Determining factors of lung tuberculosis among children in community health centers: A cross-sectional study

Tuberculosis is a serious global health issue and is particularly prevalent among children. Early and correct diagnosis is very important for effective management, but tuberculosis diagnosis in children is often difficult because the clinical symptoms are nonspecific and difficult to collect specimens. This study aims to determine risk factors based on sociodemographic, socioeconomic, nutrition and clinical status associated with tuberculosis. This study is a cross-sectional study. Children aged 0 to 14 years who came to the community health center with one or more clinical symptoms related to TB from January to December 2023 were studied. Research participants underwent two diagnostic methods, clinical and bacteriological. Positive Clinical diagnosis was found to be 72.4%. Meanwhile, bacteriologically confirmed participants were 10.3%, the remainder were unable to produce sputum. Most respondents were men (62.8%), with 66% of respondents aged ≥5 years, and those sociodemographic factors have a significant relationship with tuberculosis in children (p<0.05). This study found that 64.7% of respondents had a BMI <18.5 and significantly related to tuberculosis in children (p<0.05). Furthermore, tuberculosis incidence and malnutrition also have a significant relationship (p<0.05). Contact with active tuberculosis cases shows a significant difference in the proportion of tuberculosis and nontuberculosis cases (p=0.009), while 57.1% of children were exposed to cigarette smoke (p<0.05). Government health insurance (BPJS) was used by 70.5% of respondents who were tested positive for tuberculosis, and it is statistically significant (p<0.05). Other socioeconomic factors were also significant (p <0.05) for tuberculosis cases in children whose parents had low education at 85.9%; have 96.8% of jobs with a household income equal to or more than the provincial minimum wage (UMP) of 60.3%. In the future, these factors can be utilized to construct prediction models for clinical diagnosis in children suspected of tuberculosis, especially machine learning-based models, in addition to tuberculosis-related clinical symptoms.