Abstract
BACKGROUND: Despite its low sensitivity, microscopy remains the main method for the diagnosis of pulmonary tuberculosis in most laboratories in Ethiopia. Few studies have evaluated the performance of light-emitting diode fluorescent microscopy (LED-FM) in bleach-concentrated smear-negative sputum specimens. OBJECTIVE: This study aimed to evaluate the diagnostic performance of LED-FM for smear-negative pulmonary tuberculosis in Ethiopia. METHODS: A total of 194 adult patients with a cough lasting for more than two weeks, and who had three direct smear-negative sputum tests for Mycobacterium tuberculosis by Ziehl-Neelsen light microscopy, were included. All direct Ziehl-Neelsen-stained smear-negative sputum samples were cultured and were also visualised by LED-FM. Smears for LED-FM were performed from bleach-concentrated sputum sediment. The diagnostic performance of the LED-FM was compared to the culture method (the reference standard). RESULTS: Of the 194 smear-negative sputum specimens analysed, 28 (14.4%) were culture-positive and 21 (10.8%) were LED-FM-positive for M. tuberculosis. However, only 11 of the 21 (52.4%) LED-FM-positive patients had a confirmed tuberculosis diagnosis by culture. Light-emitting diode fluorescence microscopy (FM) had a sensitivity of 39.3% (95% confidence interval: 21.2-57.4) and specificity of 93.9% (95% confidence interval: 90.4-97.6). Ten LED-FM-positive specimens were culture-negative, and all of these specimens had scanty grading (1-19 bacilli per 40 fields on LED-FM). CONCLUSION: This study showed that implementation of LED-FM on bleach pre-treated and concentrated sputum can significantly improve the diagnosis of smear-negative pulmonary tuberculosis. However, all scanty grade, positive smears by LED-FM need to be confirmed by reference culture method.