RR:C19 Evidence Scale rating by reviewer:
Potentially informative. The main claims made are not strongly justified by the methods and data, but may yield some insight. The results and conclusions of the study may resemble those from the hypothetical ideal study, but there is substantial room for doubt. Decision-makers should consider this evidence only with a thorough understanding of its weaknesses, alongside other evidence and theory. Decision-makers should not consider this actionable, unless the weaknesses are clearly understood and there is other theory and evidence to further support it.
This study considers the accuracy of a lateral flow antigen test compared to RT-PCR for detecting SARS-CoV-2 using virus culture (TMPRSS2 culture) as a reference standard for infectivity. SARS-CoV-2 was successfully cultured from 28 of 38 RT-PCR positive samples (27 of which were also antigen test positive, for a sensitivity of 96.4% compared to 100% for RT-PCR), and of the 10 culture-negative/PCR positive samples, two were antigen test positive. The authors suggest that the lower average time post symptom onset and higher viral load in antigen-positive compared to antigen-negative samples is indicative of infectious virus isolation and supports the use of antigen-based testing as a public health intervention to help reduce community transmission of COVID-19 infection.
The study results are of interest and are a welcome addition to the body of evidence around antigen testing. The authors briefly acknowledge the potential limitations of viral culture as a proxy for infectivity towards the end of the Discussion but do not adequately consider the implications of this for their own results. In at least one large study, culture has failed to show viable SARS-CoV-2 in as much as 20% of PCR positive samples obtained in the small window of time around symptom onset when people are most likely to be infectious. Whether this failure is truly due to lack of infectivity or to failures in sampling, sample handling, or the culture process itself is unknown.
In this particular study, all participants were symptomatic for COVID-19 (all less than 7 days post symptom onset), and at least some of the 8 antigen-negative, culture-negative, but PCR positive samples must have been obtained within the first few days of symptoms. The original study documents the time post symptom onset for all PCR positive, antigen-negative samples and it would be a really useful addition to include that information in this pre-print. Other studies applying viral culture to PCR positive, antigen-negative samples have detected viable SARS-CoV-2 in between 0% and 47% of samples tested (all of which had what would be considered to be low viral loads (Ct values >25 or >30)). There is insufficient evidence to conclude that antigen tests accurately detect infectious cases, and limited evidence for tests used for asymptomatic screening.