Clinical profile, laboratory characteristics and prognostic factors in patients with miliary tuberculosis

Abstract

BackgroundMiliary tuberculosis (TB) is a severe form of pulmonary TB. It is uncommon in immunocompetent patients. In this study, we aimed to investigate features of miliary TB (clinical, biological, and radiological) and to determine factors associated with unfavorable outcomes in a population of patients in the South East Tunisia where TB remains endemic. MethodsThis is a retrospective study including patients diagnosed with miliary TB between 2006 and 2023. Factors independently associated with poor prognosis were determined by multivariate logistic regression analysis. ResultsMiliary tuberculosis (TB) accounted for 1.8% (n = 36) of all TB cases diagnosed during the study period. A notable female predominance was observed, comprising 66.6% of the cohort. The median age of patients was 47.5 ± 11.33 years. The predominant clinical manifestations included cough (88.8%), loss of appetite (77.7%), and fever (58.3%). Radiologically, a typical miliary pattern was present in 83.3% of patients, although only 36.1% had positive sputum samples on direct smear microscopy. Notably, all patients tested negative for HIV serology. Extrapulmonary TB involvement was documented in 55.5% of cases. All patients were treated with first-line anti-TB medications, and the outcome was favorable in 77.7% (n = 28) of patients. However, 16.6% (n = 6) of patients succumbed to the disease. Factors significantly associated with unfavorable outcomes included age ≥ 65 years (odds ratio (OR) = 0.39; p = 0.03), diabetes (OR = 0.13; p = 0.046), presence of fever (OR = 2.89; p = 0.01), and oxygen saturation ≤ 92% at admission (OR = 3.2; p = 0.001). ConclusionOur study identified advanced age, diabetes, fever at baseline, and low oxygen saturation on admission as significant predictors of poor prognosis in patients with miliary tuberculosis. These findings highlight the need for early identification and targeted management of high-risk individuals to improve clinical outcomes.