Diagnostic accuracy of the real-time PCR cobas® Liat® Influenza A/B assay and the Alere i Influenza A & B NEAR isothermal nucleic acid amplification assay for the detection of influenza using adult nasopharyngeal specimens

Background: Accurate detection of influenza requires diagnostic testing; however, methods such as RADTs and central laboratory-based tests are limited by low sensitivity and time constraints, respectively. Objective: To compare the performances of the cobas (R) Liat (R) Influenza A/B and Alere (TM) i Influenza A & B point-of-care (POC) assays for detecting influenza A and B viruses using fresh nasopharyngeal specimens with the GenMark Dx (R) Respiratory Viral Panel as the reference method, a FDA cleared IVD PCR test. Study design: A total of 87 samples collected in viral transport medium from adults >= 18 years of age were re-tested on both POC assays (based on the reference PCR method, 29 were influenza A and 18 were influenza B virus positive). Results: The overall sensitivity and specificity of the cobas Influenza A/B for the detection of influenza A and B relative to reference PCR was 97.9% (95% confidence interval [CI] 88.9%, 99.6%) and 97.5% (95% CI: 87.1%, 99.6%), respectively, while the sensitivity of the Alere i Influenza A & B assay relative to the reference PCR method was 63.8% (95% CI: 49.5%, 76.0%) and the specificity was 97.5% (95% CI: 87.1%, 99.6%). The individual sensitivities and specificities of the cobas Influenza A/B assay for influenza A alone and influenza B alone were comparable to those of the reference PCR method (influenza A: sensitivity of 100% [95% CI: 88.3%, 100.0%] and specificity of 98.3% [95% CI: 90.9%, 99.7%]; influenza B: sensitivity of 94.4% [95% CI: 74.2%, 99.0%] and specificity of 100% [95% CI: 94.7%, 100.0%]). For the Alere i Influenza A & B assay, the individual specificities for influenza A and B were comparable to those of the reference PCR method (98.3% [95% CI: 90.9%, 99.7%] and 97.1% [95% CI: 90.0%, 99.2%], respectively), while the individual sensitivities were low relative to reference PCR (55.2% [95% CI: 37.5%, 71.6%] and 72.2% [95% CI: 49.1%, 87.5%], respectively). Conclusion: The cobas Influenza A/B assay demonstrated performance equivalent to laboratory-based PCR, and could replace rapid antigen tests.