A novel multi-parameter scoring system for distinguishing sarcoidosis from sputum negative tuberculosis.

BACKGROUND Sarcoidosis is a granulomatous disorder of unknown cause, affecting multiple organs. Tuberculosis is the world's second most common cause of death from infectious diseases. Due to the similar clinical, radiological and histopathological pictures in sarcoidosis and tuberculosis, Mycobacterium tuberculosis has been considered as potential infectious factor. However, it remains difficult to distinguish sarcoidosis from tuberculosis, especially when sputum examinations for mycobacterium are negative. METHODS 1. to establish a scoring system for differentiating sarcoidosis and tuberculosis: We collected the risk factors, laboratory data and the data of clinical, radiographic, pathological manifestations from the 117 of sarcoidosis patients and 181 of sputum negative tuberculosis patient. And we put them into the designed form. Based on the results of univariate analysis, clinical experience and the literature, we further selected 13 variables that were more supportive to distinguish the two diseases. Finally 9 variables were selected based on logistic regression to establish the scoring systems with significant differences between the two diseases. The beta-coefficient form the logistic regression were used to calculate the weight of each variable. Four types of comprehensive scoring models were established in the end (clinical-- radiographic; clinical--radiographic--radionuclide; clinical--radiographic--pathological and clinico-radiographic--radionuclide--pathological group). Receiver operating characteristics (ROC) analysis was used to determine an optimal cutoff point for each scoring system. 2. to validate the accuracy of the established scoring system: 73 of new sarcoidosis patients and 57 of new tuberculosis patients were chosen to assess the diagnosis accuracy of the four scoring systems. RESULTS 1. we established four types of comprehensive scoring models, included clinical--radiographic; clinical--radiographic--radionuclide; clinical--radiographic--pathological and clinico--radiographic--radionuclide--pathological scoring models, the optimal cutoff values respectively were 9, 17, 18 and 22, the sensitivity and specificity of the four scoring system to distinguish the two diseases respectively were: 93.16% (109/117) and 97.79% (177/181), 92.31% (108/117) and 98.90% (179/181); 93.16% (109/117) and 98.90% (179/181); 94.87% (111/117) and 98.90% (179/181). 2. Validation of the scoring systems with 130 new patients (73 of sarcoidosis and 57 of tuberculosis):, the sensitivity and specificity of CR, CRE, CRP, CREP were 91.78% (67/73) and 87.72% (50/57), 97.26% (69/73) and 98.25% (56/57), 94.52% (71/73) and 96.49% (55/57), 98.63% (72/73) and 98.25% (56/57) respectively. CONCLUSIONS The four scoring systems established by this study can be utilized to differentiate sarcoidosis and sputum negative tuberculosis effectively. Based on the availability of clinical-radiographical/histopathological data, any of the four diagnostic scoring systems were reliable tools for differential diagnosis, with increased information leading to better discrimination.