A NY Times article published today talks about some seldom-discussed limitations of current PCR testing and highlights some of the points I made in my recent blog post urging consideration of increased used of antigen tests, especially in schools, and providing detailed explanation and support for the notion.
The article makes the following points: 1. PCR tests are very sensitive and can detect very small amounts of virus 2. When the virus amounts are low it is very unlikely the person is infectious, and possible it is even detecting dead virus fragments from previous infections 3. PCR test results take too long to get back for them to be useful for infectious disease control and community viral management, including for testing, tracing and isolation procedures necessary to open schools safely.
Their implication is that the viral quantification numbers should be reported and not just positive/negative of test, and in particular that perhaps the test should not really be reported as a “positive” below a certain level.
Another implication is the conclusion of my recent blog post: that fast antigen tests should be increasingly used for community viral management and testing, tracing and isolation decisions. This is based on the likelihood that many false negatives for antigen tests are likely low viral loads and possibly not relevant for community infection control.
My hypotheses are that:
Antigen tests (at least good ones) are sensitive enough to accurately detect viral loads above a certain threshold.
The difference in sensitivity between PCR and antigen tests is primarily for cases for which the viral loads are very low, and likely not infectious.
The false positive rate for antigen tests is low.
If we could have rigorous scientific studies confirming these hypotheses, this would motivate a major shift in our national mitigation strategies as we move towards fast antigen tests for surveillance and screening for infected individuals to quarantine. In particular, it could be used for quick assessment of whether someone with covid-related symptoms is indeed infected and likely infectious, and also for contact tracing to quickly assess which of those who have been exposed to an infectious individual are indeed positive and likely infectious and thus should be quarantined.
Incorporation of fast tests would allow very aggressive testing and tracing without having to quarantine huge swaths of people who are not infectious or even infected, a current problem of current strategies that hinder the efforts. This would allow society to function at a higher level while protecting against viral surges, e.g. allowing more businesses and schools to open while maintaining a semblance of viral monitoring and control.
The USA government just purchased 3 month supply of 150 million antigen tests. They have not said where they will be used but my great hope would be that they would use them to help keep virus under control at public schools so we can get kids back into the classroom and keep them there in communities where number of infected is sufficiently low to open.
If all of this is true, then the people who dismiss antigen tests because they are not as sensitive as PCR tests may be strongly misguided.
Side note: they really need to start counting antigen tests in the testing and case numbers in ALL STATES, and RIGHT AWAY, so we don't lose track of the viral spread as antigen tests get increasingly used. I briefly highlighted this problem in a recent short blog post, and may investigate this more deeply and report in an upcoming post.
