Abstract:

Objective Constructing nomogram based on clinical and internal thoracoscopic features of pleural effusion patients, to explore its value of diagnosis of tuberculous pleurisy. Methods A total of 206 patients treated in 4 hospitals in Beijing from 2015 to 2018 were collected in this prospective study, and of them, 129 patients were tuberculous pleurisy patients (tuberculous pleurisy group), 77 patients were pleural effusion patients (pleural effusion group), including the patients of tuberculous pleurisy and patients with pleural effusion. All the patients were from Beijing Chest Hospital affiliated to Capital Medical University (n=99), Beijing Chaoyang Hospital affiliated to Capital Medical University (n=67), Beijing Jishuitan Hospital (n=16) and Beijing Hospital (n=24). The general information, results of pleural effusion, internal thoracoscopic and pathologic characteristics were collected. The 206 patients were randomly divided into the training set (n=162) and the validation set (n=44). The predictive factors of tuberculous pleurisy were screened by least absolute shrinkage and selection operator (LASSO), regression, and the predictive nomogram was constructed to verify its discrimination, calibration and benefits, comprehensively evaluating the diagnostic and predictive ability of the nomogram. Results Among the 162 pleural effusion patients, 104 (64.2%) were diagnosed with tuberculous pleurisy. By LASSO regression, 7 predictors of high diagnosis value were selected, which were caseous granulomas, necrotic manifestations, atypical granulomas, miliary nodules, pleural adhesions, isolated nodules with thoracoscopy and adenosine dehydrogenase level (an indicator in pleural effusion examination). The AUC, sensitivity and specificity in the training set were 0.97, 92.31% and 93.42%, respectively, while in the model of the verification set, those were 0.89, 80.44% and 100.00%, respectively. Conclusion The nomogram could effectively predict the diagnosis of tuberculous pleurisy in patients with pleural effusion, it had applied diagnosis value to some extent.

Key words: Tuberculosis,pleural, Pathological conditions,signs and symptoms, Thoracoscopes, Nomogram, Diagnosis