Professor Drosten in the dilemma of the PCR test

Krankenhaus mit Bett im Flur

On 4 February 2021, the Heidelberg Local Court appointed Professor Dr. Christian Drosten as an expert in an administrative disciplinary proceeding. This was at the request of lawyer Beate Bahner.

According to Christian Drosten, “a written expert opinion is to be issued” on “the defence lawyer’s assertion that a PCR test could not prove an infection within the meaning of Section 2 of the Infection Control Act”.


Even the best PCR test cannot provide “direct pathogen detection” (RKI), let alone determine an “infection” in the sense of the IfSG, because the term assumes the presence of a “pathogen capable of reproducing” in the concerned organism. Such a thing categorically cannot be diagnosed directly, neither on its own, nor by a PCR. This is not a matter of quibbling over words or semantics, but of the legal precision that legislation and jurisprudence owe to citizens.

In other words, anyone – whether expert or layman – can in principle understand “infection” or “Covid 19 case” as they wish, and formulate their definition in such a way (vague or open) that PCRs – irrespective of the effective test design and practical handling – appear to be sufficient detection instruments. On the contrary, the legal definition of “infection” leaves comparatively little room for interpretation. In addition, the technical functioning of the PCR procedure, in general, as well as the scientific requirements for good tests and their correct handling, in particular, are also subject to judicial fact-finding (objectively applicable standards).

Expert Drosten appointed by the court

The explosive nature of the court case lies in two levels. First: So far, courts have always taken for granted the suitability of PCR tests for the detection of infections in corresponding COVID proceedings and have never questioned it (as having been clarified by experts first). Secondly: It makes a difference whether one tells the government, the media and the population one’s expert opinion in a supposedly or actually legal grey area that PCR tests are reliable (for whatever purpose) and thereby possibly believes that only the politicians are concerned with a criminal and civil liability for the legal conformity of the decisions based on PCR tests, or whether one is forced to explicitly express oneself truthfully as a publicly appointed expert on the relationship between PCR tests and “infection in the sense of the law”.

If Drosten, as one of the government’s advisors, disputes or relativizes in court the validity of PCR tests for the determination of infections within the meaning of the law, this could have far-reaching consequences for subsequent COVID trials and large parts of COVID policy. If he succeeds in getting around it in the short term, a deliberately false opinion could lead to his personal undoing in the medium or long term because it’s a criminal offence. Finding a way to avoid writing an opinion,doesn’t make a good impression either.

Tools that are of limited value on their own

It is worthwhile, then, to reassemble what is becoming known about the PCR process in general and the genealogy of COVID testing practice in particular, as well as about Christian Drosten’s actions in and contributions to this history.

In general, laboratory tests in medicine are merely aids that are of limited significance on their own and can only confirm the suspicion of specific diseases and causes of certain symptoms in conjunction with the patient’s medical history and differential diagnostics. When throat and nasal swabs are examined for cold viruses using PCR methods, their collection already contains more potential sources of error than, for example, the handling of blood samples. In the case of COVID, RNA viruses are also involved, but the PCR method only works with DNA.

Therefore, (single-stranded) RNA has to be isolated by means of an enzyme (Revers Transcriptase) into (double-stranded) DNA. (This is why the procedure is also called RT-PCR). This is preceded by lysis, which breaks down the cells and, if applicable, viruses contained in the smear into their components, shredding them to a certain extent. Even if intact and complete RNA viruses are present in the sample before the PCR technique is used, the test can only detect DNA gene fragments in the course of the PCR procedure – and only after many multiplication cycles of the material.

The conclusion on a) intact and complete RNA viruses originally present in the sample in b) reproducible numbers (viral load, infectivity according to IfSG) is thus indirect from the outset (and not direct) and depends on two factors in particular: One searches for pre-determined cDNA fragments (corresponding DNA), which are supposed to correspond to certain RNA fragments, which are attributed to the searched RNA virus as specific components. To reasonably ensure that one is not just detecting virus debris and fragments (indirectly), but can justify the suspicion of a complete virus, one would determine three target RNA genes to be found, located at the beginning, middle, and end of the virus strand. If all three target cDNA fragments then detect, the probability is quite high that an intact RNA virus was present in the smear prior to shredding by lysis.

Conclusions about the relevant presence of an effective virus as well as its ability to replicate (viral load) have something to do with the number of replication cycles needed to detect the fragments it is looking for. The more runs it takes to make the sought-after fragment visible, the less it was originally present. This is why threshold values are set for the number of multiplication cycles (Ct value). Since about September 2020, there is a consensus, including WHO, Drosten, RKI, that positive PCR test results based on more than 30 to 35 cycles are worthless or null. Positive detections below 30 to 35 cycles do indeed indicate “viruses” (really positive but not necessarily infectious), but only with less than 15 to 20 cycles would the assumption of a viral load be justified as sufficient to substantiate a suspicion of infection and infectiousness in connection with appropriate anamnesis and differential diagnosis (really positive in the sense of a possible infection/illness according to IfSG).

Scientifically and legally, only those PCR test results should have been reported as positive (suspected) cases of infection/disease in which, with regard to symptomatic people only, all three corresponding target fragments (semi-specific for SARS-CoV-2) are detected after less than 20 cycles. In addition,as the pre-test prevalence of the virus decreases in the population through testing of symptomless individuals, the probability of more false positive test results increases enormously. This is a problem of health policy or test strategy that has nothing to do with the quality of the test design, whichcannot be circumvented even by excellent tests.

The Drosten Test

In January 2020, WHO had already recommended a PCR test protocol (in several versions) for the detection of the new coronavirus as a guide for laboratories worldwide, which then became a further developed article published into the journal Eurosurveillance . The main responsible person (author as well as contact person) for the protocols and the article is Christian Drosten. From the beginning of the “Corona crisis”, doubts were articulated (e.g. by Wolfgang Wodarg) about the informative value of the “Drosten test” and about the historically first use of the PCR method for mass testing (even on symptomless persons) completely decoupled from individual anamnesis and differential diagnosis.

Then, at the end of November 2020, a consortium of scientists requested to Eurosurveillance to withdraw the Drosten study.

https cormandrostenreview

Complaints were made about conflicts of interest (some of which were not disclosed or were only disclosed later) of some authors who earn money from this PCR and/or contribute to the editorial work of Eurosurveillance which in itself lends an additional taint to the fastest peer review process (27 hours) in history, which is already questionable. More crucially, however, there are numerous meticulously demonstrated craft flaws in the test (primer design, temperatures, etc.) that a proper peer review process should have detected, and that lead to the production of false or null positive results (improperly claiming proof of infection). Later still, an Addendum which deals, among other things, with difficulties of the Drosten protocol in laboratory practice. Read here.

I will limit myself here to two shortcomings, which concern the target gene fragments and the number of cycles. They are understandable even to laymen in molecular biology or biochemistry, and have a politically consequential aftermath.

1.) The initially three, later only two targets of the Drosten protocol (dual-target test) are located between the middle and the end of the virus strand, none at the beginning. Irrespective of the question of whether the specific targets are sufficiently specific for SARS-Cov-2, it is not possible to distinguish between detected virus debris or gene segments and a complete virus.

2.) On one hand, the protocol does not specify any Ct values, neither for the question of whether a sample can still be meaningfully evaluated as positive, nor with regard to a viral load that suggests an infection or infectivity. On the other hand, it itself operates at 45 cycles. Laboratories using this protocol as a guideline according to the WHO recommendation thus had two options: Set Ct values to their own liking or use the protocol’s 45 cycles as a guide. Either way, there has never been a calibrated laboratory practice for a “positive report” (“new infection”) (but rather a chaotically non-uniform test landscape) and reports based on 40-45 cycles (which were common practice in many places) have in reality all – not only in the case of symptomless persons – been false positive or null positive.

Eurosurveillance has taken two months to reply to the criticisms. In an explanation dated early February 2021, itstates, as expected, that one sees no reason to withdraw the Drosten study. The statement does not address any of the criticised technical deficiencies, but instead falls back on an argumentation that remains general. The PCR test protocol was developed in a hurry during an emergency and was therefore as scientific as it could have been, given the limited available information at that time and the time pressure. “With more data and the knowledge evolution, laboratories have since further improved the original method, as a standard practice.”

Leaving aside the question of what the emergency should have really been in January 2020, having to develop a PCR test overnight with limited information pn a virus which was considered by Drosten ,until March. as a commonand unnoticed cold for the majority, it is worth having a look at Drosten’s effective contributions to the alleged increase in PCR quality, – originating in cooperation with the WHO – which, subsequently, are in reality to be assessed as intentional deteriorations: They concern Ct values and the number as well as specificity of target gene fragments. Read here.

Drosten and the establishment of single-gene PCR

The WHO protocol of the Drosten test of 13 January 2020 targeted three gene segments within the right half of the viral strand: RdRp (often abbreviated by laboratories as Orf[1ab]), E gene and N gene. In the second protocol of January 17, which replaces the first, only two target genes exist. The N gene is omitted, further reducing the distance between the targets (the area of the viral strand they cover). The Eurosurveillance article of 23 January then presents a mix of both protocols, declaring an N-gene test dispensable and recommending “for a routine workflow” a dual-target test for the RdRp (Orf) and E-gene. Read here.

While the Robert Koch Institute always claimed that mainly dual-target tests were used in German laboratories (bad enough: instead of 3-gene tests), it became increasingly clear in the course of April 2020 and the following months at the latest that many laboratories only tested for the non-specific E gene of all things (single tests) or, in the case of dual-target tests, interpreted the detection of the E gene alone as sufficient to assess the entire test as positive and report it accordingly as a “case”. These laboratories relied on a passage in the WHO Test Guide for Laboratories, versions of which read as follows on 2 March 2020 and 19 March 2020: (See here and here)

In areas where COVID-19 virus is widespread, a simpler algorithm could be applied, where, for example, screening by rRT-PCR of a single discriminating target is considered sufficient.”

The WHO Drosten workflow protocol in its second version dated 17 January 2020 provides for three stage testing (using two target genes) – 1. First line screening E gene. If positive, then 2nd confirmation test (RdRp). If positive, then 3rd discrimination test (RdRp). If positive, then positive overall result. The reduction of the test procedure to E-gene screening alone, as carried out by the laboratories with reference to a WHO recommendation, is in contrast to this a further deterioration in PCR quality. The complete opposite of the Eurosurveillance claiming increase in quality. As external advisors to the recommendation, the WHO gives, among others, the names of the three authors of the two WHO PCR protocols and the Eurosurveillance-article: Maria Zambon, Public Health England, UK; Marion Koopmans, Erasmus MC, Rotterdam, The Netherlands; and then only: Christian Drosten, Charité – Universitätsmedizin Berlin, Germany. Drosten may not have caused this reduction in the quality of PCR testing and interpretation practice, but it is difficult to imagine he was not aware of the weaknesses of the recommendation – and their logical consequences for test design and handling in laboratory practice. Read here

He knows what he’s doing (1)

Drosten is well aware of the unreliability of single-gene tests, indeed of the difficulties of the PCR method as a whole. He is only not interested in this when quarantine decisions are based on it or incidences are reached that are supposed to legitimize restrictions on the freedom of the population. He only reacts as soon as articles appear that question the Wuhan myth, i.e. prove by means of PCR that the coronavirus was already existing months before in France, for example. Drosten posts this in a podcast in May 2020 (Episode 40) :

A PCR test, it must be made clear, must first be regarded as doubtful as long as it is not confirmed by further PCR tests that detect the virus in other target regions of the genome. Especially in such important finding, when it is not a normal routine operation in the laboratory, where one simply wants to know, this is a standard diagnostic case: Is it now positive or negative? Then you can say: PCR is positive. We consider the patient to be infected. / Korinna Hennig: “In normal everyday life.” / Christian Drosten: “Right. But in a case like here, where you say, we rewrite the infection history of this disease and say: In reality, there was this in France and then yes probably everywhere else in the world a month earlier or even longer. And something may have been concealed or not noticed. If you want to publish such a weighty finding, you have to back it up. This would include, in addition to a second or third confirmatory PCR, also sequencing the virus, i.e. determining the entire genome sequence of the virus. This can be done if the PCRs become positive. That’s technically very easy these days.”

The site corodok comments aptly:

For Drosten, there is the “important finding”, and if you want to “publish it” as an ambitious scientist, then “you also have to back it up”, which is done by “in addition to a second or third confirmatory PCR, also sequencing the virus”: three times PCR plus sequencing, because a PCR test “is first to be considered doubtful as long as it is not confirmed by further PCR tests that detect the virus in other target regions of the genome.” At least. And then there is the “routine operation”, namely the examination of humans, “where one simply wants to know, this is a standard diagnostic case: is it positive or negative? Then you can already say: PCR is positive.” For real people, then, a single PCR is apparently enough, just as he co-decided with the WHO – even if it is actually “to be regarded as dubious,” as he knows very well. Real people simply don’t interest him, our government advisor. Accordingly, the advice. And the consequences.

In this context, let us also recall the now infamous Drosten quote on MERS in 2014 on the relationship between media and PCR testing practices in Saudi Arabia: (Read here)

It’s just that until now there was a clear case definition, a strict scheme that determined which patient was reported as a Mers case. This included, for example, that the patient had pneumonia in which both lungs were affected. But when a whole series of Mers cases suddenly appeared in Jeddah at the end of March this year, the doctors there decided to test all patients and the entire hospital staff for the pathogen. And for this they chose a highly sensitive method, the polymerase chain reaction (PCR) […], but the method is so sensitive that it can detect a single hereditary molecule of this virus. If, for example, such a pathogen flits across a nurse’s nasal mucosa for a day without her falling ill or noticing anything else, then she is suddenly a Mers case. Where before deathly ill people were reported, now mild cases and people who are actually perfectly healthy are suddenly included in the reporting statistics. This could also explain the explosion in the number of cases in Saudi Arabia. Add to that the fact that the local media has blown the whole thing out of proportion. […] I fear that the current increase is more due to increased attention. It’s no different in this country. If “Bild” or the evening news reports about an outbreak of a certain virus, the number of laboratory tests increases significantly. Simply because doctors are then sensitised and specifically look out for the pathogens that are being reported.”

In 2020, there had been no significant changes in the PCR technique compared to 2014.

But on September 30, Drosten stresses in the Tagesspiegel the “reliability of PCR tests for the SARS-CoV-2 coronavirus.” He says: “Without a full virus genome there are no virus residues.” The PCR method is “simply beyond doubt” and offers “a very watertight diagnosis”. (Read here)

No worldwide uniformity regarding the PCR test targets

There has been – Drosten knows because he was involved in it – no uniformity worldwide with regard to PCR test targets, which actually precludes comparing “infection events” between countries:

The nearly 30,000 nucleotide genome of SARS2 hosts just over a dozen open reading frames (ORFs). However, only about a handful of these are used as target sequences, and the reference laboratories of the individual countries sometimes choose different targets (see WHO protocol collection). In China, for example, these are ORF1ab and N-Gen (nucleoprotein). At the Charité in Berlin, on the other hand, the primers are directed at RdRP (RNA-dependent RNA polymerase) as well as E- and N-gene. The Centers for Disease Control in the US, on the other hand, favors targets in the N gene, while the Institut Pasteur in Paris targets two sequences in RdRP.” (

Even within a country – at least in Germany – there was no uniformity in test designs and testing practice, no binding specifications and quality controls, for example by the RKI. The laboratories could obtain kits oriented to the Drosten test or other kits from manufacturers or build their own in-house tests. What the laboratories reported positive was considered positive (even as evidence of a (new) infection or infectivity or as a “Covid 19 case”). Which target gene fragments and how many of each detected, no one wanted to know for sure. From the beginning until today, Drosten deliberately omits to point out technical undesirable developments in the mass testing of humans, which he is aware of and which he helped to bring about, because for him these are simply not “weighty findings”.

Establishment of 40 to 45 cycles

The same chaos may have prevailed with regard to the threshold values for the amplification cycles. Although it has always been a banality for connoisseurs and practitioners of the PCR method that the technique has no built-in automatic yes/no or positive/negative function, that, roughly speaking, one can get anything to glow with endless amplification cycles without real relevance, reasonable Ct values were not even discussed in relevant publications during the first months. The Drosten Protocol, worldwidely recommended by the WHO mid-January 2020, simply does not address the issue of Ct values and itself operates with 45 cycles. Even the test guide for laboratories (see above), which was developed two months later with the assistance of Christian Drosten and published by the WHO on March 19, still does not address the Ct value issue, i.e. a guideline for distinguishing positive/negative, without ambiguity. There is only one – but hidden – hint in the sense that in a completely different context a Chinese study of 29 January which is apparently the only one to date giving any account of the Ct values to use as a basis: (Read here)

“Un valore di soglia del ciclo (valore Ct) inferiore a 37 è stato definito come un test positivo, e un valore Ct di 40 o più è stato definito come un test negativo”.

On the one hand, this Ct value only differentiates between “we formally found something” (positive after less than 37 cycles = positive) and “what we are looking for is not present in a relevant form” (positive only from 40 cycles = negative) with a grey zone between 37 and 40 cycles, while no Ct value is required to determine a sufficient quantity of the gene fragments to be found (“virus” load) for a suspected infection. On the other hand, the WHO did not even include this as a rough guide in its guidelines. Thus, it was up to the discretion, the arbitrariness of the laboratories, to define how many cycles their reports of “positive cases” or “new infections” were based on. Since the WHO left the question open, as did the Drosten protocol, which operated with 45 cycles, positive reports of over 37 cycles were probably commonplace internationally (while hardly any “cases” were found in China).

After six months of “testing, testing, testing”, it took until early September 2020 triggered by a New York Times article from August 29, to see the Ct values discussed in public, also in German leading media, not only as Ct value for positive/negative cases, but even one for potentially infectious/non-infectious (“virus” load).

In the Tagesschau of 06.09. for example, it was said: (Read here)

The Ct value gives an indication of the amount of virus a patient carries. It indicates how many rounds the PCR must run before viral DNA is detected. In a patient with a lot of viral load in the body, the test often hits after 10 to 15 rounds of CT, say laboratory physicians. But if the PCR takes more than 30 rounds to detect viral material, a patient is very likely no longer contagious at all. […] A stir was caused this week by an article in The New York Times that reported that test data from Nevada, Massachusetts and New York suggest that up to 90 percent of PCR tests show Ct levels so high that patients barely had any virus left. Harvard University epidemiologist Michael Mina therefore argues that the Ct threshold should be set at 30.

For the testing practice in Germany, “the research of WDR, NDR and SZ” did not show a better picture. Here, too, “many laboratories that evaluate the PCR tests […] do not stop the analysis at a Ct value of 30, but usually only at 37 or 40 […]”. In addition, a survey of health authorities had shown “that the value is often not transmitted at all. However, this leads to the fact that without a Ct value, the health offices usually also have no indication of the degree of infectiou of a person who has tested positive. The laboratories also confirm that notification of the Ct value to the health authorities is not included in the procedure. This means that quarantine decisions by health authorities and reports of “new cases” to them and to the RKI between March and September were not made on the basis of the best possible knowledge.

Relatively suddenly, and as if it were the most natural thing in the world, the RKI and Drosten now also spoke publicly about the fact that all positive PCR test results of more than 30 cycles mean a low viral load. However, instead of setting the laboratories in future to Ct-values of maximum 30 for a mere positive interpretation and of less than 20 for a report of a suspected infection to be confirmed by anamnesis and differential diagnostics and to re-evaluate all “cases” of the past counted until then on the basis of findings data to be retrieved, they opened a discussion which, in view of the fact that for months legally relevant work was carried out with far too high cycles, represents a sophistical and pompous sideshow: (RKI and Archive)

Drosten, of all people, whose orientation protocol works with 45 cycles, now said that Ct values were too imprecise and that it was important to determine the sample volume fraction as an absolute reference value, from which more precise Ct values could be derived for the laboratories with regard to the machines they use and other differences, such as 28 cycles for one laboratory, 30 for the other. The RKI argued similarly, and Drosten promised (see source above) to work with the RKI experts to determine such a reference value in the near future. (For more detail: here and here)

In the meantime ,the laboratories were neither committed to at least a rough Ct value of 30, nor did Drosten and/or the RKI publish a sample volume proportion as a reference value until today (end of February 2021). Rather, the topic of reasonable Ct values disappeared completely from the German public again since end of September. Drosten, who just now agreed with the US-Americans that more than 30 cycles are problematic and that many positive findings between March and September were based on far more than 30 cycles, tells to the Tagesspiegel, end of last September, it is about “watertight diagnostics”, while at the same time he again refrains from any intervention to improve the obviously miserable diagnostic practice.

He knows what he’s doing (2)

Pandemic management policy is based on by broadcasting for months of fake news in the mainstream and social media about very statements that the WHO itself introduces a few months later as basic knowledge to remind to the laboratories. Overttime and again in March, April, May, June 2020, it was pointed out that PCR tests per se and alone cannot detect infections, and that poorly designed and poorly handled corona PCR tests can result in – depending on the epidemic intensity, test handling and target test population – 30 percent, 50 percent, 70 percent, 90 percent or more of reported “new infections” being false. On January 20, 2021 – one year after the Drosten PCR workflow was recommended worldwide, after months of single-gene PCR testing, of silence on Ct levels, of an international policy of action based solely on PCR test results … WHO publishes this information for “Users of in vitro diagnostic (IVD) medical devices”: (WHO)

WHO diagnostic test guidelines for SARS-CoV-2 state that careful interpretation of weak positive results is required. The cycle threshold required to detect virus (Ct) is inversely proportional to the Viral load of the patient. If the test results are not consistent with the clinical presentation, a new sample should be collected and retested using the same or a different NAT technology. WHO reminds IVD users that the prevalence of disease alters the predictive value of test results. The more prevalence decreases, the more the risk of false positive results increases. This means that the probability that a person with a positive result (SARS-CoV-2 detected) is actually infected with SARS-CoV-2 decreases with decreasing prevalence, irrespective of the claimed specificity. Most PCR tests are used as diagnostic help indicator. Therefore, healthcare providers must report each result in combination with the time of sample collection, sample type, test specifications, the clinical observations of the Patient history, the confirmed status of contacts and epidemiological information. Actions to be carried out by IVD users: […] Enter the Ct value in the report to the requesting health care provider.“ (Herv., T.M)

Confronted with this at a federal press conference by Boris Reitschuster, Drosten’s answer cannot be more audacious. (here and here).

Although also in Germany masses of symptomless people were tested (what even after RKI information 98 percent of these positive results are wrong), although in September, with Drosten taking part in the discussion, it became clear for a few days that too many cycles were being used in Germany and that the laboratories had not reported the Ct values to the health authorities, although many laboratories were demonstrably using bad single-E gene tests. Although, in the meantime, even Olfert Landt, whose company TibMolbiol distribute the Drosten test, which he co-developed, says that 50 percent of all reported “new infections” are not infections, Drosten claims that the WHO note in question should only be understood as a request to technically poorly set-up Third World laboratories to read the PCR instructions correctly. In any case, this is how he interprets the WHO note – and Health Minister Spahn seconds him saying: “That’s the beauty of our pluralistic society, that one can evaluate a piece of information differently.” (RKI information and fuldaerzeitung)

Drosten under pressure

On February 1, 2021, the Medical Journal says : “PCR detection is the standard test for diagnosing viral infections such as SARS-CoV-2. The test detects individual pathogen genes but not intact pathogens.” The same was written in May/June 2020 in the PCR instructions of the US-American CDC as well as in a Leaflet by BAG (Federal Office of Public Health – Switzerland) and Swissmedic: Read here.

“PCR (polymerase chain reaction) is a NAT (nucleic acid amplification technology) method used in modern molecular biology to amplify nucleic acids (RNA or DNA) present in a sample in vitro and then detect them using suitable detection systems. However, the detection of nucleic acid does not indicate the presence of an infectious agent. This can only be done by means of virus detection and multiplication in cell culture.”

For all the beauty of a pluralistic society, millions of quarantined people (especially the symptomless among them) may one day wonder on what scientific and legal basis each quarantine decision was actually issued. Even those suffering from the lockdown measures will increasingly want to know how many false-positive new-infection reports were involved in exceeding the 50- or 35-week incidence.

It is therefore likely to be of growing public interest how Drosten’s expert opinion answers the Heidelberg District Court’s question as to whether PCR tests can detect infections within the meaning of §2 of the Infection Protection Act. In any case, he will not be able to dismiss a judge who demands an expert opinion from him as simply as a critical journalist would do at a federal press conference.