ABSTRACT
The influenza season 2014/15 was dominated by drift variants of influenza A(H3N2), which resulted in a reduced vaccine effectiveness. It was not clear if the performance of commercial nucleic-acid-based amplification (NAT) assays for the detection of influenza was affected. The purpose of this study was to perform a real-life evaluation of two commercial NAT assays. During January-April 2015, we tested a total of 665 samples from patients with influenza-like illness using the Fast Track Diagnostics Respiratory pathogens 21, a commercial multiplex kit, (cohorts 1 and 2, n = 563 patients) and the Xpert Flu/RSV XC assay (cohort 3, n = 102 patients), a single-use cartridge system. An in-house influenza real-time RT-PCR (cohort 1) and the RealStar Influenza RT-PCR 1.0 Kit (cohort 2 and 3) served as reference tests. Compared to the reference assay, an overall agreement of 95.9 % (cohort 1), 95 % (cohort 2), and 98 % (cohort 3) was achieved. A total of 24 false-negative results were observed using the Fast Track Diagnostics Respiratory pathogens 21 kit. No false-negative results occurred using the Xpert Flu/RSV XC assay. The Fast Track Diagnostics Respiratory pathogens 21 kit and the Xpert Flu/RSV XC assay had sensitivities of 90.7 % and 100 % and specificities of 100 % and 94.1 %, respectively, compared to the RealStar 1.0 kit. Upon modification of the Fast Track Diagnostics Respiratory pathogens 21 kit, the sensitivity increased to 97.3 %. Influenza virus strains circulating during the 2014/15 season reduced the detection sensitivity of a commercial NAT assay, and continuous monitoring of test performance is therefore necessary.