COVID Tests: Sweet Little Lies…

As we are reaching the end of the second lockdown in England, I can’t help but sympathise with the prime minister, all the ministers and the lead scientists who manage the pandemic. They managed to bring themselves in the impossible situation to have shut down the economy, have dismantled every social bond at every level, have instilled agoraphobia and fear of social contact to every single person in the society, plummeted public funds to extinction, sky-rocketed the national debt and kept everyone in-house for extended periods of time and, yet, the infections (as identified by the PCR tests) keep rising.

My sympathy lasts only a second though as I consider why we observe this unprecedented disaster under the pre-text of scientific advice and practice. My real feelings about the situation are very far from sympathy but they are irrelevant to this post.

Since the very first moment of the pandemic, I got the impression that there is nothing scientific in the way it is dealt, approached and managed. There are many examples I can bring (e.g. the absurd idea that the virus survives hours and hours out of a living host, the possibility to catch it from a piece of paper or from the handle on a door, the chances to get infected by someone who just happened to walk next to you, the improbable high and dominant percentage of “asymptomatic” patients we are told, etc.). In this post, I would like to explore a grave misunderstanding (and, consequently, misuse) of the swab (PCR) tests to detect infected patients.

The second phase of managing the pandemic (what we are into at the moment) is to deploy PCR tests to the community (this is the Pillar 2 as the government calls it). Then, subsequent actions will be taken based on the number of infections the test identifies. This sounds solid and scientifically correct; no doubt about it.

What destroys the strategy is that this test is not suitable for what it is used; it can not detect the presence of COVID-19 (but it can identify the broader family of coronaviruses) and it is extremely unreliable when you do it in your living room. The first reason is beyond my expertise; the second reason includes many factors; anything from the way the personnel in the lab treats the collected cells to the way it is stored, analysed and processed.

Above all, the test is extremely unreliable because it comes with high level of falsely identified positives when the number of infections in a population are low. Let me repeat; if the infected people are not that many compared to the population that is being tested (scientists call this prevalence), the test identifies a good number of patients as positive when they are not in reality. If you are more familiar with hypothesis testings, the idea is similar to Type I and Type II errors.

This is not unique to the PCR tests. Every diagnostic method suffers from false positives. It is the nature and the limitations of the instruments and processes that introduce this error. This is well known to doctors and they take every precaution to minimise this possibility. That is why every time a flag is raised by the results of a medical test, the doctor is asking you to repeat the test in one week, one month or six months or whenever they deem necessary. This repetition will remove the random chances your case faces to fall under a false positive.

What is the case with the PCR test? How important are the false positives? Let’s work out what it means in practice to use such a tool.

Assume we test 100,000 people and there are 100 people who are really infected by the virus. Now, let’s pick a random value for the false positives. Let’s say that the level of false positives is 1.0%. So, given the population of 100,000 people and 100 really (true) infected patients, how many people do we expect the test to show as infected when they are not (false positives)?

The quick answer would be 100 * 1.0% = 1 false positive. Just 1 person; so what is the big deal?

Here is the catch. The false positives are not on the identified cases but on the tested population. This means that the false positives in our example will be 100,000 * 1.0% = 1,000 😮  :o. Wow…the test will tell us that apart from the 100 true positives we have 1,000 people flagged as infected when they are not.

Just imagine the daily briefing by the government: Today the infections have mounted to 1,100 (100 + 1,000) people.

How much wrong is this figure? Now we have 1,000 / 1,100 = 0,9091 or 90.9%—Impressive?  🙄 🙄

Wait, it becomes better now. It is not only that most of the identified cases in the tests are not really infected.

What is extremely worrying and depressing is that the scientists have no idea how much the false positive percentage of the test is. I am sure we all agree this is very dangerous situation.

Don’t take my word on this. Read the consultation report the SAGE committee received before they decided to move on with the use of PCR to test the population. You can find the report in the official website of the governmental publications (here). If you would like to read a more scientific analysis of the false positives, a good starting point is this page.

In the first page of the report, we read

It is important to remember that laboratory testing verifies the analytical sensitivity and analytical specificity of the RT-PCR tests. They represent idealised testing. In a clinical or community setting there may be inefficient sampling, lab contamination, sample degradation or other sources of error that will lead to increased numbers of false positives or false negatives. The diagnostic sensitivity and diagnostic specificity of a test can only be measured in operational conditions.

Operational false-positives and false-negatives will have significant impact in the way we respond to the COVID-19 pandemic. They will affect national surveillance, and the functioning of the UK trackand-trace programme. We have been unable to find any data on the operational false positive and false negative rates in the UK COVID-19 RT-PCR testing programme.

After this, there is not much to say; I will leave you with a song to cheer you up  😀