Rapid Antigen Test LumiraDxTM vs. Real Time Polymerase Chain Reaction for the Diagnosis of SARS-CoV-2 Infection: A Retrospective Cohort Study

Abstract

Background: Real time reverse transcription polymerase chain reaction (real time RT-PCR) testing is the gold standard for the diagnosis of SARS-CoV-2 infections. However, to expand the testing capacity, new SARS-CoV-2 rapid antigen tests (Ag-RDTs) have been implemented. Ag-RDTs are more rapid, but less reliable in terms of sensitivity, and real-life data on their performance in comparison with the real time RT-PCR test are lacking. Methods: We aimed at assessing the diagnostic performance of the third-generation antigenic swab LumiraDx™ compared with real time RT-PCR in a retrospective cohort study at the Infectious Diseases Unit of Padua. All of the patients who were consecutively tested for SARS-CoV-2 in our centre (by both real time RT-PCR and Ag-RTD LumiraDxTM) from 19 January to 30 May 2021, were included. Cycle-threshold (Ct) values of positive real time RT-PCR were recorded as well as the number of days from symptoms’ onset to testing. Results: Among the 282 patients included, 80.9% (N = 228) tested positive to real time RT-PCR, and among these, 174 tested positive also to LumiraDx™. Compared with real time RT-PCR, which is considered as the gold standard for the assessment of the presence/absence of SARS-CoV-2 infection, LumiraDx™ showed an overall sensitivity of 76.3% and specificity of 94.4%. Sensitivity increased to 91% when testing was performed <10 days from symptoms’ onset, and to 95% when considering Ct < 25. Multivariable binomial logistic regression showed that false negative LumiraDx™ results were significantly associated with high Ct values, and with further testing from symptoms’ onset. Conclusions: The results of our study suggested that the LumiraDx™ SARS-CoV-2 antigen assay may be appropriate for the detection of SARS-CoV-2 infection, especially in its early phase when the test largely meets the performance requirements of the European Centre for Disease Prevention and Control (ECDC).

Keywords: COVID-19; RT-PCR; SARS-CoV-2; accuracy; agreement; diagnosis; negative predictive value; positive predictive value; rapid antigen testing; sensitivity; specificity.

Figure 1

Flow chart summarising the RT-PCR and LumiraDx™ SARS-CoV-2 simultaneously performed test results, in accordance with the symptomatic (A) and asymptomatic (B) clinical status of the study population. Red and green boxes indicate positive (+) and negative (−) results for each testing, respectively.

Figure 2

Days from symptoms’ onset distribution among true positive (TP) and false negative (FN) tests.

Figure 3

Ct distribution among true positive (TP) and false negative (FN) tests.

Figure 4

Heat-map displaying sensitivity with respect to both Ct values and Days after symptom onset thresholds. The darker the green, the higher the level of sensitivity of the LumiraDx^TM compared to the gold standard RT-PCR.