PURPOSE: Tuberculosis (TB) is a major curable cause of lymphocyte-predominate exudative pleural effusion. The patients’ clinical characteristics and basic pleural fluid profiles may increase the possibility of tuberculous pleuritis, especially in the countries with high tuberculosis prevalence. We aim to develop a clinical score to diagnose tuberculous pleuritis and prove its diagnostic efficacy.
METHODS: Patients who presented with lymphocytic exudative pleural effusion in King Chulalongkorn Memorial Hospital were retrospectively analyzed and then classified into 2 groups 140 with TB and 187 with non-TB. Patients with transudative effusion or effusion with unidentified causes were excluded. Demographic data and pleural fluid profiles were collected. Multivariate analysis was used to select clinical variables which were significantly different between TB and non-TB group. Variables beta coefficients were used to define a numerical score. We further used the score in another set of patients with pleural effusion (n=100) to prove its diagnostic yield in terms of sensitivity and specificity.
RESULTS: 327 eligible patients were included into the study. We found 10 independent variables to predict tuberculous pleural effusions including male gender, age < 25 years, fever, pleuritic chest pain, right-sided pleural effusion, loculation, straw color or yellow turbid pleural effusion, specific gravity > 1.029 and lymphocyte > 86% in pleural effusion. The coefficient of these variables were 2.5, 1.5, 3.5, 1.0, 1.0, 1.5, 2.5, 2.5, 2.0, and 2.0 respectively. Using the score more than 14 in the diagnosis of TB pleuritis revealed a sensitivity of 75% and a specificity of 80%.
CONCLUSIONS: A clinical score was developed to diagnose tuberculous pleuritis in the setting of chronic lymphocytic exudates and showed a moderate diagnostic yield.
CLINICAL IMPLICATIONS: Clinical score based on clinical features and pleural effusion may help physicians to make a treatment decision and avoid overuse of high cost investigation for diagnosis of TB pleuritis, especially in the areas of high tuberculosis prevalence.
DISCLOSURE: The following authors have nothing to disclose: Kamon Kawkitinarong, Sitthipong Dumrongpiwat, Chanchai Sittipunt
No Product/Research Disclosure Information