PCR Tests: Not Perfect But Not Bad Either

Zoonotic diseases are diseases that are transmitted from animals to humans and they have been in existence (as any search will yield) since the origin of Homo sapiens 200,000 years ago or, roughly, since Adam.¹ Given a globally increasing population density, where the current total is 7.9 billions, one ought to anticipate an increase in zoonotic diseases of which coronavirus 2 (SARS-CoV-2) is a typical example.

A characteristic of all viruses is that they mutate. The current Covid-19 is in its fourth mutation as per the Greek alphabet : alpha, beta, gamma and now delta. Another characteristic of viruses is that they seem to become more virulent with each mutation. It is an arms race after all. One might consider the process a the quintessential vindication of Charles Darwin and Alfred Wallace.

For any given test, including jury trials, the result of the test may be factually correct or incorrect. In the case of the test being correct a negative result conveys no disease (or whatever) existing and similarly for a positive result. In the case of the test being factually incorrect a negative result may occur when one (in fact) does does have the illness and similarly for a positive test when one does not have the illness or AIDS or Cov-19 or is pregnant or a jury convicts someone who is (in fact) innocent.

A means of detecting a coronavirus is a Polymerase Chain reaction or PCR test. PCR tests are utilised to identify the presence of viral coronovavirus RNA which is detectable prior to antibodies forming or symptoms of Covid-19 being present.² Thus the tests can determine whether or not someone has the virus very early given an infection. PCR gives a reasonable ability to detect infection but some studies place the incidence of a negative result when one does have Covid-19 at 30% or just under a third of all tests. Such a 'failure' is known as a Type II error or the guilty person goes free' in the case of a jury. A positive result when one does not have the illness is known as a Type I error. The Virology Journal[^3] reported a 5/52 or close enough to 10% Type II rate (sometimes reported as a 'false negative'). In other words, five of the fifty two cases did in fact have Cov-19 from the 52 negative results.

Arevalo-Rodriguez et al.³ place the rage of Type II error rate (or false-negative rate) from 0.018 to 0.58 or from 2% to almost 60% with a working rate, given the conclusion of 54% which does seem high. Such is life. We don't live in a perfect world. Moreover, some people may be asymptomatic where their infection does not yield any symptoms at all but these "carriers" are able to transmit a virus to others.

Nevertheless, let's consider the more panoramic view. The PCR test, irrespective of Type I and Type II errors, is a point in time statement; not unlike a company balance sheet (as at a given date). It is entirely possible, everything else being equal, that the person who did not have the virus when tested acquired the virus the following day. All viruses require an incubation period with most returning a result after 14 days of infection. About 1% of the population will return a result some time after 14 days if infected. There are no guarantees.

What is novel is that the governments about the world are endeavouring the manage the virus whereas the task is the prerogative of the country's health department with the government merely taking advice from its Health department. It is beyond the length and scope of this article to illustrate, globally, that electoral considerations are influential.

We might argue as to whether smallpox has been eradicated or merely no cases in recent decades. All viruses are able to remain dormant outside of a host. Allowing for the doubtful case of smallpox viruses cannot be eliminated. The best that can occur is that the antibodies in hosts become able to cope with successive variants of the virus; which is quite a different matter.

It is not in the "interests" of the virus to kill the host but it is in the interest of the virus not to be outdone by the antibodies and hence the case of increasing virulence; even in the case of bacterium, against antibiotics, for that matter. The good news is that no virus has eliminated Homo sapiens for more or less the same reason that Homo sapiens have never eliminated a virus (we can argue over smallpox).

The crude theoretical rate for deaths and permanent disability is about 2.5% of 40% of a given population or or about 1% of the population or about 50,000 in the case of New Zealand. The observed (or practical) rate for Sweden is 1/10 of the theoretical. Indeed the death rate in Sweden is 1.3% of infection. If we were to assume another 1.3% for permanent disability the odds are remarkably good for recovery.

The over 60s are semi-vulnerable but the principal incidence is with the age 80 plus (who are likely to die in any event).⁴

By way of a conclusion, there is doubt with all medical tests in terms of Type I and Type two errors. Such is merely a case of not living in a perfect world. As a general guide the tests tend to improve over time. In the case of Covid-19 we are here for "the long haul". Given government involvement October 2022 is going to appear synonymous with October 2021 and possibly October 2023. The vaccinated need not worry. It is the anti-vaxers that will ‘Darwin-ise’ themselves out of existence.

[^3] https://doi.org/10.1371/journal.pone.0242958