Statistical-Pharmacoepidemiology and Decision Tree Analysis of Sputum Smear-Positive Tuberculosis with Comorbid Diabetes Melitus Patient in Intensive Phase

• Main Author: Rahmatina, Erba
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/45536
• Institution: Institut Teknologi Bandung

Tuberculosis (TB) remains as one of the leading infectious causes of death in the world. Diabetes Melitus (DM) is a risk factor of TB. DM can affect some of TB treatment outcomes. The high prevalence of DM and TB in Indonesia makes a higher risk of TB-DM comorbidity. This study was intended to identify the factors and probabilities of TB treatment success or failure in TB patients with comorbid DM. This study was a retrospective cross-sectional study in new case pulmonary TB patients with sputum smear-positive with or without comorbid DM by assessing the patients’medical records in Salamun Hospital, Bandung from Januari 2010 to October 2014 and assessing the previous studies’data. The subjects were divided into two groups : the experiment group (TB with comorbid DM) and control group (TB without comorbid DM). Data was statistically analyzed and the probabilities of TB treatment outcomes were determined by using decision tree analysis. Sputum conversion at the end of intensive phase and length of TB treatment were statistically different for the two groups (P = 0.002 and P = 0.005). Types of antidiabetic drugs had a statistically significant relationship with the sputum conversion at the end of intensive phase (P = 0.009), fasting blood glucose (P = 0.000), and post prandial blood glucose (0.001). Age, fasting blood glucose, and BMI had a statistically significant relationship with the length of TB treatment (P = 0.038; P = 0.014; P = 0.004). Post prandial blood glucose had a statistically significant relationship with BMI (P = 0.000). Based on the probability analysis, fasting blood glucose and types of antidiabetic drugs affected the success or failure of TB treatment. Oral antidiabetic drug had the highest probability in having patients with 6 month of TB treatment with controlled fasting blood glucose or controlled post prandial blood glucose and normal BMI. In all of blood glucose and BMI status, the combination between insulin and oral antidiabetic drugs had the highest probability in having patient with 6 months of TB treatment, followed by the monotherapy of oral antidiabetic group and monotherapy of insulin group. In unknown BMI status, the probability of TB length treatment varied, but in known BMI status, most patients had TB treatment for 6 month which highest probability was in patient with normal BMI. Types of antidiabetic drugs, fasting blood glucose, and BMI affected the outcome of TB treatment, especially the length of TB treatment.