A preprint in Medrxiv by a Dutch group looked at five commercially available rapid antigen essays on samples from 1754 people who had already tested positive for the coronavirus with a standard PCR test, which is sensitive, but slow.
The two most sensitive tests detected the virus more than 97% of the time, whereas the least sensitive test detected about 75% of cases. All samples came from people with COVID-19 symptoms who have high levels of coronavirus.
The researchers cautioned that the rapid antigen test might be less effective at picking up the presence of the virus in people with low levels of virus.
Rapid antigen tests provide results within 30 minutes but not all of the tests on the market are equally effective at detecting the virus. In the study, they showed that detection limits of five commercially available rapid detection tests, RDTs, differed considerably impacting the translation into the detection of infectious cases.
Rapid detection and isolation of new cases, combined with contact tracing and quarantine, has become a critical pillar of the global efforts to reduce SARS-CoV-2 infection.
The purpose of the test, trace and isolate strategy is to stop the transmission chain and reduce the impact of COVID-19. Epidemiological modelling suggests that aggressive TTI combined with physical distancing and the use of personal protective equipment when physical distance is not achievable could suppress a second wave without the need for a prolonged lockdown.
The case numbers have increased in recent weeks across Europe and the capacity to keep up with the traditional test, trace and isolate strategy is reaching its limits. There is therefore an urgent need for optimising available resources.
Currently case diagnosis relies on PCR swab testing, which has limitations in terms of time to result and for which scaling up capacity is impaired and hampered by the scarcity of critical reagents. In addition, while highly specific, the sensitivity of RT PCR combined with prolonged shedding of low amounts of viral RNA for weeks may lead to positive test results following clinical recovery long after a person is infectious.
Furthermore, screening of persons without symptoms may yield weak positive test results, raising questions about how to handle such cases.
Ideally, screening of cases would be based on testing for infectivity, but such cell culture based assays have long turnaround times and therefore are not suitable for rapid screening, which is essential to the success of test, trace and isolate.
Rapid antigen detection tests, RDTs, have recently entered the diagnostic market. Compared with RT PCR, they are relatively easy to produce, cheaper, easy to use with faster turnaround times, and depending on the assay, without the need for dedicated equipment or high level laboratory capacity.
The widespread and frequent use of such tests has recently been proposed as a solution to the safe reopening and return to the pre-pandemic social interactions, but to our knowledge, there is no data about the performance in detecting infectious cases. The authors assess the potential impact of introduction of RDTs in a testing strategy in the Netherlands, in which the majority of testing is done through drive through test stations.
Detection limits of five commercially available rapid detection tests were determined using serial dilutions of freshly harvested SARS-CoV-2 virus stock. They collected fresh nasal and nasopharyngeal samples from people presenting at the drive through test station with a range of CT values and were tested in parallel with PCR and RDT samples. Samples were also inoculated onto Vero E6 cells to assess the correlation between PCR/antigen rapid test results and infectiousness for persons with mild symptoms, to supplement published data on this relationship for hospitalised patients and patients with mild symptoms. They looked at the profile of positive patients in the drive through test stations, and selected data from 223 COVID-19 index cases who had a complete background of date of symptom onset, date of sample collection, date of diagnosis, etc.
Data on time since symptom onset showed that the majority of persons reported at the testing station within the first week at symptom onset, which was in line with national recommendations for testing. The viral load distribution for the complete patient group was also similar and therefore an assumption was made that there was a similar profile of cases. Results were available for the same day for 33% of cases, the next day for 55%, after two days for 11% and after three days for 1%.
Infectiousness was assessed by including the matched viral cultures and PCR of 78 randomly selected individuals who were diagnosed with SARS-CoV-2 in the drive through testing station. The study determined the probability of being infectious for all patients tested in the drive through station using a logistic regression analysis. Interestingly, the comparison suggested that those notes from outpatients were more often cell culture positive than those from hospitalised cases.
The most sensitive RDT is 97.3% of infectious individuals would expect to be detected. This unfortunately decreased to 92.73% and 75.53% for two assays respectively.
This preliminary analysis suggested major differences in the rapid detection antigen tests regarding suitability for tracking most infectious cases.
Assuming that the implementation of rapid tests will lead to reporting of the results on the same day, followed by contact by public health official in all cases, the proportion of cases with optimal start of contact tracing can increase from 33% to 75.53% when using the least sensitive assay to 97.3% for the most sensitive assay.
The use of the rapid antigen test for screening of individuals offers the potential for rapid identification of those individuals at greatest risk of spreading the infection. This study tested this reasoning based on real life data and posed a basis for discussion on choices for assays and testing algorithms.
The advantage of faster time to resolve and therefore initiation of contact tracing is a great added benefit of rapid detection tests. In the example from the study, routine application of RDTs will increase the proportion of suspect cases who received their test results the same day from 33 to 97%. The shortening of testing delay was a critical determinant of success of contact tracing strategy and shortening from three to one day can push expanding outbreaks into suppression with a reproductive number below one.
The analysis showed, however, that antigen RDTs differed greatly in their ability to detect infectious cases, therefore requiring careful validation before routine application. In the study setting, people were tested relatively soon after onset of disease when viral loads are at their peak thus ensuring higher sensitivity of the RDTs. A more challenging application is the use of rapid detection tests when testing persons without symptoms as a strategy to reopen society after lockdown.
In the absence of knowledge of time since exposure, negative predictive values are difficult to assess, the risk of false negatives is higher than in symptomatic persons that may be using physical distancing and is currently recommended globally.
Nonetheless, when used judiciously, RDTs offer hope to improve containment by more rapid isolation and contact tracing of the most infectious individuals.
The London General Practice is assessing all available COVID-19 testing and reflects on them in their clinical governance meetings. As soon as a test is determined to be specific and sensitive, the practice will include them in its everyday use.
Dr Paul Ettlinger
The London General Practice