Direct comparison of Xpert Xpress, FilmArray Respiratory Panel, Lumipulse antigen test, and RT-qPCR in 165 nasopharyngeal swabs
Yosuke Hirotsu, Makoto Maejima, Masahiro Shibusawa, Yume Natori, Yuki Nagakubo, Kazuhiro Hosaka, Hitomi Sueki, Kenji Amemiya, Miyoko Hayakawa, Hitoshi Mochizuki, Toshiharu Tsutsui, Yumiko Kakizaki, Yoshihiro Miyashita, Masao Omata
Abstract
BACKGROUND: The nucleic acid amplification test (NAAT) and antigen test are approved diagnostic tests for COVID-19. In this study, we aimed to investigate the assay performance of two NAATs (Xpert Xpress SARS-CoV-2 and FilmArray Respiratory Panel) and a quantitative antigen test (Lumipulse). METHODS: One hundred and sixty-five nasopharyngeal swabs were subjected to Xpert, FilmArray, Lumipulse, and RT-qPCR assays. RESULTS: Of 165 samples, RT-qPCR showed 100 positives and 65 negatives. The Xpert had an overall agreement of 99.4% (95% confidence interval [CI]: 96.7-99.4%), sensitivity of 99% (95% CI: 96.8-99%), and specificity of 100% (95% CI: 96.6-100%). FilmArray had an overall agreement of 98.8% (95% CI: 95.9-98.8%), sensitivity of 98% (95% CI: 95.6-98%), and specificity of 100% (95% CI: 96.3-100%). Lumipulse had an overall agreement of 95.5% (95% CI: 91.8-95.5%), sensitivity of 92.3% (95% CI: 89.2-92.3%), and specificity of 100% (95% CI: 95.5-100%). The κ coefficient showed excellent agreement between each test and RT-qPCR. There was a high correlation between Xpert Ct values, RT-qPCR Ct values, viral loads and antigen level. CONCLUSIONS: Xpert Xpress and FilmArray Respiratory Panel exhibited an equivalent performance. The Lumipulse antigen test was slightly less sensitive than the NAATs, but showed high assay performance except for samples with low viral load. The Xpert Xpress, FilmArray Respiratory Panel and Lumipulse antigen tests offer rapid sample-to-answer data, allowing random access detection on automated devices.