JLE

Annales de Biologie Clinique

MENU

Retrospective analysis of the performance of the SARS-CoV-2 rapid antigen detection test compared to the reference RT-PCR test Volume 79, issue 2, Mars-Avril 2021

Figures



  • Figure 1

  • Figure 2

  • Figure 3

  • Figure 4

Tables

Authors
1 BioLittoral-Biogroup, Plateau technique la Bastide, Sanary-sur-Mer, France
2 BPO-BioÉpine-Biogroup, Plateau technique Chocolaterie, Levallois-Perret, France
3 UMR 1173 Inserm, Université Paris-Saclay - UVSQ, Montigny-le-Bretonneux, France
4 BioEsterel-Biogroup, Plateau technique de Mouans-Sartoux, France
5 CAB-Biogroup, Plateau technique des Poteries, Strasbourg, France
6 Biolam-LCD-Biogroup, Saint-Denis, France
7 Dyomedea-Neolab-Biogroup-Plateau technique de la Sauvegarde, Lyon, France
* Correspondance

Background: Discovered in 2019 in the region of Wuhan, China, the Severe Acute Respiratory Syndrome coronavirus 2 (SARS-CoV-2) rapidly established itself as a major pathogenic agent of morbidity and mortality. France has implemented a strategy to fight this virus which relies essentially on widespread RT-PCR virological testing in order to isolate positive patients. Antigenic tests have recently been made available to help the diagnostics. We have conducted a retrospective study to determine the sensitivity of these antigenic tests, comparing them to the reference RT-PCR method. Method: Between December 7, 2020 and January 31, 2021, each patient we received in our laboratories for an RT-PCR test was enrolled. Out of 271,649 patients, 4,881 had been submitted to an antigenic test (TDR) in the preceding 24 hours. Comparing the data resulting from both tests, we established the sensitivity and the specificity of the antigenic tests. For our analysis we included the parameter of symptom and/or the value of Cycles threshold (Ct) in our parameters. Results: The sensitivity of the TDRs compared to all the positive RT-PCR tests is 56%. We further demonstrate the correlation between the symptom duration and the reduction of the nasopharyngeal viral load. Based on this data, we have established that the sensitivity of the TDRs decreases very rapidly after symptom onset, contrary to the estimated viral load in the RT-PCR. Indeed, less the 24 hours after clinical symptom onset, the sensitivity of the TDRs decreases from 74% to 60%. By including the Ct value in our parameters, we have established that, despite a high viral load and clinical symptoms since 7 days or less, the sensitivity of the TDRs is 66%. Although, a high number of asymptomatic patients among carriers of SARS-CoV-2, we have estimated a specificity of 93% for our test. Conclusions: Performance in terms of sensitivity and specificity of the TDR, as assessed in practice, are inferior to those given by the manufacturer, which raises several questions. What is the impact of falsely negative results for patients carrying a high viral load? Are the implemented measures sufficient to prevent the epidemic?