Predictors of smear non-conversion among new-treatment pulmonary tuberculosis: a single center case-control study in Indonesia

Abstract

BACKGROUND Previous studies concluded predictors of smear non-conversion pulmonary tuberculosis (TB) globally as well as in Indonesia. However, there is a limited data in hospital setting. This study aimed to explore predictors of smear non-conversion pulmonary TB in hospital setting. METHODS A case-control study was conducted to explore predictors of smear nonconversion among new-treatment pulmonary TB in Dr. Kariadi General Hospital from 2017 to 2019. Number of cases and controls have been determined consecutively. Data were collected from secondary data accessed in medical records and directly from subjects. Non-conversion status in the case group was defined as a persistent sputum smear-positive after 2 months of intensive phase of treatment. The subjects’ characteristics i.e., age, sex, BMI, education level, occupational status, and predictors of smear non-conversion, i.e., patient’s compliance, smoking status, alcohol consumption, presence of drugs side effects, health care access, first acid-fast bacilli (AFB) smear grading, diabetes mellitus (DM), housing condition, housing density, and household income were observed. Chi-square test and binary logistic regression were used. RESULTS 35 subjects were determined in the case group while 76 subjects were the control group and involved in the final analysis. Age, sex, first AFB smear grading, smoking status, housing condition, housing density, and DM were involved in the model of logistic regression. DM (OR = 3.4; 95% CI = 1.19–10.00) and first AFB smear grading (OR = 11.2; 95% CI = 3.86–33.00) were concluded as the predictors of smear nonconversion. CONCLUSIONS DM and first AFB smear grading were the predictors of smear nonconversion among new-treatment pulmonary TB subjects.