mass index, level of neutrophil, level of lymphocytes, erythrocyte sedimentation rate, and anti-tuberculosis medicines such as rifampicin, isoniazid, pyrazinamide, ethambutol, and streptomicin. There are seven classes are obtained by classification using CHAID. The result from analysis of CHAID method shows that the factors related to the occurrence of type 2 DM on TB patients are age, body mass index, gender and pirazinamid (Z). Based on CHAID method, a classification of the occurrence of type 2 DM on pulmonary TB patients is obtained, which is pulmonary TB patients who are ≥ 40 years old, have body mass index ≥ 25 kg/m ² , and the most patients are male (Class-4 ^th ).