Intended for healthcare professionals

Rapid response to:

Feature Briefing

Covid-19: How the UK is using lateral flow tests in the pandemic

BMJ 2021; 372 doi: https://doi.org/10.1136/bmj.n287 (Published 04 February 2021) Cite this as: BMJ 2021;372:n287

Rapid Response:

Low sensitivity of rapid antigen tests: should we use saliva instead of nasopharyngeal samples in the process of the test validation?

Dear Editor,

Since November 2020 BMJ has published at least three articles regarding currently commercially available rapid antigen tests.

According to Iacobucci et al, the sensitivity of rapid antigen detection (RAD) averaged 76.8%, but in the case of a self-trained person, the sensitivity dropped to 58% (Iacobucci et al 2020). At the same time, a preprint of the pilot study was published by the University of Liverpool in December 2020. The result of the two-week trial showed that the commercially available rapid antigen test had a sensitivity of less than 50% (48.89%) (Wise, 2020; Liverpool pilot, 2020)

Unfortunately, these studies create a bias regarding the use of RAD in clinical practice. Most often, for RAD tests, nasal specimens from COVID-19 patients are used. The data published in your journal shows that currently available commercial rapid antigen tests have severe limitations:

- It requires additional well-trained personal;
- The procedure isn’t patient-friendly;
- Sensitivity isn’t proper enough to promote patients screening

We agree that the nasopharyngeal tests may not be a better solution for RAD, based on the current research. It is worth asking: what if the commercially available tests have severe limitations due to the process of the test validation on nasopharyngeal samples (NPS)? Should we use other sampling-collection techniques to increase the sensitivity of the assay?

For these purposes, BMJ has published an article: “Lateral flow tests need low false positives for antibodies and low false negatives for virus” in a letter form. The letter illustrates the limitation of currently available rapid tests that were performed on nasopharyngeal swab specimens (BMJ, 2020). The simplicity of sampling and homogeneity of the sample itself are significant parameters to consider, and other sources, such as saliva, have been considered as alternative specimens for COVID-19 testing (McCormick-Baw et al., 2020; Williams et al., 2020). Several RAD tests for saliva are currently in the development stage.

Rapid Antigen tests are typically cheap, return results in minutes, and reveal an active infection, similarly to genetic tests. The test results of RAD can be interpreted without specialized instruments and usually are available within 30 min. Hence, RAD tests can relieve the workload in diagnostic hospitals and laboratories and improve the turn-around time. In the case of country border crossing locations, such as airports, the rapid diagnostic tests may promote high-quality screening, allowing to trespass the border, re-establishing the interstate logistics during the State of Emergency. These advantages accelerate further scientific debates regarding the validation of the test by choosing the proper sampling methodology. For example, the rate of RT-PCR detection of SARS-CoV-2 in patients with COVID-19 is as high as 93% in bronchoalveolar lavage fluid, 72% in sputum (derived from saliva), and 63% in nasopharyngeal swabs, while it is only 32% in pharyngeal swabs and 29% in the stool (Wang et al., 2020). The question is: should we use saliva instead of NPS in the process of the test validation?

Due to the unwell NPS procedure, scientists are looking for alternatives to nasopharyngeal samples. In October 2020, the study compared NPS and saliva testing using commercially available RAD tests (Agulló et al 2020). Unfortunately, the use of saliva as a biosample showed lower sensitivity compared to nasopharyngeal swabs. Nevertheless, Torres et al showed promising results in a saliva-based PCR test (Torres et al, 2020). The same result was obtained by our colleagues from the co-research group from the Latvian National Research Programme using saliva-based RT-PCR (Gavars et al, 2020). It is also worth mentioning that commercially available tests are validated on nasopharyngeal samples, thus creating a bias in decision making regarding the use of saliva as a testing sample.

Controversially to the previous statement, our research group has been validating an antigen test that primarily uses saliva as a source of SARS-CoV2 antigens.

We agree that the saliva-based PCR test has significantly higher sensitivity compared to antigen detection. This result can be explained by the nature of the reaction of gene amplification. The FIND study shows that rapid antigen detection by commercially available NPS-based RAD tests is about 20% less sensitive than NPS-based PCR (FIND, 2020). Despite that fact, our data shows that the saliva-validated antigen test may potentially increase the sensitivity of the commercially available tests (currently unpublished data). The saliva-validated rapid antigen detection has several advantages compared to commercially available NPS-based RAD and saliva-based PCR tests by:

- Easy-to-use concept (Comparing to PCR)
- Short test period (comparing to PCR)
- Patient-friendly (comparing to NPS)
- No need for well-trained personnel (comparing to NPS tests)

Regular coronavirus tests usually target spike protein, both with poly- or monoclonal antibodies. However, in intact viral particles, a bigger part of this protein is hidden within the capsid or heterotrimers. On average, one particle contains about 300 copies of monomeric spike protein or specific epitope for monoclonal antibodies to target. On the other hand, there are about 2000 copies of membrane protein, which is also available for detection by specific polyclonal antibodies.

Competing interests: No competing interests

Competing interests: No competing interests

01 March 2021
Svjatoslavs Kistkins
Endocrinology resident
Dr Ilya Dukhovlinov, professor Valdis Pirags
Paul Stradins Clinical University Hospital
Pilsonu iela 13, Riga, Latvia