Sensitivity and potential utility of SARS-CoV-2 rapid antigen and nucleic acid amplification tests in the context of an elimination approach

AIM: To assess the sensitivity and potential utility of five RATs and the IDNow, Liat and Oxsed nucleic acid amplification tests (NAATs) in our population. METHOD: 39 retrospective and contrived SARS-CoV-2 positive samples were tested in parallel by standard RT-PCR and RAT. A second group of 44 samples was tested by standard RT-PCR, rapid RT-PCR and two isothermal NAAT assays. Limit of detection was compared at RT-PCR cycle thresholds for all assays. RESULTS: We found that the Cobas Liat RT-PCR had 100% concordance with conventional RT-PCR, whereas the sensitivity of other rapid NAAT assays was less at lower viral loads indicated by Cts >30 (p=0.042) and the RATs at Cts >25 (p <0.001). When applied to New Zealand testing scenarios, IDNow or Oxsed NAAT could miss up to 12% and RATs up to 44.3% of COVI D-19 cases compared with the RT-PCR currently used at our laboratory. CONCLUSION: We found that the POC Cobas Liat, a platform that delivers a sample answer in 20 minutes, demonstrated equivalent performance to standard RT-PCR. However, the RATs and isothermal NAAT assays demonstrated reduced sensitivity, limiting their utility in New Zealand's currently very low prevalence setting.