Photo: Fæ on Wikimedia Commons

How do the Covid-19 saliva tests stack up? – Expert Reaction

Calls are mounting for more saliva testing to be used in New Zealand, after the latest border incursion. 

There are two tests being used in New Zealand currently. SalivaDirect is developed by the Yale School of Public Health and is part of an ESR trialShield is developed by the University of Illinois, and is being used in Auckland Airport by Rako Science.

The SMC asked experts to comment on this announcement. 

Dr Nikki Freed, Senior Lecturer in the School of Natural and Computational Sciences, Massey University, comments:

“The use of saliva for routine COVID-19 testing has some strong advantages. Simply put, it is an easier, non-invasive sample that patients can self-collect. This can reduce exposure for public health staff and reduce patient discomfort. This is an especially good tool for people who need to get tested routinely, such as border staff and health care workers.

“The disadvantage is that saliva testing is slightly less sensitive than the standard COVID-19 test in New Zealand (that uses Nasopharyngeal swabs). Less sensitive means that a person will need more copies of the virus to be detected, so saliva-based tests might miss a very small fraction of people who are just starting to get sick or who are just at the end of COVID-19 illness.

“The data provided in the preprint from the SHIELD saliva test developed at the University of Illinois (the test that Rako Science is using) claims it can detect “500-1000 viral particles per mL”. This is an excellent level of detection, however, some “best in class” tests using nasopharyngeal swabs claim to have even higher sensitivity, detecting as few as 100 viral particles per mL. The SalivaDirect method being tested by ESR claims to have a detection limit of 6000 to 18000 viral particles per mL.

“Finally, a recent review of 16 studies looked at the performance of saliva testing versus nasopharyngeal swabs, and found that both test types perform fairly similarly.
“So what does this all mean? It is clear that using saliva or nasopharyngeal swabs for PCR-based tests can detect a very low amount of the virus that causes COVID-19. All tests have a higher chance of detecting someone who is sick when they are used frequently. The use of saliva for routine COVID-19 testing has some strong advantages that should not be ignored – it is an easier, non-invasive sample that patients can self-collect. This can reduce exposure for public health staff and reduce patient discomfort, making it easier for more people to get tested more often. Saliva testing is an especially good tool for people who need to get tested routinely, such as border staff and health care workers. Any questions about lower sensitivity of these tests can be overcome if these high-risk individuals are tested every few days.

“However, a major disadvantage of using even a slightly less sensitive test in New Zealand, is that missing just one sample that is positive can have major impacts on the elimination strategy. I believe this is why we are not seeing it rolled out just yet for the wider public.”

Note: Dr Freed has been awarded HRC funding to research rapid diagnosis and genome sequencing to follow the coronavirus outbreak.

No conflict of interest declared.

Professor David Murdoch, Dean and Head of Campus, University of Otago, Christchurch, comments:

“There has been a lot of recent discussion about saliva testing for COVID-19, particularly in relation to its potential role in the testing of border workers. Saliva testing simply refers to the testing of saliva rather than a nasal swab, which has been the standard sample type used in COVID-19 testing so far. The actual method used to test saliva is usually the same PCR method used to test nasal swabs.

“The main advantage of saliva is it can be self-collected without the discomfort of a nasal swab. This can be particularly helpful for people who need to be tested regularly, like border workers.

“However, the testing of saliva may require additional processing steps in the laboratory in order to best prepare the sample for testing. This can affect laboratory workflows and efficiency. Newer methods are being developed to avoid additional processing steps, some of which may require little technical expertise and enable testing outside a laboratory.

“There are now many published studies assessing the accuracy of saliva testing for COVID-19. Although many different methods and study populations have been used in these studies, there is a consistent indication that PCR testing of saliva has a similar accuracy to PCR testing of nasal swabs.

“There is now clear evidence that saliva testing for COVID-19 may play a useful role in the frequent testing of individuals, as long as resourcing and laboratory workflow issues can be adequately addressed.”

Conflict of interest statement: Member, COVID-19 Vaccine Strategy Taskforce, NZ Government. Member, COVID-19 Vaccine Strategy Scientific and Technical Advisory Group, NZ Government. Member, Advisory Group, Vaccine Alliance Aotearoa New Zealand (VAANZ). Independent Member, Clinical Trials Steering Committee, University of Oxford COVID-19 Vaccine. Member, COVID-19 Expert Advisory Network, Ministry of Health.

Associate Professor Arindam Basu, College of Education, Health & Human Development, University of Canterbury, comments:

“A systematic review published [recently] reveals that saliva testing and nasopharyngeal swab testing have a similar level of accuracy at detecting positive cases. The saliva tests overall, when pooled together were 83.2% accurate, this indicates that on an average such a test will miss about 17% true positive cases. But the nasopharyngeal swab tests are similar at around 85% accurate, suggesting that overall, when all things are considered, they will miss about 15% true positive cases.

“So, even though on average when all results are pooled together, it does appear that the saliva tests perform slightly worse than nasopharyngeal swab tests, they are not all that different, as their 95% confidence intervals or credible intervals overlap (74.7% – 91.4% for saliva, 76.2% – 92.4% for nasopharyngeal). Hence, overall, in terms of how accurate they are, there is little to choose from saliva and nasopharyngeal swab tests. The advantage of saliva tests is obviously their ease of obtaining samples.

“Of the two different saliva tests being used in New Zealand, if you compare the performance, limit of detection, and convenience between SalivaDirect and Shield tests, the administration and principles are identical, and both employ innovative ways to cut down on the time of processing. SalivaDirect was able to detect with 94% accuracy true positive results returned by a comparable nasopharyngeal swab. Shield was able to detect true positive results with 88.9% accuracy. This indicates that the performance of Shield is roughly similar to what is expected of all saliva tests, and not much different from SalivaDirect. SalivaDirect reported a lower detection limit of about 6 SARS-CoV-2 copies/μL, so roughly about 6,000 copies/millilitres while Shield put that figure around 500 per millilitre.

“In final analysis, it may be said that:

  1. Saliva tests are as good as nasopharyngeal swab tests when it comes to their head to head diagnostic performance, but
  2. Saliva tests are more convenient.
  3. SalivaDirect and Shield are very similar with respect to their performance profiles, except that SalivaDirect may be more sensitive than Shield. Shield has shown that their detection limit is lower, however it is uncertain if this difference will help to make a decision about preference of one over the other, as saliva load of virus also tends to be higher and either method (SalivaDirect or Shield) will equally detect the viruses.

“Finally, it must be acknowledged that while the results of the systematic review and SalivaDirect research was assessed on the basis of peer reviewed publication, the research on Shield is not peer-reviewed and therefore open to bias.”

No conflict of interest declared.