Abstract
PURPOSE: To evaluate the performance of TB-LAMP in the diagnosis of TB empyema using pleural tissue specimens obtained during pleural decortication. METHODS: Using the clinical records and the different diagnostic test results of patients who underwent pleural decortication in a TB-designated hospital over 3.5 years, we calculated the sensitivity, specificity positive predictive, and negative predictive values of the pathology, MGIT 960 culture, and TB-LAMP obtained by using pleural tissue specimens against the etiologic diagnosis and composite clinical reference standard (CCRS) as the reference standards. RESULT: A total of 304 patients' records were extracted. All these patients had gone through pleural decortication. When the etiologic diagnosis was used as the reference, the sensitivity of TB-LAMP in identifying TB empyema was 77.8% (compared to 10.6% of MGIT 960 P < 0.05). The sensitivity of MGIT 960, pathology, and TB-LAMP was 8.2%, 77.7%, and 67.2% against CCRS as the reference; and the specificity of the three was 100.0, 100.0, and 96.2% against the same standard. A combination of pathology and TB-LAMP would increase the sensitivity and specificity to 84.7 and 96.0%. Using TB-LAMP to diagnose TB empyema using pleural tissue samples obtained from pleural decortication was faster with satisfactory performance. CONCLUSION: TB-LAMP has great potential in faster and more accurate diagnosis of TB empyema. Our findings provide insights for optimizing diagnostic algorithms for TB empyema.