Are PCR Tests Reliable to Detect Sars-CoV-2?

How a PCR test works. To perform a PCR test, a human sample is required. In the case of Sars-CoV-2, a smear is taken from the throat, which is accessed through the mouth or nose.

If the smear contains whole viruses or virus fragments, then it is possible to isolate the genetic material from the smear. In the case of the Sars-Cov-2 virus, it is RNA. First, an enzyme translates this RNA into DNA. Next, the DNA must be amplified to obtain a sufficient quantity to allow detection.

This amplification requires small pieces of DNA, called primers, which attach to the virus genome and copy it. In this way, more and more DNA chains are created in several passages. The name of the test is derived from it: PCR stands for Polymerase Chain Reaction. If the smear does not contain any virus, no copy is possible. On the other hand, if the sample contains the desired viral genome, the number of copies increases continuously with each passage.

Finally, the DNA copies at the end of the chain are made visible using fluorescent substances. The result is represented by a curve which shows the increase in DNA content as a function of the number of replications performed.

Does the PCR test detect infections? A PCR test can detect whether certain sequences of the genetic material of the Sars-CoV-2 virus are present or not in a sample. Most of the tests used detect two different sections of the virus genome, some even three. The higher this number, the more reliable the result.

But the argument that detection of these pieces of genetic material is not necessarily synonymous with infection is increasingly being used. The experts themselves do not seem to agree on this question. A glance at the leading German medical site Pschyrembel helps to clarify the situation, because an infection is defined there as:

“The entry of pathogenic microorganisms, such as bacteria or viruses, into an organism with subsequent colonization and reproduction.”

As soon as the viral RNA is detected, the first part of the definition is fulfilled: the virus has entered the human body. This is usually done passively, for example through the mucous membranes. But what does the test say about the second part of the medical definition? Has the detected virus already multiplied and therefore infected its carrier?

In principle, it is possible that a virus can be detected that has not yet multiplied. But Jürg Böni, head of the Department of Diagnostics and Development at the Institute for Medical Virology at the University of Zurich, estimates that these results represent only around 0.1%.

“The amounts of virus are then so low that the result is weakly positive at best.”

In such a case, the result should be checked. This can be done either by re-analyzing the sample or by re-testing the affected person. However, these weakly positive results do not appear as positive cases in the statistics, Jürg Böni recalls.

“At the University Hospital Zurich, we only communicate results that are clearly positive. In these cases, we find up to several billion copies of the viral genetic material ”.

Such concentrations are not possible without multiplication of the virus and therefore acute infection. However, the PCR test does not tell us whether the focus of infection is still active, or whether the person is still contagious. It is not known whether the RNA sections detected are only fragments or a complete virus. A positive result without infection is therefore basically possible, but it is not counted as a case.

When can the virus be detected? Usually, the test is positive one to two days before symptoms appear and up to two to three weeks after. The viral load in the body is highest when symptoms start. The test detects an infection, even if the person has no symptoms. However, these asymptomatic cases – contrary to previous assumptions – only account for around 20% of all infections, according to a recent study from the University of Bern. This figure is based on an analysis of 79 studies published between January and June of this year and includes data from more than 6,500 patients.

The diversity of PCR tests. There is not just one but countless different PCR tests from different manufacturers. The first PCR testing protocol was developed by a team led by German virologist Christian Drosten at the German Infection Research Center of Charité in Berlin in January 2020. The University of Zurich has also used this protocol in a light form adapted in the initial phase of the health crisis.

Soon after, commercial tests from China appeared on the market. However, these had isolated problems of impurities and were not approved for diagnostic purposes in the European Union. In the meantime, however, a large number of different PCR tests have been tested and CE marked. Jürg Böni confirms this:

“Most labs have now moved from homemade testing to commercial testing.”

Each laboratory is free to choose the test it wishes. But he is obliged to validate the reliability of the test, explains Jürg Böni. It is only when a test is reliable, i.e. correctly identifying people with and without Sars-CoV-2, that it is used in practice. The test results are evaluated using control samples which are systematically analyzed during each PCR.

Circular essays. To check the correct functioning of the various tests and the cleanliness of the work of the various laboratories. The scientific association Instand eV carried out circular tests in April 2020. The independent German institute sent positive and negative samples of Sars-CoV-2 to 463 laboratories in 36 countries.

The results are there: Regardless of the genetic area examined, the tests showed between 98.9% and 99.7% correctly attributed positive results and between 97.8 and 98.6% correctly attributed negative results. However, this is an average value for all laboratories – some tests are more reliable, others less. “These cross-comparisons help laboratories see where they stand,” says Jürg Böni. And to assess whether another test would not give more reliable results… In this way, the accuracy of the test can be continuously adjusted.

Double control. In addition to internal laboratory checks, the Federal Office of Public Health (OFSP) has introduced a double check of results. At the start of the pandemic, all laboratories sent their samples to the National Reference Laboratory in Geneva for testing. However, as the number of samples quickly exceeded the capacity of the national laboratory, other large centers, such as the University of Zurich, joined in to serve as control bodies as well.

Are positive cases increasing because more tests are done? It’s mathematical: the more we test, the more infected people we detect. But in terms of proportions, the increase we see today is not only linked to the increase in the number of tests as is sometimes claimed.

Two facts prove it:

• So far there have been three phases of increased testing in Switzerland: At the start of the pandemic in March, then in June and finally the current period. The second phase, in June, is particularly significant: the tests have increased, but not the number of positive cases.

• The real increase in the number of infections is demonstrated by the constantly increasing proportion of positive tests compared to the total number of tests performed. Just a month ago, between three and four percent of tests were positive. Currently, that figure is eight to ten percent. And the trend is on the rise.

A big difference with the first wave. Currently, the number of cases is increasing every day. However, one factor is decisive in comparing the current figures against those of the spring: the number of unreported cases. According to a study by the University of Bern, this factor was ten in March. Today, in contrast, it is only two or three, as epidemiologist Christian Althaus explained in an interview with Tagesanzeiger.

This can be explained by the testing capacity, which is much lower in the spring. The study is only preliminary and has not yet been peer reviewed. However, if these figures are correct, around 10,000 people would have been newly infected each day in March. At present, that figure would be only two to three thousand people per day.

This article was originally published by our editorial partner Higgs.ch.