9 July Antigen detection part 2

Fri, 07/10/2020 - 19:12

9 July Diagnostic tests with focus on rapid antigen detection


  1. A review in Frontiers and another in New microbes and New Infections for your orientation in the state of the art of COVID-19 diagnostic tests in general.


  1. Diao in medRxiv used an fluorescent immunochromatographioc strip on nasopharyngeal swabs and compared with RT-PCR for N protein. They found an overall specificity of 100 % and a sensitivity of 68 % (negative predictive value = 32 %; accuracy = 72 %) in RT-PCR positive (Ct < 40) patients.  The sensitivity, negative predictive value and accuracy were much higher (98 %, 97 %, 99 %) in samples with high viral load (Ct value < 30).  There is a remarkably high percentage of positive results in urine as well.   Clearly, since only the PCR positive samples were tested, this may not present a realistic view.


  1. Porte et a in the Lancet Infect Dis use a commercial N antigen test (Bioeasy Biotechnology Co., Shenzhen, China) on naso- and oro-pharyngeal swabs from 127 symptomatic patients, of whom 82 were RT-PCR positive. A high sensitivity of 94 % and accuracy of 96 % is found (with a specificity of 100 %). Here also sensitivity was much higher (100 %) in samples with high viral load and lower if the symptoms were already present for > 1 week.  Nevertheless the false negative samples had a relatively low Ct value (high VL) in RT-PCR


  1. The paper by Mak et al in J Clin Virol provides a much bleaker picture on another commercial rapid antigen diagnostic test (RAD) from BIOCREDIT.  RAD was 1,000 fold less sensitive than viral culture and 100,000 fold less sensitive than RT-PCR. The RAD test detected between 11.1 % and 45.7 % of RT-PCR-positive samples from COVID-19 patients. As shown in Fig 1, they used various respiratory samples (naso-pharyngeal, sputum, saliva) and found the same trend in all sample types.


  1. Nagura-Ikeda in JCM compares various molecular techniques and 1 rapid antigen test (ESPLINE) on self-collected saliva from patients, who were RT-PCR positive on either naso- or oropharyngeal swabs.   Of the 103 saliva samples, viral RNA was detected in 50.5–81.6% of the specimens by molecular diagnostic tests and an antigen was detected in 11.7% of the specimens by the rapid antigen test. 


  1. Finally some news on the further development of the “Belgian” rapid test by Coris, now called “Respi-Strip” by Mertens et al in Frontiers in Medicine.  The evaluation was done on 328 N-P samples and showed a sensitivity of 57 %, sensitivity of 99.5 % and an accuracy of 83 % as compared to standard RT-PCR. Clearly, you need a  a high VL as the CT cut off was at 22.



  • The rapid tests are very specific, but not very sensitive, with the remarkably exception of the Bioeasy kit, evaluated in Chile. 
  • Self-collected saliva further decreases sensitivity to an unacceptable low value of 11 %.


Therefore, most of the presently available rapid antigen tests can be used to quickly identify the patients with the highest viral load in naso- or oro-pharyngeal samples at the gate of the hospital, but they cannot rule out COVID-19 and they should not be used with self-collected saliva .