Performance of Nucleic Acid Amplification Tests in Patients with Presumptive Pulmonary Tuberculosis in Taiwan

Several nucleic acid amplification tests (NAATs) for detection of Mycobacterium tuberculosis (TB) complex (MTBC) are available in Taiwan; however, their performances may differ and have not been extensively evaluated. Therefore, we aimed to explore the accuracy of NAATs overall followed by comparison between platforms commonly used in Taiwan.

This study enrolled presumptive pulmonary TB patients with NAATs throughout Taiwan. The diagnostic performance of smear microscopy and NAATs was assessed using sputum culture as a reference standard. To investigate the performance of NAATs in excluding non-tuberculous mycobacteria (NTM), we quantified the false-positive proportion of NAATs in patients infected with NTM.

Of the 4126 enrollees, 860 (20.8%) had positive NAATs. The sensitivity and specificity of NAATs were 83.2% and 96.7%, respectively, compared to 81.5% and 55.3% for smear. There was no significant difference in sensitivity between the NAATs and smear; however, the specificity of smear was significantly lower than that of the NAATs [difference 41.4%, 95% confidence interval (CI) 39.6-43.2%]. There was no significant difference in sensitivity among Roche Cobas Amplicor Mycobacterium tuberculosis assay (Amplicor), Xpert MTB/RIF assay (Xpert) and in-house polymerase chain reaction (in-house PCR) (82.2% versus 83.8% versus 82.4%); however, in-house PCR was significantly less specific than Amplicor (difference 5.3%, 95% CI 2.4-8.2%) and Xpert (difference 5.8%, 95% CI 3.1-8.5%). The sensitivity of NAATs among smear-negative cases was 33.1% (95% CI 26.0-40.3%). In-house PCR had a significantly higher false-positive rate among specimens that were culture positive for NTM than Amplicor (7.7% versus 0.3%; difference 7.4%, 95% CI 3.4-11.5%) and Xpert (7.7% versus 0.7%; difference 7.0%, 95% CI 2.9-11.0%).

The NAATs overall had a relatively high sensitivity and specificity in detecting MTBC while Amplicor and Xpert performed better than in-house PCR in excluding NTM. Our findings will be useful for the development of national policy.